LATERITE AND LATERITIC SOILS

AND OTHER PROBLEM SOILS

AFRICA

A n Engineering Study
for

AGENCY FOR INTERNATIONAL DEVELOPMENT

AID/csd-2164

LYON ASSOCIATES, INC.

I
Baltimore, MaryIand, U. S.A. 1
I

BUILDING & ROAD RESEARCH INSTITUTE,

t Kumrsi, Ghana
-
- -.
i

1
BESTAV/4!LABLE COPY :
LATERITE AND LATERITIC SOILS
AKD OTHER PROBLEM SOILS

AFRICA

DEVELOPMENT OF ENGINEERING STAND-

ARDS AND CRITERIA FOR THE USAGE OF
LATERITE, LATERITIC SOILS, AND OTHER
PROBLEM SOILS IN CONNECTION WITH THE
CONSTRUCTION OF ROADS, HIGHWAYS AND
AIRFIELDS

An Engineering Study
for
AGENCY FOR INTERNATIONAL DEVELOPMENT
AID/csd-2164
June, 1971

LYON ASSOCIATES, INC.

Baltimore, Maryland, U. S.A.

B'CI'ILDING & ROAD RESEARCH INSTITUTE,

Kumasi, Ghana
 TABLE OF CONTENTS

Page No .
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix-x-xi

Chapter 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Chapter 2. Engineering Implications of Tropical Weathering and Laterization . . . . . . . . . . . . . . . . . . . . 5
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Weathering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Profiles of Weathering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Hydrated and Dehydrated Laterite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Chapter 3. Regional Setting and Genetic Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Summary of Physiography and Geology of Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Geology of Ghana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Soil Forming Factors and Profile Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Chapter 4. Literature Review on the Engineering Characteristics of Lateritic Soils . . . . . . . . . . . . . . . . . . 29
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Chemical. Mineralogical and Physico-Chemical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 29
Physical and Engineering Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Density and Compaction Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Consolidation and Permeability Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Utilization of Lateritic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Stabilization of Lateritic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Chapter 5. Engineering Properties of Tropically Weathered African Soils . . . . . . . . . . . . . . . . . . . . . . 39
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Soil Grouping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Correlation of Physical and Engineering Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Effects of Drying on Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Summary of Findings Regarding Major Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Collapsing Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Chapter 6. Test Procedures and Pretreatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Durability and Aggregate Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Conclusions and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Chapter 7. Air Photo Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Identification of Laterites and Other Tropical Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Color Photography for Present Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Description of Plates 1. 2 and 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Description of Plates 4. 5 and 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Chapter 8. Suitability of Current Engineering Classification Systems for Tropically Weathered Soils . . . . . . . . . . . 83
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
 Applicability of Existing Classification Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Chapter 9. Stabilization of Selected African Lateritic Gravels and Soils . . . . . . . . . . . . . . . . . . . . . . . 87
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Procedures for Lime and Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
pHMethod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Sand Stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Asphalt Stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Asphalt Stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Chapter 10. Proposed Pavement Test Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Selection of Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Base Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Load Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Regional Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Number of Test Lengths Within One Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Lengths of Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
. Construction Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Testing Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Chapter 11. A Survey of Pavement Design Procedures in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Design Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Material Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Chapter 12. Pavement Condition Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Analysis and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Detailed Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Conclusions and Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Chapter 13. Pavement Design Procedures and Specifications for Tropical Soils . . . . . . . . . . . . . . . . . . . . 143
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Design Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Design Charts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Chapter 14. Applications of Tropical Soils in Construction Other Than in Roads or Highways . . . . . . . . . . . . . 151
Building Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Foundations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Slopes in Lateritic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Earth Dams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Chapter 15. A Review of the Literature on Tropical Black Clays and Black Cotton Soils . . . . . . . . . . . . . . . 163
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Occurrences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Morphological. Chemical and Mineralogical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Environment and Formation of Black Tropical Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Geotechnical Properties of Tropical Black Clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Engineering Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Chapter 16. Geographical Distribution and Physical Properties of African Black Clay Soils . . . . . . . . . . . . . . 183
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Geographical Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Test Procedures and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Discussion of Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Chapter 17. Design Considerations for Roads over Tropioal Black Clays . . . . . . . . . . . . . . . . . . . . . . 199
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Recent Design Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Recommended Design Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Chapter 18. Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Appendix for Chapter 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Laboratory Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Discriminant Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix for Chapter 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Method of Test for the Preparation of Sensitive Soils . . . . . . . . . . . . . . . . . . . . . .
Error Significance in One-Point Liquid Limit Test . . . . . . . . . . . . . . . . . . . . . . .
Methods for Sampling and Testing of Mineral Aggregates. Sands and Fillers . . . . . . . . . . . . .
Soil Swelling Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix for Chapter 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stabilization Test Data-BRRI Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . .
Stabilization Test Data-Asphalt Institute . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix for Chapter 12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Determination of Load-Deflection Characteristics of Pavement System by Means of the Deflection Beam
Pavement Condition Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Factorial Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Typical Compaction Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix for Chapter 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Expansion Pressure of Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Procedures for Sample Compaction and Testing with the Swell Index Device . . . . . . . . . . . .
 LIST OF ILLUSTRATIONS
Fig. Page
No. TITLE N o.

A Generalized Map of the Tectonic Basins of Africa with Intervening Swells. Plateaus a nd Rift Valleys . . . . . . . 10
Rift Valley System in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Geological Map of Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . in rear Jacket
Solubility in Relation to pH for Some Components Released by Chemical Weathering . . . . . . . . . . . . . . 15
Classification of the Principal Clay Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Sketch of Formation of Weathering Mantle in Areas that are Tectonically Inactive . . . . . . . . . . . . . . . 18
Moisture Regions of Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Mean Annual Rainfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Water Balance Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Typical Cross-Section for Ferralitic Soil Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Typical Cross-Section for Ferruginous Soil Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Vegetation Map 0.f Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Soils Map of Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . in rear Jacket
Location of Soil Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Ferruginous Soil Group Gradation Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Ferrallitic Soil Group Gradation Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Ferris01 Soil Group Gradation Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Location of the Tropical Red Soils on the Casagrande Chart . . . . . . . . . . . . . . . . . . . . . . . . 48
Relationship Between Maximum Dry Density and Optimum Moisture Content . . . . . . . . . . . . . . . . . 49
CBR vs Plastic Index Percentage X Passing No . 200 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
CBR vs Plastic Index Percentage X Passing No . 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
CBR vs AASHO Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Suitability of Material as Base Course on the Basis of Maximum
Dry Density at AASHO Modified Density and Fineness Index of CBR Specimen .
Typical Grading Envelope of Lateritic Gravels in Group 1 . . . . . . . . . . .
Typical Grading Envelope of Lateritic Gravels in Group 2 . . . . . . . . . . .
Typical Grading Envelope of Lateritic Gravels in Group 3 . . . . . . . . . . .
Typical Grading Envelope of Lateritic Gravels in Group 4 . . . . . . . . . . .
Relation Between Maximum Dry Density/Plasticity Index and CBR Values. Group 1
Relation Between Maximum Dry Density/Plasticity Index and CBR Values. Group 2
Relation Between Maximum Dry Density/Plasticity Index and CBR Values. Group 3
Relation Between Maximum Dry Density/Plasticity Index and CBR Values. Group 4
Phase Relationship of Halloysite . . . . . . . . . . . . . . . . . . . . . .
Flow Diagram for Laboratory Testing . . . . . . . . . . . . . . . . . . . .
Comparison Between Air Dried and Oven Dried Liquid Limits . . . . . . . . . .
Comparison Between As-Received and Soaked Liquid Limits . . . . . . . . . .
Relationship Between ASTM and British Standard Liquid Limit . . . . . . . . .
Dry Density/Moisture Content Relationships after Different Methods of Drying . .
Harvard Miniature Moisture-Density Relationships . . . . . . . . . . . . . . .
Relationship Between Harvard and Modified AASHO Densities . . . . . . . . .
Relationships Between Harvard and Modified AASHO Moisture Contents . . . . .
Effect of Number of Blows on Harvard Density . . . . . . . . . . . . . . . .
Coarse Durability vs . CBR . . . . . . . . . . . . . . . . . . . . . . . . .
Coarse Durability vs . Aggregate Impact Value . . . . . . . . . . . . . . . . .
Durability Coarse vs . Specific Gravity Coarse . . . . . . . . . . . . . . . . .
Fine Durability vs . Specific Gravity-Fine . . . . . . . . . . . . . . . . . . .
Coarse Durability vs . Water Absorption . . . . . . . . . . . . . . . . . . .
Fine Durability vs . Water Absorption . . . . . . . . . . . . . . . . . . . .
Aggregate Impact Value vs . CBR . . . . . . . . . . . . . . . . . . . . . .
Aggregate Impact Value vs . Specific Gravity-Coarse . . . . . . . . . . . . . .
Aggregate Impact Value vs . Water Absorption . . . . . . . . . . . . . . . .
Relation Between Swelling and % Passing No . 200 Sieve . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Relation Between Swelling and Liquid Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Relation Between Swelling and Plasticity Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Relation Between Swelling and Linear Shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
CBR vs . AASHO Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
CBR vs . Unified Soil Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
CBR vs . Group Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Ghana Location of Stabilization Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Flow Chart for Lime and Cement Stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Maximum Density-Optimum Moisture Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Maximum Density-Optimum Moisture Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Stabilization Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Dry Density and Unconfined Compressive Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Comparison Between Split-Tensile and Unconfined Compressive Strengths-Cement . . . . . . . . . . . . . . . 105
Comparison Between Split-Tensile and Unconfined Compressive Strength-Lime . . . . . . . . . . . . . . . . 105
Comparison Between Compressive Strength and CBR-Cement . . . . . . . . . . . . . . . . . . . . . . . . 106
Comparison Between Compressive Strength and CBR-Lime . . . . . . . . . . . . . . . . . . . . . . . . . 106
Effect of Sand on the Moisture-Density Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Variation of Atterberg Limits with Percentage of Sand . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Test Section A, Variable Strength. Uniform Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Test Section B. Uniform Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Test Section Plan View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Flexible Pavement Design Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Pavement Condition Survey Locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Measurement of Pavement Deflection Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Deflection Characteristics of Inside and Outside Wheel Paths . . . . . . . . . . . . . . . . . . . . . . . . 127
Relationship of Deflection Parameters for Inside and Outside Wheel Paths . . . . . . . . . . . . . . . . . . . 127
Relationship between MRD and Subjective Pavement Rating . . . . . . . . . . . . . . . . . . . . . . . . 131
Relationship between Tan 6 and Subjective Pavement Rating . . . . . . . . . . . . . . . . . . . . . . . . 131
Relationship between Maximum Rebound Deflection and Rating at Various Traffic Loadings . . . . . . . . . . 132
Relationship between Slope of Deflection Curve and Rating at Various Traffic Loadings . . . . . . . . . . . . . 132
Factorial Experiment-Thickness. Soil Type and California Bearing Ratio vs . Deflection . . . . . . . . . . . . . 134
Factorial Experiment-Thickness and Compaction vs . Deflection . . . . . . . . . . . . . . . . . . . . . . 135
Coefficients of Relative Strength of Base (a, ) and Subbase (a3) . . . . . . . . . . . . . . . . . . . . . . . 136
Maximum Rebound Deflection vs .Pavement Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Slope of Deflection Curve vs . Pavement Structure (SN) . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Base Course Coefficient vs . Base Course CBR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Subbase Coefficient vs . Subbase CBR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Subgrade Coefficient vs . Subgrade CBR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Pavement Deflection vs . Structural Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Relationship Between Maximum Rebound Deflections and Pavement Performance at Various Traffic Loadings . . . 140
Structural Number and Allowable Deflection for Given Traffic Loadings . . . . . . . . . . . . . . . . . . . 140
Design Curves for Three Levels of Traffic and Six Inch Base Course . . . . . . . . . . . . . . . . . . . . . 144
Design Curves for Three Levels of Traffic and Eight Inch Base Course . . . . . . . . . . . . . . . . . . . . 145
Relationship of Liquid Limit and CBR at Various Percentages of Material Passing No . 200 Sieve . . . . . . . . . 148
Relationship of Plastic Index and CBR at Various Percentages of Material Passing No . 200 Sieve . . . . . . . . . 148
Gradation Limits for Surface Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Critical Height and Slope Angle for Lateritic Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Distribution of Black Clay Soil in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Outline of Weathering in the Tropical Soils of the Hawaiian Islands . . . . . . . . . . . . . . . . . . . . . 171
Progressive Types of Clay Development in Hawaiian Soils under a Climate having Alternating Wet and Dry Seasons . 172
Progressive Types of Development in Hawaiian Soils under Continuously Wet Climate . . . . . . . . . . . . . . 172
Moisture-Density-CBR Relationship (CBR Swell Shown below CBR Curve) . . . . . . . . . . . . . . . . . . 189
Relationships Between Expansion-Pressures and Dry Densities at Various Initial Moisture Content . . . . . . . . 190
Effect of Surcharge Load on Volume Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Plasticity Chart, African Black Clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Relationship Between Compacted Densities and Liquid Limit . . . . . . . . . . . . . . . . . . . . . . . . 193
Relationship Between Compacted Density and PIX Percentage of Fines . . . . . . . . . . . . . . . . . . . . k94
Relationship Between Optimum Moisture Content and Plastic Limit . . . . . . . . . . . . . . . . . . . . . 194
Potential Expansiveness of Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Relationship Between Free Swell and Atterberg Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Relationship Between PVC and Atterberg Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Relationship Between Free Swell. PVC and Group Index . . . . . . . . . . . . . . . . . . . . . . . . . . 197
CBR Design Curves Revised by Israel P.W.D. for Various Traffic Categories and a Wheel Load of 6.0 Tons . . . . . 201
Percent Differences Between Liquid Limit and 1-Point Water Content Determination . . . . . . . . . . . . . . 229
Loading Device. Two Axle Truck with Rear Dual-Tired Wheels . . . . . . . . . . . . . . . . . . . . . . . 247
Deflection Beam . Similar to Benkelman Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Position of Beam for Measuring Maximum Rebound Deflection . . . . . . . . . . . . . . . . . . . . . . . 248
Rebound Deflection Measured at Various Distances from the Point of Loading . . . . . . . . . . . . . . . . . 248
 LIST OF TABLES
Table Page
No . TITLE No .

Mobility of the Common Cations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Weathering Products in Relation to Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Physical and Chemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Typical Soil Test Results for Ferruginous Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Typical Soil Test Results for Ferrallitic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4041
Typical Soil Test Results for Ferris01 Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Average Soil Properties- Ferrallitic. Ferrisols and Ferruginous . . . . . . . . . . . . . . . . . . . . . . 4344
Average Soil Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Changes in Engineering Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58-59
Test Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63-64
Recommended Modifications of Test Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Change in Atterberg Limits When Drying at Various Temperatures and Times . . . . . . . . . . . . . . . . . 67
Aggregate Impact Strength of Quartz Lateritic Gravel Mixtures . . . . . . . . . . . . . . . . . . . . . . . 73
Airphoto Interpretation of Tropical Soils- Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Breakdown of the Unified Soil Classification and AASHO Systems for Lateritic Soils . . . . . . . . . . . . . . 84
Properties of Soils Investigated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Design Content of Lime and Cement Stabilized Materials . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Stabilization Test Data. Stabilizer Content Determined from pH Test . . . . . . . . . . . . . . . . . . . . . 109
Material Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119-120
General Basis for Studies of One Pavement Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
General Basis for Relationships Among Variables for One Pavement Type . . . . . . . . . . . . . . . . . . . 124
Laboratory Testing Schedule for Pavement Condition Survey Field Samples . . . . . . . . . . . . . . . . . . 125
Results of Structural Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
General In-Situ Conditions of Base. Subbase and Subgrade . . . . . . . . . . . . . . . . . . . . . . . . . 130
Coefficients of Base. Subbase and Subgrade Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Selection of Thickness of Subbase. Low Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Selection of Thickness of Subbase. Medium Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Grading Requirements for Laterite and Lateritic Gravels . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Recommended Criteria for Subbase Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Recommended Criteria for Base Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Results of Compaction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Selection of Soil for Soil-Cement Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Engineering Properties of Lateritic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154-155
Recommendation for Stable Slope Angles of Western Malaysia . . . . . . . . . . . . . . . . . . . . . . . 157
Properties of Some Earth Dams in Tropical Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Distribution of Dark Clay Soils in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164-165-166
Chemical Composition of the Clay Fraction of Some Margalitic Soils. Indonesia . . . . . . . . . . . . . . . . 168
Partial Chemical Analysis of South African Black Clays . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Partial Chemical Analysis of the Clay Fraction of a Soil Profile at Ata. India . . . . . . . . . . . . . . . . . . 169
Properties of Tropical Black Clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173-174
Classification of Indian Black Cotton Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Chemical and Mineralogical Properties. Tropical Black Clays . . . . . . . . . . . . . . . . . . . . . . . . 184
Physical Properties. Tropical Black Clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Swelling Properties. Tropical Black Clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Data for Making Estimates of Probable Volume Change for Expansive Soils . . . . . . . . . . . . . . . . . . 196
Test Data on Lime Stabilization of Black Clay Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Value of Factor F with Depth D from the Relation D = 20 log F . . . . . . . . . . . . . . . . . . . . . . . 203
Liquid Level Test Error Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Stabilization Test Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Properties of Congo Lateritic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Compaction Characteristics of Congo Laterite Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Summary of Stabilization Test Data on Congo Laterite Soils . . . . . . . . . . . . . . . . . . . . . . . . 244
Summary of Stabilization Test Data on Congo Laterite Soils . . . . . . . . . . . . . . . . . . . . . . . . 245
Traffic Data for Road Sections Tested . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
 ACKNOWLEDGEMENTS

The study on lateritic soils and other problem soils of members of the Public Works Department in Ghana also
Africa was undertaken between September, 1968 and provided valuable assistance in sampling and testing during the
November, 1970 in Kumasi, Ghana, by Lyon Associates, pavement condition survey. Mr. K.A. Afriyie, Senior Technical
consulting engineers and the Building and Road Research Officer of the Kumasi Public Works Department Laboratory
Institute. Lyon Associates was under contract with the U.S. handled all the laboratory testing and his technicians per-
Agency for International Development, while the Building and formed all of the field testing and measurements required
Road Research Institute represented the Ministry of Works during the survey. Chemical and mineralogical anaIyses were
and Housing of the Ghana Government. Supplementary also performed by Mr. Bampo-Addo of the Soil Research
laboratory testing and mineralogical and chemical analyses Center, Kwadaso. The cooperation of all these individuals is
were performed at the University of Illinois and at various gratefully acknowledged.
Public Works Laboratories and Universities in Africa. The U.S. AID staff provided invaluable assistance and
The Lyon Associates staff included Dr. W.J. Morin, Project cooperation throughout the course of the study. Mr. J. Ashby
Director, Mr. P.C. Todor and Mr. R.S. Levinson, Project Williams, Chief of the Transportation Division, Office of
Research Engineers. Mr. Todor was responsible for the road Central Engineering, Washington, contributed many useful
condition survey and the collection and analysis of the ideas and suggestions and assisted the conduct of the study in
properties of lateritic soils. Mr. Levinson was responsible for many ways. Mr. Robert Rose, Public Works Engineer, U.S.
much of the laboratory testing, testing procedures, stabi- AID, Ghana, contributed considerable time and effort in
lization studies and assembled the information on other uses assuring that administrative and technical details progressed
of lateritic soils. smoothly. Mr. James Baird, Public Works Engineer, also
The Building and Road Research Institute staff included greatly assisted in the later stages after he replaced Mr. Rose in
Dr. J.W.S. de Graft-Johnson, Director; Mr. H.S. Bhatia, October, 1970.
Principal Research Engineer; Dr. M. D. Gidigasu, Mr. S. L. Many staff members of U.S. AID and American Embassy
Yeboa and Mr. A. K. Castel, Research Engineers; Mr. J. K. personnel in the various African countries were extremely
Ayetey, Engineering Geologist; and Messrs. A.A. Hammond, cooperative and helpful during the extensive travel that was
S.H.K. Bani and A.G. Hornsby-Odoi, Senior Technical performed. These are too numerous to mention here, but their
Officers. Mr. Philip Seaman and Mr. James Meyer, Peace Corps assistance made possible many informative interviews and the
Volunteers, contributed considerable assistance in statistical collection and shipment of many samples. Without this
analyses and in programming much of the data for the IBM assistance, far fewer countries could have been visited.
computer at the University of Science and Technology in A number of countries participated in this study locally,
Kumasi. In addition the B.R.R.I. provided the supporting either through the Public Works Department or Iocal Uni-
junior technical staff. A list of the reports which formed the versities. The Ivory Coast Public Works Department assisted in
Building and Road Research Institute's contribution to the the pavement condition survey. Mr. Boni, Director General
project are found at the end of this section. and Mr. D. Mandelbaum, Chief of the PWD Laboratory, also
The board of consultants for the research program included provided much useful information. Several samples were tested
Professors Ralph Peck, Edward Cording, Don Deere, James in Sierra Leone. Mr. B.B. Ibrahim of Fourah Bay College
Eades, Herbert Ireland and Thomas Thornburn, all of the performed the physical properties tests and stabilization
University of IIlinois. Each of the consultants travelled to studies on these soils. His w r k , published as "Lateritic Soils
Ghana at least once and Profs. Peck, Eades and Thornburn also Study, Sierra Leone Participation," Engineering Research
travelled t o other African countries. Professor Eades per- Publication, Research Paper No. SM 1, Department of Engi-
formed the mineralogical analyses and other chemical and neering, Fourah Bay College, University of Sierra Leone, 1970,
stabilization studies at the University of Illinois. Professor has been incorporated in this report. The Malawi PWD ran a
Thornburn analysed the aerial photographs and prepared the number of tests on residual soils of that country and Mr. D.C.
report covering this experimental aspect of the study. The Snelgrove wrote a report on these results. The Public Works
consulta~ltsprovided valuable guidance throughout the course Department in West Cameroon generously assisted in special
of the study and critically reviewed most sections of this sampling of the volcanic soils that occur in that country.
report. Special black cotton soil samples were provided by Ato Abate
The staff of the Central Materials Laboratory of Ghana, Sahlu of the Imperial Highway Authority, Addis Ababa,
Public Works Department, Accra, specifically Mr. C.M. Pant, Ethiopia and by Dr. Anthony Madedor of the Nigerian Public
Chief Materials Engineer, and Messrs. H.R. Aggarwal, R.N. Works. Mr. Bal Bhogal of the Civil Engineering Department of
Kapoor and B. Singh, all Principal Technical Officers, provided the University of East Africa in Nairobi made available
chemical analyses and considerable supporting data. Staff research conducted at that institution on black clays of Kenya.
Tabulated data on soil properties were prepared for this Engineer (Roads), Supernumerary; Mr. N.B. Onduto, Chief
project by Mr. Lecaillon of Travaux Publics and Centre Materials Engineer; and Mr. B. Harley, Chief Materials En-
Experimental de Recherches et d'Etudes du Batiment et des gineer, Supernumerary.
Travaux Publics, Gabon and Mr. G. Liautaud, of CEBTP, In Malawi: Mr. Pullinger, Deputy Secretary of the Ministry
Ghana and Paris. Mr. Liautaud aIso provided a number of of Works and Supplies; Mr. Brind, Director of Roads; Mr.
reports on his work in West and Central Africa and Sidney Garaghty, Design Engineer; Mr. Joseph Prendergast,
Madagascar. His study of cut slopes in Ghana is discussed in Chief, Materials Office of the Roads Department; Mr.
Chapter 14. Habgood, Director of the Geological Survey of Malawi; Mr.
Many people were interviewed during the course of this David Keith, Regional Engineer; and Mr. Cliff Wintle, Regional
study and provided a great deal of information on local design Materials Engineer.
practices, soil types and specifications. These are listed in the In Tanzania: Mr. Hardyal Singh, Engineer-in-Chief, Ministry
order they were interviewed. of Roads and Aerodromes; Mr. Channe Singh, Chief of the
In Nigeria: Mr. S.B. Audifferan, Director of Engineering, Soils Section; and Mr. Hoare, DeLeuw Cather and Co.
Ministry of Works and Transport; Mr. Abeke, Assistant In Uganda: Mr. A. Pollard, Chief Materials Engineer,
Director of Engineering; Dr. A. Madedor, Director of the Ministry of Works and Communication.
Materials and Research Laboratory; S.S. Sidhu, Senior Ex- In Sudan: Mr. Curbow, Project Engineer, C.H. Leave11 and
ecutive Engineer, Jos; B. Branko, Soils Engineer, Jos; J.S. Co.; Mr. Khalafalla, Deputy Under Secretary, Ministry of
Soutar, Senior Works Superintendent, Jos; G. Puddephatt, Works; and Mr. Sayed Bedawi A-laz, Chief Engineer for Roads.
Senior Executive Engineer of Planning, Kano; P.S. Jain, Soils In Mali: Mr. Mbo, Director-General, Public Works; Mr. Sako
Engineer, Kano; E.A. Taiwo, Chief Engineer, Ibadan; Mr. Abdoulaye, Chief of the Road Department; Mr. S. Sissoko,
Olugbekan, Deputy Engineer, Ibadan; and Dr. J.O. Shojobe, Chief of the Public Works Laboratory; Mr. H. Diallo, Director
University of Lagos. of Roads and Bridges; and Mr. D. Volcy, Engineer, Roads and
In Cameroon: Mr. Boumnack, Secretary of State for Public Bridges.
Works; Mr. M. Silliaux, Head of the Research Section, Public In Ivory Coast: Mr. Boni, DirectorGeneral, Public Works;
Works; Mr. Fenze, Assistant Director of Public Works; Mr. Mr. D. Mandelbaum, Director, Public Works Laboratory and
Laroche, Director of the Public Works Laboratory; Mr. CEBTP; Mr. Atlan, Assistant Director, and Mr. G. De Speville,
Sendze, Director, Public Works Department, West Cameroon; Project Manager, Kaiser Engineers.
Mr. Palma, Senior Executive Engineer for the West Cameroon In Democratic Republic of the Congo: Mr. Henrron,
Ministry of Works; and Mr. Garcia, Engineer in the Public Director, Public Works Laboratory, Kinshasha.
Works Laboratory. In Chad: Messrs. E. Beethoven, King and Gavaris; Mr.
In Liberia: Mr. Gabriel Tucker, Secretary of Works; and Mr. Dyme, Assistant Chief of the Technical Services Division,
Warren Coonred, (U.S.G.S.), Geological Survey of Liberia. Public Works; and Mr. Irving Licht and Mr. Rocher of the Lake
In Gambia: Mr. Jack Moran, Director, Department of Chad Basin Commission; and Mr. Denis of the U.N. Food and
Public Works; Mr. F.R. Ewer, Senior Executive Engineer; Mr. Agricultural Organization.
J. Sheperd, Bathurst District Engineer; and Messrs. H. Hamblin In Central African Republic: Mr. Y. Boulvert, Head of
and L. Watson of the Crown Agency, U.K. Research, Office de la ,Recherche Scientifique et Technique
In Senegal: Mr. J.P. Serfass, Director of the Public Works Outre-Mer, Bangui.
Laboratory. In Gabon: Mr. C. Marius, Pedologist, ORSTOM; Mr.
In Sierra Leone: Mr. Ibrahim, Fourah Bay College, Civil Leclerq, Principal Engineer, Public Works; and Mr. Lecaillon,
Engineering Department; Mr. Jarret-Yaskey, Professional Director, Public Works Laboratory and CEBTP.
Head, Public Works Department; Mr. Jones-Dove, Chief In Burundi: Com. N. Damiem, Director General, Ministry
Engineer; and Mr. R.D. Thorpe, Laboratory Supervisor. of Public Works; Mr. K. Noel, Chief, Soils Laboratory; Mr.
In Upper Volta: Mr. M. Magnien, Director of the Public Mendanhal, Maintenance advisor; Mr. A. Mdezako, Director,
Works Laboratory. Geology and Mines Division; Mr. Tedeschi, Director,
In Niger: Mr. P. Dieppois, Public Works Department; Mr. A.M.S.A.R. (Contractor); and Mr. J. Dewey, Agricultural
n s s e of CEBTP, Abidjan; and Mr. Rex Anderson, Wilbur Engineer.
Smith and Associates. In Zambia: Mr. D.E.R. Jones, Chief Engineer, Ministry of
In Ethiopia: Ato Shifenaw Bizuneh, Chief Engineer; Transport, Power and Works-Roads Branch; Mr. B.W.
lmperial Highway Authority; and Ato Abate Sahlu, Materials Coultous, Executive Engineer; Mr. G.S. Auby, Technical
and Research Engineer. Officer, Mr. W. Strude, Resident Engineer; Mr. H.T.J. Nel,
In Kenya: Mr. Bal Bhogal, Senior Lecturer in Highways and Senior Road Inspector; and Mr. T.U. Yager, Soil Scientist,
Traffic Engineering; University of Nairobi; Mr. S.J. Mbugua, Mount Makula Agricultural Research Center.
Engineer-in-Chief, Ministry of Works; Mr. F.E.A. Nderitu, In Malagasay Republic: Mr. E. Rahalison, Chief of Central
Deputy Chief Engineer (Roads); Mr. Gordon Manners, Chief Technical Services, Ministry of Equipment; and Mr. Sabatier,
Director of the Public Works Laboratory and CEBTP. One-point Liquid Limit Procedure as Applied to Soils
The Asphalt Institute in College Park, Maryland performed in Ghana. BRRI Project Report SM No. 3, January
all of the laboratory testing and analysis for the asphalt 1969. This report is summarized in Chapter 6.
stabilization studies. Four samples were shipped t o them from The Effect of Pretreatment on the Atterberg's Limits
Congo, Niger and Dahomey. Their assistance in this project is of Ghanaian Soils. BRRI Project Report SM No. 4,
thankfully acknowledged. June 1970. This report is incorporated in Chapter 6.
Finally, special acknowledgement is given to those in large A Study on the Engineering Properties of Some
research institutions outside of tropical Africa who con- Lateritic Gravels from Kumasi District. BRRI Project
tributed generously of their time during interviews and Report SM No. 5, March 1970. This report is
critically reviewed certain sections of this report. Dr. R.S. summarized in Chapter 5.
Millard, Deputy Director of the Road Research Laboratory,
The Effect of Pretreatment on the Compaction
U.K. and members of the Road Research Laboratory staff
Characteristics of Laterite Soils. BRRI Project Re-
were interviewed and kindly provided details on much of their
port SM No. 6, January 1970. This report is
work on tropical soils, as well as valuable advice during the incorporated in Chapter 6.
course of the study. Mr. A. Remillon, of CEBTP, Paris and
many members of his staff in charge of local laboratories in The Engineering Characteristics of Laterite Soils
many French speaking countries made available many in- (Literature Review). BRRI Project Report SM No. 7,
August 1970. This report was condensed to form
formative reports and data accumulated through many years
of work in West and Central Africa. Professor J. Jennings, Mr. most of Chapter 4.
G. Poole, and Mr. A.B.A. Brink of the University of The importance of Soil Profiles to the Engineering
Witswaterand; Mr. S.H. Kuhn, Director, National Institute for Studies of Laterite Soils of Ghana. BRRI Project
Road Research, South Africa and Dr. P. Louden, of his staff; Report SM No. 8. The geology of Ghana was
and Mr. A.A.B. Williams and Mr. G. Donaldson of the National condensed to form part of Chapter 3.
Institute for Building Research were interviewed and parts of The Durability and Strength Characteristics of Lat-
the report were reviewed by members of these organizations. eritic Aggregates of Ghana. BRRI Project Report SM
Their comments stimulated many improvements. Other en- No. 9. This report is summarized in Chapter 6.
gineers in South Africa were also interviewed which made A Study on the Engineering Characteristics of Lat-
possible the assimilation of some of the work performed in eritic Gravels of Ghana. BRRI Project Report SM No.
that country. Mr. Elda de Castro of the Civil Engineering 10, August 1970. This report is incorporated in
Laboratory in Lisbon provided reports on work performed in Chapter 5.
the Portuguese colonies. Mr. H. Tateishi, Hawaiian Department Stabilization of Lateritic Gravels using Sand. BRRI
of Highways provided test details he devised for the moisture- Project Report SM No. 11. This report is summarized
sensitive soils of that state. in Chapter 9.
Listed below are the papers contributed by the B.R.R.I.
Engineering-Geological Study of some Ghana Lat-
These may be obtained by writing directly to the B.R.R.I.,
eritic Soils and their Profiles. BRRI Project Report
U.S.T, P.O. Box 40, Kumasi, Ghana.
SM No. 12.
Cement Stabilization of Laterite Soils (Literature
I. The Formation and General Characteristics of Lat- Review). BRRI Project Report SM No. 13.
erites and Lateritic Soils (Literature Review). BRRI Proposed Pavement Test Sections. This report is
Project Report SM No. 2, January 1969. included as Chapter 10.
 CHAPTER 1

INTRODI

The countries of tropical Africa, almost all recently tropical areas, should be studied in a thorough research
independent, share one common very basic need, adequate program. It was believed that a major research effort would
road facilities for transportation and communication. The provide a satisfactory engineering definition, classification and
existing road networks were largely laid out by the former utilization criteria for these materials to insure adequate but
governing administrators and understandably designed t o not overly conservative design and construction.
develop each colony individually. These networks were basi- Accordingly, a contract was awarded in June, 1967 to Soil
cally adequate for developing colonies. The interior was and Pavement Consultants of Southeast Asia for Phase I of the
opened up to commerce and agricultural and mineral resources program. Phase I was conducted in Bangkok, Thailand with
could be developed. the assistance of the Royal Thailand Department of Highways.
With independence the situation is changing. Regional The area under study was southeast Asia, with emphasis on
networks of roads are required. Commerce and com- Thailand.
munication between countries is essential t o stimulate regional The contract for Phase I1 was awarded in June, 1968 to
development. Feeder roads are urgently needed t o allow Lyon Associates, Inc. Phase [I was based in Kumasi, Ghana
development of agricultural potential. Existing roads have to and the area studied included all of tropical Africa, with
be strengthened and improved t o withstand the present and emphasis on Ghana. The Building and Road Research In-
projected volumes of traffic. In fact, among these three stitute, of the Ghana Council of Scientific and Industrial
requirements, it is difficult t o assign priorities. Research, actively participated in the study with their staff
It is therefore essential that the limited funds available be engineers and geologist. They provided laboratory technicians
utilized in the most efficient manner. The designs must be the and assistants as well as laboratory and office facilities.
most economical for the soil conditions and the traffic loads. The American staff of Lyon Associates was initially made
This point is readily conceded. Two others however should be up of two engineers in Ghana, where virtually all of the
emphasized because they are so often neglected. Proper investigation was performed. A third engineer was added about
maintenance must be exercised as a never-ending effort. Too half-way through the project. The Building and Road Research
many road facilities have been lost or seriously impaired Institute also increased the available personnel, both senior
because funds have been diverted to more obvious, tangible and junior staffs.
works. All road facilities, no matter how well designed, require It was decided that the study should be truly regional,
adequate maintenance. The second factor is construction incorporating as many countries in Africa as time and travel
control. This principle is also fundamental, but often violated. would allow. In this particularly important aspect, the scope
Many countries do not have adequate staff, trained personnel, o f work was carefully followed from the beginning. The
or necessary equipment to verify that the construction consultants and staff engineers for Lyon Associates considered
materials and techniques are following the design. In order to an in-depth study of the Ghanaian lateritic materials of limited
insure that allocated funds are not wasted on sub-standard value to other tropical African countries. This regional study
construction, control and supervision must have highest was largely possible because of the amount of work already
priority. performed in Ghana by the Building and Road Research
Roadway design practices in the tropics have evolved, by Institute. Without this background of locally developed
and large, from American and European experience. These knowledge, it would not have been possible to exert as much
procedures were devised through years of experience, but with effort on regional investigations.
temperate-zone soils. The adequacy of these procedures have The objectives of the program were well described in the
never really been evaluated for the climatic conditions, the U.S. AID scope of work. These objectives are summarized
soils, and the traffic patterns of tropical Africa. Organizations below with additional comments when limitations were
such as the Road Research Laboratory of the United King- advisable or necessary or where the scope of work was
dom, and the Centre Experimental de Recherches et d'Etudes expanded.
du Batiments et des Travaux Publics of France have made 1 . To develop an engineering definition of laterite and
major contributions in modifying the usual empirical pro- lateritic soils and other problem tropical soils. The major
cedures and these modifications have been adopted with, emphasis was on lateritic soil and tropical black clays, often
generally, considerable success. termed black cotton soils. The black clays are undoubtedly the
Nevertheless, the question of the available materials, their most difficult soils for road building in Africa. Collapsing soils
quality and their adequacy in road structures has never been are also covered lightly, principally by reviewing the recent
fully investigated. The United States Agency for International literature and giving locations of soils believed t o be sus-
Development decided that lateritic materials, so plentiful in ceptible to collapse.
 2. T o propose a suitable classification system for lateritic through 1 4 deal with tropical and lateritic soils and Chapters
materials that adequately describes their engineering p r o p 1 5 through 1 7 deal with tropical black clays. A review of the
erties. pertinent literature is provided for both types of soils and the
3 . Evaluate the engineering characteristics of these soils physical and engineering properties of both are also described.
and determine the range of properties for each soil clas- T h e genesis or laterization process is discussed in detail in a
sification. separate chapter. Recommended test procedures and de-
4 . T o determine suitable test procedures that properly scriptions of modified or non-standard tests are provided. The
evaluate these materials. results of stabilization studies are reported for both lateritic
5. To study the test means and materials for stabilizing soils and tropical black clays.
sub-standard lateritic and other problem soils of Africa. One of the most important aspects of this study was the
6. To develop suitable specifications, design standards and pavement condition survey, described in Chapter 12. This
criteria for the satisfactory use of these materials in highways study allowed the evaluation of the material specifications and
and airfields. This study was limited t o highways because the provided a basis for roadway design incorporating the most
design procedures, etc. are essentially similar for airfields and significant parameters.
specifications and standards proposed for the one could be Such a detailed analysis was not possible on other problem
adopted for the other. soils, but design procedures and considerations have been
7. T o evaluate the adequacy of these lateritic materials by included for roads over tropical black clays, Chapter 17. These
assisting in the construction of test roads. Construction funds procedures should facilitate the construction of adequate
in Ghana precluded the construction of test roads. This was roads my minimizing volume change or swelling t o tolerable
not considered particularly detrimental, however, since test limits.
roads are difficult t o design and construct in such a way that Most of the data collected during this study is included
all extraneous variables are held constant and only those under either in the body of the report o r in the appendices. Various
investigation isolated and varied for evaluation. Usually a methods of showing the data are employed. Some are
number of test sections are required t o evaluate such variables tabulated whereas others are only shown graphically. Com-
as subgrade soil, climate and traffic conditions. Consequently puter printouts have also been reproduced. The various
it was felt that "tests o n roads" were superior t o "test roads." appendices also include details on tests, descriptions of the
This program of tests on roads was conducted as the pavement statistical analyses and computer programs.
condition survey and necessarily limited t o Ghana with The usage manual has been printed separately in booklet
supplementary locations in Ivory Coast. form and includes many illustrations. This should facilitate its
8 . T o appraise the various methods presently in use in use both in the field and in the laboratory. Test procedures
Africa for pavement and roadway design. The pavement and descriptions are straight-forward and should be com-
condition survey provided a suitable means for evaluating prehensible t o junior and inexperienced technicians and
these procedures and proposing modifications or changes. engineers who are required t o set up and operate laboratory
9 . T o prepare a usage manual for laboratory and field use facilities in the field.
for lateritic materials. The manual was originally considered as
a useful summary of the project for senior engineers and CONCLUSIONS
technicians. This was subsequently altered when a more The following general conclusions summarize the more
important audience was visualized, namely, junior technicians important aspects of the study.
and inexperienced engineers who are unfamiliar with soil
testing practices. The soil testing manual originally designed by 1. All tropically weathered red residual soils were
Harvard University, under the direction of A. Casagrande and included in this study. Some of these were true
S. Wilson, for the U.S. War Department in 1945 (TB 5-253-I), lateritic soils under the chemical criterion. Others are
and subsequently adopted by the American Road Building presently undergoing or have undergone laterization,
Association, was considered ideal for this purpose. Large but leaching of the silica and bases is incomplete.
sections were copied with added material generally limited t o There is n o practical means of distinguishing between
testing procedures peculiar t o lateritic materials. the two. Further, the engineering properties of both
10. T o study the potential application of lateritic soils t o groups are quite similar. Most lateritic soils contain an
other engineering construction. This was necessarily limited t o appreciable clay fraction which generally displays
reviewing the literature and recounting the practical ex- properties that are similar t o those of other tropical
perience of the consultants and engineers in a number of soils which are only partially laterized.
countries. The other applications considered were: earth dams,
slopes, foundations and building materials. 2. The more descriptive terms ferruginous soil, ferrallitic
soil, and ferrisol are recommended in lieu of lateritic
The report is divided into 1 8 chapters with each covering a soil. The geographical distribution of each of these
specific topic included in the scope of work. Chapters 2 soil groups is shown o n the Soils Map, Fig. 13.
 3. Variations in properties among the three groups 7. Each point on the compaction curve should be
should be established in each country. Possible determined with new material. This prevents ex-
variations within each group (among the subgroups) cessive breakdown during compaction, which occurs
should also be investigated. when the same material is reused. The soil should be
"cured" overnight at its compaction moisture
Correlations were established among various properties of
content.
the red tropical soils. The most significant of these are the
relationship between: (1) liquid limit and plasticity index: (2) 8 . A minimum of three points should be determined to
optimum moisture and Modified AASHO maximum density; establish the CBR accurately at the optimum mois-
(3) CBR and AASHO classification; and (4) for four groups of ture content or any other particular moisture
lateritic gravels from Ghana CBR and maximum dry density content.
divided by PI. General, but useful, relationships were es-
9. To convert the Harvard miniature density to Modified
tablished between (1) CBR and PI multiplied by the per-
AASHO density, the equation:
centage passing the No. 200 sieve; and (2) CBR, maximum
MDD Modified = 16.0 + .926 MDD HM may be used.
density and fineness index.
To convert Harvard optimum moisture to Modified
4. The following useful relationships were established: AASHO OMC, the following equation may be used:
a) PI = 0.71 LL - 8.50 for ferruginous soils OMC Modified = ,358 - .797 OMC HM
PI = 0.57 LL - 3.62 for ferrallitic soils An experimental study on the applicability of natural
PI = 0.50 LL - 1.50 for ferrisols color and color infrared air photos was conducted and a few
and preliminary conclusions are possible. Both natural and infrared
PI = 0.58 LL - 4.54 for all tropical red soils colors have advantages over black and white in revealing subtle
b) MDD = 160 - 2.78 OMC differences in the condition of the vegetation and such
C) CBR = 129 - 7.5 (AASHO) differences are often soil-related. Infrared color is preferable to
where: (AASHO) = The numerical value of the natural color in the identification of vegetation which is
AASHO classification system, i.e., l=A-1-a; 2=A-1-b; diseased or starved for lack of nutrients or water. However,
12=A-7-6 natural color photos are more useful for observing tonal
d) Fig. 20 may be used to estimate the range of changes in the surface soil.
CBR values from the PI and percentage of fines.
10. A more complete study of the applicability of color
e) The suitability of material for use as base course and color infrared air photos in identifying tropical
can be estimated from Fig. 23, which shows the
soils is recommended. The possibilities of distin-
range in CBR values from the maximum density
guishing among the soil groups and locating indurated
and fineness index.
horizons and gravel deposits should be investigated.
5. Special pretreatments and test procedures are re-
The suitability of current engineering classification
quired when ferrallitic soils or soils over volcanic
systems was analyzed for tropically-weathered soils. The
rocks are tested. These procedures can be ignored
AASHO system distinguishes a greater number of soil groups
only if it is certain that the soil will be dried during
among those tropical soils tested than does the Unified system.
construction. If drying is not feasible, the laboratory
testing should be performed on the samples in the 11. The AASHO classification system can be applied as
as-received condition. Otherwise, the laboratory re- well with tropical soils as it can with temperate zone
sults will apply only to the altered material and not soils.
to the material that obtains in the field.
12. Cement stabilization is a practical means of improving
6. The one-point liquid limit test can be used for lateritic gravel unless the clay content and plasticity
tropical soils. index are greater than 15.
The equation: 13. Lime is generally a better stabilizing additive for
improving lateritic gravel whose clay content and
plasticity index are greater than 15. Cement in
combination with lime is required when lime alone
will not sufficiently increase the strength.
can be applied, where: W = moisture content 14. Asphalt can be a practical additive for stabilizing
N = The number of blows. sandy tropical soils.
tan B = The slope of the flow curve of
blows versus the moisture con- A design has been prepared for proposed test sections to
tent on a semi-log plot. evaluate the effectiveness of cement stabilized base courses. In
one section, a uniform thickness of stabilized base with compaction. Atterberg limits and gradation are
cement contents varying from 0% to 8% is suggested. Three satisfactory values for insuring adequate pavement
thicknesses of base course should also be evaluated in a second performance.
section, all at one cement content of 4%. 22. Figs. 1 1 2 and 1 1 3 show that modifications in the
A pavement condition survey was performed in Ghana Atterberg limits or gradation are possible provided
and Ivory Coast to evaluate lateritic and tropical n~aterialsas that the minimum CBR value for the base course class
structural components of existing roads. (design charts, Figs. 110 and 11 1) can be attained. If
15. There is a strong relationship between maximum the percentage of fines is low, then the L L (and PI)
rebound deflection and the slope of the deflection can be increased without causing a reduction in CBR.
basin for both wheel paths. Similarly, if the L L (and PI) values are low, then a
larger percentage o f fines can be allowed.
16. Both deflection parameters can be used to predict
pavement performance or serviceable life (Figs. 9 7 23. Modified AASHO compaction should be specified.
and 98). However, the possibility that excessive fines may be
produced should be investigated. Further research is
17. CBR and thickness of the structural component, and recommended.
the degree of compaction are the most important
factors in reducing deflection. Compaction below The literature on tropical black clay and black cotton soils
90% Modified AASHO density results in excessive was reviewed. Most of the black clays have formed residually
deflection (within the structural course). That de- over basic rocks, particularly basalts. These soils generally
flection is highly dependent upon CBR and thickness contain large clay contents. Montmorillonite is the pre-
of the structural component is illustrated in Fig. 107. dominant clay mineral in all cases where mineralogical analyses
were performed. When the black clays are desiccated, or
18. Subbase courses must have CBR values of 2 0 or partially desiccated, and exposed t o water, swelling occurs.
greater in order t o be effective in reducing deflections Expansion pressures can be considerable.
(Fig. 105).
The properties of the African black clays are similar t o
19. As-molded, rather than soaked, CBR values can be those reported in the literature o n black cotton soils and also
applied particularly in the more arid regions (fer- to other expansive soils. The expansion pressures vary con-
ruginous soils) if adequate drainage can be provided. siderably, depending on the clay and on the test employed.
Design charts were developed for base course thicknesses
24. Potential expansiveness of a black clay soil can be
of six and eight inches. A CBR value of 3 0 was adopted for the
estimated from its PI and clay content by utilizing
subbase, although other values greater than 2 0 can be readily
Fig. 127.
applied t o devise other design charts. Different curves are given
for various values of base course CBR which, with light traffic, 25. The free swell test and PVC test are recommended as
can be as low as 50. Three charts are provided for each base routine tests when investigating tropical black clays
thickness, covering three ranges of traffic loading. or other expansive tropical soils.
Material specifications that are appropriate to the traffic 26. Moisture changes in black clay subgrades can be
were determined from the results of the pavement condition minimized by:
survey. The specifications apply t o subbase, base and surface a) Stripping the vegetation and allowing the soil to
materials. Further modifications to the Atterberg limits and fallow for a year. Adding a granular cover is
grading requirements are also provided which are permissible particularly effective in increasing the moisture
as long as the CBR requirements are fulfilled. content of the subgrade.
20. It is recommended that the design charts, Figs. 110 b) Compacting the subgrade at Standard AASHO
density and optimum moisture content.
and 11 1 be adopted for roadway design in tropical
Africa. The findings of the pavement condition c) Providing adequate drainage facilities, and by
locating drainage ditches well away from the
survey should be verified in the individual countries,
sl~oulders.
as additional data and local conditions may warrant
modifications t o the design charts. Nevertheless, it is 27. In all the black clays tested, increasing the com-
felt that these design procedures should be adopted pactive effort t o Modified AASHO density did not
now and doing so would resblt in more satisfactory increase the bearing capacity. This should be verified
and more economical facilities. for individual projects.
21. The specifications tabulated in Chapter 1 3 are also 28. Lime stabilization is a practical means of increasing
recommended. The values listed for CBR, degree of bearing capacity and reducing expansion pressure.
 CHAPTER 2
ENGINEERING IMPLICA TIONS OF TROPICAL WEA THERING
AND LA TERIZA TION

INTRODUCTION Those aspects of the phenomena of tropical weathering of

The engineering behavior of soils, whether formed under primary significance to engineers are summarized in the
arctic, temperate or tropical conditions, is determined by remainder of this chapter. In succeeding chapters scientific
certain physical characteristics designated as engineering prop- background is introduced where required in connection with
erties. Once the engineer has ascertained which properties are specific engineering questions.
the significant ones with respect to a particular type of
construction at a particular site, he may at least in principle
perform suitable tests to evaluate the properties quantitatively WEATHERING*
and may base his design on the findings. This fundamental Rocks when exposed at or near the earth's surface find
procedure is at the root of all sound geotechnical design. themselves in a physical and chemical environment often quite
In practice, the apparently straightforward fundamental different from that in which they were originally formed. The
procedure can rarely be used without modification. Since minerals which constitute the rocks may react chemically with
determination of the engineering properties is usually expen- the rainwater, groundwater, and dissolved solids and gases of
sive, index properties, simpler and cheaper to evaluate but the new near-surface environment t o form new minerals which
indicative of the engineering characteristics, are investigated. are more nearly in equilibrium with the surface conditions.
On the basis of these investigations, materials having similar Some materials may be carried away by solution in ground-
behaviors are classified into separate categories. Experience water. The end result of these changes is t o convert the upper
may then be organized and used as the basis for design in new portion of the rock into a residual debris more soil-like than
localities or on new projects associated with soils of similar rock-like in character and with chemical, mineralogical, and
classification. physical properties entirely different from those of the
It has long been appreciated that the engineering classifica- original rock.
tion of soils is greatly facilitated by taking into account the The changes which take place in the rock by these
soil-forming processes by which nature has created the various near-surface processes are encompassed by the geological term
types of soil conditions. Similar combinations of soil-forming weathering. One form of weathering, denoted as physical,
processes in different parts of the world have been found to includes the effects of such mechanical processes as abrasion,
lead t o materials of similar index properties and similar expansion, and contraction. Physical weathering produces
engineering characteristics. Therefore, in any study such as the end products consisting of angular blocks, cobbles, gravel, sand,
present one, the identification of localities in which the silt, and even clay-sized rock flour. The mineral constituents of
soil-forming processes have been similar is a prime objective. all these products are exactly like those of the original rock.
The consequences of tropical weathering have been of Chemical weathering, on the other hand, results in the
overriding significance in the formation of the soils of decomposition of the rock and the formation of new minerals.
Equatorial Africa. Hence, no system of classification or no The near-surface agents of weathering which enter into
attempt to identify the significant engineering characteristics chemical reactions with the primary minerals of the rock
can succeed if not based on an appreciation of tropical include water, oxygen, carbon dioxide, and organic acids
weathering processes. These processes have been the subject of derived from vegetation. The various chemical processes
a vast amount of research and are described in an impressive include hydration, hydrolysis, oxidation, solution, and car-
body of literature. bonation. All of these may operate simultaneously, some more
For the most part, however, the literature of tropical rapidly than others and some more effectively in the alteration
weathering is concerned with the chemistry, morphology, and of one mineral than another.
mineralogy of the process. The complexity of the subject is The chemical changes operating in the primary minerals of
reflected not only in the extent of the literature, but also in the rocks in temperate or semi-tropical zones tend to produce
wide differences of opinion even over such basic questions as end-products consisting of clay minerals predominantly repre-
the definition of laterite. The civil engineer may easily find sented by kaolinite and occasionally by halloysite and by
himself diverted from his more limited aims by the complexity
of the literature. The detailed scientific knowledge is relevant *This section is taken largely from P. Jiminez-Quinones, "The
Compaction Characteristics of Tropically Weathered Soils", Ph.D.
t o the civil engineer only in the same way as knowledge of the Thesis, University of Illinois, 1963, and from lectures by Prof. Don U.
chemistry of cement is relevant to the development of the Deere in his regular courses at the University of Illinois and as
principles and practice of reinforced concrete design. condensed in guest lectures at Harvard University in 1968.
hydrated or anhydrous oxides of iron and aluminum. Quartz such as the presence of a dike or a contact, or because of
experiences slight solution but remains essentially unchanged, differences in the degree of jointing or extent of shearing. The
although it may undergo some comminution. latter features increase the permeability and the depth to
Chemical weathering is favored by warm humid climates, which weathering can proceed.
by the presence of vegetation, and by gentle slopes. Thus, Similar weathering profiles develop under similar conditions
tropical and subtropical regions of low relief with abundant as reflected by the soil-forming factors mentioned above.
rainfall and high temperatures are the most susceptible to Hence, a convenient means for regional description of soil
chemical alterations. Deep, strongly leached red, brown, and properties for engineering purposes is the determination of the
yellow profiles are manisfestations of the effects of severe engineering characteristics of the various members or zones in
chemical weathering. typical profiles of weathering. Such investigations constitute a
Under conditions favorable to tropical weathering, the framework by means of which information and further
weathering processes may be so intense and may continue so experience can be coordinated.
long that even the clay minerals, which are primarily hydrous
aluminum silicates, are destroyed; in the continued weathering HYDRATED AND DEHYDRATED LATERITE
the silica is leached and the remainder consists merely of In continuously wet regions the end-products of the process
aluminum oxide such as gibbsite, or of hydrous iron oxide of laterization are likely t o be characterized by high natural
such as limonite or goethite derived from the iron. This water contents, high li'quid limits, and irreversible changes
process is known as laterization. upon drying. Inasmuch as limonite and geothite are hydrated
The extent to which a residual soil has been laterized may oxides of iron, and gibbsite is a hydrated oxide of aluminum,
be measured by the ratio of silica, SiO,, remaining in the soil drying alters the characteristics of the material markedly. The
(except for discrete pebbles of free quartz that may remain) t o plasticity decreases and the grain-size increases such that much
the amount of F e 2 0 3 and A1203 that has accumulated. The of the clay-sized material agglomerates t o the size of silt. When
silica: sesquioxide ratio dealing with such materials, the engineer may need t o
determine the index properties and the engineering properties
on the basis of tests performed at the natural water content
and again at various degrees of air drying in order t o evaluate
has served as a basis for classification of residual soils. Ratios fully the range of properties associated with the physical
less than 1.33 have sometimes been considered indicative of conditions that may prevail on the job. Serious engineering
true laterites, those between 1.33 and 200 of lateritic soils, and construction problems have arisen through failure t o
and those greater than 2.00 of nonlateritic tropically weath- recognize materials likely t o experience irreversible changes as
ered soils. a consequence of desiccation.
Determination of even the relatively simple silica: ses- On the other hand, in regions subject t o distinct wet and
quioxide ratio requires a chemical analysis not likely to be dry seasons, the products of laterization are likely to be
available to the civil engineer. Furthermore, experience has characterized by low natural water contents, low plasticity,
indicated that the ratio is by no means a unique indicator of and the presence of concretions and cemented horizons. The
the index properties or engineering characteristics required by alternate upward and downward movement of water causes
the engineer. Many other factors are of greater importance. concentration of the iron or of the alumina. Laboratory tests
run from the natural water content or from the air-dried state
PROFILES OF WEATHERING lead to essentially the same results.
The alteration of rock by the processes of chemical The difference in behavior of the hydrated and dehydrated
weathering takes place progressively through a series of events forms of lateritic soils deserves attention because of the danger
and stages which result in a profile of weathering. Those who of being misled by reports in the literature in which the
have worked with tropically weathered residual soils have significance of the distinction may not be realized.
noted the frequent occurrence of an upper clayey zone a few
feet thick, underlain by a silty or sandy zone which, in turn,
passes through a very irregular transition into weathered and
finally into sound rock. The thickness of each member of the CONCLUSION
profile varies greatly from site to site because of the Progress in dealing with tropically weathered soils will
complexity of the interrelationships among the controlling depend primarily on the selection of appropriate physical
soil-forming factors of rainfall, temperature, time, character of properties indicative of engineering behavior and the estab-
parent rock, topography and vegetation. Moreover, at a given lishment of suitable tests for their investigation. These tests
site there may be great differences in the depth and thickness may be expected to differ in some respects from those that
of the weathered rock within lateral distances of only a few have been found suitable in temperate regions. The results of
feet. The differences arise because of differences in lithology, such investigations can be systematized by the identification
and physical testing of the various materials in typical profiles evident that an almost unlimited variety of such materials,
of weathering. The profiles may be grouped and classified, at having at least slightly different chemistry, morphdogy, and
least as a first approximation, on a regional and local basis physical properties, can be found. The differences, often
according t o the similarities or differences among the impor- subtle, have led t o much controversy over terminology and
tant soil-forming factors. definition. To avoid such controversy, entirely sterile from the
The setting of geotechnical problems in Equatorial Africa is engineering point of view, it would be preferable t o refer to all
described in Chapter 3 as a background for all subsequent such materials simply as reddish tropically weathered rnate-
discussion. Chapter 4 is concerned with those aspects of soil rials, irrespective of the details of the degree of their
genesis needed t o establish and make use of procedures for a weathering, and irrespective of the presence or absence of
general classification of soils and profiles of weathering in ferruginous crusts or concretions. Even transported soils
various parts of Africa. This information is followed by a cemented by ferruginous matter would be included. Never-
summary of what is currently known about the engineering theless, because of the widespread use of the term laterite, in
characteristics of the products of laterization and of other this report all such reddish tropically weathered materials are
deeply weathered tropical soils, as disclosed in the literature for convenience referred t o as lateritic soils. This usage,
and as determined in the present investigation. therefore, carries no implication with respect t o the definition
The foregoing discussion indicates that all the products of or properties of laterite as such. The distinctions useful for
tropical weathering with red or reddish color have much in engineering purposes are based on differences in physical or
common, with respect t o both origin and properties. It is also engineering properties.
 CHAPTER 3

REGIONAL SETTING AND GENETIC CLASSIFICA TION

SUMMARY OF PHYSIOGRAPHY AND into the Mediterranean from Africa; the Senegal, Niger and
GEOLOGY OF AFRICA Volta Rivers in West Africa; the Congo of Central Africa; and
the Zambesi of southeast Africa.
A brief description of the physiography and geology of
Africa is provided for background information. A basic Geology
knowledge of the geology of the continent is necessary to
The geology of tropical Africa is, generally speaking, as
determine the effect of parent material and time of weathering
unvaried as the physiography. The major part of the continent
on soil development. Similarly, a knowledge of the physi-
has been emergent and relatively stable since Precambrian.
ography is required to evaluate the effects of topography and
The geological factors of interest are the rock types that
drainage.
occur, their spatial arrangement, and their geological history.
The rock types are of interest because of their influence on the
Physiography soil that has developed. The structure is important in a
Africa is the second largest continent, after Eurasia. It has regional sense because broad features such as basins and rift
an area of 11,700,000 square miles, and has been described by valleys affect topography, drainage and climate. Finally, the
D'Hoore (1964) as "a huge crystalline mass surrounded by a geological age of the parent rock and its eilvironmental llistory
sedimentary fringe." through geological time have a direct bearing on the soil
The major part of the continent consists of broad swells or development.
plateaus separated by wide basins, Figure 1. Precambrian rocks Stratigraphy. The rock types that occur in Africa are a
outcrop in about one third of the continental area, on the result of the stable nature of the continent. The large expanse
plateaus and swells. The basins are both tectonic and topo- of Precambrian rock indicates a long history of erosion and
graphic features. They are filled with thick deposits of denudation. Continental-type deposits are generally limited to
continental sediments derived from the surrounding highlands. the large tectonic basins. The limited extent of marine deposits
The surface of the African continent is, on the whole, is evidence that only small areas were ever submerged, and
unusually level. Mountains of folded sedimentary rocks occur most of these areas, with the exception of the Sahara, were
in the extreme north where the Atlas Mountains form part of limited to the continental margins.
the Alpine Mountain system, and in the extreme south in the The generalized geological map (Figure 3, rear jacket)
Cape Ranges. Within the tropical zone, there are only a few shows the stratigraphy of Africa. The wide expanse of
mountainous areas. These are the Guinea Highlands in Guinea, Precambrian exposures is apparent. The continental sediments
Sierra Leone and Liberia; the Togo Mountains in Togo and occupy the basins illustrated in Figure 1 . The Precambrian,
Dahomey; the Cameroon Highlands; the Marra Mountains of although exposed over one-third of the total continent,
western Sudan; the Ethiopia and Somalia Plateaus; and other actually covers about one-half of the tropical portion. The
plateaus and volcanic mountains associated with the Rift basement complex extends from Capetown to the Northern
Valley. 28th parallel, well within the Sahara Desert, and from Senegal
The continental plateau slopes from elevations of over to Somalia. Over half of the island of Malagasay is
4,000 feet in the southeast to 1,000 feet in the Sahara. A Precambrian.
number of erosional surfaces have resulted from repeated Furon (1963) states that the Precambrian can be divided
uplifting of the continental block. D'Hoore (1964) states that into four cycles. The Lower Precambrian is essentially granites
mid-Tertiary surfaces are the most extensive; they are char- and gneisses, strongly folded. The Middle Precambrian consists
acterized by many ferruginous crusts. mainly of schists, quartzites and marbles. The Upper and
The Rift Valley system of eastern Africa, Fig. 2, is the most Terminal Precambrian are composed predominantly of sand-
important structural and physiographic feature of tropical stones with some schists, linlestones and dolomites. All cycles
Africa. It extends circuitously, with several branches, from are represented in all parts of the continent. The boundary
Mozambique through Ethiopia to the Red Sea. Holmes between Precambrian and Cambrian is not always clearly
(1965) states that the total length of the Rift Valley to the defined chronologically and the stratigraphy has been cor-
Red Sea exceeds 1,800 miles. The total distance is over 3,000 related matching late Precambrian formations in one area w~tli
miles if the continuation through the Red Sea, Dead Sea and Cambrian in another. Tillites bear evidence to a glacial period
Lake Tiberias is included. at the end of Precambrian and a major unconformity is
One-third of the continent has no drainage to the oceans or accepted by Furon as structural evidence.
Mediterranean Sea. Most of the remainder is drained by a few Paleozoic sediments are widespread in northern Africa. but
major rivers; the Nile, the only perennial stream which flows some of the various formations extend south into western
FIG.1 A GENERALIZED MAP OF THE TECTONIC BASINS OF AFRICA WITH INTERVENING SWELLS,
PLATEAUS AND RIFT VALLEYS ( o f t e r Holmes, 1965)
CARTE GENERALE DES BASSINS TECTONIQUES AFRICAINS AVEC LEURS BOMBEMENTS,
PLATEAUX ET FOSSES (d'apres Holmes, 1965)
Africa. The Cambrian includes conglomerates, limestones,
schists, sandstones and some volcanics. The sandstone extends
as far south as Mali. The Ordovician, made up mainly of LEGEND
,
sandstone, extends t o Guinea and Ghana. Silurian shales are
widespread in Guinea and also occur in Ghana. Devonian
limestones and sandstones occur as far south as the coast of
Ghana. Continental Carboniferous deposits are limited t o areas
north of the central Sahara. Paleozoic sedimentary deposits
cover over half of Ghana and parts of Togo and Dahomey;
large areas of Mauritania, Mali, Senegal and Guinea are also
covered (D'Hoore, 1964). The only other significant remnants
of early Paleozoic formations are limited t o South Africa.
A glacial period occurred during Upper Carboniferous and
glacial deposits of tillites occur in south Africa and within the
Congo Basin, as well as in other parts of the southern
henlisphere. This was followed during Permian and Triassic
times by the deposition of a deep succession of continental
deposits called the Karroo System after the type locality in
South Africa. These deposits are found throughout the
southern part of the Continent, in ~ g l a ~ a ands a ~in tectonic
troughs of eastern Africa (Furon, 1963). The lower beds are
alluvial in origin, followed by clayey shales and topped with
sandstones and shales (D'Hoore, 1964).
The Triassic culminated with volcanic activity. Basalt flows
occurred in northern Africa and t o a much larger extent in
FIG 2 R l F T VALLEY SYSTEM IN AFRICA (after Mohr,1962)
southern Africa. Basic intrusive bodies are found in western L E SYSTEME DU GRAND RIFT AFRICAIN (d'apres Mohr,l96Z)
Africa, in Senegal, Sierra Leone, Liberia and Ma!i.
Both continental and marine Jurassic sediments were which have caused some investigators t o conclude that
deposited in northern and eastern Africa. Continental con- laterization was an "ancient" phenomenon. It is now realized
glomerates, sandstones, marls, limestones, arkoses, mudstones that the pedogenic process is still active at the present day.
and clays are exposed in northern Africa, the Sahara and in Important volcanic activity commenced in the Cretaceous
Tanzania (D'Hoore, 1964). in western Africa and in the Tertiary in eastern Africa. Flood
The "Continental intercalaire" defined by Kilian (1931) basalts were extruded in Ethiopia and eastern Africa as far
and cited by Furon (1 963) includes the continental sediments south as Malawi. These flood basalts were associated with the
deposited between a period of tectonic activity in Carbon- uplift in the eastern part of the continent that preceded the
iferous-Permian time and a marine transgression in Lower down-faulting of the extensive Rift Valley system. Cretaceous
Cretaceous time. The continental Jurassic deposits make u p a and Tertiary basalts were extruded in t h e Cameroons, Nigeria,
part of this "Continental intercalaire." The upper part is made Tibesti (northern Chad) and the islands in the Gulf of Guinea.
up of both continental and marine sediments and is more The Quaternary deposits are varied and widespread. They
widespread. Limestone, sandstone and gypsiferous clays were have been summarized by D'Hoore (1964) into four
laid down when seas invaded the Sahara from the north and groups: (1) eolian sand deposits-found in the Sahara and
from the Gulf of Guinea in Nigeria. Marine deposits also occur southwest African deserts and the Kalahari extending north
in Tanzania, Somalia and Ethiopia. into the Congo Basin; (2) alluvial deposits of various ages,
Tertiary limestone, marls, red sandstone and conglomerates some of which have been altered by wind and surface erosion;
were found in northern Africa and in the Sahara. Fluvial and (3) colluvial mantles, often thick enough t o have complete
eolian sandstones, and locally, lenses of lacust rine limestone profile development and often including talus and debris from
occur south of the equator (D'Hoore, 1964). Tertiary deposits ferruginous crusts; and (4) volcanic rocks, usually basaltic, a
of various sediments were laid down along much of the coast continuation of the volcanic activity that began in the
of western Africa and parts of eastern Africa. Tertiary, associated with the Rift Valley system, the chain of
Remnants of Tertiary surfaces exist today in southern west volcanic mountains in equatorial Guinea and Cameroon, and in
Africa and in central Africa which are capped with ferruginous smaller, isolated areas of central Nigeria, Tibesti, Hoggar
crusts. As these Tertiary surfaces have been eroded, crustal (southern Algeria) and central Libya.
debris, being more resistant, has remained in areas with more Structure. The oldest fractures occurred during the Pre-
recent geological deposits. These Tertiary remnants were cambrian. According t o Furon (1963, p. 76): "These
recognized as having been formed by pedogenic processes, movements are manifestations of immense stresses causing
enormous undulations (upwarps and basins) and fractures faults that lie within the Cameroonian trough. The volcanic
(faults, horsts and grabens)." He states the movements within mountains of Mount Cameroon, Mount Manengouba and
this basement complex have controlled the erosion and cycles Mount Bambouto erupted along this line, as did the volcanic
of sedimentation in the covering sedimentary materials. Furon islands of Fernando Po, Principe, Sao Thome and Annobon.
believes the major faults occurred in two directions: SW-NE Furon believes this line extends to the northeast, passing into
and NW-SE. He includes the various sections of the African the Chad Basin toward the Tibesti region in northern Chad.
Rift system in this directional pattern. Many of the older Volcanic activity also began in the Cretaceous at Mount
Precambrian faults have been rejuvenated by later movements. Cameroon. The mountain, which is over 13,000 feet in
The eastern section of the continent has been subjected t o a elevation, is still active. The latest activity occurred in
long period of volcanism, uplifting and faulting, which has February, 1959 (Furon, 1963). All of the volcanic mountains,
culminated in the high plateaus, the African Rift system and including the volcanic islands are composed mainly of basalts.
its associated volcanic rocks and mountains. The oldest fault The plateaus and basins of Africa have been subjected to
scarps are of Karroo age (Holmes, 1965), but the most repeated uplift and erosion through geologic time. These are
important movements were in Tertiary and Quaternary time. described in detail by Holmes (1965). The exposed rocks of
Eruptions of the basalts started in early Tertiary, prior to the swells and plateaus are Precambrian which were formed at
uplifting of the plateaus. Renewed volcanic activity coincided considerably lower elevations. Chad Basin, containing thick
with further uplifting in mid-Tertiary. Major faulting also deposits of continental sediments, is an arid region without
occurred in the Rift system at this time. The present external drainage. Despite this, Lake Chad is not saline,
topographic form results from additional uplifting and faulting although evaporation is high; it drains underground to oases
in Quaternary time, in midPleistocene (Mohr, 1967). The last about 450 miles northeastward in the southern Tibesti region,
phase is still in progress (Holmes, 1965). perhaps along Furon's tectonic line. The Kalahari and Cubango
The African Rift system can be separated into four sections basins in the southern hemisphere are other examples of
(Figure 2): internal drainage; both drain into brackish swamps.
1. The Malawi section with its several forks extending The Congo basin includes thick continental sediments of
south from Malawi as far as the Zambesi River in Karroo age overlain by sandy soils. These sediments dip inward
Mozambique. from the surrounding swells and plateaus, as a result of
2. The Western Rift extending from Lake Tanganyika Tertiary uplift and warping. The Congo River has incised the
through Lakes Kivu, Edward and Albert. Between Lake western rim of this basin at Stanley Pool where it flows
Edward and Lake Albert lies the uplifted block of through a series of cataracts to Matadi, the port for ocean
Ruwenzori with its "Mountains of the Moon" which vessels on the Congo River.
include peaks of over 16,000 feet. These rocks are The Sudan basin is also of interest. The large areas of
Precambrian and correspond to much lower rocks on the sediments in this basin are largely lacustrine. The lake level
adjacent plateau. rose until it breached the northern rim, along the present
3. The Eastern Rift, also called the Gregory Rift, east of course of the Nile, forming the series of cataracts between
Lake Victoria in Kenya. Khartoum and Aswan. The southern section of this basin,
4. The Ethiopian section extending from Lake Rudolf where the White Nile flows northward, contains a large
through Ethiopia t o the Afar region where it separates swampland of floating vegetation known as the Sudd.
into the Red Sea section and the Gulf of Aden section. Others basins are the Somalia, with thick sediments of
The widths of the various Rift Valley sections are Jurassic, Tertiary and Quaternary age; the Karroo basin of
remarkably uniform, varying only from about 4 0 to 60 South Africa now elevated as part of the southern plateau; and
kilometers (Holmes, 1965). the El-Juf or Taoudene basin of West Africa. These are shown
in Figure 1.
The depths of the Rift Valleys vary considerably. Holmes
lists the following elevations: 2,150 below sea level in the The swells and uplands adjacent to the basins include many
Tanganyika trough; 3,400 feet above sea level in L k e Kivu; areas with ferruginous crusts. These are usually on Tertiary
and 2,000 feet in Lake Albert. The elevations of the eastern surfaces or colluvium from Tertiary surfaces. D'Hoore (1964,
Rift vary from 600 feet to over 2,000 feet. Depths of the p.26) states, "If it is admitted that the greater part of the
Ethiopian section range from below sea level in the Afar region outstanding features of present day African geomorphology,
to 5,300 feet north of Lake Zwai. such as the alpine foldings of the north, the great faults, and
Furon (1 963) recognizes "in Cameroon a genuine trough of the marked subsidence of the continental depressions date
Rift Valley type extending northeastwards from the Gulf of back toward the end of the Tertiary, one must believe that
Guinea towards Chad. It is bordered on both sides by large during the Tertiary the African relief was even more level than
sub-vertical faults whose throw reaches 1,000 m." This now. Thus very unusual topography diminished the possibilities
faulting began near the end of the Cretaceous. Furon also of external drainage and engendered soil formation where
recognizes a tectonic line, made u p of a number of parallel temporary or permanent water-logging played a predominant
role, so long as the climate was not too arid. It is in such Geological Survey of Ghana.
conditions that the phenomena of mobilization, precipitation The main formations consist of Precambrian rocks, all more
and accumulation of free iron are most intensive." or less metamorphosed and folded, with their associated
differentiates and basic intrusives. The remainder of the
GEOLOGY OF GHANA country consists of the Palaeozoic Voltaian Series of sand-
Because of the concentration of studies of the present stone, shale, mudstone, conglomerate, limestone and tillite.
program in Ghana, the foregoing generalized geological de- Small areas on the coast are underlain b y flat gently dipping
scription of Africa is supplemented in the following paragraphs sediments of Devonian to Recent.
by a more detailed review synthesized from reports of the The table of succession is shown below:-

Table of Succession
Time Divisions Local Name and Description
Quaternary Recent Unconsolidated clays and sands of
lagoon, delta and littoral areas.

Tertiary ( Upper Partly consolidated red continental

deposits of sandy clay and gravel.

Marine sediments, sandstone, glau-

conitic sandstone, clay and shale,
limestone, oil sand.

Mesozoic
1 Upper
Jurassic
'Amisian' freshwater series of
bouldery and sandy clays and
conglomerates.

I
Devonian 'Sekondian' marine series of
sandstones, shales, black, sulphu-
rous shales.

'Accraian'. Middle Devonian only,

sandstone and shale.

'Voltaian', sandstone, shale, mud-

Ordovician stone, conglomerate, limestone,
Cambrian tillite.

'Buem formation', Folded and meta-

morphosed.

Togo Series sediments. Not granitized

or Tarkwaian intruded b y granites.

Precambrian
-I Middle 'Birrimian'. Geosynclinal sediments
and volcanics, partly granitized and
greatly intruded b y g~anites.

'Dahomeyan'. Massive crystalline

gneisses and migmatites representing
thick series of argillaceous,
calcareous and arenaceous sediments,
Note: Unconformity. - but with few schist remnants.
 Rocks of the Dahomeyan system, which underlie the Accra mudstones, sandstones, arkoses, conglomerate, tillite and
Plains as well as the Southern Volta Region, extend through limestone.
Dahomey to Nigeria. The system consists mainly of massive Covering only small areas along the coast, Devonian, Upper
crystalline granulites and gneisses which have suffered meta- Jurassic and Cretaceous-Eocene formations are often obscured
morphism, and migmatites with subordinate quartz schists, by Upper Tertiary sands and peneplain residuals and by raised
biotite schists and other sedimentary remnants. beach and lagoon deposits.
The Birrimian system is made u p of thick geosynclinal The Devonian rocks underlie the Takoradi area with
sediments with volcanic rocks predominant in the upper part. patches extending along the coast to Cape Coast and also in
The sediments which have been metamorphosed to phyllites or Accra. The rocks are a marine series of reddish sandstone and
slate were made u p of argillaceous materials often mixed with shale reaching 10,000 feet in thickness. The Takoradi shales
greywackes and tuffaceous types containing more or less are black and sulphurous. The Amisian beds (Upper Jurassic)
carbonaceous, ferrous, manganiferous o r calcareous materials. outcrop at Saltpond and in a few localities towards Winneba.
The composition of the lavas varies from basalt to rhyolite The rocks are made u p of bouldery fresh-water beds of
with basic types being the commonest. The rocks are fragments of Precambrian and Voltaian rocks. The Cre-
granitized in parts but only slightly altered in others and have taceous-Eocene rocks appear at the eastern and western tips of
been intruded by batholithic masses of granite. The granites southern Ghana where the edges of marine sedimentary basins
have been divided into two types on the basis of their cross the frontiers of Ghana. These are made up o f alternating
containing more potassium than sodium on one hand and sands, clays and limestones.
more sodium on the other. The two types are known as Cape After the Cretaceous-Eocene marine sediments had been
Coast Granite G I and Dixcove Granite G 2 . tilted and uplifted, their surfaces suffered leveling due to
The Tarkwaian system is made u p of a series of shallow erosion of streams. Upper Tertiary red limonitic sandy pebbly
water sediments, sandstones, phyllites and conglomerates clay, with a bed of quartz conglomerate at the base, was
derived from the Birrimian. It reaches a thickness of between deposited on the bevelled surface at several points near the
8,000 feet and 10,000 feet in the Tarkwa area. Areas of Ghana coast; the thickness reaches about 100 feet in a number of
underlain by the Tarkwaian are far smaller than those by places. In the Keta area this continental deposit obscures most
Birrimian; the largest lies between Konongo and Tarkwa in the of the Cretaceous and Eocene rocks. This is also the case in the
forest zone. Axim area in the southwest part of the country.
The Togo series consists of metamorphosed arenaceous and
agillaceous sediments now represented by indurated sandstone, SOIL-FORMING FACTORS AND
quartzite, quartz schist, sericitic and hematitic quartz schists, PROFILE DEVELOPMENT
shale, phyllite, sericite schist, talc mica schist and a little
Soil Forming Factors
limestone. These rocks form the Akwapim and Togo ranges;
the harder rocks such as the quartzites and similar rocks form Pedologists have agreed that the products of rock weath-
the hills and the shale and phyllites form the intervening ering are determined by certain physical conditions designated
valleys. as the soil-forming factors. These include climate (precipi-
The Buem formation underlies a large area of the country tation and temperature), topography (drainage), vegetation,
west of the Akwapim Togo range including Kpandu, Jasikan, parent rock, and time. They are of importance in the present
the Hohoe area, and the region northwest through Kadjebi to study because they bear on the possibility of classifying the
Dadiase. The Buem formation consists of a thick series of soils on a rational genetic basis. The influence of the major
shales, sandstone and volcanic rocks with subordinate lime- factors is considered in the following paragraphs.
stone, tillite, grit and conglomerate. The shales, which form The mechanism of breakdown during chemical weathering
the lowest beds, are overlain by the sandstone series. Con- of the parent mineral structures has been discussed in detail b y
glomerate and tillite, interbedded with calcareous and fer- Loughnan (1 969), who draws two conclusions:
ruginous shales and volcanic rocks, overlie all. The volcanic 1. Hydrogen ions play an important part in the breakdown
rocks consist of both lavas and bedded tuffaceous rocks with of the silicate minerals. The hyrogen ions are produced
agglomerates which are interbedded with shale, limestone and by the hydrolyzing
- - action of the mineral surfaces, when
sandstone. mineral surfaces combine with (OH)- radicals. Hy-
Covering about two-fifths of Ghana, the Voltaian formation drogen ions easily penetrate into crystal structures
occupies the central and eastern parts of the Northern Region where their large charge-to-radius ratio upsets the
and parts of the Upper Region, as well as the northeastern part charge balance within the crystal structures.
of Ashanti and the Eastern Region. Steep scarps, marking the 2. Aluminum and highly mobile cations are retained in the
northern, western, and southern limits of the Voltaian, secondary mineral products.
overlook the Precambrian below. The Voltaian rocks, thought Lougllnan (1969) has shown the solubility of some
to be of Palaeozoic age, are made u p o f a series of shales, components released by chemical weathering in relation to pH;
 FIG. 4 SOLUBILITY IN RELATION TO pH FOR SOME COMPONENTS RELEASED BY CHEMICAL WEATHERING
( after Loughnan, 1969 ).
SOLUBlLlTE DE QUELQUES ELEMENTS LIBERES PAR ALTERATION CHlMlQUE EN FONCTION DU pH
(d1apr6s Loug hnan, 19 69).

the relation is reproduced in Fig. 4. Within the normal pH Eh is dependent on (1) accessibility of oxygen; (2) organic
range of ground water, which is between 4 and 9 (Mason, matter; and (3) climate and topography. "A hot well-drained
1952), calcium hydroxide, magnesium hydroxide and the environment favors oxidat ion through the rapid destruction of
alkalis are completely soluble but TiO, , FeOH3, and A1, O3 organic matter and the considerable lowering of the water
are insoluble. The solubility of silica is constant through this table, whereas cool, poorly-drained environments promote
range of pH, but it is low. Loughnan concludes that the only accumulations of organic matter and reducing conditions. The
components whose solubility is influenced by variation in the oxidation potential for the transition of ferrous t o ferric iron
environmental pH are Ti(OH)4, CaC03 and Fe(OH), . The falls withn the range anticipated for natural environments and
solubility of quartz is approximately one-tenth that of consequently both states are common. Since a considerable
combined silica (Krauskopf, 1958). Alumina is soluble only in difference in solubility exists between the two, the mobility of
an acid environment when the pH drops below 4, such as in iron is greatly influenced by the prevailing Eh of the
soils with abundant organic matter. Alumina under these weathering environment." (Loughnan, 1969, p. 43-44)
conditions is taken into solution and migrates either down the The mobility of the common cations is listed in Table 1.
profile or to regions with higher pH where precipitation and The clay minerals that form are principally of four
enrichment occur. Ferric hydroxide is precipitated at pH 3 and groups: kaolin, halloysite, montmorillonite and illite. Each
is normally retained in the soil profile. Ferrous hydroxide, group has a number of nlineralogical variations reflecting
however, precipitates at a much higher pH. Therefore, in a differences in mineral structure or chemical composition. Most
reducing environment where the pH is acidic, the ferrous iron of these are relatively unimportant insofar as engineering
is leached; in an oxidizing environment, the ferrous iron is behavior is concerned. One exception is metahalloysite. A
converted to the insoluble ferric state. The oxidation- classification of the principal clay minerals is given in Fig. 5.
reduction potential (Eh) decreases as pH increases. Conse- The clay minerals are essentially hydrous aluminum sili-
quently, oxidation of iron takes place more readily in an cates. Hydrous silicates of magnesium and iron are less
alkaline environment. common.
 The parent rock determines in a negative sense which clay TABLE 1
minerals form initially. Kaolinite cannot form, for example,
unless alumina is present. The parent soils may be a dominant Mobilities of the Common Cations
factor or only of transitory significance (Keller, 1956).
(after Loughnan, 1969)
Weathering in the warm humid climate common in the tropics
is rapid and the influence of the parent material is usually
shortlived. However, as Keller has pointed out, the parent 1. Ca*, Mg*, r\rY- readily lost under leaching conditions
material must provide the constituent elements of which the 2. K+- readily lost under leaching conditions
clay minerals are formed. but rate may be retarded through
Tlle climate largely controls the kind and density of th,e fixation in the illite structure.
vegetation, as well as the organic products resulting from its 3. Fe*- rate of loss dependent on the
decay. In hot dry climates, vegetation is sparse, and organic oxidation-reduction potential and
matter is quickly destroyed through oxidation (Grim, 1968). degree of leaching.
In hot humid climates, abundant vegetation may result in 4. si4'- slowly lost under leaching conditions.
accumulations of organic matter, and, with a shallow water 5. ~i~'- may show limited mobility if released
table, much of the weathering zone may be permanently from the parent mineral as
saturated and in a reduced state. Ti(OH)4 ; if in TiO, form,
The effect of rainfall on chemical composition and mineral- immobile.
ogy was examined in connection with data presented by 6. Fe3'- immobile under oxidizing conditions.!
Alexander and Cady (1962) for West Africa lateritic soils. As 7. ~1~'- immobile in the pH range of 4.5 - 9.5.
rainfall increases, the amount o f kaolinite decreases, as do
both silica-alumina and silica-sesquioxide ratios, whereas the temperatures of around 25°C. It has, however, not been
amount of gibbsite increases. Generally speaking, the lower the possible t o establish any definite thermal criteria for the
silica-sesquioxide ratio, the higher the percentage of gibbsite. formation of laterites.
Low pH values (3.5 - 5.5) may also increase the rate of The effect of temperature in tropical weathering is rather
weathering and "the residue becomes progressively enriched in indirect, its first effect being t o help either the destruction or
minerals containing a high proportioil of alumina (kaolinite, accumulation of humus. Mohr and Van Baren (1954) have
halloysite, gibbsite, and boehrnite) and titania (anatase and shown the relation between these two and temperature, and
rutile). Ferric oxide minerals (hematite and goethite) may from the results it is noted that at soil temperatures above
become dominant constituents if the rate of destruction of the 30°C. humus destruction resulting from bacterial activity is
organic matter is right." (Loughnan, 1969, p. 69). predominant, whereas below 20°C. humus accumulation is
A series of weathering environments may exist within one predominant. Thus, in tropical lowlands, a bare soil which is
profile. Fig. 6, reproduced from Loughnan, shows the various adequately moistened by rain will exlubit a flora which is
mineralogical components and environmental conditions with predominantly bacterial whilst in a mo~intainousarea with a
depth in various climatic zones. dense forest vegetation, at an elevation where the soil
Temperature. Of the primary factors which influence the temperature is about 2OUC,the soil will show an accumulation
soil forming process, climate is considered the most important. of humus. rhis is important from the point of view of
Moist tropical climates provide the essential degree of weatller- clay-mineral type and content of final product.
ing for the formation of iron and aluminum oxides and Precipitation. The climate of the African continent varies
leaching of soluble silica that is characteristic of laterite and from the very humid zones of the equator through the
lateritic soils. sub-humid, semi-aird, arid and Mediterranean climates at the
Temperature plays an important role in both physical and extreme north and south of the continent (D'Hoore, 1964).
chemical reactions that occur during the soil forming process. The general moisture regions are shown in Fig. 7. In
Each rise of 1 0 " ~increases
. the rate of chemical reactions by a transitional zones, relief plays an important role in the soil
factor of two t o three. Africa is the hottest continent on earth. forming process. The drainage conditions, which are reflected
The mean annual temperature falls below 10°C. only in the by relief, control the amount of moisture available for
highest parts and over the greater part it is above +20°C. for percolation, an important factor in soil formation (Ode11 and
nine months of the year (D'Hoore, 1964). Dijkerman, 1967).
The occurrence of laterite is prevalent where temperatures The mean annual rainfall over the African continent is
are warm or are believed t o have been warm at the time of the shown in Fig. 8 (Thompson, 1965). It varies from below 4
formation. Though there have been few studies on the thermal inches (100 mm) in the arid zones t o about 130 inches (3,200
conditions of laterite formations, Crowther (1 930) demon- mm) over the Niger River delta area of Nigeria and beyond to
strated that the SiO, /A], O3 ratio increases as temperature the west coast of the Cameroon and over Liberia, Sierra Leone
rises at constant humidity. Maignien (1966) suggested that and Guinea. The area of tropical red soils generally has over 24
contemporary lateritic soils have developed at mean annual inches (600 mm) of rainfall per year.
 SEMI -DESERT TROPIC

ET DU PODSOL SEMI -DESERT

2400 - E T DESERT
200 - TOC

EVAPORATION

..'.................

FRESH ROCK
ROCHE FRAICHE
ROCK DEBRIS (SAPROLITE), LITTLE ALTERED CHEMICALLY
DEBRIS ROCHEUX (SAPROLITE I, PEU A L T ~ R ~CHIMIQUEMENT
S
HYDROMICA- MONTMORILLONITE -BEIDELLlTE ZONE
ZONE DE LIHYDROMICA,DE LA MONTMORILLONITE ET DE LA BElDELLlTE

a KAOLINITE ZONE
ZONE DE LA KAOLINITE

m]] S O L ARMOR, F e f 1 3 + AI2O3

CUIRASSE , Fez%+ AIz03
ZONE OF INCREASED MOBILITY OF ALL COMPONENTS
ZONE DE MOBILITE ACCFVE q TOUS LES ZEMENTS.
FIG. 6 SKETCH OF FORMATION OF WEATHERING MANTLE IN AREAS THAT ARE TECTONICALLY INACTIVE
(after Strakhov, 1967 ).
SCHEMA DE FORMATION DE LA COUVERTURE D ALTERATION DANS DES REGIONS SANS ACTlVlTE TECTONIQLIE
( d l apr'ec Strakhov, 1967)

Lateritic soils are believed t o always correspond to climates not be important at the initial stages of weathering when the
in which the wet period is warm; this applies to semi-humid parent rock is relatively impermeable. After the initial stages
tropical and equatorial climates. On the other hand, sub- of weathering the soil that results may be more permeable and,
tropical climates in which precipitation occurs in the cold at this stage, conditions of topography and the relation
season do not seem t o favor laterization even if the tempera- between rainfall and evaporation may result in various types of
t ure is above 20°C. water movement in the soil. These factors affect the final
There seems t o be no upper limit of precipitation under weathering product. The movement of water may be:
which lateritic soils may form. In lower Guinea, for example, 1 . Continuous Downward Movernent: This occurs in
where the climate is tropical, lateritic soil is well defined at an regions with a continuously wet climate where even in
annual precipitation of more than 240 inches (6,000 mm). the dry season it is never really "dry." Precipitation
With regard to incrustation itself, it seems to be closely related may be less in the dry season than in the wet but the
t o precipitation insofar as it affects the water regime of the n~onthlyrainfall is still between about 2.5 in. and 4 in.
soils. (60 mm and 100 mm). Under s u c l ~conditions fresh rain
More important perhaps than mean annual rainfall in the water is constantly being added t o the soil liquid with
soil forming process is the moisture diagram of a given area. the result that leaching of the material subjected to
The moisture diagram considers precipitation, distribution, weathering proceeds continuously.
evaporation and transpiration as shown in the water balance 2. Alternating Downwurd ~Vovernentand Cessation: This
diagram, Fig. 9. movement of water occurs if a wet season is followed
A plentiful supply of water in the form of rains is essential by an extremely dry season. In such a case the surface
for cllemical weathering in the tropics. However, the type of of the soil layers is so much above the water table that
movement in the soil is also important from the point of view the ground water does not reach the surface even by
of mineral content of the clays. The movement of water may capillary action during the dry season. The small
 'HUR

LEGEND LEGENDE
VERY HUMID
TRES HUMIDE
HUMID
HUMIDE
MOIST SUBHUMID
SUBHUMIDE
DRY SUBHUMID
SUBHUMIDE SEC
1-1 SEMI-ARID

I J

FIG.7 MOISTURE REGIONS OF AFRICA (after Oxford Regional Economic Atlas of Africa, 1 9 6 7 )
REGIONS HUMIDES ET S E C H E S D ' A F R I Q U E (d'opres Oxford Regional Economlc Atlas of Africa, 1967)
 15O lo0 5O 0' 5O lo0 15O 20° 25O 30 O 35O 40° 4 5" 50° "E
I I I I I I I -r- I I I I I I

ON ON

35" - - 350

30° - - 30°

25O - - 25-

200 - - 200

15O -

lo0 -

50 -

o0 - o0

5" - - 50

lo0 - - lo0

15' -

200 - INCHES MILLIMETRES

POUCES MILLIMETRES

ABOVE 130 ABOVE 3 2 0 0

250 - Au DESSOUS DE 130 AU DESSOUS DE
65 - 130 1600 -3 2 0 0
30 - 65 800 - 1600
16 - 10 400 -800
30° - - 30°
8 - 16 200 -4 0 0
4 - 8

BELOW 4 1 BELOW 100

- 350
s50 - EN DESSOUS DE 4 EN DESSOUS DE 100

Mercator Projection

O S O S
I I I I I I I I I I I I I I

15' lo0 5O O0 5O IO" 15' 20° 2 5' 30' 35O 40° 4 5O 50° 'E
A

FIG. 8 MEAN ANNUAL RAINFALL (after Thompson, 1965 1

PLUVlOSlTE ANNLlELLE MOYENNE (d'aprbs Thompson, 1965
 amount of rain that may fall does not penetrate this although forest cover seems to prevent the hardening. Nodular
layer above the water table and is quickly evaporated. forms are very common in forest areas and less widespread in
Water movement then practically ceases during the dry the savannah zones.
season. Leaching thus stops until the wet season sets in Parent Material. Results of investigations all over the world
again. have shown that lateritic soils are found overlying various
3 . Alternating Downward and Upward Movement: This kinds of rocks. They have been found over basic igneous rocks
happens when climatic conditions are as in (2) but such as basalt, norite, and diabase as well as over acidic rocks
when the ground water level is sufficiently high t o such as granite, granulites, gneisses, phyllites, and sericitic
permit water t o ascend by means of capillary force t o schists. Stephens (1946) and Marbut (1932) have reported the
the uppermost layers in the zone of evaporation. The occurrence of lateritic soil on shales, sandstone and other
surface soil is thus continuously enriched by a fresh sedimentary rocks including limestone, and in alluvium in
supply of subsoil water. Thus, there is downward association with a groundwater table.
movement during the rainy season and upward move- The effect of the parent rock material is more pronounced
ment during the dry season. If leaching is greater than at the initial stages of weathering than at the later stages when
accumulation, ultimate results will be like in (1) and (2) the conditions of weathering predominate. Climatic influences,
but if accumulation is greater than leaching the ultimate for example, active on a bare solid rock will produce a
results will be as in (4). different reaction from that produced when a freshly ejected
4 . Ascending Water Movement: In this case movement is volcanic ash is attacked by atmospheric forces.
exclusively in the upward direction. This results in an The recent work of Loughnan (1969) has thrown consider-
accumulation of salts in the upper surface. This able light on the chemical weathering of various rock types.
phenomenon occurs when the temperature is so high His findings are summarized in this section.
and rainfall so low that evaporation constantly pre- 1. Basic crystalline rocks, such as basalts, gabbros,
dominates over precipitation. dolerites and metamorphic rocks with low silica content, are
5. Intermittent Ascending Movement: This occurs when composed of the minerals olivine, pyroxenes, and calcic
there are long, very dry seasons alternating with brief plagioclases with accessory magnetite and ilmenite. These are
periods of rain. Precipitation is greater than evaporation the least stable minerals.
but not sufficiently high t o cause a real downward flow Olivine may weather directly t o montmorillonite, without a
of water. The very dry surface layers are wetted so that serpentine phase, and pyroxenes follow the weathering se-
for a time the upward flow is halted. quence shown below.
Vegetation. A forest vegetation was once considered neces- Magnetite and ilmenite are easily oxidized in weathering to
sary for formation of indurated laterite, but more detailed produce hematite, maghemite and goethite. Titanium is
observations have shown that, while laterites occur in rain- released.
forest zones, well developed laterites are most commonly 2. Acid crystalline rocks such as granites, granite gneisses,
found under a low forest and hard surficial laterites are a porphyries and rhyolites contain principally potash-rich and
very common feature in open savannahs adjacent t o the forest soda-rich feldspars and quartz, with amphiboles and micas.
zones. Humbert (1948) has illustrated the occurrence of These feldspars are more resistant t o chemical attack than are
laterites in an open savannah that was gradually replacing the calcic plagioclase feldspars of the basic rocks. They usually
forest, a condition believed t o exist where a dry season is weather t o kaolinite or halloysite.
prominent. Alexander and Cady (1962) have reported the Quartz being resistant t o leaching, does not weather to new
change of soft laterite t o hard form in Guinea a few years after mineral products. It is slowly dissolved. At ambient temper-
a forest was cleared. Aubert (1950) has also confirmed the atures the solubility of quartz is between 7 and 14 parts per
hardening of laterites following a change from forest t o million. This is considerably greater than the solubility of
savannah caused by man's activities. alumina, ferric oxide and titania at pH 7.
The conclusion drawn from the available literature on this The weathering sequence for the amphiboles is believed to
subject is that indurated laterites are most extensive in be similar t o the pyroxenes. Micas initially weather t o
savannah areas and that they form also in forest zones, vermic~lliteor montmorillonite and later alter t o kaolinite and

Partial loss of ( CHLORITE ) Complete ( KAOLINITE )

( MONTMO- ) loss of ( FERRlC )
( RILLONITE ) Ca*, Mg* ( OXIDES )
PYROXENE- Mg*,Ca*,Fe" ( CALCITE )- Partial ( ANATASE )
( FERRIC ) loss of
Oxidation ( OXIDES ) Si0.2
of Fe* ( ANATASE )
gibbsite. prevent the formation of more mature horizons, more defi-
3. Argillaceous sediments include all sedimentary rocks cient in kaolinite and silica.
containing clay minerals such as shales, claystones, argillaceous There is general agreement among various authors that
sands and sandstones, and low-grade metamorphic rocks such indurated laterite is commonly associated with flat areas and
as argillites and slates. The most abundant clay minerals in gently undulating or elevated plateaus and very rarely with
argillaceous rocks are illites. These are found in fine-grained steep slopes.
sediments of marine and freshwater origin.
Montmorillonites are not particularly common in sediment- Profile Development
ary rocks. They are initially formed during the weathering of
basic crystalline rocks, but they are altered relatively easily to CLASS1 FICATION
other clay minerals such as kaolinite or halloysite. For the purposes of this study the various tropically
Kaolinite is the most stable of the clay minerals and is a weathered soils are broadly differentiated on a genetic basis,
common constituent of continental and near shore sediments. determined by the soil-forming factors. The three major
It is the final clay mineral phase that forms in the weathering groups of significance have been defined by D'Hoore (1964).
sequence. It does alter to gibbsite (clay-size, but not a true His formal definitions are as follows:
clay mineral) with continued leaching, but this reaction is Ferruginous Tropical Soils. Ferruginous tropical soils gener-
completely reversible: gibbsite with resilification reverts to ally have ABC profiles. A marked separation of free iron
kaolinite. Halloysite is "either rare or unknown in sedimentary oxides is frequently observed, which may be either leached out
rocks." (Loughnan, 1969, p. 108). of the profile or precipitated within the profile in the form of
The major part of the African continent is geologically spots or concretions. The reserve of weatherable minerals is
extremely old. Much of the area is underlain by Precambrian often appreciable. The siltlclay ratio (2012 microns) deter-
rocks. In contrast, only a few crystalline rocks are Tertiary or mined by repeated dispersion, sedimentation and separation of
younger, and these are largely volcanic extrusives of Ethopia, the supernatant suspension is generally above 0.15. The clay
East Africa and Cameroon. Hence, the influence of the parent mineral is mostly kaolinitic but often contains small amounts
rock on the properties of lateritic soils in Africa is, in general, o f 2: 1 lattice clays; gibbsite is generally absent. The S i 0 2 /
a minor one. A1203 ratio is near 2 or slightly higher while the SiO2/R2O3
Nevertheless, the effect of parent rock appears t o influence ratio is always below 2.
many of the soils younger than Tertiary. Halloysite, of either The soil profiles are seldom thicker than 250 cm. Included
variety, is only reported over volcanic rocks in Africa, in the definition are the following sub-headings:
although it has been demonstrated this mineral forms over all 1) Ferruginous tropical soils on sandy parent materials - Ja
types of igneous and metamorphic rocks. Apparently 2) Ferruginous tropical soils on rocks rich in ferromagne-
halloys~te alters t o the more stable kaolinite structure with sian minerals - J b
time. This would explain why halloysite is found only over 3) Ferruginous tropical soils on crystalline acid rocks - Jc
volcanics in Africa. Outside of sediments, which do not 4) Ferruginous tropical soils, not differentiated - Jd
produce halloysite during weathering (Loughnan, 1969), vol- Ferrallitic Soils. Ferrallitic soils are often deep and the
canic rocks are the only exposures younger than Tertiary. The horizons are only slightly differentiated. There is little or no
volcanic rocks also occur in plateaus and mountainous regions reserve of weatherable minerals. The siltlclay ratio determined
wherc rainfall and humidity are sufficiently high t o retain the by repeated dispersion, sedimentation and separation of the
hydrated rorm in some of these soils. The Sierra Leone and supernatant suspension is generally less than 0.25 in horizons
other West African occurrences are in soils over basic intrusives B and C. Clay minerals are predominantly of the 1: 1 lattice
believed t o be of Triassic age. This age may be incorrect; but in type and are mostly associated with large quantities of iron
these areas, rainfall and humidity are high, and alteratiun to minerals, although they are generally associated with hydrated
kaolinite may be retarded. minerals of aluminum, gibbsite, one of the crystalline forms, is
Topography and Drainage. Topography controls the rate of not an essential constituent though frequently present. The
weathering by determining the amount of water that moves S i 0 2 / A 1 2 0 3 ratio is sometimes near 2 but usually below.
through the weathering zone. The percentage of rainfall Included in the definition are the following sub-headings:
deflected as surface run-off depends largely on the slope. The I ) Ferrallitic soils, dominant color yellowish-brown. on
steepness of the slope and the quantity of runoff determine loose sandy sediments - La
the amount of erosion that occurs. In the mountainous regions 2) Ferrallitic soils, dominant color yellowish-brown, on
of eastern Africa, for instance, erosion has not allowed the more or less clayey sediments - Lb
development of deep soil profiles. In western Africa, raw 3 ) Ferrallitic soils, dominant color yellowish-brown, not
mineral soils are found in the mountainous regions also, but differentiated - LC
there is also a wide band of weathered materials. These usually 4 ) Ferrallitic soils, dominant color red, on loose sandy
lie in areas of subdued topography, where erosion is not great sediments - Li
enough t o prevent profile development but is great enough to 5 ) Ferrallitic soils, dominant color red, on rocks rich in
 ferromagnesian minerals - Lm ration and transpiration as shown in the e.uplanatory text of
6) Ferrallitic soils, dominant color red, not differentiated - Fig. 9. The regions in which ferruginous soils occur show three
Ln moisture cycles including periods when I ) there is no water in
7) Humic ferrallitic soils - Ls the soil (water deficiency); 2) the water needs exceed
8) Ferrallitic soils with dark horizon - Lt precipitation (moisture utilization); and 3) the precipitation
9) Yellow and red ferrallitic soils on various parent exceeds water needs (moisture recharge).
materials - Lx The mean annual rainfall in the ferrallitic soil zones range
Ferrisols. Ferrisols have a profile closely resembling that of from 6 0 inches to over 160 inches (1 500 millimeters t o over
ferrallitic soils, often with a structural B horizon with 4000 millin~eters). In terms of a moisture region (Fig. 7)
aggregates having glossy surfaces. These surfaces are qot ferrallitic soils occur in moist sub-humid t o very humid
necessarily coatings of clay but may be due t o the presence of regions. The moisture diagram in these zones shows a complete
mixed alumina-silica gels. The reserve of weatherable minerals cycling of moisture conditions including periods when I) there
is generally low but may exceed 10 percent of the 50-250 is no water in the soil, 2) the water needs exoeed precipitation,
micron fraction. The silt/clay ratio (2012 microns) determined 3) the precipitation exceeds water needs and, 4) the precipi-
by repeated dispersion, sedimentation and separation of the tation exceeds both water needs and soil capacity (water
supernatant suspension is generally above 0.20 over igneous surplus). This water surplus appears t o be a distinctive
and metamorphic rocks. The clay fraction consists almost hydrological feature of ferrallitic soils and with tropical
entirely of kaolinite, free iron oxides and amorphous gels temperature would account for the little reserve of weather-
sometimes with small quantities of 2: 1 lattice clays and able minerals associated with the ferrallitic soil group.
gibbsite. The Si0,/A1,03 ratio is near 2, mostly slightly The mean annual rainfall in the ferrisol soil zone ranges
below. Ferrisols differ from the other soil groups in that from 5 0 inches t o 110 inches (1250 millimeters t o 2750
surface erosion prevents normal profile development and millimeters) and, in terms of moisture regions, from moist
forces the profile t o develop at depth in the less weathered sub-humid t o humid.
parent material. Information is lacking, but the moisture diagram is assumed
In the definition are the following sub-headings: t o be similar t o that given for the ferrallitic soil group, except
1) Humic ferrisols - Ka that there may be a shorter period when precipitation exceeds
2) Ferrisols on rocks rich in ferromagnesian minerals - Kb both the soil needs and capacity.
3) Ferrisols, not differentiated - Kc A general correlation exists among the various soil groups
Remillon ( 1967) has shown a general land form relating to and the vegetation map of Africa, Fig. 12. Generally, moist
the development of ferruginous soils and t o the development forests (rain forests) at low and medium altitudes cover
of ferrallitic soil. These are shown in Figs. 1 0 and 11. The ferrallitic soils while woodlands and savannahs (relatively
typical cross-section as shown for the ferruginous soil in- moist types) cover the ferrisols. Wooded and grass-covered
cludes: Horizon A as the weathered zone; Horizon B as the steppes and savannahs cover the ferruginous soils.
total accumulation zone: Horizon C as the zone of leaching of The environments of formation and the weathering pro-
the source rock and the liberation of the constituent elements; ducts of the major soil groups are given in Table 2. Mean
and Horizon D as the bedrock. The development of the values of pH are given for all soils tested during this study. The
ferruginous profile is based o n a defined wet season and dry mineralogy given is from x-ray diffraction analysis performed
season where the duration of the dry season is several months. for this study as well as that reported by Loughnan (1969).
The typical cross-section as shown for the ferrallitic soil One group of ferrallitic soils display properties considerably
includes: I) unaltered rock; 2) beginning of weathering; 3) different from all the other ferrallitic soils. This group is thc
weathering and kaolinite concentration zone; 4 ) zone of sandy ferrallitic soils (La in D'Hoore's classification). The areal
relative concentration of gibbsite and goethite; 5) leaching extent of this group is shown on the Soils Map (Fig. 13, rear
zone; 6 ) zone of total concentration of iron oxides and jacket). The causes of the abnormal sand content are the
hydroxides; and 7 ) zone of sandy-clay colluvium. The parent materials-either coastal sands or continental Quater-
development of the ferrallitic profile is based on wet tropical nary t o Recent sandy sediments of the interior basins. These
or equatorial climates that are characterized by high rainfall are in general restricted t o thin strips along the coasts and t o
with one or several dry seasons lasting for a period of only one large parts of the Congo and Kalahari Basins. The parent
t o three months. materials for these soils are geologically young and contain a
The mean annual rainfall in the ferruginous soil zones range high percentage of' resistant quartz. Climatic effects have not
from 2 4 inches t o 72 inches (600 millimeters t o 1800 had sufficient time t o chemically alter the constituents and
millimeters). In terms of a moisture region (Fig. 7), fer- dissolve much of the quartz.
ruginous soils occur in semi-arid t o moist sub humid regions. One subgroup of the ferruginous soils, those over sandy
More important perhaps than mean annual rainfall in the soil parent material (designated Ja by D'Hoore) is shown sepa-
forming process is the moisture diagram of a given area. The rately on the Soils Map. These deposits are generally found
moisture diagram includes precipitation, distribution, evapo- only along the coastal margins, where the parent material
 WEATHERED ZONE
TOTAL ACCUMULATION ZONE
ZONE OF LEACHING OF THE SOURCE ROCK AND T H E
LIBERATION OF CONSTITUENT E L E M E N T S
BED ROCK
I UNALTERED ROCK
2 START OF ALTERATION
3 WEATHERING AND CONCENTRATION OF KAOLlNlTE
4 ZONE OF R E L A T I V E CONCENTRATION OF G l B B S l T E
AND GEOTHITE
5 ZONE OF LEACHING
6 ZONE OF TOTAL CONCENTRATION OF THE OXIDES AND
HYDROXIDES OF IRON
-
7 ZONE OF CLAY SAND COLLURIUM

ZONE D'ALTERATION
ZONE DmACCUMULATION ABSOLUE
ZONE DE L E LESSIVAGE DE L A ROCHE DE DEPART E T
ENTRAINEMENT DES ELEMENTS CONSTITUANTS
ROCHE DE FOND
I ROCHE S A l N E
2 DEBUT DE LIALTERATION
3 ALTERATION E T CONCENTRATION DE K A O L l N l T E
4 ZONE DE CONCENTRATION RELATIVE D E GlBBSlTE
E T GOETHITE
5 ZONE DE LESSIVAGE
6 ZONE DE CONCENTRATION ABSOLUE DIOXYDES E T
D'HYDROXIDES DE FER
7 ZONE DE COLLUVIONNEMENT SABLES ARGILEUX

FIG 10 TYPICAL CROSS- SECTION FOR FERRALLlTlC SOlL FORMATION ( a f t e r R E M l L L l O N 1 9 6 7 )

COUPE TRANSVERSALE T Y PIQUE DE FORMATION D'UN SOL FERRALLITIQUE ( d'apres REMILLION 1 9 6 7 )

WEATHERING ZONE
TOTAL ACCUMULATION ZONE
ZONE OF LEACHING
BED ROCK

ZONE D'ALTERATION
ZONE D'ACCUMULATION
ZONE DE LESSIVAGE
ROCHE S A l N E

FIG I I TYPICAL CROSS - S E C T I O N FOR FERRUGINOUS S O l L FORMATION ( a f t e r Remillion, I 9 6 7 )

P R O F I L E N TRAVERS T Y P E ASSOCIE A L A FORMATION D'UN S O L FERRUGINEUX (dlaprbs Remillion, 1 9 6 7 )
 TEMPERATE AND SUBTROPICAL GRASSLANDS
' I,, ' PRAIRIE TEMPEREE E T SURTROPICALE M ERCATOR PROJECTION
P R O J E C T I O N MERCATOR

FIG. 12 VEGETATION M A P O F A F R I C A ( a f t e r A E T F A T , 1959)

CARTE DE LA VEGETATION EN AFRlQLlE (d'apres AETFAT, 1959)
 TABLE 2

Weathering Products in Relation To Environments

Environment pH (mean) Eh Behavior of Metallic Ions Mineralogy

Nonleaching, hot Alkaline Oxidizing Some loss of alkalies, Partly decomposed parent
rainfall 0-12 in iron present in ferric state minerals. Illite, chlorite
montmorillonite, and mixed
layered clay minerals,
Hematite, carbonates and
salts. Organic matter
absent or sparse.

FERRUGINOUS- Oxidizing Loss of alkalies and Hematite, goethite, and

Moderate leaching alkaline earths. Some loss kaolinite. Organic
hot climate of silica. Concentration of matter absent or sparse.
rainfall 24-72 in. alumina, ferric iron and
titania

FERRISOLS Oxidizing Loss of alkalies, alkaline Predominantly kaolinite

Intense leaching earths and silica. Concen- with goethite; hematite
hot climate tration of alumina, ferric and gibbsite less common-
rainfall over 50 in. oxide and titania. Organic matter sparse to
abundant.

FERRALLITIC Oxidizing Loss of alkalies, alkaline Gibbsite, goethite with

Intense leaching to earths and silica. Concen- kaolinite. Organic
hot climate Reducing tration of alumina. Iron matter may be abundant.
rainfall over 60 in. may be lost under reducing
Eh and concentrated under
oxidizing Eh.

consists of Quaternary to Recent deposits of marine sands, or D'Hoore, J. L., 1964. Soil map of Africa scale 1 to 5,000,000,
along the northern and southern limits for lateritic soils. The explanatory monograph. Comrn. for Tech Co-op in Africa;
eolian sands of the Sahara form the parent material in the Publ. 93: 205 p.
northern basins. In the South the eolian sands of the Kalahari Furon, R., 1963. The geology of Africa. Oliver and Boyd,
desert similarly form the parent material. Ltd., London: 377 p.
Grim, R. E., 1968. Clay mineralogy, 2nd Ed. McGraw-Hill,
REFERENCES New York: 596 p.
AETFAT, 1959. Vegetation map of Africa, Oxford University Holmes, A., 1965. Principles of physical geology, 2nd Ed.
Press. Konald Press Co., New York: 1288 p.
Alexander, L. T., and Cady, J. G., 1962. Genesis and Humbert, R. P., 1948. The genesis of laterite. Soil Sci., vol.
hardening of laterite in soils. U.S. Dept. Agr. Tech. Bull. 65: 281-290.
1282: 90 p. Keller, W.'D., 1957. The principles of chemical weathering.
Aubert, G . , 1950. Observations sur la degradation des sols et la Lucas Bros., Publ., Columbia, Md.
formation de la cuirasse lateritique dans le nord-ouest du Kilian, C., 1931. Des principaux complexes continentaux du
Dahomey. Trans. Fourth Int. Cong. Soil Sci., Amsterdam; Sahara. C. R. Soc. Geol. Fr.: 109-111.
~ 0 1 . 3 : 123-128. Krauskopf, K. B., 1959. The geochemistry of silica in
Crowther, E. M., 1930. The relationship of climatic and sedimentary environments. In: Silica in sediments. Sym.
geological factors to the position of soil clay and the Soc. of Econ. Paleon. and Miner., vol. 7: 4-18.
distribution of soil types. Proc. Royal Soc., vol. 107: 1-3. Loughnan, F. C., 1969. Chemical weathering of the silicate
 minerals. American Elsevier Publ. Co., Inc., New Observ., Haile Selassie I Univ.. Addis Abaha, No. 1 1: 65 p.
York: 154 p. Oxford Regional Economic Atlas, 1965. Africa, Oxford Univ.
Maignien, R., 1966. Review of research on laterites. National Press: 164 p.
Resources Research IV, United Nations Educ., Sci., and Remillon A., 1967. Les recherches routieres entreprise en
Cult. Org. (UNESCO), Paris: 148p. Afrique d'expression Franaise. An. de 1'Inst. Tech. du
Marbut, C. F., 1932. Morphology of laterites. Proc. Second Batiment et des Travaux Publics, No. 231-232: 366-388.
Int. Cong. Soil Sci., vol. 9: 76-80. Stephens, C. G., 1946. Pedogenesis following the dissection of
Mason, B., 1952. Principles of geochemistry. Wiley and Sons, lateritic regions in southern Australia. C.S.I.R.O., Australia,
Inc., New York: 276 p. Bull. No. 206.
Mohr, L. C. J. and Van Baren, F. A., 1954. Tropical soils. Thompson, B. W., 1965. The climate of Africa. Oxford Univ.
I
Interscience Publ., New York: 498 p. Press: 132 p.
Mohr, P., 1967. The Ethiopian rift system. Bull. of Geophys.
 LITERATURE RE VIEW ON THE ENGINEERING
CHARACTERISTICS OF LATERITIC SOILS

relative t o the other chemical components. The amount of

INTRODUCTION
alumina o r iron oxides is an important factor in differentiating
In many Countries of Africa and Asia, lateritic soils are the aluminous and ferruginous varieties. The base (alkalis and
traditional material for road and airfield construction. Though alkaline earths) are almost absent in lateritic horizons, except
a good deal of literature is available on lateritic soils and in some ferruginous crusts developed in alluvium and some
several excellent reviews have been prepared o n the subject concretionary horizons in ferruginous tropical soils. Other
(Maignien, 1966; Sivarajasingham, et a]., 1962; Florentin and common lateritic constituents are manganese, titanium,
L'Heriteau, 1952; Prescott and Pendleton, 1952; and Du- chromium and vanadium oxides.
Preeze, 1949), there has generally been very little discussion Chemical analyses do not usually reveal the origin, nature
on the engineering behavior of laterites. Until the U.S. Agency o r even the composition of laterites or lateritic soils. For
for International Development sponsored the study on laterite example, physically similar laterites may have different
and lateritic soils in Southeast Asia (Vallerga, et a]., 1969), the chemical composition and chemically similar laterites may
engineering investigations have been isolated studies. This has display different physical properties (Maignien, 1966).
been due t o the absence of uniform methods for the The mineralogical composition is considered to be more
preparation of samples and testing and the variable nature of important in explaining the physical properties of indurated
lateritic soils. As a consequence, the drawing of any rational laterite and lateritic soil. Maignien (1966) has divided the
conclusions on the engineering properties of lateritic soils has mineralogical constituents into major elements, which are
been very difficult. essential t o laterization, and minor elements which d o not
This review is devoted t o engineering studies on lateritic effect the laterization process. The major constituents are
soils of Africa. The perusal of the available data on lateritic oxides and hydroxides of aluminum and iron, with clay
soils gives the impression that the red color seems t o have been minerals and, to a lesser extent, manganese, titanium and
accepted by most authors as the most important property by silica. The minor constituents are residual remnants o r clastic
which these soils could be identified. Other obviously signifi- materials.
cant basic physical properties such as texture, structure, Iron occurs in many forms. The most common minerals are
consistency, etc., often were ignored. It is also noted that lack goethite or limonite and hematite. Maghemite, magnetite and
of uniformity in pretreatment and testing procedures (result- ilmenite have also been reported.
ing from association with different standards in different parts Alumina is common as gibbsite. Boehmite and amorphous
of Africa) makes it difficult to compare even textural data o n forms are also fairly common, but diaspore and corundum are
the same soils. It is noted that three major factors influence relatively rare. Titanium minerals include rutile, anatase and
the engineering properties and field performance of lateritic ilmenite.
soils. They are: The clay mineral most common in lateritic soils is kaolinite.
1. Soil forming factors (e.g. parent rock, climatic- Halloysite is also reported. lllite and montmorillonite are rare.
vegetational conditions, topography and drainage con- A study of some unpublished information and results of
ditions). analyses at the Ghana Geological Survey Department indicate
2. Degree of weathering (degree of laterization) and texture that due t o intense chemical weathering processes almost all
of soils, genetic soil type, the predominant clay mineral the unstable primary minerals are either partially o r complete-
type and depth of sample. ly broken down and transformed into secondary minerals. The
3. Pretest treatments and laboratory test procedures as well most common primary minerals encountered are quartz, mica
as interpretation of test results. (muscovite) feldspar (microcline) and hornblende. The particu-
lar mineral available depends upon the parent rock type. The
In the review emphasis is given to these factors insofar as chief clay mineral generally observed is kaolinite, with some
they affect the engineering properties, material selection and mica, sericite and illite.
classification, design and construction control procedures, The secondary minerals resulting mainly from the lateriza-
field performance and stabilization of lateritic soils. tion processes are gibbsite, goethite, limonite and hematite.
Neither manganese nor titanium minerals were observed in
CHEMICAL, MINERALOGICAL ARID significant amounts. The chemical and mineralogical character-
PHYSICO-CHEMICAL CHARACTERISTICS istics of all lateritic soils exhibit some similarities (Clare and
A distinctive feature of laterite and lateritic soils is the O'Reilly, 1960; Terzaghi, 1958; Matyas, 1969; Ruddock,
higher proportion of sesquioxides of iron and/or aluminum 1967; Morin and Parry, 1969; Baldovin, 1969; and Bawa,
1957). There is very little information on the physico-chemical Another factor which has been found to affect the
characterisitics of lateritic soils. A few data published on this sedimentation test is the method of drying. Oven-Jried
subject have been confined to the pH, organic content, and lateritic soils were found to give the least amount of clay
calcium carbonate content (Clare and O'Reilly, 1960). The fraction, as compared to air-dried or as received (natural
conclusions reached show that the lateritic medium is either moisture content) samples (Moh and Mazhar, 1969). The
neutral or acidic and the organic content is generally low, decrease in the clay content was accompanied by an increase
below 2%. The calcium carbonate content is either very small in silt and sand fraction contents as a result of the cementation
or lacking in most lateritic soils profiles (Gidigasu, in and coagulation of the clay particles by free iron oxide into
preparation). clusters (Terzaghi, 1958). The variation in the grading of
The distribution of such properties as the pH, base lateritic gravels with the method of manipulation is also widely
exchange capacity, hygroscopic moisture content, calcium reported (Novais-Ferreira, 1965 and Nascimento et al., 1959).
carbonate content, organic content and the chemical compo- In the study of the particle-size distribution of lateritic
sition of soil can form a basis for the understanding of soils, three sources of confusion were noted. The first
behavior of lateritic soils as a material of construction confusion arises from the belief by some authors, e.g. Bawa
(Gidigasu, in preparation). The significance of the physico- (1957), that lateritic soils represent a group of materials that
chemical properties t o the success or otherwise of soil can be defined within a specific range of particle-size distribu-
stabilization can hardly be over-emphasized. Lack of data of tion. The second source of confusion seems to arise out of
physico-chemical properties is also a major factor hindering attempts by some authors to confine the word laterite to
the understanding of the soil-water relationships in lateritic concretionary lateritic gravels. The third source of confusion
soils. arises out of the attachment of unnecessary importance to the
soil color.
PHYSICAL A N D ENGINEERING PROPERTIES Nascimento et al. (1959) have suggested an interesting
General papers on the properties of laterite and lateritic lithological classification of lateritic soils as follows:
soils have been prepared by De Medina (1964); Clare and Lateritic clays <0.002 mm
O'Reilly (1960); Bawa, (1957); Nixon and Skipp (1957); Lateritic silts =0.002 - 0.06 mm
Remillon (1 955); Christophe (1 949); and Fruhauf (1 946). Lateritic sands ~ 0 . 0 6- 2 mm
Data on occurrences in various parts of the world are reported. Lateritic gravel =2 - 60 mm
Selected data on African soils are given in Table 3. Laterite stones
and cuirasse X 0 mm
Texture In this way the textural significance of "laterite" is
Experience with soils in stable temperate zones has revealed dispensed with.
that particle-size distribution exerts great influence on the Studies on indurated laterite reveal that fairly good
engineering properties of soils. It is also one of the most assessment of their engineering properties and field perform-
important properties by which soils can be easily identified ance as road materials (aggregates) can be obtained from test
and classified on the basis of simple field and laboratory tests. procedures established for natural rocks and aggregates
Consequently great importance has also been accorded to (Ackroyd, 1960; Novais-Ferreira and Correia, 1965; and de
particle-size distribution in dealing with lateritic soils. Graft-Johnson et a]., 1969).
Recent studies have revealed however, that lateritic soils are The most important differentiating factor with regard to
strikingly different from temperate zone soils in terms of indurated laterite seems to be hardness. Attempts have been
genesis and structure. Their concretionary structure as com- made to group these materials and some success has been
pared to the dispersed temperate zone soils have necessitated attained. The summary of the results so far obtained will be
modifications t o mechanical or grading tests (Remillion, 1967; discussed under appropriate sections together with those of
1955). other lateritic soils.
Consistent reports of variations in the particle-size distribu- One of the main characteristics of lateritic gravels and
tion with methods of pretreatment and testing have been gravelly soils is the high content of fines. Consequently, such
widely reported on laterite soils. Schofield (1957) found that materials d o not fit into the existing temperate-zone
wet sieving increased the silt and clay fraction from 7 t o 20% classification systems for coarse-grained soils. The grading
as compared t o the dry sieving. It has been found that sodium curves often have flat platforms between 4 mm and 0.5 mm
hexametaphosphate generally gives better dispersion of the revealing the absence of this fraction. Studies on lateritic
fine fractions. It was also found, for example, that using gravels by de Graft-Johnson et al. (1969) among others, have
sodium oxalate on a halloysitic clay from Kenya gave between shown that the grading, though important for identification
20 and 30% clay fraction, while the sodium hexametaphos- purposes, cannot alone form the basis for grouping lateritic
phate gave as high as between 40 to 50% clay fraction for the gravels in terms of mechanical properties. The strength of the
same soil (Quinones, 1963). aggregates was found to be an important factor. On the basis
 TABLE 3- PHYSICAL AND CHEMICAL PROPERTIES

HAYTER & CAIRNS

OE GRAFT-JOHNSON Lateriteclays Ghana Muscovite 96.6-91.0 25.0-2.8 52-4 12-4 0.5-6.0

BHATIA & GlOlGASU Biotite
(1967, 1969) Granite
Shales & Ghana Compaction
Lateritic soil Phy llites 104-110 16.22 14-54 8-18 (Gravelly clays)
Subbase gravel 118-130 11.16 2085 15-64 " " (Subbas)
Basa " Granite, gneiss 130.138 6.10 68-120 60.74 " " (Bare)
of studies of lateritic aggregates in Nigeria, Ackroyd (1960) soils may be explained in terms of their genesis.
distinguished hard, medium and weak lateritic gravels. It was The available literature on the genesis of lateritic soils has
also established that the strength of the aggregates is mainly a emphasized the physical (hardening) and mineralogical aspects
function of the degree of maturity of the lateritic concretion- of the laterization processes, but has almost completely
ary particles and the predominant sesquioxide in the aggre- ignored the physico-chemical aspects which involves the
gates (Novais-Ferreira and Correia, 1965). Soils with hard coating of the soil particles by iron'oxide and alumina.
aggregates generally have stable grading curves while weak According to Moh and Mazhar (1969), the increasing degree
gravels give different grading curves with different methods of of laterization results in an increase of the thickness of free
pretreatment. iron oxide coating of the soil particles. The soil particles later
The problem of particle-size distribution of the fine-grained coagulate into large clusters with subsequent reduction of
lateritic soils is more complex than those of the gravels and specific surface. It seems that much of the differences between
gravelly soils. Information on the grading of fine-grained soils temperate-zone and lateritic soils may be explained on the
is rather scanty, because they have been least studied. The silt basis of this coating and coagulation of the soil particles.
and clay contents reported vary from 12% (Winterkorn and
Chandrasekharan, 195 1) to over 82% (Clare and O'Reilly, Atterberg Limits
1960).
One reason for the wide variation in these soils is the Remillon (1967) grouped western and central African
tendency to form aggregations or clusters of clay-minerals and lateritic soils on the basis of climatic conditions and their
clay-size mineral fragments. This tendency has been noted by relation to the consistency parameters. He distinguished the
many authors (Tateishi, 1967; Quinones, 1963; Newill, 1961; ferrallitic soils formed at the wet forest regions (rainfall above
Clare and O'Reilly, 1960; Terzaghi, 1958; Hirashinla, 1948; 1200 mm per annum) from the ferruginous soils of the dry
and Fruhauf, 1946). Hirashima (1948) found that the Hawai- savannah regions (rainfall below 1200 mm per annum). Clare
ian clays become granular upon drying. (1960) also found some correlation between the consistency
Halloysite is particularly susceptible to aggregation. Robert- characteristics and the rainfall distribution for the Zambian
son (in Terzaghi, 1958) found that the halloysite in the soils.
Sasumua clay does not deflocculate like kaolinite or mortmor- The available data on lateritic clays of western Africa and
illonite even in efficient dispersing agents. Aggregate sizes up those of eastern Africa indicate that the parent material and
to 10 microns were measured. He concluded that the clay is the climatic conditions influence the nature of the predomi-
strongly aggregated in its natural state and the aggregates will nant clay mineral and hence the levels of plasticity of the soils
not lose their strength unless deflocculated which is unlikely containing these minerals. Montmorillonitic tropical clays in
to occur (in this case, in a dam fill). Halloysite soils often fall Ghana plot above the A-line while the halloysitic clays over
below the A-line on Casagrande's plasticity chart, indicating volcanic ash plot below the A-line although both are very
they are silt-like in character (Clare and O'Reilly, 1960, plastic materials. Micaceous granite soils from Ghana and
Terzaghi, 1958). Liberia both plot below the A-line as soils of medium
Tateishi (1967) has proposed the "aggregation index" as a plasticity. Soils over dolerites from Guinea, and over schists
measure of the change in physical properties of a tropical soil from the Ivory Coast, plot below the A-line while soils over
due to dehydration. He defined this index as the sand schists and phyllites in Ghana generally plot above the A-line.
equivalent value of the oven-dried sample divided by the sand Examination of coarse grained soils from different parts of
equivalent of the soil in its natural state. The sand equivalent Africa, however, revealed a general uniformity of plasticity
test measures the relative percentage of fines in a soil (AASHO characteristics; most of these materials plot above the A-line
Designation T-176-65). Tateishi classified soils with an aggre- and have liquid limits generally below 60%. Ruddock (1967)
gation index of 1 as those which display no difference in observed that reliable Atterberg limit test results could not be
properties with dehydration; an index of 2 would indicate a obtained with some micaceous silty soils in Ghana because the
moderately sensitive soil. plastic paste instead of shearing during the liquid limit test,
merely slid in the cup. The position of the sample in the
laterite profile was also found to influence the plasticity
Structure characteristics (Baldovin, 1969). Because the formation of a
One major characteristic of all soils which influences other lateritic soil profile involves differential weathering as well as
engineer properties is the structure. By virtue of the processes movement and deposition of dissolved materials, the variation
of laterization, lateritic soils have acquired peculiar structural of the plasticity characteristics with depth cannot be predicted
characteristics. Though it is an over-simplification to identify even in two similar profiles on different topographical sites.
all temperate zone soils with dispersed structure and lateritic The influence of the pretreatments and testing procedures
soils with concretionary structure (Remillon, 1967), lateritic on the plasticity characteristics have been widely studied and
soils may be on the whole more concretionary than most discussed. The variations in test results due to pretreatment
temperate zone soils. The concretionary nature of the lateritic and testing procedures have made the interpretation of test
results very difficult. Remillon (1955) states that mixing the there are no clear conclusions and some findings are conflict-
soils with water during the testing procedure causes the ing. A perusal of data on the density and compaction
breaking up of the fine particles and also deflocculation. characteristics of these soils shows that the factors which
Terzaghi (1958); Clare and O'Reilly (1960); Newill (1961); influence these properties can be separated into two groups.
Quinones (1963); and Frost (1967) all found the limits change The first group of factors derive from the genesis of the soil
with drying and with manipulation. Moh and Mazhar (1969) while the second group derive from pretreatment procedures
and Novais-Ferreira and Meireles (1969) found less significant before testing.
changes. Newill states (p. 303), "when liquid limit tests were The most important genetic factor is texture. The densities
carried out the aggregations of clay particles were broken of indurated laterites are generally above 140 pcf. with natural
down by the manipulation involved; this led to difficulties in moisture contents between 0%. and 6%. For the gravels the
obtaining consistent values for the liquid limit. The treatment maximum dry densities range between 115-145 pcf and the
applied to the soil also created problems in relating the results optimum moisture contents between 6%- 16%; for gravel-
of laboratory tests to the behavior of the soil in the field, sand-clays and for sand-clays the corresponding densities are
where it does not necessarily receive equivalent treatment." 100-120 pcf and the optimum moisture contents are 16% to
24%. The maximum dry densities are generally below 100 pcf
Activity with optimum moisture contents above 24% for the clays. The
Skempton's colloidal activity determined as the ratio of the variations as shown are also due to different compactive
plasticity index of the clay content in the fines is generally low efforts.
for lateritic soils. Because the predominant clay minerals in The moisture-density relations for typical micaceous sandy
lateritic soils are of the kaolinite group these soils are known soils developed over granite in Liberia, Ivory Coast and Ghana
t o be inactive or normal (activity less than 1.25). The colloidal were determined by Quinones (1963); Florentin et al. (1957);
activity for some Ghanaian soils ranges between 0.4 and 1.0 and Gidigasu (1969). The low natural densities of the soils of
while that for non-residual soils and tropical clays is generally the residual granite soils of Ivory Coast were shown to result
above 1.0 (Gidigasu, 1969). Nixon and Skipp.(1957) report a from the high degree of weathering of the parent rock. The
range of activity of 0.50 to 1.75 for 18 tropical red soils. decrease in density between that of the parent rock and the
Mellier and Phlipponnat (1969) show an activity of 1.8 for weathered product was found t o be over 50%. In contrast,
Gabon. The low activity of most lateritic soils may be due to another aspect of the effect of weathering on the density is
the mode of weathering which involves the coating of the soil illustrated by the fact that some laterite rocks found in Ghana
particles with sesquioxides, resulting in the suppression of the have also gained in density over 30% in relation t o the parent
surface activity of the clay particles. rock.
Another genetic factor identified with the compaction
Specific Gravity characteristics is the position of the sample in the soil profile
(depth of sample). The significance of this factor is elaborated
The available data indicate that specific gravities vary not upon below but it is associated with depth of desiccation in
only with the textural soil groups but also within different lateritic soils.
fractions. In the first place lateritic soils have been found to Numerous researchers have investigated the effect of
have very high specific gravities of between 2.6 to 3.4 (de pretreatment and testing procedures both in Africa and
Graft-Johnson and Bhatia, 1969). For the same soil, gravel elsewhere. The effects of the method of drying on the
fractions were found to have higher specific gravities than fine compaction characteristics have been investigated for soils
fractions due to the concentration of iron oxide in the gravel varying in texture from clays (Newill, 1961; Tateishi, 1967) to
fraction, while alumina is concentrated in the silt and clay gravels and gravelly soils (Brand and Hongsnoi, 1969). It was
fractions (Nascimento et al., 1959; Novais-Ferreira and shown that oven-drying always gives the highest maximum dry
Correia, 1965). The specific gravity has been used as a measure densities and lowest optimum moisture contents, while soils at
of the degree of maturity (laterization) by some authors (e.g. natural moisture content give the lowest maximum dry
Ackroyd, 1960). It is common to see specific gravities densities and highest optimum moisture contents. The effect
reported for the gravel and fines separately. The average of the of drying temperatures on the compaction of lateritic and
two values can be assumed t o be more representative of the related soils was also investigated (Frost, 1967; Novais-Ferreira
specific gravity for the whole soil. and Meireles, 1969) and the same trend of conclusions was
observed.
DENSITY AND COMPACTION
CHARACTERISTICS Quinones (1963) found that Hawaiian soils display irrevers-
Most of the results available on the density and compaction ible changes with drying which are either caused by the
characteristics of African lateritic soils have limited applica- dehydration of the hydrated oxides of iron and alumina or of
bility because no systematic attempt has been made to the mineral halloysite. Compaction curves developed from the
consider all the factors in any single study. The result is that air-dried or oven-dried condition display higher maximum
densities and lower optimum moistures than compaction all soft concretions that do not perform satisfactorily.
curves developed by determining points as the soil dries from Remillon (1967) used a field penetration test to determine
its natural moisture content. Such differences can be signifi- the strength-depth variations for the ferrallitic soil and
cant for soils with in-situ moisture contents well above the ferruginous soil profiles in western and central Africa. The
optimum. The density determined from a dried sample is, in variations in strength with depth are due to variation in the
effect, for a different material. Brand and Hongsnoi (1969) degree of laterization and homogeneity in the profiles.
noted less significant changes in Thailand soils. The strength and hardness of indurated laterite is a function
The effect of hydrated halloysite on the compaction curves of chemical composition, age and homogeneity. The iron-rich
has been investigated by Newill (1961). Soil from Sasumua, indurated laterites are harder than alumina-rich ones (Nasci-
Kenya which had been dried produced a higher dry density mento et a]., 1959). The greater the homogeneity and the less
than one which had not been dried. However, he considers the the content of foreign bodies, the harder the material; older
difference illusory and due to the loss of interlayer watej in indurated laterites are generally harder than younger ones.
changing hydrated halloysite to metahalloysite in oven-drying The aggregate tests (Aggregate Impact, Los Angeles Abra-
the moisture samples. When corrected for the loss of interlayer sion), have been extensively used on the concretionary lateritic
water, both curves are in close agreement. It is nevertheless gravels in Africa (Novais-Ferreira and Correia, 1965 ; De
desirable to perform compaction tests from the natural Castro, 1969; de Graft-Johnson et al., 1969; Ackroyd, 1960).
moisture content for those soils which display irreversible The results of various studies have shown that aggregate tests
changes with drying. Compaction curves so developed include could be very useful guides for the selection of these materials
the effect of the loss of interlayer water and will be valid either as base for road construction or as aggregate for
criteria for controlling compaction during construction. concrete. The most popular test for assessing the strength of
The breakdown of weak concretionary lateritic gravels laterite gravels remains the CBR (Remillon, 1967; Ackroyd,
during compaction also affects the compaction characteristics. 1959; Meireles, 1967).
Whether higher or lower densities are obtained after compac- Triaxial compression tests have been reported on a few
tion depends on whether the breakdown leads to improvement lateritic soils. Baldovin (1969) found that at water contents
in grading or deterioration. The degree of breakdown may be a below optimum, the angle of friction of a quartz-rich lateritic
function of the degree of maturity (laterization) and the soil from Cameroon was about 40". Remillon (1967) states
physical characteristics of the particles (e.g. angularity, flaki- that most lateritic soils display angles of over 40". Trow and
ness, etc.). The conclusions reported on the effect of compac- Morton (1969) report an angle of 36.5" (with 4 psi cohesion)
tion on the particle breakdown have revealed that the reuse of for a Dominican Republic lateritic soil.
concretionary lateritic soils should be discouraged (Remillon,
1967). CONSOLIDATION AND
PERMEABILITY CHARACTERISTICS
Strength
- Properties The consolidation and permeability characteristics of tropi-
Strength tests on lateritic soils vary mainly with texture. cal lateritic soils have been recently summarized (Little, 1967;
Crushing tests, field penetration tests (Remillon, 1967) aggre- de Graft-Johnson and Bhatia, 1969). The compressibilities are
gate and durability tests (Ackroyd, 1960; Nascimento et a]., generally moderate with the modulus of compressibility
1959; Novais-Ferreira and Correia, 1965) are used for indur- ranging between 1 x 1(r3 to 1 x 1(r2 sq. ft./ton. The
ated laterite and concretionary gravel. The CBR, compression, permeabilities show wide variations, being generally higher for
shear, penetration and weathering tests are used for the the residual soils and less for the clayey and compacted
gravelly and fine grained soils. materials. The higher the molding moisture content the lower
Studies in Angola by Nascimento et al, (1959) have shown the permeability for compacted soils.
that though the crushing strengths are generally high they all
fall below 100 kg/cm2 (L.N.E.C., 1969). Ackroyd (1960) UTILIZATION OF LATERlTlC SOILS
classified concretionary gravels of western Nigeria on the basis Bawa (1957) has stated, "The suitability of any type of
of the crushing strength of the 10 percent size. laterite soil deposit can usually be judged by a systematic
Novais-Ferreira and Correia (1965) have developed a study of the different engineering characteristics of the soil,
method of evaluating the hardness of lateritic concretions. relevant to a given structure." On the basis of this statement it
Their procedure is to divide the sum of the percentages can be generally said that the performance of any soil type
retained on the 1-inch;314-inch; 112-inch; 318-inch; No. 4; No. under given traffic and environmental conditions depends on
10; No. 40; and No. 200 sieves (which they define as the many factors. Much of the literature discussing the perform-
granulometric modulus) for the natural material by the sum ance of lateritic soils as materials of construction fails to make
of the percentages retained after 200 revolutions in a Los any reference to:
Angeles abrasion machine without a charge. They term this the 1. The engineering characteristics of the soils, their
"hardness index" and claim a minimum value of 80 excludes selection and method of placement under pavements.
 2. The nature, pattern, weight and intensity of existing LL = 25% and PI = 6% have been relaxed to 40 and
and predicted traffic. 15 respectively, provided the swelling does not
3. Climatic and drainage conditions. exceed 10%.
3. The requirements for maximum values of LL = 35%
These three groups of factors are closely related to field
and PI = 4 t o 9% for surface courses have been
performance. Obviously it is not only important to say that a
relaxed respectively t o 40% and 6 - 15%, provided the
soil type is performing badly but also t o be able t o explain
swelling does not exceed 10%.
why the situation is so.
Airfields constructed on lateritic soils have been described
Highways and Airfields in a few papers. Nixon and Skipp (1957) and Dreyfus (1952)
provide general papers on various lateritic soils. Da Silva
A number of authors have described the use of lateritic soils (1967) gives the details on 4 airfields in Angola with lateritic
in highways. Remillon (1967, 1955) and Christophe (1949) soil base courses that are performing satisfactorily. As with
have given details of highway construction in western and highways, lateritic soils of good quality are satisfactory for
central Africa. De Medina (1964) has given details of lateritic airfield construction. The same specifications can be applied.
soils in Brazil and West Africa. French experience has been
reported in papers in Construction (La Technique Moderne), Dams and Foundations
September 1955: on corrugations by B.C.E.E.O.M.; on later-
itic soils of Gabon, Cameroon and Central African Republic by There is relatively scanty information on the behavior of
Frejacques and Remillon; soils problems around Dakar, Sene- lateritic soils as foundations for structures. Experience in
gal by Dreyfus; on lateritic gravels from Senegal and Cameroon Ghana shows that this is a function of the texture of the soil,
by Cans; on the utilization of laterite in Cameroon by Chirat; the maturity of the lateritic soil and the drainage conditions.
and other papers on specific road projects Clare and O'Reilly Generally lateritic soils perform badly only when they have
(1960) have provided data on a number of tropical red clays been subjected to rough handling, involving destruction of
used for highways in Africa and some Caribbean islands. structure, or subjected t o poor drainage condition. Indurated
Wiggington (1967) described the use of lateritic gravel in laterite horizons are often very good foundation supports even
Thailand. Millard (1962) has described some of the problems for heavy structures, provided, of course, that the thickness is
associated with road building in the tropics. great enough. Analyses of bearing capacity and settlements
Generally speaking, good quality lateritic soils perform based on the standard theories established for the temperate-
satisfactorily in highways. Specifications need not be as zone soils have generally given good field results.
stringent as with temperate zone soils. Plastic limits are usually Though very little information is available on the perform-
higher and swelling lower than in other soils with comparable ance of lateritic soils as earth structure, experience with the
clay content. However, inferior lateritic soils have sometimes red volcanic halloysitic clays in Kenya (Terzaghi, 1958)
been employed which have not been adequate. showed that these soils could be used as earth dam materials.
The distribution and engineering properties, as well as Some applications of lateritic soils in dam construction have
specifications for highway utilization, of lateritic soils from been described by Little (1967).
Nigeria have been covered in detail (Clare and Beaven, 1962; Other applications include hammer foundations (Prakash
Ackroyd, 1967, 1963, 1960). The stipulated grading specifi- and Basavanna, 1969) and pile foundations (Chin, 1969). Vold
cations for the base and surfacing materials have been (1967) has applied a bilateral load test (a plate bearing test
summarized by Ackroyd (1959). These grading and plasticity per f o r m e d horizontally in a trench) t o determine
limits are based on the ASTM and British Road Research bearing capacity.
Laboratory specifications. Since the CBR method of pavement
design is generally used, the strength requirements are based STABILIZATION OF LATERlTlC Sol LS
on CBR tests. A minimum CBR of 80% after 24 hours soaking
Stabilization of lateritic soils has been successful with both
for mechanically stable road-base material, using the Ghana
cement and lime, although some soils do not show a marked
method of compaction, is adopted in Ghana and Nigeria.
improvement. Satisfactory results with cement have been
Specifications regarding the use of concretionary aggregates
reported by Moh and Sheikh (1968); Moh et al., (1967);
are based on the studies by Ackroyd (1960) as a result of
Santos (1969); De Medina (1964); Gregg (1963); Clare and
which gravels are rated according to strengths of the
O'Reilly (1960); Road Research Lab (1958); Jaouen (1955);
aggregates.
and Frejacques and Remillon (1955). Only partially successful
In Angola, ASTM Specification D1241-551 is generally
results have been reported by Wiggington (1967 and Winter-
accepted with the following reservations (De Castro, 1969):
korn and Chandrasekharan (195 1).
1. In base and sub-base course materials the Los Angeles Lime has not been used as extensively as 'cement for
abrasion should not exceed 65%. stabilizing lateritic soils. Millard (1962) and Clare and O'Reilly
2. The requirements for maximum values of liquid limit, (1960) consider both equally effective. Biszewski (1963)
reports significant improvement with lime. Hayter and Cairns soils for use as road foundations in northern Rhodesia.
(1966); Cregg (1963); Cartmell and Berg (1958); and Scho- Dept. Sci. and Ind. Res., (U.K.), Road Res. Lab. Overseas
field (1957) give details on a number of lateritic soils Bull. No. 9: 28 p.
satisfactorily stabilized with lime. Chin, F. K., 1969. The point resistance of model piles in
Lateritic soils have been stabilized with other additives. remoulded lateritic clay. Proc. of Spec. Session on Eng.
Muktabhant and Ongskul(1969) stabilized a lateritic soil from Prop. of Lateritic Soils, Seventh Int. Conf. Soil Mech. and
Thailand with sand, but report an optimum sand content of Found. Eng., Mexico City, Vol. 1 : 177-187.
52% for maximum dry density. Santos (1969) successfully Chirat, C., 1955. L'utilization des laterites pour la construc-
stabilized Philippine lateritic soils with volcanic cinders. tion des routes bitumees. Construction, vol. 10: 95-105.
Asphalt has been utilized successfully for stabilization accord- Christophe, L., 1949. Les laterites et leur utilization comme
ing to Nixon and Skipp (1957), and with only partial success materiau de construction des routes. Rev. Cen. des Routes,
by Winterkorn and Chandrasekharan (1951). Actual construc- et Des Aerodromes, No. 212: 3 8 4 0 .
tion experience with asphalt stabilization has not been Clare, K. E., 1960. Roadmaking gravels and soils in central
reported. Africa. Dept. of Sci. and Ind. Res., (U.K.), Road Res. Lab.
Overseas Bull. No. 12: 4 3 p.
REFERENCES Clare, K. E., and Beaven, P. J., 1962, Soils and roadmaking
materials in Nigeria. Dept. Sci. and Ind. Res., (U.K.), Road
Ackroyd, L. W., 1967. Formation and properties of concre- Research Lab. Tech. Paper No. 57: 54 p.
tionary and non-concretionary soils in western Nigeria. Clare, K. E., and O'Reilly, M. P., 1960. Road construction
Proc. Fourth Reg. Conf. for Africa on Soil Mech. and over tropical red clays. Conf. on Civil Eng. Problems
Found. Eng., Capetown, Vol. 1 : 47-52. Overseas. Inst. Civil Eng.: 243-256.
Ackroyd, L. W., 1963. The correlation between engineering Da Silva, C. M. X. A., 1967. Some applications of laterites in
and pedological classification systems in western Nigeria air fields. Proc. Fourth Reg. Conf. for Africa on Soil Mech.
and its implication. Proc. Third Reg. Conf. for Africa on and Found. Eng., Cape Town, Vol. 1 : 83-85.
Soil Mech. and Found. Eng., Salisbury, Vol. 1 : 115-118. De Castro, E., 1969. A swelling test for the study of lateritic
Ackroyd, L. W., 1960. Notes on the crushing strength of some soils. Proc. of the Spec. Session on Eng. Prop. of Lateritic
western Nigerian concretionary gravels and their selection Soils, Seventh Int. Conf. on Soil Mech. and Found. Eng.,
for use as a building material. Min. of Transport (Nigeria), Mexico City, Vol. 1 : 97-106.
Tech. Paper 6. De Craft-Johnson, J.W.S., Bhatia, H.S., and Cidigasu, D.M.,
Ackroyd, L. W., 1959. The engineering classification of some 1969. The engineering characteristics of the lateritic gravels
western Nigerian soils, and their qualities in road building. of Ghana. Proc. of the Spec. Session of Eng. Prop. of
Dept. of Sci. and Ind. Res. (U.K.), Road Res. Lab. Overseas Lateritic Soils, Seventh Int. Conf. Soil Mech. and Found.
Bull. No. 10: 32 p. Eng., Mexico City, Vol. I : 117-128.
Baldovin, C., 1969. The shear strength of lateritic soils, Proc. De Craft-Johnson, J. W. S., and Bhatia, H. S., 1969.
of the Spec. Session of Eng. Prop. of Lateritic Soils, Engineering properties of lateritic soils. General Report,
Seventh Int. Conf. on Soil Mech. and Found. Eng., Mexico Spec. Session on Eng. Prop. of lateritic soils, Seventh Int.
City, Vol. 1: 129-142. Conf. Soil Mech. and Found. Eng., Mexico City, vol.
Bawa, K.S., 1957. Laterite soils and their engineering charac- 1: 117-128.
teristics, Proc. Amer. Soc. Civil Eng., Jour. Soil Mech. and De Medina, J., 1964. Les laterites et leur application a la
Found, Div., Vol. 83: 1-15. technique routiere. Rev. Cenerale des Routes, No.
Biszewski, E., 1963. The effect of time on the strength of 382: 81-94.
lateritic gravel stabilized with hydrated lime. Proc. T h r d Dreyfus, J., 1955. Les laboratories de sols d7A.0.F. Construc-
Reg. Conf. for Africa on Soil Mech. and Found. Eng., tion, Vol. 10: 25-28.
Salisbury, Vol. 1 : 63-67. Dreyfus, J., 1952. La laterite. Rev. Cen. des Routes et
Brand, E. W., and Hongsnoi, M., 1969. Effects of method of Aerodromes, Paris, June.
preparation on compaction and strength characteristics of Du Preez, J. W., 1949. Laterite: a general discussion with a
lateritic soils. Proc. of the Spec. Session on Eng. Prop. of description of Nigerian occurrences. Bull. Agr. Congo
Lateritic Soils, Seventh Int. Conf. on Soil Mech. and Belge, No. 40: 53-66.
Found. Eng., Mexico City, Vol. 1 : 107-116. Florentin, J., L'Heriteau, C., and Farhi, M., 1957. Etude de
B.C.E.E.O.M. (Bureau Central d'Etudes pour les Equipments quelques propietes physico-mecaniques d'echanitillons de
d'0utre-Mer), 1955. Les procedes de lutte contre la tole laterites eluviale. Fourth Int. Conf. Soil Mech. and Found.
ondulee. Construction, Vol. 10: 14-18. Eng., London, Vol. 1 : 28-3 1.
Cans, C., 1955. Les chaussees en tout venant lateritiques. Frejacques, J., and Remillon, A., 1955. Stabilisation des couch
Construction, Vol. 10: 60-68. de roulement et utilization des materiaux d'apport. Con-
Cartmell, H. S., and Bergh, A. O., 1958. Lime stabilization of struction, Vol. 10: 71-79.
Frost, R. J., 1967. Importance of correct pretesting prepar- properties of some African lateritic residual soils. Proc. of
ation of some tropical soils. Proc. First Southeast Asian Spec. Session on Eng. Prop. of Lateritic Soils, Seventh Int.
Conf. on Soil Eng., Bangkok: 43-53. Conf. on Soil Mech. and Found. Eng., Mexico City, Vol.
Fruhauf, B., 1946. A study of lateritic soils. Proc. High. Res. 1: 53-63.
Board, Vol. 26: 579-589. Muktabhant, C., and Ongskul, S., 1969. Stabilization of
Gidigasu, M. D., 1969. The highway geotechnical properties of lateritic soil with sand. Proc. of Spec. Session on Eng. Prop.
Ghana soils. Unpublished Ph.D. Thesis, State Technical of Lateritic Soils, Seventh Int. Conf. Soil Mech. and Found.
University, Warsaw. Eng., Mexico City, Vol. 1: 153-164.
Gidigasu, M. D., in preparation. The importance of soil genesis Nascimento, U. De Castro, E. and Rodrigues, M. 1963, 1964.
in the engineering classification of Ghana soils. Engineering Swelling and petrifaction of laterite soils. Proc. Third
Geology. African Reg. Conf. on Soil Mech. and Found. Eng.,
Gregg, J. S., 1963. The performance of several stabilized roads Salisbury, Vol. 1.
and some concepts on design. Proc. Third Reg. Conf. for Nascimento, U. et al. 1959. As laterites do ultramar Portu-
Africa on Soil Mech. and Found. Eng., Salisbury, Vol. guese. Laboratorio Nacional de Engenharia Civil, Lisbon,
1 : 77-79. Memoria No. 14 1.
Jaouen, J., 1955. Chaussees en sols-ciment. Construction, Vol. Newill, D., 1961. A laboratory investigation of two red clays
10: 56-59. from Kenya. Geotechnique, vol. 12: 302-318.
Hirashima, K. B., 1948. Highway experience with thixotropic Nixon, J. K., and Skipp, B. A., 1957. Airfield construction on
volcanic clay. Proc. High. Res. Board, Vol. 28: 481494. overseas soils, Part 5: Laterite. Part 6. Tropical red clays.
Laboratorio Nacional De Engenharie Civil (Lisboa), et al., Proc. Inst. Civil Eng., Vol. 8: 253-292.
1969. Portuguese studies on engineering properties of Novais-Ferreira, H., and Meireles, J. M. F., 1969. The influence
lateritic soil, Proc. of the Spec. Session on Eng. Prop. of of temperature of humidification o f . the geotechnical
Lateritic Soils, Seventh Int. Conf. on Soil Mech. and properties of lateritic soils. Proc. of the Spec. Session on
Found. Eng., Mexico City. Vol. 1: 85-96. Eng. Prop. of Lateritic Soils, Seventh Int. Conf. on Soil
Little, A. L., 1967. The use of tropically-weathered soils in the Mech. and Found. Eng., Mexico City, Vol. 1: 65-74.
construction of earth dams. Proc. Third Asian Reg. Conf. Novais-Ferreira, H., and Meireles, J. M. F., 1968. Some
on Soil Mech. and Found. Eng., Haifa, Vol. 1: 35-41. methods of laterite study for application to roads and
Maignien, R., 1966. Review of research on laterites. Natural airfield, L.E.A. Luanda.
Resources Research IV, United Nations Ed. Sci. and Cult. Novais-Ferreira, H., and Correia, J. A., 1965. The hardness of
Org. (UNESCO), Paris: 148 p. laterite concretions and its influence in the performance of
Matyas, E. L., 1969. Some engineering properties of Sasumua soil mechanics tests. Sixth Int. Conf. Soil Mech. and Found.
clay. Proc. of Spec. Session on Eng. Prop. of Lateritic Soils, Eng., Montreal, Vol. 1: 82-86.
Seventh Int. Conf. on Soil Mech. and Found. Eng., Mexico Prakash, S., and Basavanna, B. M., 1969. In-situ properties of
City., Vol. 1: 143-151. laterite for hammer foundations. Proc. of Spec. Session on
Meireles, J. M. F., 1967. Trial pavements of laterite base Eng. Prop. of Lateritic Soils, Seventh Int. Conf. Soil Mech.
course. Proc. Fourth Reg. Conf. for Africa on Soil Mech. and Found. Eng., Mexico City, Vol. 1: 165-175.
and Found. Eng., Cape Town, Vol. 1. Prescott, J. A., and Pendelton, R. L., 1952. Laterites and
Mellier, G., and Phlipponnat, G., 1969. Utilisation routiere lateritic soils. Commonwealth Bur. Soil Sci. Tech. Com-
des laterites-Gabon. Proc. of the Spec. Session on Eng. mun. No. 47: 5 1 p.
Prop. of Lateritic Soils, Seventh Int. Conf. on Soil Mech. Quinones, P. J., 1963. Compaction characteristics of tropically
and Found. Eng., Mexico City, Vol. 1: 199-207. weathered soils. Unpublished Ph.D. thesis, Univ. of
Millard, R. S., 1962. Road building in the tropics. Jour. Appl. Illinois: 134 p.
Chem., Vol. 12: 342-357. Remillon, A., 1967. Les recherches routieres entreprise en
Moh, Z. C., and Mazhar, F. M., 1969. Effects of method of Afrique d'expression Francaise. An. de 1'Inst. Tech. du
preparation on index properties of lateritic soils. Proc. of Batiment et des Trav. Publics., No. 231-232: 366-388.
the Spec. Session on Eng. Prop. of Lateritic Soils, Seventh Remillon, A., 1955. Stabilization of laterite soils. High. Res.
Int. Conf. on Soil Mech. and Found. Eng., Mexico City, Board. Bull. No. 108: 96-101.
Vol. 1: 23-35. Road Research Lab., 1958. Notes on the cement treatment of
Moh, Z. C., and Sheikh, A. S., 1968. Effect of sulphate on Ghana soils. Dept. Sci. and Ind. Res. (U.K.), Road Res.
cement stabilized lateritic soils. Proc. of the Southeast Lab. Overseas Bull. No. 8: 24 p.
Asian Reg. Conf. on Soil Eng., Bangkok: 485495. Ruddock, E. C., 1967. Residual soils of the Kumasi District.
Moh, Z. C., Chin, Y. P., and Ng, S. C., 1967. Cement Geotechnique, vol. 17: 359-377.
stabilization of lateritic soils. Proc. Third Asian Reg. Conf. Santos, J. C., 1969. Utilization of lateritic soils for construc-
on Soil Mech. and Found. Eng., Haifa, Vol. 1 : 4 2 4 6 . tion purposes. Proc. of the Spec. Session on Eng. Prop. of
Morin, W. J., and Parry, W. T., 1969. Mineralogy and index Lateritic Soils, Seventh Int. Conf. on Soil Mech. and
 Found. Eng., Mexico City, Vol. 1: 189-197. Soil Mech. and Found. Eng., Mexico City, Vol. i : 75-84.
Schofield, A. N., 1957. Lime-stabilization of nodular clayey Vallerga, B. A., Shuster, J. A., Love, A. L., and Van Til, C. J.,
pea-laterite in Nyasaland. Dept. Sci. and Ind. Res., (U.K.), 1969. Engineering study of laterite and lateritic soils in
Road Res. Lab. Overseas Bull. No. 3: 15 p. connection with construction of roads, highways and
Sivarajasingham, S., Alexander, L. T., Cady, J. G., and Cline, airfields. Phase I-Southeast Asia (Thailand). Final Report
M. G., 1962. Laterite, In: Advances in Agronomy, Vol. 14, for U.S. Agency for International Development, Contract
Academic Press, London: 1-60. AID/csd-18 10. Soil and Pavement Consultants, Oakland,
Tateishi, H., 1967. Basic engineering characteristics of high- Calif.: 165 p. and appendices.
moisture tropical soils. Paper presented at the West. Assoc. Vold, R. C., 1967. The bilateral load test and its application to
State High. Off., Honolulu, July 12: 19 p. lateritic soils. Proc. Third Asian Reg. Conf. on Soil Mech.
Terza&, K., 1958. Design and Performance of the Sasumua' and Found. Eng., Haifa, Vol. 1 : 47-50.
dam (with Appendix: A mineralogical analysis of Sasumua Wiggington, W. B., 1967. Problems in the use of iron-rich
clay, by R.H.S. Robertson). Proc. Inst. Civil Eng., Voi. 9: laterites in the roadway structure-S.E. Thailand, Proc.
369-395. Third Asian Reg. Conf. on Soil Mech. and Found. Eng.,
Trow, W. A., and Morton, J. D., 1969. Laterite soils at Haifa, vol. 1 : 5 1-54.
Guardarraya, La Republica Dominicans, their development, Winterkorn, H. F., and Chandrasekharan, E. C., 195 1. Laterite
composition and engineering properties. Proc. of Spec. soils and their stabilization. High. Res. Board, Bull. No.
Session on Eng. Prop. of Lateritic Soils, Seventh Int. Conf. 44: 16-29.
 CHAPTER 5

ENGINEERING PROPER TIES OF TROPICALL Y WEATHERED AFRICAN SOILS

IN'TRODUCTION Investigational Program

The engineering properties of red, tropically weathered The pedological classification and distribution of these
African soils are presented in this chapter. Inasmuch as the groups have been used as a basis for sampling throughout
only criteria for distinguishing between lateritic materials and Africa. The extent of this investigation is shown in Figure 14.
those only partially laterized are chemical and mineralogical The use of the D'Hoore map made possible the assessment of
properties, there are no practical means from the engineering large regions and permitted the extrapolation of the detailed
point of view. Both types of materials are utilized in highways engineering soil study in Ghana to other locations in Africa
and other engineering applications, usually without dis- having similar soil and climatic conditions on a limited number
tinction. The properties of both are sufficiently similar that of samples outside of Ghana.
separating them has little engineering significance. The ex- The examination of physical and engineering properties of
ceptions are those soils which change properties upon drying. tropical red soils included samples obtained during field trips
Such soils are usually highly laterized, ferrallitic soils which in Ghana and in other African countries. The sampling outside
occur in areas with high rainfall. However, certain halloysitic of Ghana was supplemented by data obtained from the record
soils which are only partially laterized also display large files of the local Public Works Departments of the countries
changes in properties with drying. visited. Additional information was obtained through various
Pedological descriptions and maps of the soils were utilized agencies that sampled and tested soils upon request. The
during sampling and testing in the present study as these were Ghana data included results of the physical and engineering
the only means of determining sample locations and areal test conducted during the pavement condition survey.
distribution of the various soil types. It was found that the
pedological classifications of lateritic soils are reflected in their Generalized Test Results
engineering properties and, because of this, the pedological The test results have been analyzed and are presented
terms were adopted as being more descriptive than "lateritic." maintaining the D'Hoore classification as a basis of descrip-
In the remainder of this report, the terms ferrallitic and tion. The results of the laboratory test are given in the
ferruginous soils and ferrisols are used unless the soil type is appendix. Typical test results for the various soil groups are
unknown. shown in Tables 4, 5 and 6. As indicated they represent a
general distribution throughout Africa. The average values are
SOIL GROUPING given in Table 7. The grading envelope for the major soil
A generalized pedological soils map of Africa has been groups are shown in Figures 15, 16 and 17.
presented by D'Hoore (1964). The greater part of the
information was collected under the auspices of the Institut DISCUSSION
National pour 1'Etude Agronomique au Congo (INEAC), the The average values of the various properties, i.e., liquid
Junta de Investigacaes do Ultramar, the Office de la Re- limit, plastic index, gradation, compaction and CBR, indicate
cherche Scientifique et Technique Outre-Mer (ORSTOM), that the ferruginous soils, ferrallitic soils and ferrisols represent
and the local Agricultural Services. The laterites and lateritic distinct soil groups with a characteristic range of properties. A
soils discussed in the present study are included in the broad discriminant analysis revealed that the groups were distinguish-
groups designated as ferruginous tropical soils, ferralitic soils, able on the basis of physical and engineering properties. The
and ferrisols. An appreciable part of the effort of the present discriminant analysis and an abbreviated program are described
study has been to ascertain the extent to which the engi- in the Appendix. The computer program is also included. The
neering properties of soils in each of these groups differ ferruginous soils are lower in plasticity and the compacted
significantly from those of the others. densities at AASHO Modified compaction are higher, with
higher CBR values, than the other two groups. The ferrallitic
soils are similar, to some extent, to the ferrisols, exhibiting
Distribution of Groups higher plasticity and lower CBR values than ferruginous soils.
The results of the study are summarized in a modified The average index properties of the major subgroups are
D'Hoore map, Figure 13 (rear jacket). With respect to the shown in Table 8. There are major differences between soils of
continent as a whole, the ferralitic soils cover the largest area, the ferruginous group and the soils of the ferrallitic and
5,338,000 k m 2 . Next in areal extent are ferruginous tropical ferrisol groups. The data indicate that there is an apparent
soils (3,295,000 km2). Least extensive are the ferrisols similarity in the physical properties of the ferrallitic and
(91 7,000 km2). ferrisol soil groups.
 TABLE 4

Typical Soil Test Results for Ferruginous Soils

AASHO MOD SOAK

COUNTRY D'HOORE AASHO GI LL PL PI 1 314 112 318 4 10 40 200 OMC MDD CBR

Ghana BcJd A-2-6 0 26 15 11 45 200 12 130 150

- . --.

Senegal Ja A-2-7 0 39 20 19 95 91 80 68 46 33 27 200

- - .- -

Upper Volta Jc A-2-6 1 38 14 24 100 97 88 80 66 51 38 25 14 117 35

- - - -- - -

Niger Ja A-2-4 0 21 11 10 97 89 79 73 64 55 40 25 9 134 22

Tanzania Ja A-2-6 6 34 19 15 100 100 100 100 100 100 93 61

Kenya Jb A-2-7 0 45 31 14 100 100 97 94 88 52 40 28 19 52

P
Uganda Jd A-2-6 2 38 17 22 100 100 99 96 83 61 51 34 13 19
0 -- - - --
.-

Sudan Ja A-2-4 0 21 12 9 100 100 100 100 100 98 57 27

Gambia Jd A-2-6 0 36 16 20 98 77 60 53 42 34 28 22

TABLE 5

Typical Soil Test Results for Ferrallitic Soils

AASHO MOD SOAK

COUNTRY D'HOORE AASHO GI LL PL PI 1 314 112 318 4 10 40 200 OMC MDD CBR

Ghana Lx A-6 6 38 18 20 100 100 100 100 95 81 67 47 14 120 15

.--

Liberia Lx A-2-7 2 56 29 27 100 95 84 72 57 41 36 27

Gabon Li A-2-4 0 35 18 17

Sierra Leone Ab Lm A-2-6 1 55 31 24 100 98 95 90 68 37 29 27

 6 OZ I SI 8E EP IS 19 68 001 001 001 PZ IE SS E S-L-V 9x98 !1W
3 a a ~ 3 ~ 0 002 OP 01 P 8 / ~ Z/I P/E I 1d ?d ?? 13 OHSVV BXOOHJI AX1Nft03
nvos aow OHSVV
6E SS ZL S8 66 00l 001 001 PZ 12 SP P L-Z-V 17 Aawoqea
3 saw 3 ~ 0 002 OP 01 P 811 Z/I PIE I 1d ?d 77 13 OHSVV B X O O H ' ~ AX1Nfl03
nvos aow OHSVV
FIG 14 LOCATION O F SOIL S A M P L E S
LlEUX D'ORIGINE DES E C H A N T I L L O N S DE SOL
 TABLE 7

Average Soil Properties, Ferrallitic, Ferrisols and Ferruginous

ATTERBERG LIMITS
FERRALLITIC FERRISOLS FERRUGINOUS
Ghana Ghana Ghana

TT (mean) 42 24 19 46 23 23 31 18 14
a (standard 12 7 7 11 6 7 11 3 9
deviation)
N (number of (41 (41 (41 (46) (46) (46) (47) (47) (47)
samples)

Other Countries Other Countries Other Countries

LL PL PI LL PL PI LL PL PI

GRADATION
FERRALLITIC FERRISOLS FERRUGINOUS
Ghana Ghana Ghana
- - -
Sieve x a N Sieve x u N Sieve x u N

Other Countries Other Countries Other Countries

- -
Sieve X u Sieve x u Sieve
 TABLE 7 (Cont.)

Average Soil Properties, Ferrallitic, Ferrisols and Ferruginous

AASHO MOD1 Fl ED COMPACTION - CBR

FERRA LLITIC FERRISOLS FERRUGINOUS
Ghana Ghana Ghana
OMC MDD CBR OMC MDD CBR OMC MDD CBR

Other Countries Other Countries Other Countries

OMC MDD CBR OMC MDD CBR OMC MDD CBR

W
LL
z
z
0 loo
cn
Z
w$ 90

0e
I0
W
80

70
>
cn m
z 60
a
a W
'='L 50
LL
cn
W 4o
0 I-
Z
cn W 30
0
cr
0, 20

Z 10
W

W 0
0 100 50 20 10 5 2 I 0.5 0.2 0.1 0.05 0.02 0.01 5&
a
I-
mm
z
W
U I ASTM E - 11-61
a
3
!
3,- h-' ; 0.5at 4
I
10 25
!
40 60
I
100 200

0 I I I I t - - - r--7BRITISH 8s-410-62
a 2, 2, ;- 0!5" 5 7 22 36 60 100 200

I I I I I I FRENCH AFNORX-11-501
38 31 28 24 El 2 0

FIG 15 FERRUGINOUS SOIL GROUP- GRADATION LIMITS

GRANULOMETRIE DES SOLS FERRUGINEUX
6 mm
Z
W
U
u t
;-& I 1 I ASTM E-11-61
3 k 015u I4 10 25 40 60 100 200

-
I
2 5" F- 05"

FIG 16
f

38
FERRALLITIC
7-7
I
31

SOIL
d2 j6

28

GROUP
do
I
24
IdO

- GRADATION
I
2t)0

I
21 2 0
BRITISH

FRENCH
85-410-62

AFNORX-11-501

LIMITS
GRANULOMETRIE DES SOLS FERRALLlTlQUES

There are no major differences within the ferruginous soil interrelationships between the liquid limit, plastic limit and
subgroups. The differences that exist between the subgroups minus 200. There are considerable variations in the CBR values
indicate that although they have distinguishable properties, for each of the soil groups as given in Tables 4 through 6. The
there are degrees of association. CBR of a tiopical red soil is highly dependent upon the
There are no major differences within the ferrisol sub- percentage of gravel in the material; this will be discussed in
groups. The Ka and Kb subgroups are similar t o each other and the next section.
are also different from soils of other groups.
The ferrallitic soil group has one subgroup which shows
major differences from other soils within the group. This is CORRELATIOIU OF PHYSICAL AND
La - yellow ferrallitic soil on loose sandy sediments. This ENGINEERING PROPERTIES
subgroup is widespread in the Congo and Cubango Basins. Correlations of Atterberg limits, gradation and CBR values
With few exceptions these soils are restricted in the northern have been found t o exist among various temperate zone soils.
hemisphere t o the sandy deposits of the coastal fringe. The The physical and engineering properties of the tropical soils
distribution of this soil group is shown on the soil map. have been similarly analyzed in the present study by means of
Other than the subgroups discussed, the analyses suggest regression analyses t o establish the existence of useful relation-
that the soils within the major groups (subgroups) are ships t o assist the engineer in the selection of materials for
different, but exhibit an association within their major groups. highway construction on the basis of routine classification
The previous results have shown that the pedological defini- tests.
tions of ferruginous, ferrallitic or ferrisols represent distinct The first relationship t o be examined is that between liquid
soil groups with a characteristic range of properties and limit and plasticity index of the tropical red soils. The relation
W
0:
3
W
0:
W
LL
z I00
z
o 5 90
5w 2 80
r
- 3
n
70
g&
60
U)
z a
2w LL
50

40
:: I-

",: 30

w 20
g
z 10
W
w 0
W 100 50 20 10 5 2 I 0.5 0.02 0.01 5& 2"- I
3 rnrn
z
W
0
[L t , , 7 ASTM E - 11-61
3 3" 2,' ot.5'. I4 1
I0 2b 5 do 60 I 0I 0 200
0
-
a I , I I 1 I BRITISH BS-410-62
3,. ;a' i', d5,' 5 7 22 36 60 100 200
I I I I I FRENCH AFNORX-11-501
38 31 28 24 21 2 0

FIG 17 F E R R I S O L S O I L GROUP- GRADATION LIMITS

G R A N U L O M ~ T R I E DES F E R R I S O L S

is customarily shown on the Casagrande plasticity chart. The MDD = 160 - 2.78 OMC
results are shown in Figure 18. The general equation of the Standard deviation o = -+ 5.5 1bs
line of "best fit" through regression is as follows: Coefficient of correlation r = - 0.84 1
PI = 0.58 LL - 4.54 Number of samples N = 81
Standard Deviation
Relationships between CBR (at OMC-MDD, using AASHO
Coefficient of
modified compaction with a four day soaking period) and
correlation
gradation and plasticity index are shown in Figs. 20 and 21.
Number of samples
There is no well defined relationship between the CBR and the
The equations of the individual soil types are as follows: product (plasticity index times percent passing the No. 200 or
40 sieves). The results can be presented in a broad envelope
PI = 0.7 1 LL - 8.50 (ferruginous)
which indicates higher CBR's at lower values of the product.
PI = 0.57 LL - 3.62 (ferrallitic)
In both groups of data there is an apparent limit for
PI = 0.50 LL - 1.50 (ferrisols)
ferruginous soils and ferrisol-ferrallitic soils.
The positions of these soils are generally above the A-line; the A conservative estimation of CBR is possible, assuming
exceptions being samples obtained in Liberia, Sierra Leone, Modified AASHO compaction can be obtained in the field,
Cameroon and Ghana. The majority of these are ferrallitic or from the data shown in Fig. 22. The soaked CBR values have
ferrisols formed over basic rocks. The implications of this were been plotted against the AASHO classifications. The general
discussed in Chapter 3. equation given is:
The relationships between moisture, density, index proper-
ties and CBR have also been examined. The relationship CBR = 129 - 7.5 (AASHO)
between optimum moisture and maximum dry density at Standard deviation o=*32
Modified AASHO compaction is shown in Fig. 19. The general Coefficient of correlation r = - 0.607
equation is as follows: Number of Samples N = 63
 TABLE 8

Average Soil Characteristics

PERCENT PASSING
LIQUID LIMIT PLASTIC INDEX NO. 200 SIEVE

NUMBER
GROUP OF TESTS

FERRUGINOUS 34 10 16 7 30 17

Ja 33 12 16 7 29 13

Jb 43 3 18 5 18 14

Jc 39 10 18 8 36 19

Jd 33 7 17 6 20 14

30 8 13 7 27 13

FERRISOLS

Ka

Kb

Kc

FERRALLITIC

La

LC

LC

Lm

Ln

Lx

LxHb

AsHa

DATA FROM SAMPLES TESTED IN KUMASI AND EXTRACTED FROM ROADWAY SOIL SURVEY REPORTS FROM
VARIOUS COUNTRIES IN AFRICA.
 W
U)

s;
W X
00
3

32. $ 5
0
3
K K
;=
a a
K K
LOU)
bd
EL?
K K
$J%
J
EE
LLlL
KK
LLLL
kk
LLLL
 0 2 4 6 8 10 !2 14 16 18 20 22 24

OPTIMUM MOISTURE CONTENT (Ole)

TENEUR EN EAU OPTIMUM ('lo)

FIG 19 RELATIONSHIP BETWEEN MAXIMUM DRY DENSITY

AND OPTIMUM MOISTURE CONTENT.
DENSITE SECHE MAXIMUM EN FONCTION DES TENEUR
EN EAU OPTIMUM.

A FERRUGINOUS SOILS
- S O L S FERRUGINEUX
F E R R A L L I T I C SOILS
SOLS FERRALITIQUES
FERRISOLS
FERRISOLS

500 1000 1500 2000

PLASTICITY INDEX X PERCENT PASSING 200
INDICE DE PLASTlClTE X POURCENTAGE PASSANT LE TAMlS 2 0 0
FIG 2 0 CBR VS PLASTICITY INDEX x PERCENTAGE PASSING No. 200
C 0 R ET INDICE DE PLASTlClTE x POURCENTAGE PASSANT LE TAMlS 2 0 0
 W
A FERRUGINOUS SOILS
SOLS FERRUGINEUX
~ F E R R A L L I CSOILS
SOLS FERRALLlTlWES
~FERRISOLS
FERRlSOLS

V)

a
CO
U
0
0 500 1000 1500 2000
PLASTICITY INDEX X PERCENT PASSING NO. 4 0
INDICE DE PLASTICITE x POURCENTAGE PASSANT LE TAMIS 4 0
FIG.21 CBR VS PLASTICITY INDEX x PERCENTAGE PASSING No.40
CBR ET PRODUIT INDICE DE PLASTlClTE x POURCENTAGE PASSANT LE TAMIS No40

where:
(AASHO) = The numerical value of the AASHO
classification system, i.e.,
l=A-1-a; 2=A-1 -b
12=A-7-6
A conservative estimation of CBR is:
CBR =
- X
- o
where: X = average CBR value
o = standard deviation

There is only about a 15% chance of the CBR being lower than
estimated, since o represents approximately 65% of the data.
A relationship which could be used as a guide in the
selection of base course materials is shown in Fig. 23. It has
been generally accepted that base course materials should have
a CBR in excess of 75, under the moisture conditions which
will prevail during the expected life of the pavement. The data
shown in Fig. 23 have been used t o classify the soils into three
suitability groups: (1) soils which can be used as base course
materials under moisture conditions which exceed the
optimum moisture content ;(2) soils which can be used as base
A A SHO CLASSIFICATION course materials under moisture conditions which will not
CLASSIFICATION AASHO
F10.22 CBR VS AASHO CLASSlFlCAT ION
exceed the optimum moisture content; and (3) soils which do
C B R EN FONCTION DE LA CLASSIFICATION AASHO not have a suitable CBR value under either condition.
 A - A' LIMITS OF SOIL TESTED
B - B' LOWER LIMIT OF SUITABLE BASE
MATERIAL (SOAK GBR GREATER
THAN 7 5 )
G - G' LOWER LIMIT OF SUITABLE MATE-
RIAL-IF GOOD DRAINAGE ( A S
MOULDED GBR GREATER THAN 7 5 )
G - G' UPPER LIMIT OF NON-SUITABLE
MATERIAL (GBR LESS THAN 7 5 )

A - A' LlMlTES DES SOLS ETUDIES

B - B LIMITE INFERIEURE DU MATERIAU
PROPRE A LA CONSTRUCTION DE
LA GOUGHE DE BASE (GBR IMBIBE
SUPERIEUR A 7 5 )
G - G' LlMlTE INFERIEURE DU MATERIAU
SI L E DRAINAGE EST SOIGNE (GBR
SANS IMBIBITION SUPERIEUR A 7 5 )
C - C LlMlTE SUPERIEURE DU MATERIAU
INADEOUAT GBR INFERIEUR A 7 5 )

FINENESS INDEX
INDICE DES FINES

FIG 23 SUITABILITY OF M A T E R I A L AS BASE COURSE ON THE BASIS OF MAXIMUM DRY DENSITY, AT AASHO
MODIFIED DENSITY, AND FINENESS INDEX OF CBR SPECIMEN
CHOlX ENTRE MATERIAUX PROPRES E T IMPROPRES A UNE UTILISATION EN COUCHE DE BASE E N FONCTION
DE LA DENSITE SECHE MAXIMUM APRES COMPACTAGE AASHO MODlFlE E T DE L'INDICE DE FINESSE DE
SPECIMEN DU CALIFORNIA BORING RATIO

The relationship is shown as a function of maximum dry where: C1 = 72.5 Group 1

density at Modified AASHO density and the fineness index. = 88.0 Group 2
The fineness index in this case is the accumulated percentage
= 35.5 Group 3
passing the 1", 3/4", 1/2", 3/8", No. 4 , No. 10, No. 40 and No.
20C sieves of the laboratory CBR sample prior t o compaction. = 59.2 Group 4
The soils which conform to the first suitability group are ydmax = maximum dry density
enclosed by the points A, B, B', A. The second group is PI = plasticity index
enclosed by B, C, c', B', B, and the unsuitable soils by A', C,
and Cz = 7.5 Group 1
C' , A'.
The gravels in Ghana have been grouped on the basis of = 37.34 Group 2
particle size distribution. Four groups have been suggested, = 10.69 Group 3
Figs. 24, 25, 2 6 and 27. A discriminant analysis showed a = 19.5 Group 4
significant difference between the four groups in terms of
maximum dry density, optimum moisture content and CBR The relationships between CBR and maximum dry densi-
values. The different groups show a linear relationship between tylplastic index for the gravel groups are shown in Figs. 28,
the fineness index and the maxiumum dry density. The 2 9 , 3 0 and 31.
fineness index is the accumulated percentage passing the 3/4",
3/16", No. 8 , and 200 sieves divided by the number of sieves EFFECTS O F D R Y I N G O N PROPERTIES
in the nest (6). Two factors in some types of lateritic soils cause changes in
In relating these groups in CBR the following expression properties with drying. These are: (1) the tendency to form
was established: aggregations on drying, and (2) the loss of water in the
C B R = C 1 Log- (ydmax) - C2 hydrated minerals. Although these are related, the phenomena
PI
are different.
W
a
3
wa
W
LL
z
I00
z
0
p$ 90
w a
z W3 80

k'z 7 0
Y, 60
z
a
CK
w
gz gn
LL
fl')
W
40
n I-
Z
W 30
E
g 20

Z 10
W

8 O100 50 20 10 5 2 1 0.5 2"- I

2 mm
Z
W
0
iK I , I I I I 1 I ASTM E - 11-61
3 3" 2" 1" 05" 4 I0 25 40 60 160 200
0
a I , , , I I I 1 I 1 BRITISH BS-410-62
3 - 2 " 1" 05" 5 7 22 36 60 100 200
I I I I I 1 1 FRENCH AFNORX-11-50!
38 31 28 24 21 2 0

FIG.24 TYPICAL GRADING ENVELOPE OF LATERlTlC GRAVELS IN GROUP l

ENVELOPPE TYPIQUE DES COURBES GRANLILOMETRIQUES DE GRAVIERS
LATERITIQUES DU GROUPE 1

U
a I , . I ASTM E - 11-61
3 3" 2" ' 015" 4 10 25 40 60 100 200
8 I I BRITISH 0s -410 - 6 2
3, 2" 1" 0'5" 5 7 22 36 60 100 200
I I I FRENCH AFNORX-11-50!
38 31 28 24 21 20
FIG.25 TYPICAL GRADING ENVELOPE OF LATERlTlC GRAVELS IN GROUP 2
ENVELOPPE TYPIQLIE DES COURBES GRANULOMETRIQUES DE GRAVERS
LATERITIQUES DU GROUPE 2
W

w5IX I
STONE
GRAVEL
GRAVEL
SAND I s1LT ]CLAY
W SAND SILT CLAY
LL
CAILLOUX GRAVIER SABLE
z I00
LlMON ARGILE

z
0% 90
V)
z2
gr a0

70
w >
0 m
60.
V)
z
2~ 59
(3

Y;
LL
40
i
w I-
0 2
w 30
* I

g
V)

- -
20
-- - - -- - -- . --
Z lo
W
0
100 50 2 0 10 5 2 I 0.5 0 2 01 005 0.02 0.01 5" 2fi I
3 mm
z
W
U
lr 3 , . , ! ASTM E - 11-61
3 3" 2" 1" 05" 4 I0 25 40 60 100 200
0
n I BRITISH 85-410-62
3" 2" 1" 05" 5 7 22 36 60 100 200
I I 7 FRENCH AFNORX-11-501
38 31 28 24 21 2 0

FIG.26 TYPICAL GRADING ENVELOPE OF LATERlTlC GRAVELS IN GROUP 3

ENVELOPPE TYPlQUE DES COURBES GRANULOMETRlQiJES DE GRAVIERS
LATERITIQUES DU GROUPE 3

W
PT
3
W
P:
W
LL
z 100
z
0 I- 90
z
z
I
(3
z
Z?
n
O0

w > 70
O m
m 60
Z a
2 %5 0
a LL

V) 40
w I-
nZ
W 30
m 0
n lr
5w 20
n a
Z 10
W

W 0
(3 100 50, 2 0 10 5 2 1 0.5 02 01 0.05 002 001 5" p
mrn
Z
W
0
cr 4 I c ASTM E - 11-61
3 3" 2" I 05" 4 10 ;5 40 $0 100 200
0
n I , I

3" 2" 1" 0.5" 5 7 22 36' 60 100 200 BRITISH 8 s - 410 - 6 2

t-- I 7 FRENCH AFNORX -11-501

38 31 28 24 21 20

FIG.27 TYPICAL GRADING ENVELOPE OF LATERI'TIC GRAVELS IN GROUP 4

ENVELOPPE TYPIQUE DES COURBES GRANULOMETRIQUES DE GRAVIERS
LATERITIQUES DU GROUPE 4
 ad M A X 1
(- PI

FIG 28 RELATION BETWEEN MAXIMUM DRY DENSITY/

PLASTICITY INDEX AND C B R VALUES, GROUP I 6 MAX
(PI d
RELATION ENTRE DENSITE SECHE MAXIMUM-
INDICE DE PLASTICITE E T LES VALEURS FIG 29 RELATION BETWEEN MAXIMUM DRY DENSITY/
DU CBR, GROUPE I PLASTICITY AND C B R VALUES, GROUP 2
RELATION ENTRE DENSITE SECHE MAXIMUM-
INDICE DE PLASTlClTE ET LES VALEURS
DU C 0 R, GROUPE 2

(w%,
PI

(6, MAX FIG 31 RELATION BETWEEN MAXIMUM DRY DENSITY/PLASTI-

CITY INDEX AND CBR VALUES, GROUP 4
PI
RELATION ENTRE DENSITE SECHE MAXIMUM-INDICE
FIG 3 0 RELATION BETWEEN MAXIMUM DRY DENSITY/ DE PLASTICITE ET VALEURS W CBR, GROUPE 4
PLASTICITY AND C B R VALUES, GROUP 3
RELATION ENTRE DENSITE SECHE MAXIMUM
INDICE DE PLASTlClTE ET LES VALEURS
-
DU C B R, GROUPE 3
 The tendency to form aggregations was perhaps first Halloysite has been identified in African soils and in these
described by Hirashima (1948) and later, in more detail by it causes property changes with drying. The Kenyan Sasumua
Terzaghl (1 957), for the red clays at Sasumua Dam in Kenya. clay first described by Terzaghl (1957), is the best docu-
The dam is located in an area of volcanic rocks whose soils mented. Newill (1961), in his detailed laboratory investigation
correlate well with the ferrallitic group. The high angle of of two red clays from Kenya, determined that air drying
internal friction (34"), and low compressibility were consid- increased the maximum density and lowered the optimum
ered abnormal because of high clay content (70%-100%) and moisture content of a Sasumua sample with hydrated halloys-
high liquid limit. Terzaghi explained this anomalous behavior ite, but did not significantly affect a sample from Kabete with
by assuming that the clay existed in clusters of interconnected metahalloysite. The change in density and optimum moisture
clay particles. The clay was considered strongly aggregated, a content was attributed to the interlayer water in the hydrated
state unlikely to change during construction. halloysite. Lamb and Martin (1955) proposed a correction for
The aggregating effect of the iron oxide in the Sasumua the interlayer water in halloysite, but the percentage of
clay was later demonstrated by Newill (1961). He removed the halloysite must be known.
iron chemically and found the clay aggregations had been Halloysite exists in two forms, hydrated halloysite,
dispersed. These did not reform as had been the case after A12 O 3 2 S i 0 , - 4 H 2 0 a n d m e t a h a l l o y s i t e ,
mechanical manipulation. The Atterberg limits were deter- A12O3 2Si02*2H20. The stability curve for halloysite, shown
mined before and after chemically removing the free iron in Fig. 32, indicates that this mineral forms only under water
oxide. The liquid limit and plasticity index increased consider- or in high humidity. With a loss of water in a drier
ably. environment, it readily alters to metahalloysite which can be
The dehydration of other clay materials also causes changes difficult to distinguish from kaolinite. It is likely .that much of
in properties. Tateishi (1 958) described variations in the values the kaolinite reported in connection with the weathering of
of maximum density tests depending on whether the tests
were performed after air drying or without air drying, or by
determining the points backward along the moisture-density
curve as the soil dries. This was confirmed by Quinones (1963)
on a variety of lateritic soils from Hawaii, Puerto Rico and
Liberia.
The clay materials most susceptible to changes of properties
with drying are considered by Frost (1967) to be allophane,
halloysite and gibbsite. These will be discussed individually.
Allophane is a very unstable mineral which contains
variable silica and water. The formula is A1203*Si02- n H 2 0 . It
is usually reported over recent volcanic extrusives and may be
a precursor of halloysite (Aomine and Wada, 1962). There
appears to be no doubt that allophane can lose much of its
attached water readily and would change properties in doing
so. Gradwell and Birrell (1954) state that allophane apparently
affects the physical properties of New Zealand clays formed
over volcanic ash because of its extremely fine particle size and
high water content. They cite a soil with allophane as the
predominant clay mineral which displays a liquid limit of 225
when undried and 100 when air dried prior to testing.

Allophane was not reported in any of the samples collected TEMPERATURE 'C
TEMPERATURE 'C
for this study, including a few samples of volcanic soils from
Ethiopia, Kenya, and Cameroon. It is possible that the
pretreatment prior to X-ray analysis removed the allophane or FIG 3 2 PHASE RELATIONSHIPS OF HALLOYSITE; UPPER
that the crystallinity was too poorly developed to defract CURVE IS VAPORIZATION CURVE OF WATER; LOWER
x-rays. However, none of the volcanic soils displayed the CURVE IS BOUNDARY BETWEEN HALLOYSITE (4H20)
drastic changes of properties with drying. If allophane exists in AND HALLOYSITE (2H20) (after Bates, 1952)
the African soils it must be restricted to those over volcanics in RAPPORTS DES PHASES DE L'HALLOYSITE; LA
the high rainfall areas, such as the high volcanic mountains of COURBE SUPERIEURE REPRESENT LA COURBE DE
East Africa or in southwestern Cameroon. Because of the VAPORISATION DE CEAU; LA COURBE INFERIEURE
restricted size of these areas, it is not considered a general E S T L A LlMlTE ENTRE LHALLOYSITE (4H20) E T
problem. CHALLOYSITE (2H20) (d'apras Bates, 1952)
 feldspars is actually metahalloysite which has formed by the changes can all be attributed to aggregation caused by
desiccation of hydrated halloysite (Loughnan, 1969). The dehydration of the iron minerals. Near-surface soils have been
hydrated form dehydrates when the relative humidity drops aggregated naturally during pronounced dry seasons. The
below 40% or the moisture content drops below 10% (Lamb depth of seasonal desiccation of course varies, but in Ghana
and Martin, 1955). Therefore, hydrated halloysite can exist this is apparently 5 feet or more.
only in a permanently moist environment. Changes in ferrisols are usually minor also, except over
Halloysite in a red clay has been reported on the volcanic volcanic soils which contain hydrated halloysite. In the
extrusive rocks on the island of Fernando Po (Salas, 1963), majority of ferrisols, changes can be attributed to aggregation
where changes of property with drying were also noted. The of the iron minerals.
form is undoubtedly hydrated because of the changes in Changes in properties with air drying can be significant in
properties upon drying and because of the high natural construction. Laboratory testing, if performed on air-dried
moisture content, over 40%. samples only, m y produce test results on altered soils. If the
Metahalloysite was identified in the following African soils material will not be similarly air-dried during construction,
during the course of this study. These were soils over basic erroneous values for the Atterberg limits, gradation, and
volcanic rocks from West Cameroon, Ethiopia and Kenya. A density may be applied. However, it is believed on the basis of
soil over volcanic ash in southern Tanzania was not analyzed t h s study that such materials are limited to: (1) ferrallitic
but is believed t o contain halloysite. It was reported that soils in areas of over 6 0 inches of rain per year and where
laboratory tests did not reveal the properties of the soil which air-drying during the construction season may not be feasible;
were evident during construct ion. or (2) t o areas with volcanic soils such as Ethiopia, East Africa,
Halloysite can be considered a problem only when it exists Cameroon, Fernando Po and over the basic rocks along the
in its hydrated form w h c h is the only form associated with West African coast. In these areas, it is strongly recommended
changes of properties. The mineral can occur only over basic that the susceptibility t o change with air-drying be investig-
rocks, the intrusives of West Africa and extrusives of Camer- ated. The appropriate laboratory procedures are described in
oon and eastern Africa. In these areas the susceptibility of the the following chapter. If changes with drying occur that are
soils to moisture changes should be ascertained. significa~tand air-drying cannot be assured during construc-
Gibbsite is a common clay material in lateritic soils, and tion, then test results on as-received samples should be applied.
Frost (1967) attributes changes in properties to this mineral in
soils from New Guinea and New Zealand. Gibbsite has the SUMMARY OF FlRlDlNGS REGARDING
chemical formula AI(OH)3. Possibly loss of hydroxyl radicals MAJOR GROUPS
occurs in moist soils with air-drying or perhaps gibbsite has
attached, but not crystalline, water which it may lose on air Ferruginous soils occur in the more arid extremes for
drying. However, it is doubtful if this would be of sufficiently lateritic soils, in areas with pronounced dry seasons. They have
large quantity to cause large changes in properties. It is formed over all rock types: igneous, metamorphic and sedi-
believed that changes reported by Frost are not due only to mentary. Dehydration of the iron minerals and less abundant
gibbsit e, but to associated hydrated halloysite and allophane. aluminum minerals has occurred in all except the soils which
Gibbsite was identified in only a few African soils. These are sufficiently moist that seasonal dehydration does not
are all ferrallitic soils. This agrees with D'Hoore (1964) in h s occur. Ferruginous crusts are common which are hard and
definition of ferrallitic soils, and also with the findings of durable. Gibbsite does not occur or is uncommon. Similarly
w

Delvigne ( 1 965). halloysite or allophane do not occur, even over volcanic rocks.
Considerable testing was performed on ferrallitic soils to The absence of these minerals and the state of oxidation
determine if changes in properties occur. Admittedly, some of (dehydration) precludes significant changes of properties with
these samples may not have been of sufficient depth to include air-drying. The fact that these soils display lower Atterberg
only material that has not been subjected t o seasonal air limits, higher densities and CBR values could all be explained
drying. However, many were recovered from auger borings to by dehydration and aggregation. Hydrated minerals would be
depths of 9 feet or more. Results of these tests are shown in expected to display greater plasticity and liquid limits.
Table 9. Most of the changes reported are not particularly Ferrallitic soils occur in the more humid extremes for
significant, and are mostly limited to areas of over 6 0 inches lateritic soils and in areas with dense vegetation. These soils
annual rainfall. Gibbsite, or partially hydrated gibbsite may be have also formed over all rock types. Gibbsite is a common
the cause of these changes or aggregation with dehydration of clay material and other hydrated forms of alumina occur as
both iron and alumina minerals may be the cause. well as hydrated iron minerals. Halloysite is fairly common
over volcanic rocks and to a lesser extent over the relatively
Changes in properties in ferruginous soils were only small bodies of basic intrusives found in West Africa. Changes
detected in a few soils and these changes were minor. All of in properties with drying occur due to hydrated halloysite
these were from Ghana, below depths of 5 feet. Gibbsite is not (generally restricted to continually humid areas) and, to a
a normal weathering product in ferruginous soils, so these lesser extent due to gibbsite. Ferruginous crusts or hardpans
are also relatively common. These have not been subjected to oedometer at its natural moisture content and loaded u p t o 2
complete dehydration as have the crusts in ferruginous regions tsf before introducing water. The sudden drop in void ratio is
and consequently have not attained the hardness or durability. indicative of the potential collapse under that particular load
However, nodular concretions in relatively shallow horizons intensity.
can be hard and as durable as those in ferruginous soils. Highway construction may not impose loads of sufficient
Ferrallitic soils can also be expected t o harden o n exposure. intensity t o cause collapse. However, Knight and Dehlen
Dehydration is normally prohibited by the dense vegetative (1963) describe one such road failure resulting from soil
cover, but with exposure t h s process is no longer prevented. collapse. The only means of preventing collapse in such cases is
Ferrisols have formed over all types of rocks in inter- t o densify the soil by vibratory means. Jenriings (personal
mediate t o high rainfall areas where erosion has kept pace with communication) believes a depth of 6 feet should be explored
profile development. Kaolinite predominates, but goet hite also in soils suspected of being collapsible.
occurs. The degree of hydration of the clay materials is high, Occurrences and experiences have been reported in south-
causing the properties t o be similar t o ferrallitic soils. Changes ern Africa by Brink and Kantey (1961); Jennings and Knight
in properties with dehydration are considered minor when due (1957); Knight (1963); and Novais-Ferreria and Meireles
to aggregation of the iron minerals, but can be significant in (1967). Collapsing soils are known t o exist in northwestern
volcanic rocks when hydrated halloysite is a constituent. Nigeria. Other soils that are believed t o be susceptible include
the eolian sands of the Congo Basin and the weathered granite
COLLAPSING SOILS o f Madagascar. The possibility of collapse in these areas, as
Collapsing soils, or soils that decrease in volume with the well as others with eolian deposits or residually weathered
addition of water, are more of a problem in Africa than is granite, should be investigated.
generally realized. Little work has been reported on such soils
in Africa outside of Nigeria and South Africa. This section is
CONCLUSIONS
only a brief review of this subject; most of the information The pedological definitions as presented by D'Hoore (1964)
was obtained from Dudley (1 970), Jennings (personal com- have been reviewed and proposed as a means of identification
municat ion), and Brink (personal communication). of the red tropical (lateritic) soils.
A wide variety of deposits have reportedly produced The ferruginous, ferrisols and ferrallitic soils have been
collapsing soils, including eolian, colluvial, alluvial and re- sampled throughout Africa. Ranges of index and engineering
sidual. The eolian and residual deposits are believed t o be the properties have been given for each of the soil groups.
most common in Africa. A discriminant analysis revealed that the ferruginous soils
The amount of collapse varies considerably depending on are distinctly different from ferrisols and ferrallitic soils.
the initial density, type of material, moisture content, load, Ferrallitic soils and ferrisols display a small but distinguishable
etc. Susceptible soils are those that are loose (high void ratios) difference in properties.
with relatively low moisture contents. The soils in southern The significance of the sub-headings which D'Hoore has
Africa that are subject t o collapse are eolian sands, found in included have been examined. The analysis suggests that the
the Transvaal, Botswana and South West Africa, and residual soils within the major groups (sub-groups) are different, but
soils formed over granite. In the eolian deposits, clay, exhibit an association within their major groups.
principally kaolinite, was formed as the sand weathered and Correlations of the physical and engineering properties have
much of the clay remained in place. As the weathered sand been presented. A number of the relationships can be utilized
dried or partially desiccated, the clay became concentrated in assisting the engineer in selecting materials for hghway
between the sand grains and formed randomly oriented, construction on the basis of Atterberg limits, compaction and
flocculated clay "bridges." This structure is capable of gradation.
supporting the overburden load at various moisture contents, Dehydration causes permanent changes in soil properties in
but collapses with added load. Residual deposits over granite some lateritic soils, either through the loss o f water of
have formed in well drained areas with high rainfall. In South crystallization or by the aggregation of soil grains as the
Africa these are ferrallitic soils formed along the eastern part hydrated iron and aluminum minerals dry or oxidize. Fer-
of the country. The high rainfall causes weathering and rallitic soils or soils over basic rocks are the most susceptible.
leaching of the soluble material and, because drainage is
unimpeded, the dissolved material is carried away. Details on REFERENCES
the causes of the phenomenon are provided b y Dudley (1970). Aomine, S., and Wada, K., 1962. Differential weathering of
Collapsing soils can be identified by means of standard volcanic ash and pumice, resulting in formation of hydrated
laboratory tests. One method is t o determine if the natural halloysite. Amer. Mineralogist, vol. 47: 1024-1048.
void ratio is higher than the void ratio at the liquid limit. This Brink, A. B. A., and Kantey, B. A., 1961. Collapsible grain
indicates a soil that is susceptible t o changes in structure with structure in residual granite soils in South Africa. Proc.
saturation. Another method, used by Jennings (1965), makes Fifth Int. Conf. on Soil Mech. and Found. Eng., Paris, vol.
use of the consolidation test. A sample is placed in an 1: 661 - 9 1 4 .
 TABLE 9

Changes I n Engineering Properties

AIR DRIED AS RECEIVED AIR DRIED AS RECEIVED AIR DRIED AS RECEIVED AIR D. AS R.
SAMPLE
- -
NO. SOIL DEPTH LL PL P1 LL PL PI A1 10 40 200 10 40 200 MDD OMC CBR MDD OMC CBR AASHO AASHO

LIB-3 Lx 3' NO CHANGE

LIB-4 Lx 3' 59 25 34
LIB-5 LC 3' 76 39 37
LIB-6 LC 12' NO CHANGE
N-7-6 BcJd 10' 45 20 25
GS-1 Kc 3' 44 19 26 1.4 64 60 47 88 84 64 111 17 22 105 21 10 A-7-6 A-7-6
1-5 Jc 3' 28 10 5 32 25 22
1-2 Jc 3' 42 20 5 43 30 15
1-14 Jc 3' 34 13 5 40 19 11
G2-1 Kc 3' NO CHANGE 24 12 3 47 44 40
G3-1 Kc 3' NO CHANGE 34 29 23 45 40 33
G3-3 Kc 3' NO CHANGE 8 6 6 42 22 15
G5-1 BcJd 3' 43 14 6 50 33 25
G6-3 Kc 5' 46 20 26 20 22
GH9-5 Jc 5' 47 21 26 1.1 68 46 27 46 37 30
(310-1 Kb 3' 49 29 20 1.3 100 98 76 100 97 79
GI 0-2 Kc 3' 46 20 26 0.895 21 15 12 38 26 16
GI 0-3 KC 9-1 0' 47 33 14 2.4 100 67 50
uv-1 JC 0-4' 41 18 23 1.4 82 71 58 80 68 58 112 19 20 103 22
UV-1 A JC 4-8' 45 23 22 88 78 64 89 83 70
DAH-3 Li 4' 51 30 21 1.1 72 55 39
ul SL-1 LcNa 3' 70 32 38 1.2 69 67 60 73 72 68
O3 SL-2 Ab Lm 2' 60 31 29 37 29 27 32 22 20
CB-2 Jd 3' 36 20 16 1.4 34 28 22 30 27 24
SEN-I Ja 2' 41 21 21 1.0 33 27 22 33 24 21
SEN-3 JaJd 2' 25 14 11 34 29 18 53 39 25
SEN-6 Ja 2' 40 21 10 44 47 27 46 45 42
GI 1-1A Lx 4' 46 18 28 1.6 64 56 41 64 56 37
Cll-2A Lx 2' 31 17 14 35 30 24 46 40 30 130 10
Gll-3A Lx 3' 40 28 13 39 34 28 45 40 30
Gll-3B Lx 5' 54 31 23 69 61 54 91 83 67
Gl l-3D Lx 8' 52 38 14 65 58 47 77 71 52
G11-4A Lx 4' 40 17 23 81 67 47 95 85 53
Cll-4B Lx 8' 40 15 25 1.0 45 36 25 61 55 29
Gll-4C Lx 20' 35 26 9 98 72 37 96 75 36
Gll-5A LC 1.5' 65 32 33 1.8 46 45 40 61 60 58
Gll-5B LC 74 33 41 86 80 75
Gll-5C LC 69 38 32 68 59 50 66 65 62
C11-8A Kh 3' 50 19 31 1.0 27 20 14
Gll-6B J 4' NO CHANGE 67 63 45 97 92 82
Gll-6C J 6' NO CHANGE 74 67 40 99 92 55
Gll-8B Kh 5' 42 44 18 81 78 64 NOCHANGE
Gll-3C Lx NO CHANGE 68 55 50 72 66 59
Gll-2B LX 2-6' NO CHANGE 68 66 57 89 86 74
G11-1B Lx 6' NO CHANGE .96 51 45 32 54 47 31
BURUNDI 2 BcLsLt 1' NO CHANGE
BURUNDI 1 BcLsLt 10' NO CHANGE
GB-1 Li 29 13 16
(39-2 Jb 3' 49 26 23
GH9-1 Jc 6' NO CHANGE
G8-2 LC NO CHANGE
Gl l-6A J 2' 36 20 16
NIA-1 Ja 1' 23 15 8
 TABLE

Changes I n Engineering Properties

- - -

AIR DRIED AS RECEIVED AIR DRIED AS RECEIVED AIR DRIED AS RECEIVED AIR D. AS R.
SAMPLE NO. SOIL DEPTH %M LL PL PI LL PL PI A1 10 40 200 10 40 200 MDD OMC CBR MDD OMC CBR AASHO AASHO

KABALA S.L.
BATKANU S.L.
TORMABUM S.L.
FREETOWN S.L.
GSS-1
Gss-2

Jc
BcJd

Lrn
Lx
Lrn
Lx
Lx
Lx
Lx
Jc

BcJd

GSS-32 BcJd
Delvigne, J., 1965. Pedogenese en zone tropicale. Office Knight, K., and Dehlen, G., 1963. The failure of a road
R e c h . Sci. Tech. Outre-Mer (ORSTOM), Dunod, constructed on collapsing soil. Proc. Third Reg. Conf. for
Paris: 177 p. Africa on Soil Mech. and Found. Eng., vol. 1: 3 1-34.
D'Hoore, J . L., 1964. Soils map of Africa, scale 1 to Lamb, T. W., and Martin, R. T., 1955. Composition and
5,000,000, explanatory monograph. Comm. for Tech. engineering properties of soil (111). Proc. High. Res. Board,
Co-op. in Africa, Publ. 93: 205 p. vo1. 34: 566-582.
Dudley, J. H., 1970. Review of collapsing soils. Amer. Soc. of Loughnan, F. C., 1969. Chemical weathering of the silicate
Civil Eng., Jour. of Soil Mech. and Found. Div., vol. minerals. American Elsevier Publ. Co., Inc., New York:
96: 925 - 947. 154 p.
Frost, R. J., 1967. Importance of correct pretesting prep- Newill, D., 1961. A laboratory investigation of two red clays
aration of some tropical soils. Proc. First Southeast Asian from Kenya. Geotechnique, vol. 12: 302-318.
Conf. on Soil Eng., Bangkok: 43-53. Novais-Ferreira, H., and Meireles, J . M. F., 1967. On the
Gradwell, M. and Birrell, K. S., 1954. Physical properties of drainage of muceque-a collapsing soil. Proc. Fourth Reg.
certain volcanic clays. N.Z. Jour. Sci. and Tech., vol. Conf. for Africa on Soil Mech. and Found. Eng., vol.
36: 108-122. 1: 151-155.
Hirashima, K. B ., 1948. Highway experience with thixo t ropic Quinones, P. J., 1963. Compaction characteristics of tropically
volcanic clay. Proc. High. Res. Board, vol. 28: 481-494. weathered soils. Unpublished Ph.D. Thesis, Univ. of Illi-
Jennings, J. E., 1965. The theory and practice of construction nois: 134 p.
on partially saturated soils as applied t o South African Salas, J. A. J., 1963. Note on a halloysite red clay from
conditions. In: Engineering effects of moisture changes in Fernando Po Island. Proc. Third Reg. Conf. for Africa on
soils. Texas A&M Press: 345 - 363. Soil Mech. and Found. Eng., Salisbury, vol. 1: 85-88.
Jennings, J . E., and Knight, K., 1957. The additional set- Tateishi, H., 1967. Basic engineering characteristics of high-
tlement of foundations due to a collapse of structures of moisture tropical soils. Paper presented at the West. Assoc.
sandy soils on wetting. Proc. Fourth Int. Conf. on Soil. State High. Officials, Honolulu, July 12: 19 p.
Mech. and Found. Eng., London,vol. 3: 316 - 319. Terzaghi, K., 1958. Design and performance of the Sasumua
Knight, K., 1963. The origin and occurrence of collapsing dam (with appendix: a mineralogical analysis of Sasumua
soils. Proc. T h r d Reg. Conf. for Africa on Soil Mech. and clay, by R. H. S. Robertson). Proc. Inst. Civil Eng., vol.
Found. Eng., vol. 1: 127-130. 9: 365-395.
 CHAPTER 6

TEST PR OCEDURES AND PR ETR EA TMENT

INTRODUCTION to testing. An aggregation index value greater than 2 indicates

It is known that some lateritic soils show changes in the susceptibility of the soil t o changes in physical properties
physical properties when tested under different conditions. In when dried.
Africa experience indicates that changes are limited to Samples tested for the four pretreatment methods were
ferrallitic soils in high rainfall areas and t o areas with volcanic prepared in the following manner:
soils. T h s has been discussed in the preceeding chapter. One of As Received (AR) - at natural moisture content
the goals of this project was t o evaluate the effects of various Soaked (S) - immersed in water for 24 hours
pretreatments, and t o recommend test procedures more Air Dried (AD) - dried to constant weight under
applicable to lateritic soils. normal temperatures
Four different pretreatment methods were applied t o a (23" t o 3 0 " ~ )
number of the samples tested in the laboratory. These Oven Dried (OD) - dried in an oven for 24 hours
methods were air drying (AD), oven drying (OD), soaking (S), at 110" k5"
and as received (AR) or at the natural moisture content. The The OD and AD samples were prepared in accordance with
effect on the Atterberg limits of drying at various temper- ASTM D421-58, "Dry Preparation of Soil Samples for
atures and for various periods was also studied. The change in Grain-Size Analysis and Determination of Soil Constants." For
compaction characteristics of lateritic soils when subjected to OD samples, section 3 was modified from air drying to oven
drying and recompaction was also investigated. drying. The S and AR samples were prepared in accordance
A number of test procedures were tried t o determine their with ASTM D2217-66, "Wet Preparation of Soil Samples for
applicability t o lateritic soils. The Potential Volume Change Grain Size Analysis and Determination of Soil Constants,"
(PVC) and Portuguese swell tests were run t o evaluate the Procedure B. T h s procedure provides that the sample be kept
swelling potential of lateritic soils. The Harvard miniature at a moisture content equal to or greater than the natural
compaction procedure was compared with the modified moisture content. Heating in an oven is one method suggested
ASSHO compaction procedure. The aggregation index was t o evaporate excess water. This should not be attempted on
used t o determine the susceptibility t o change of physical lateritic soils because the samples may harden or aggregate.
properties. A study of the one-point LL of lateritic soils was The State of Hawaii has adopted a test method for the
also performed. The durability and aggregate strength of preparation of sensitive soils which irreversibly change prop-
gravels were studied to determine the suitability of these erties when dried before testing. This method is presented in
aggregates as potential construction materials. A summary of the Appendix. Hawaii recommends siphoning or decanting t o
testing procedures and recommended modifications is remove excess water. In some cases where the excess water will
presented. not become clear, as with soils of high clay content, or where
fines can be lost during decanting, it is suggested that the
PROCEDURES excess water be filtered off. Buchner funnels were successfully
To determine the physical characteristics of the African used to filter free water from the S and AR samples during this
lateritic and black cotton soils, 200 different samples were study .
subject to a total of about 2000 tests. The stabilization and All tests were run in accordance with ASTM standards
road condition tests are not included in these totals. To listed in Table 10. The applicable ASSHO and British Standard
facilitate the selection of samples for the various pretreatments designations are also listed. Recommendations are given in
and subsequent tests, a special system of sample handling was Table 11 for modifications of standard procedures when
adopted. This is shown as a flow chart on Fig. 33. testing sensitive soils.

Sample Preparation Atterberg Limits

The susceptibility t o change of physical properties due to The change in Atterberg limits of lateritic soils has been
dehydration was investigated by the aggregation index as discussed in Chapter 4. Submitting a number of samples to the
proposed by Tateishi (1967). The aggregation index is the four pretreatment methods showed that significant changes in
ratio of the oven dried sand equivalent value to the sand properties occur in only certain lateritic soils: ferrallitic soils in
equivalent value at the field moisture content. The sand high rainfall areas and over volcanic rocks. The normal
equivalent test was run in accordance with ASTM D24 19-69 Atterberg limit procedures were followed in performing the
except for the field moisture sample which was not dried prior tests, except for the met hod of preparation. Excessive manipu-
 LEGEND
SAMPLE PREPARATION
P R E P A R A T I O N DE L ' E C H A N T I L L O N
A CALCULATION
CALCUL

0
A. L .
TESTS
ESSAl
ATTERBERG LIMIT T E S T S
GRAD E S S A l DES L l M l T E S D'ATTERBERG
SOAK -SWl EEVTE - 1 0 I S.A. SlEVE ANALYSIS
A N A L Y S E AU TAMIS
HYD HYDROMETER
HYDROMETRE
'GRAD GRADATION
GRANULOMETRIE
SlEVE SAND EQUIVALENT
EQUIVALENT DE SABLE
AGGREGATION INDEX
I N D I C E D'AGREGATION
HARVARD MINIATURE
HARVARD MINIATURE
POTENTIAL VOLUME CHANGE
CHANGEMENT P O T E N T I E L DE V O L U M E
C.B.R 0 CALIFORNIA BEARING RATIO
CALIFORNIA BEARING R A T I O

[---l-l /["h3
M.D. MOISTURE DENSITY
I AS RECEIVED RELATION D E N S I T E - TENEUR E N E A U
U .C. UNCONFINED COMPRESSION
COMPRESSION S I M P L E
SPECIFIC GRAVITY
SlEVE DENSITE
CHEM CHEMICAL T E S T S
E S S A I S CHlMlQUES
MINERALOGICAL T E S T S
E S S A I S MINERALOGIQUES
U S. SlEVE SIZE
TAMIS NORME U . S

SlEVE
A. L .

F I G . 33 F L O W DIAGRAM FOR L A B O R A T O R Y T E S T I N G
SUITE DES E S S A I S D E L A B O R A T O I R E

lation during testing leads to crumbling of the soil structure tests during the course of t h s study. The common complaint
and disaggregation; both consequences produce fines which was that the tests are too prone to variations in results with
result in higher values. The mixing time was kept to a manipulation and lateritic soils are particularly susceptible to
minimum, generally about 5 minutes for each point. This changes.
mixing time is about normal for most soils. Various tests as possible substitutes were investigated,
A number of engineers have criticized the Atterberg limit either new or modified from more or less standard tests.
 TABLE 10

Test Standards

Applicable Applicable
ASTM AASHO British
Designation Test Standard Standard

Dry Preparation of 1377

soil samples Part 2 Sec. 4

Par ticle-size Analysis 1377

Test 6

Liquid Limit 1377

Test 2A

Plastic Limit and 1377

Plasticity Index Test 3

Shrinkage Factors

Moisture Density
Relations-Standard

Specific Gravity TlOO 1377

Test 5

Moisture Content 1377

Test 1A

Wet Preparation of Soil

Samples

Bearing Ratio

Materials for Soil

Aggregate Subbase-Base
and Surface courses

Density of Soil in Place by 1377

the Sand-Cone Met hod Test 12

Unconfined Compressive
Strength

Expansion Pressure

Sand Equivalent Value

Determination of pH of
Aqueous
Solutions with Glass
Electrode
 TABLE 10 (Cont.)

Test Standards

Applicable Applicable
ASTM AASHO British
Designation Test Standard Standard

Specific Gravity
of Coarse Aggregate

California Coarse and

Fine Durability Test.
Calif. 229-6

Determination of
Aggregate Impact Value

Soil Swelling Test

L.N.E.C. E200-67

Unconfined compression tests on samples compacted at Complete disaggregation should not be attempted as t h s does
Modified and Standard AASHO densities and "Pocket Penetro- not normally occur with lateritic soils.
meter" values were compared with the standard liquid limit
test results. No significant correlations were obtained. TABLE 11
A dye-absorption test was examined, in an attempt to
devise a completely new procedure. The various clay minerals Recommended Modifications of Test
absorb dyes in varying amounts. The clay minerals definitely Procedures
influence the liquid limit, as well as the other Atterberg limits.
Consequently, one would expect a correlation between dye ASTM
absorption and liquid limit. However, natural soils are usually Designation Test Modification
made up of more than one clay mineral plus other clay-size
materials. Lateritic soils in particular contain various iron and D l 221 7-66 Wet Preparation Use Procedure B.
alumina minerals, oft en clay-size. Do not heat in oven
Results of the dye-absorption tests correlated fairly well to evaporate excess
with the standard liquid limit test results, but erratic variations water
were common. Attempts at refining the test procedure only
introduced more tedious processes and special equipment. A D422-63 Part icle-size May be performed on
spectrophotometer was finally utilized to determine the Analysis wet sample. The dry
percentage of methylene blue dye absorbed. weight used in the
The test was finally abandoned as impractical. The main computation for sieve
objective would be to produce a new test which is descriptive analysis and hydrometer
and reproducible, yet relatively simple. The dye-absorption should be computed
procedure did not satisfy these requirements. from this initial wet
It was concluded that the liquid limit test, like all index weight and the moisture
tests, can be only partially descriptive. The liquid limit test content of the sample.
should define only the liquid limit, and the present standard
adequately provides this definition. Because lateritic soils are D423-66 Liquid Limit Run from wet t o dry
susceptible to breakdown with manipulation, the test proce-
dures should be more rigidly controlled. A nominal mixing D424-59 Plastic Limit Run from wet t o dry
time of 5 minutes should be adopted to prevent undue
breakdown. If the test is to be definitive, it should reflect the D l 55 1-70 Moisture-Density Use Note 1. Run from
true liquid limit of the material as it is utilized in construction. Relations wet to dry.
 LEGEND
L EGENDE
1

10 I

0
0 10 20 30 40 50 60 70
LlQUlD L I M I T (OVENDRIED) % LlQUlD LIMIT (SOAKED) %
LIMITE DE LIQUIDITE(SECHAGE' A ~ E T U V E ) O / ~ LIMITE DE LIQUIDITE (APRES' IMBIBITION)%

FIG.35 COMPARISON BETWEEN AS RECEIVED AND SOAKED LlQUlD LIMITS

FIG. 34 COMPARISON BETWEEN AIR DRIED AND OVEN DRIED LlQUlD LIMITS.
COMPARAISON ENTRE LlMlTES DE LlQUlDlTE DES ECHANTILLONS
COMPARAISON ENTRE LlMlTES DE LlQUlDlTE APRES SECHAGE A NONTRAITES E T APRES IMBIBITION

Pretreatment "the height t o which the cup that comes in contact with the
The four methods of pretreatment resulted in slight changes base is exactly 1 cm above the base." This point is not the
in the liquid limit between the AD and OD and between the bottom of the cup. B.S. No. 1377 states "when the cup is
AR and S but these changes were w i t h n the normal raised t o its maximum height the 1 cm gauge will just pass
reproducible range of the liquid limit. Liu and Thornburn between it and the base." This means that the B.S. method
(1964) reported a range of limits of approximately 2% using allows the cup t o be raised approximately 1 mm higher than
an experienced operator under uniform laboratory conditions. the ASTM method.
Normally, 5 t o 6 percent would be the average range of The liquid limit of 30 soils were run with both settings to
reproducible results. The AD liquid limit was about 5 percent demonstrate the effect of this difference in cup height. The
higher than the OD liquid limit. The AR liquid limit was about results of t h s comparison are shown on Fig. 36. For the same
3 percent lower than the S liquid limit. The differences are not soil the BS liquid limit should be slightly lower than the ASTM
large compared t o the reproducibility of the liquid limit test. liquid limit. However, this was not the case. Some of the
The OD results gave an extreme limit of the change in liquid ASTM liquid limits were lower than the BS liquid limits. It can
limit, not actually encountered under field conditions. The be seen that there is not a great difference in the results from
comparison of liquid limits is shown in Figs. 34 and 35. the 2 methods. The differences are within the standard
Because of these relatively small differences, subsequent deviation of the liquid limit test. Therefore, the results from
testing of moist soils was performed only on the AD and AR the BS test are within the same reproducible range of the
condition. liquid limit obtained from the ASTM method. Ibrahim (1970)
A soil was considered t o be in the moist state if its field found very little or no difference in the liquid limit resulting
moisture content was about 1 3 percent. Some soils sampled from the two methods.
dry were soaked to distinguish reversible trends. However,
there were no significant changes in the physical properties of
most soils when soaked.
Effect of Various Temperatures and Drying Periods
ASTM vs British Standard A special study was initiated t o compare the effect of
It was discovered halfway through the testing program that various temperatures and drying periods on the Atterberg
the locally used British Standard (BS) liquid limit determina- limits. In addition t o the AD and AR pretreatment, the
tion differed slightly from the ASTM standard. In the samples were oven dried at 5 0 " ~and 1 0 5 " ~for 6 and 24
adjustment of the liquid limit device the ASTM D423-66 states hours at each temperature. The changes in the Atterberg limits
 Waterways Experiment Station (1949) and was given in the
form:
LL=W(-) N tanB
25
where:
W = moisture content
N = the number of blows
tan B = the slope of the flow curve of blows versus
moisture content on a semi-log plot.
The Waterways Experiment Station found a mean value for
tan B of 0.121 with a standard deviation of 0.032. Norman
(1959) found tan B was 0.092 for 504 soils from Great Britain
and overseas. The standard deviation was 0.032. For 484
Canadian soils, Eden (1 959) found tan B t o be 0.108 with a
standard deviation of 0.036. Vaswani et al. (1 960) found tan B
t o be 0.078 for soils from India. The ASTM method D423-66
recommends using a tan B of 0.12. The value of tan B
determined during the initial study of Ghanian soils was 0.1 16
with a standard deviation of 0.044.
The second phase of the study included ferrisols, fer-
ruginous, ferrallitic soils from Ghana and other parts of Africa.
The results were as follows:
FIG. 36 RELATIONSHIP BETWEEN ASTM AND BRITISH STANDARD
LIQUID LIMIT
Soil tan B Standard deviation
RELATION ENTRE LlMlTES DE LlQUlDlTE SUIVANT L E S NORMES
ASTM E T LES NORMES BRITANNIQUES. Ferruginous 0.148 0.07
Ferrisols 0.140 0.05
Ferrallitic 0.153 0.07
are presented in Table 12. The data reveal that the largest All Soils 0.150 0.08
change that takes place in the liquid limit and plasticity index
occurs from the AR and AD condition. Most liquid limits The standard deviation in the percent error was 2.4 percent.
decreased from 12 t o 15 percent of the AR value. When the It can be seen that tan B in the second study was found to
samples were dried in the oven at the different temperatures be higher than previously reported. The difference between
and times, the decrease in the liquid limit and plasticity index the three groups is not significant when the standard deviation
was small. Three soils showed large decreases from 50°C when is considered. The standard deviations are higher than those
heated to 100°C. When the air dried samples were soaked, the previously reported. The mean value from the initial study of
liquid limit did not increase from the air dried liquid limit tan B = 0.1 16 is closer t o those previously reported, but when
except for one soil (Tamale 9), which is an example of a considering the large standard deviation, there is little meaning
reversible change. The aggregations formed during drying were in these differences. The significance of error in the one-point
easily dispersed during the soaking period. The aggregations liquid limit test is presented in more detail in the Appendix.
were apparently due t o only partial dehydration of .the iron The error inherent in the liquid limit test itself outweighs
minerals, and thus, were dispersed during soaking. any normal deviation of tan B. If tan B were zero, the standard
deviation in the percent error would be 2.6 percent. Thus, if
One-Point Liquid Limit the range of blows is kept between 20 and 30, tan B could be
At the beginning of the program a study was performed to assumed t o be zero and still lead to results within the
determine if the one-point liquid limit test could be adopted reproducible limits of the liquid limit test. For all practical
for lateritic soils. A total of 565 Ghanaian lateritic soils had cases almost any reported value of tan B could be used in the
been investigated by BRRI, prior to the present program. one-point liquid limit. The recommended value for tan B of
Later, 564 of the liquid limit tests run during the present 0.12 given in the ASTM procedures would be adequate for
program on soils from Ghana and other African countries were determining the one-point liquid limit for lateritic and
used in another study. These 564 test results were compared temperate soils.
separately for each lateritic soil group (i.e., ferruginous,
ferrallitic and ferrisol) and with all groups combined. Mechanical Analysis
The one-point LL method has been described or investi- Particle size analyses were performed on the AD and OD
gated by a number of authors, and various formulas have been samples in accordance with ASTM D422-63. Wet sieve analyses
proposed. The most widely used was developed by the U.S. were performed on AR and S samples on the portion retained
on the No. 10 sieve. The dry weights used in the computations with a size less than 1 micron. Here again, as in the liquid limit
for the gradation were calculated by using the wet weight and test, complete deflocculation or disaggregation is not mean-
moisture contents of the samples before sieving. The moisture ingful with tropically weathered soils. The dispersant should
content of the portion passing the No. 10 sieve was used to be standardized as well as the dispersing time.
determine the dry weight for the hydrometer analysis. It was decided to use sodium hexametaphosphate as a
Remillon (1967) states that it is nearly impossible to dispersing agent as recommended in ASTM. All samples were
completely deflocc~~late lateritic soils. Terza& (1958) found dispersed for 15 minutes, the longest time recommended in
that when the Sasumua clay was dispersed with sodium the ASTM procedures.
oxolate the percentage of particles smaller than 2 micron The test results disclosed very little change in the gradations
ranged between 20 and 30 percent and when dispersed with between the OD and AD and between the AR and S samples.
sodium hexametaphosphate ranged between 40 and 50 per- The differences between the AD and AR samples were not
cent. The mineralogical investigation showed that the clay often appreciable. Those soils that displayed significant dif-
contained between 7 0 and 100 percent clay mineral particles ferences are discussed in Chapter 5.

TABLE 12

Change in Atterberg Limits When Drying At

Various Temperatures and Times

As Air Soaked 6 24 6 24
Soil Received Dried 24 hours hours hours hours hours
at 50°C at 50°C at 105°C at 105°C

OKODEE LL
PL

BEKWAI LL
PL

TAMALE 9 LL
PL
CBR led to considerable breakdown of the gravel size particles. New
After the maximum density and optimum moisture were material should be used for each point on the compaction
determined, a CBR sample was prepared at the optimum curve.
moisture content in accordance with ASTM D 1557-70 Method
D. The sample was allowed to soak for 4 days under a 10 Harvard Miniature Compaction
pound surcharge. Swell readings were taken before and after The Harvard miniature compaction procedure was adopted
soaking. The CBR values were determined in accordance with because of the expense in transporting large size samples. A
ASTM D1883-67. tentative procedure for performing the Harvard Miniature
The CBR value is dependent on the molding moisture Compaction test was proposed by Wilson (1970).
content. Multi-point CBR curves were prepared for the After some preliminary studies of the Harvard miniature
Roadway Condition Survey - Chapter 12. It was observed compaction procedures it was decided to use a 4 0 pound
that a slight increase in the molding moisture content above spring with 1 0 blows on each of 5 layers. The results of the
the optimum moisture content would drastically reduce the maximum density-optimum moisture determinations are pre-
CBR values. A minimum of 3 points should be performed to sented on Fig. 38.
determine the CBR accurately at the optimum moisture T h s procedure appears to be a useful test to obtain the
content 01. at any particular moisture content. maximum density and optimum moisture when there is not
sufficient soil t o run the modified compaction test. There
Compaction seems to be a good relationshp between the maximum
densities and OMC's determined by the Harvard compaction
A special study was undertaken to understand how the
apparatus and the ASTM modified method as shown on Figs.
method of preparation could affect the compaction character-
39 and 40. To convert the Harvard to the ASTM modified
istics of lateritic materials. Fresh and re-used samples were
maximum density the equation
tested AR, AD and OD. The samples were compacted in
accordance with the Ghana compactive effort which uses a 10 MDD Modified = 16.0 + 0.926 MDD HM
pound rammer falling 18 inches with 25 blows on each of 5
layers in a 6 inch mold. The AD and OD samples were may be used. To convert the Harvard OMC t o modified
prepared in accordance with ASTM Dl 557-70 Method D. The
AR samples were tested from the wet t o dry state starting OMC-Modified = 0.1 23 + 0.802 OMC HM
from the natural moisture content, or adding moisture when
the natural moisture content was less than the optimum
moisture content.
The results of this study indicate that the compaction
characteristics of lateritic soils change after various pretreat-
m e n t ~ Generally,
. OD samples have higher maximum densities
and lower optimum moistures than the AD or S samples. A
similar relationship was reported by Grady (1949) who
observed that significant variations in the compaction char-
acteristics of Hawaiian lateritic soils appeared t o be related to
the natural moisture content. The drier the initial state, the
higher the maximum density and the lower the optimum
moisture content.
There was significant increase in the maximum density and
a decrease in the optimum moisture content when the same -
sample was re-used. T h s is illustrated in Fig. 37 which is
typical of most lateritic gravels. Similar results were also GRAVEL NO. I
GRAVILLON NO. I
reported by Johnson and Sallberg (1962). When the re-used
FERRISOL
samples were subjected to the same pretreatment as fresh FERRISOL
samples, the highest maximum densities and lowest optimum RE-USED SAMPLE
moisture contents were obtained with the OD soil. During
/ ECHANTILLON R E U T I L I S ~
1 I

compaction the gravel size particles tended t o break down to

sand size particles and thus improved .the gradation of the MOISTURE CONTENT '10
TENEUR E N EAU '10
sample and increased the maximum density. The amount of
breakdown of coarse gravel particles varied, but generally was DRY DENSITY/MOISTURE CONTENT RELATIONSHIPS
AFTER DIFFERENT METHODS OF DRYING
less for soils compacted at the natural moisture content than
RELATION DENSITE SECHE -TENEUR E N EAU
for air or oven dried soils. In all cases, re-using the same sample APRES DIFFEREN'TES METHODES DE SECHAGE
 140
I LEGEND $
40
LEGEND
LEGENDE
FERRUGINOUS
A FERRUGINEUX
FERRALLlTlC
O FERRALLITIOUE

401b SPRING ".

0 I0 COUPS / COUCHE
5 LAYERS

OPTIMUM MOISTURE CONTENT ( HARVARD COMPACTION ). %

TENEUR EN EAU OPTIMUM (COMPACTAGE DU HARVARD), %
10 15 20 25 30 35 40
FIG. 4 0 RELATIONSHIPS B E T W E E N HARVARD AND MODIFIED AASHO
O P T I M U M MOISTURE CONTENT, %
MOISTURE C O N T E N T S
TENEUR E N E A U OPTIMUM. /'a
RELATION ENTRE TENEURS EN EAU OPTIMUMS DES PROCEDES
HARVARD E T AASHO MODIFIE.
F16.38 HARVARD MINIATURE M O I S T U R E - D E N S I T Y RELATIONSHIPS
RELATION ENTRE TENEUR E N EAU E T DENSITE DU HARVARD
MINIATURE.

Both regressions had a correlation coefficient of 0.888. even a decrease in maximum density when the blows were
To determine if a closer correlation could be obtained with increased (see Fig. 41).
modified compaction, a number of tests were performed with
20 blows of the Harvard apparatus per layer instead of 10 DURABl LlTY AND AGGREGATE STRENGTH
blows. There appears t o be a limit on the number of blows to The durability test measures the relative resistance of a soil
which lateritic soils can be subjected with a corresponding to producing clay particles when the coarse and fine fractions
increase in the maximum When the were are subjected to mechanical degradation. The aggregate impact
increased from 1 0 to 20 per layer no significant change test measures the strengthofaggregate.
occurred in the maximum density. In a few cases there was

FERRUGINOUS

FERRALLITIPUE

90
I << i H A R ~ RDE
PROCEDE D;MpAcTloN PROCEDURE^
COMPACTAGE DU HARVARD

4 0 1b SPRING
RESORT DES 4 0 LIVRES
LAYERS
COUCHES
BLOWS / LAYER
COUPS/ COUCHE
80
70 80 90 100 110 120 130 140 150
MAXIMUM D E N S I T Y ( H A R V A R D COMPACTION) p c f
DENSITE MAXIMUM (COMPACTAGE DU HARVARD) LIVRE'S / PIED^
""
80 9 0 100 110 120 130 140 150 160
FIG.39 RELATIONSHIP BETWEEN HARVARD AND MODIFIED AASHO -
HARVARD COMPACTION ( 2 0 b l o w s / layers) p c f -
DENSITIES COMPACTAGE HARVARD ( 2 0 COUPS/COUCHE ) LIVRES/ PI E D

RELATION ENTRE DENSITES DES PROCEDES HARVARD ET AASHO FIG.41 E F F E C T OF NUMBER OF BLOWS ON HARVARD DENSITY
MODlFlE . INFLUENCE DU NOMBRE DE COUPS SUR LA D E N S I T E
 LEGEND
-
" -
A A
The majority of the samples selected for these tests were
LEGENDE
A FERRUGINOUS 0
from the condition survey.
FERRUGINEUX
The general description of the equipment and method for
70 '~~~%~I!~~~I$Es
the durability test is found in the State of California,
Department of h b l i c Works, Division of Highways Material
Manual, Test Method 229-C. The method was used throughout
this study with the following modifications.
1. The washing vessel specified in the test method is no
longer commercially manufactured. Consequently, three U.S.
Standard sieves were welded together with a sieve bottom pan
to form a washing vessel 8 inches in diameter and 8 % inches
high.
2. A standard Tyler sieve shaker with a revolution of about
300 cycles per minute and a stroke of 2 inches was used rather
than the modified shaker recommended in the test procedure.
3. The sample was air dried instead of oven dried since it
was discovered earlier in the study that there were significant
differences in the durability values for both melhods. The
exposure to heat caused partial induration which resulted in
higher durability values for the oven dried samples. The air
dried sample durability would be more reflective of the actual
field conditions.
The aggregate impact test was performed in accordance
COARSE DURABILITY ( D C ) with the British Standard Specification No. 812, 1960
DURABILITE DES ELEMENTS GROSSIERS
presented in the Appendix.
FIG.42 COARSE DURABlLlTY VS. C BR The specific gravity and absorption of the coarse aggregate
DURABILITE DES ELEMENTS GROSSIERS
EN FONCTION DU C BR were determined in accorance with ASTM C127.
Coarse Durability (D,) values ranged from 5 t o 76 and the
fine durability (Df) values ranged from 1 6 to 90. These values
indicate quality that ranges from very weak t o good. The State
of California has a minimum requirement of 4 0 for base
aggregate. The results of this study show that 8 0 percent of

70 r- - r --T
~ - - -
y;;E
f, FERRUGINOUS
FERRUGINEUX
the gravels tested gave values above 40. The durability values
did not correlate well with mechanical or strength properties.
When the Coarse Durability is compared with the CBR values,
as shown on Fig. 42 there is a wide scatter of results. This
relationship suggests that the durability of the aggregate does
not contribute to the CBR strength. It is interesting to observe
that the coarse durability as shown on Fig. 4 3 also does not
correlate with the aggregate impact value (AIV).

The specific gravity and absorption tests indicate a rela-

tionship with coarse durability. The fine durability shows a
relationship only with absorption and not with specific
gravity. The coarse durability increases as the specific gravity
increases. This is shown on Fig. 44. The fine durability
compared with specific gravity is shown on Fig. 45. As the
water absorption decreases, the fine and coarse durability
increase, Figs. 4 6 and 47. With a durability value of 40, the
water absorption for coarse durability would be around 3%
percent and about 6 percent for fine durability. The higher
0 10 20 33 40 50 60 73 absorption for the fine durability would be due t o the larger
COARSE D U R A B I L I T Y (DC)
DURABILITE DES E L E M E N T S GROSSIERS surface area of the finer particles. As would be expected, a
FIG.43 COARSE DURABILITY VS. AGGREGATE IMPACT VALUE high specific gravity and low absorption rate would indicate a
DURABILITE DES ELEMENTS GROSSIERS EN FONCTION DE L A RESISTANCE
DES AGREGATS AU CHOC. sound aggregate and soil.
 I10
LEGEND
LEGENDE

100 .
FERRUGINOUS
FERRUGINEUX , , FERRUGINOUS
FERRUGINEUX

FERRALLITIOUES
-
V)
0 FERRISOLS
FERRISOLS
W

z
0
I a
W lO
n
m m
$a 4
:
22
w l

10 20 33 40 50 60
COARSE DURABILITY (DC)
DURABlLlTE DES ELEMENTS GROSSIERS
FIG.46 COARSE DURABILITY VS. WATER ABSORPTION
DURABlLlTE DES ELEMENTS GROSSIERS EN FONCTION DE
LA QUANTITE D'EAU ABSORBEE.

relationship with the CBR as shown on Fig. 48. The AIV

showed a fair correlation with specific gravity. As the specific
SPECIFIC GRAVITY (COARSE)
gravity increased, the AIV decreased linearly. From this
DENSITE (ELEMENTS GROSSIERS) relationship, shown on Fig. 49, it can be seen that with an AIV
FIG.44 DURABILITY COARSE VS. SPECIFIC GRAVITY C O A R S E
WRABlLlTE DES ELEMENTS GROSSIERS EN FONCTION DE LEUR of 30 or less, the specific gravity is about 3.0. The AIV
DENSITE
increased proportionally with the absorption, as shown on Fig.
The values of AIV, which ranged from 17 to 54, indicate a 50. At an AIV of 30 the absorption was about 4 percent. As
range of aggregate quality from excellent to poor. Generally, the absorption increased, the soundness of the aggregate
an AIV of less than 3 0 indicates excellent quality, whereas decreased.
over 50 indicates poor quality. The data show that -the quality Gravels derived from granitic bedrock usually contain
of all except one gravel ranged from excellent to fair, with 40 quartz fragments and concretionary nodules in varying
percent having an AIV less than 30. The AIV did not show a

I
i
FERRUGINOUS 1

0
10 20 30 40 50 60 70 80
SPECIFIC GRAVITY (FINE) FlNE DURABILITY (DF)
DENSITE (FINES)
DURABlLlTE DES F I N E S
FlG.45 FlNE DURABILITY -
VS. SPECIFIC GRAVITY FINE FIG. 47 FlNE DURABILITY VS. WATER ABSORPTION
D U R A B l L l T E DES F l N E S E N FONCTION M L E U R D U R A B l L l T E D E S F I N E S E N FONCTION D E L A
DENSITE QUANTITE D'EAU ABSORBEE.
 amounts. A study was performed on .the aggregate impact
A FERRUGINOUS strength of quartz-lateritic gravel mixtures. The results of this
FERRUGINEUX
study are presented on Table 13. It can be seen that as the
70 1 0FERRALLlTlC
FERRALLITIQUES /----L----
I
1 4 3 I
I
quartz content increases the AIV also increases. This is true for
1 0 FERRISOLS
FERRISOLP all of the mixtures tested. This means that in terms of hardness
the lateritic gravels are t o be preferred t o quartz gravels. The
AIV is a function of the degree of maturity of the lateritic
gravels and their relative content in the natural gravel.

EXPANSION
The Potential Volume Change (PVC), CBR Swell, and
Portuguese swell tests were adopted to measure the expansion
of lateritic clays. The PVC test is a procedure used by the
United States Federal Housing Authority (Lambe, 1960) to
measure the swell pressure due to a change in the moisture
content of a soil. The swell index pressure, measured after a
2-hour soaking period, is correlated with volume change, swell
pressure and PI. The PVC test procedure is described in the
Appendix to Chapter 16.
The Portuguese swell test apparatus was developed by the
National Laboratory of Civil Engineering in Lisbon (1969),
DeCastro (1969), to introduce swelling as a new parameter
among the consistency limits. This test procedure is described
in .the Appendix.
On the basis of studies on typical African lateritic soils, the
0 10 20 30 40 50 60
National Laboratory and DeCastro were able to associate field
AGGREGATE IMPACT VALUE - %
RESISTANCE AU CHOC D E S AGREGATS-So performance with swelling. Soils with swell not exceeding 10
FIG. 48 AGGREGATE IMPACT V A L U E VS. C B R percent were generally found t o give satisfactory performance.
RESISTANCE AU CHOC D E S AGREGATS E N
It was found possible to relax the recommended requirements
FONCTION DU C B R . (ASTM D1241-68T) for base course materials if the swelling

LEGEND
LEGENDE
FERRUGINOUS
n FERRUGINEUX
n OFERRALLlTlC
L.1 FERRALLITIQUE
a 0
o FERRISOLS
FERRISOLS

Cs 0

q 0 : ; .

pppppp

SPECIFIC GRAVITY (COARSE)

DENSITE (ELEMENTS GROSSIERS)

FIG.49 AGGREGATE IMPACT VALLIE VS. SPECIFIC GRAVITY - COARSE

RESISTANCE A U CHOC AGREGATS E N FONCTION DE L A D E N S I T E
DES ELEMENTS GROSSIERS
 TABLE 13 70

Aggregate l mpact Strength of Quartz A FERRUGINOUS

FERRUGINEUX
60 FERRALLITIC
(Ferrisol over granitic bedrock)
8
6.a?
- 5 0
AGGREGATE IMPACT VALUE % Ey
Lateritecontent 10W0 75% 50% 25% 0% 5 J

3
Quartz Content WO 25% 50% 75% 10W0 X ,4 0
0 0
g2
Sample 1 20 O
. - 22.8 - 27.0 0 9

:
3

LI 3 0
2 22.5 - 25.5 - 27.0
z;
0

v, a
+

3 25.0 27.0 28.0 29.3 30.5 w

V) 0 20
a u

4 25.1 - 29.4 - 31.6

0 t I
I
6 24.0 28.5 32.0 34.0 35.5 0 2 4 6 8 10

WATER ABSORPTION, O/o

7 24.7 29.0 32.0 35.5 36.0 EAU ABSORBEE, O/O

FIG. 5 0 AGGREGATE IMPACT VALUE VS. WATER ABSORPTION

8 25.5 - 31.5 - 34.8 RESISTANCE AU CHOC DES AGREGATS EN FONCTION DE LA
QUANTITE D'EAU ABSORBEE.

plasticity index about 28.

10 39.3 41 O
. 42.5 43.3 44.0 There is a good relationship between Portuguese swell and
linear shrinkage. An increase in the linear shrinkage had a
38.0 39.2 40-0 40-7 42.0 corresponding increase in swelling, Fig. 54. At 10 percent
swelling the linear shrinkage was 6 percent.
did not exceed 10 percent. The LL and PI for base and
sub-base course could be increased from 35 t o 40 and 10 to 15 CONCLUSIONS AND RECOMMENDATIONS
respectively. The allowable percentage of the minus 200 The pretreatment given to lateritic material should depend
fraction could be increased by 1.4 times the recommended upon the actual conditions that will be encountered in the
ASTM values. field. For this reason, samples should not be oven dried before
The Portuguese swell and PVC tests show that lateritic clays testing since the soils generally will not experience such severe
are not highly expansive. The majority of the soils tested by temperature in the field. Generally, samples may be air dried
the PVC method were in a non-critical or marginal category. prior to testing, provided the index tests show that significant
Most of the soils tested by the Portuguese swell apparatus had irreversible changes do not occur. Such soils should be tested
less than 10 percent swell. Although these two tests both in a wet state.
showed that lateritic soils are not too hghly expansive, there The difference between the British Standard and ASTM
was no other correlation between the results of the two determinations of liquid limit does not appear to be beyond
methods. the reproducible limits of the test itself. An effort should be
The test results show there is a correlation between the made to reconcile the two procedures.
Portuguese swell and the percent passing the No. 200 sieve, The one-point liquid limit test is effective in determining
Fig. 51. The swelling increases as the percentage passing the the liquid limit of lateritic soils by using the formula:
No. 200 sieve increases. At 10 percent swelling the value of
minus No. 200 is 72, which is h g h for most lateritic soils. LL = W(25) N tan B
The Atterberg limits also show a correlation with Portu- When the number of blows is between 20 and 30, tan B can be
guese swell, Figs. 52 and 53. There was a gradual increase in assumed t o be zero and the results will still be within the
swelling as the liquid limit and plasticity index increased. At accuracy of the liquid limit test. If more accurate results are
10 percent swelling the liquid limit was about 50 and the required, tan B ~ 0 . 1 2can be used. The one-point method
 20 40 60 80 100
PASSING NO. 2 0 0 SIEVE
%
ELEMENTS PASSANT L E TAMlS U.S. 200, %

FIG. 51 R E L A T I O N B E T W E E N S W E L L I N G A N D % PASSING
NO. 2 0 0 S I E V E
R E L A T I O N E N T R E GONFLEMENT E T L E POURCENTAGE % P L A S T I C I T Y INDEX
DES E L E M E N T S PASSANT L E T A M l S U.S. 2 0 0 INDICE D E PLASTICI'TE
FIG.53 RELATION BETWEEN SWELLING AND
PLASTICITY INDEX.
should not be used on soils containing an appreciable amount
of organic matter. RELATION ENTRE G O N F L E M E N T E T INDICE
Sodium hexametaphosphate appears to be the best agent DE PLASTlClTE
for dispersing most lateritic soils. However, this agent may not
be sufficient to deflocculate all of the clay minerals in
halloysite soils. Mineralogical investigations would then be
desirable. If the soil cannot be dried in the field before compaction
When a compaction test is performed on lateritic soil, care then the laboratory curve should be run from a wet t o dry
should be taken to avoid significant breakdown of the nodular state. If the field moisture content is below optimum moisture
concretions and soil particles. Each point on the compaction content, then additional moisture should be added to some
curve should be run with new material and the material samples to bring the moisture content above the optimum
allowed to cure overnight prior to making the moisture-density moisture content. If the field moisture content is not at least
test.

ij? " '2

5
w
5
--1
I
w 3
-J 8
L
g
a
ij?

20 40 60 80 100
01 I I
1 I

% LIQUID LIMIT 2 3 4 5 6 7
L l M l T E DE LlQUlDlTE '10 '10 LINEAR SHRINKAGE
FIG. 52 R E L A T I O N B E T W E E N SWELLING AND LIQUID LIMIT RETRAITE LlNEAlRE
FIG.54 RELATION B E T W E E N SWELLING AND LINEAR SHRINKAGE
RELATION ENTRE GONFLEMENT E T L l M l T E D E
LlQUlDlTE R E L A T I O N ENTRE GONFLEMENT E T RETRAIT LlNEAlRE
13 percent, the soil is too dry to run from wet to dry. The Mexico City, vol. 1 : 97-106.
CBR value is extremely dependent on the molding moisture Eden, W. J., 1959. Use of one-point liquid limit procedure.
content. To determine the CBR accurately at optimum Symp. on Atterberg Limits. Amer. Soc. of Test. Mat., STP
moisture content, at least 3 points should be run. No. 25: 168p.
The Harvard Miniature Compaction can be used when there Grady, J. O., 1949. Discussion on effect of re-using soil on
is insufficient soil to run a standard or modified compaction moisture-density curves. High. Res. Board, Proc. No.
test. When performing this test, a 40 pound spring with 10 29: 488-490.
blows on 5 layers should be used to correlate with modified Ibrahim, B. B., 1970. Laterite soils study, Sierra Leone
compaction. Modified maximum density is equal to 16.0 + participation. Eng. Res. Publ., Res. Paper No. SM 1, Dept.
0.926 (Harvard miniature maximum density). The modified of Eng., Fourah Bay College, Univ. of Sierra Leone,
optimum moisture is equal to 0.358 + 0.797 (Harvard Freetown.
Miniature OMC). Johnson, A. W. and Sallberg, J. R., 1962. Factors influencing
The Durability test does not appear t o be a valid test to compaction test results. High. Res. Board, Bull. No.
determine the quality of African lateritic soils and gravels. 319: 148p.
Valerga et al. (1969) concluded that the Durability test Laboratorio Nacional de Engenharia Civil (Lisboa), 1969.
correlates well with observed performance in Thailand. The Portuguese studies on engineering properties of lateritic
roadway condition survey, Chapter 12, disclosed no cor- soils. Proc. of the Spec. Session on Eng. Prop. of Lateritic
relation between the durability and performance .of the Soils. Seventh Int. Conf. on Soil Mech. and Found. Eng.,
materials in Africa. The best relationship with performance Mexico City, vol. 1: 85-96.
was displayed by CBR and .thickness. If there were a Lambe, T. W., 1960. The character and identification of
relationshp between durability and performance, a relation- expansive soils (soil PVC Meter). Fed. Housing Admin.
ship between durability and CBR could be expected. No such Report FHA-701.
relationship was found to exist. The durability also does not Liu, T. K. and Thornburn, T . H., 1964. Study of the
correlate with the aggregate impact strength. Reproducibility of Atterberg Limits. High. Res. Board
The range of Aggregate Impact Value results indicates that Record 63: 22-30.
most lateritic gravels will perform well in road pavements. T h s Norman, L. E. J., 1959. The one-point method of determining
was verified in Chapter 12. The AIV test is a simple procedure the value of the liquid limit of a soil. Geotechnique, vol.
to determine the resistance of aggregate to dynamic loading. 9 : 1-8.
Further studies are needed to define more accurately the Remillon, A., 1967. Les recherches routiers entreprise en
empirical limits of the applicability of this test to actual Afrique d'expression Franqaise. Ann. de 1'Inst. Tech. du
performance of pavements. It has been established there is a Biitiment et des Trav. Publics. No. 23 1-232: 366-388.
relationship of specific gravity and water absorption to Tateishi, H., 1967. Basic engineering characteristics of high
durability and AIV. With further studies it may be possible to moisture tropical soils. Paper presented at the West. Assoc.
predict more accurately the mechanical and strength prop- State High. Off., Honolulu, July 12: 19p.
erties from these simple physical tests. Terzaghi, K., 1958. Design and performance of the Sasumua
The PVC test has limited value to measure expansion dam. Proc. Inst. Civil Eng., vol. 8 : 369-395.
accurately but can be used as a quick and easy procedure to Vallerga, B. A., Shuster, J . A., Love, A. L., and Van Til, C. J.,
determine if a soil is critically expansive. The swell index 1969. Engineering study of laterite and lateritic soils in
cannot be reasonably predicted by the ordinary index proper- connection with construction of roads, highways and
ties of the soil airfields. Phase I - Southeast Asia (Thailand). Final Report
The Portuguese swell test shows a relationship with the for U.S. Agency for International Development. Contract
Atterberg limits, shrinkage limit and percentage passing the AID/csd- 18 10. Soil and Pavement Consultants, Oakland,
No. 200 sieve. There was no relationship between the results Calif.: 165p. and Appendices.
of this test and of the PVC. The results of this study (Chapter Vaswani, S. V. K., Gogna, M. L., and Sood, S. C.,
12) showed that the soils that had swelling less than 10 1960. Simplified one-point method of liquid limit test of
percent performed well in road pavements. However, as most soils-its applicability to Indian soils. Bull. of Indian Nat.
of the soils tested were in this category, the test could not be Soc. of Soil Mech. and Found. Eng., No. 7.
critically evaluated. Waterways Experiment Station, 1949. Simplification of liquid
limit test procedure. W. E. S. Tech. Memo. No. 3-286.
Wilson, S. D., 1970. Suggested method of test for moisture-
REFERENCES density relations of soils using Harvard compaction appara-
DeCastro, E., 1969. A swelling test for the study of lateritic tus. In: Special procedure for testing soil and rocks for
soils. Proc. of the Spec. Session on Eng. Prop. of Lateritic engineering purposes. A.S.T.M. Spec. Tech. Publ.
Soils, Seventh Int. Conf. on Soil Mech. and Found. Eng., 479: 101-103.
 CHAPTER 7

AIR PHOTO INTERPRETATION

Lateritic soils also include fine-grained materials, however.

In recent years, considerable attention has been given to the Some of these materials may occur naturally in a moist
development of methods through which identification of earth condition and show the property of hardening to a rock-like
materials can be carried out over a fairly large area quickly and mass or coarse-grained aggregations upon drying. Such soils are
economically. Once the identification has been made much useful as an inexpensive source of construction block material
information concerning the engineering properties of the in some parts of the world. Other lateritic materials differ little
materials may be inferred. Although the information provided from temperate zone sands, silts and clays. Thus, terrain
it usually qualitative and seldom serves directly as the basis for investigation for engineering purposes is not as much con-
engineering design of a project, it provides an ideal basis for cerned with the identification of lateritic material per se as
extrapolating supplemental information obtained from field with the identification of particular kinds of lateritic materials
surveys and physical testing of materials. Where a fairly large having engineering value or presenting special problems.
area of terrain is being examined, as in selecting the location Liang (1964), writing from an engineering viewpoint,
for a highway project or in searching for sources of suitable proposed a classification of tropical soils into eleven groups
aggregates, an evaluation of the entire terrain through airphoto (Table 14). Among the lateritic soils he included laterite crust,
interpretative techniques may eliminate a large number of sites laterite gravel, laterite (hardening upon exposure), and lateritic
from further consideration and assign priorities to the remain- soil (including bauxite). The latter term was used in a much
ing ones worthy of field investigation. more restricted sense than that adopted in the present report.
While aerial photography is often of prime importance as a Other important groups were red and brown clays, desert soils,
medium for terrain investigation, it is by no means the only and black clays. Although these classifications differ somewhat
tool used by the investigator. A familiarity with the area under from those presented earlier in this report, they are based on
consideration is first obtained through publications on the terrain characteristics which can be identified in aerial photo-
geology, geography, agronomy or forestry of the region. Of graphs (Table 14). Liang (1964) was not able to use color
great usefulness are reports and maps presenting lithology, photography in his investigation but noted that, "Color plays a
agricultural soil characteristics, road systems, land use pat- very important role in the identification and interpretation of
terns, or characteristics of the vegetative cover. Usually some lateritic soils and the red, brown and yellow soils in the
correlation can be discovered between a reported mapping unit tropics. Color photography has a special advantage, particu-
and the appearance of the area represented by the mapping larly when the ground is not hidden by vegetation, and should
unit in aerial photography. The photos then provide a guide to be explored further."
the more accessible occurrences of these units of interest
which will then be located and investigated in the field. IDENTI FlCATlON OF LATERITES
Usually, the information obtained can be extended to other AND OTHER TROPICAL SOILS
areas where similar patterns or appearances occur in the The highly weathered lateritic soils occur in a part of the
photography. It is to be expected that this will be the normal world where most other soils have high clay contents. Thus,
procedure followed in much of Africa, particularly if quite these soils containing concretionary material are distinctive
detailed and relatively specific information is desired. because of their generally coarser texture, greater per-
Engineering interest in the properties of soil materials meability, and lack of ability to retain water and nutrient ions.
centers about such characteristics as gradation, permeability, Areas of such soils are usually used agriculturally only for poor
strength, deformation under load and change in properties pasture or are not farmed at all. Surrounding areas may be
with wetness. Information on these properties of African soils forested but the coarse-grained or indurated area supports only
is generally unavailable or poorly organized. The scarcity in scattered scrubby trees and sparse grass. The general char-
many parts of the tropics of sound rock or coarse granular acteristic of high permeability applies even to well-indurated
material suitable for use as base course or construction lateritic crusts.
aggregate enhances the engineer's interest in the hard crusts or True laterite crust, LC (Table 14) often appears in association
gravel inclusions in some lateritic soils because they provide with the remnants of ancient peneplain surfaces. Dowling
suitable sources of coarse aggregate. If they can be identified ( I 968) describes several occurrences in northeastern Nigeria
such accumulations of concretionary material have the advan- where the soils have developed over very different geological
tages of being at or close to the surface of the ground and parent materials. Liang (1964) states that this association is a
furthermore to be more resistant to tropical weathering than general characteristic of lateritic crusts and cites examples with
the primary rock material quarried at some depth below the illustrations in Brazil, Uganda and Australia. Nearly all of the
surface. peneplain remnants appear to be of rather small size; at most a
 TABLE 14

Airphoto Interpretation of Tropical Soils-Summary (after Liang, 1964)

Soil Groups Topography (land Form) Drainage Erosion Grav Tone Vegetation Remarks

1. Laterite Crust Cap rock (relatively flat Little surface Little; rockfalls Light Grass or low
LC hilltops): Ledges on drainage on edges of cap shrubs
slopes

2. Laterite Gravel Flat hilltops to rolling Little surface Little; gullying Light Grass or low shrubs
LG ground drainage erosion starts
on slopes below
the gravel
layer

3. Laterite (Har- Subdued to rolling ground Little surface Little erosion Light Variable; General absence of land-
dening upon intervened by flat val- drainage Plantations; slides; suitable for golf
exposure) leys at high to inter- Swidden agriculture course or cemetery develop-
LH mediate slopes ment; possibility of exca-
vation pits

4. Lateritic Soil Rolling ground, at high Variable; less Variable; less Light to medium Variable; Plantations; Termite hills are common
LS to intermediate slopes surface erosion with Swidden agriculture
drainage with more advanced common; Intensive
more advanced stage of and agriculture in high
stage of and deeper later- rainfall area
deeper later- ization
ization

5. Red and Brown Variable; all forms Considerable Gullying Medium to dark Variable; intensive
Clays except in basins surface drain- erosion; depth agriculture common
CR age develop- of gullying
men t indication of
depth of soil
to rock; slides

6. Black Clays Gentle lower slopes and Little surface Little erosion Dark Intensive agriculture Quality of clay depends
CB flat depressions drainage; high o n source of material
water table and mode of deposition

7. Desert Soils Variable Dry drainage Wind erosion Light General absence
SD channels

8. Saline Soils Low, coastal land; water Tidal drainage Little erosion Light patches (dark in General absence
SS table near surface channels special cases)

9. Swamps Low, flat land; water Tidal drainage Little erosion Dark Mangrove and other
SW table at surface channels swamp vegetation

10. Alluvium Low, flat land bordering Little surface Little erosion Light to medium Recent-little vegetation;
SA stream drainage Old-intensive agriculture

11. Rock and Thin Rolling to hilly, at Variable Variable Light to medium Variable; no agriculture
Soils high slopes development
SR
few square miles in area. Bates (1962) suggests that as erosion present in the subsoil a foot or two below the surface in
encroaches upon the peneplain remnant, permitting substantial well-drained areas where the vegetative cover is grass and shrub
lateral flow of groundwater, segregation of iron and aluminum rather than forest. Near the upper convex portion of hillsides,
oxides occurs. The iron is transported in solution to the a concentration of gravel analogous t o a laterite crust may be
exposed sides of the remnant where it is precipitated as the found. If a crust is present, an accumulation of detached scree
groundwater enters a strongly oxidizing environment. The material may be present. Apparently, concentrations of gravel
interior of the remnant is found to be less consolidated and to are least likely to be found near mid-slope. Near the bottoms
contain less iron oxide and more aluminum oxide than the of slopes, soil aggregations may be enlarged or cemented by
outside fringe. This observation is confirmed by Brammer additions of iron oxide precipitated from groundwater. If the
(1962) in discussing the agricultural soils of Ghana. Ap- underlying material is quite impermeable, the enriched mate-
parently, laterite crust LC occurs typically as a ring surround- rial may remain permanently moist and become Liang's
ing a flat-topped peneplain remnant rather than as a caprock laterite hardening on exposure, LH. At the bottom of the
covering the peneplain surface. Because of the high per- slope, deposits containing lateritic detritus are usually buried
meability of the surface soil on the remnant the airphoto beneath finegrained colluvial or sheet wash material or by
frequently shows only sparse vegetation and the absence of alluvium in a major valley.
surface drainage features. Slopes leading downhill away from Even those lateritic soils which do not contain appreciable
the mesa usually present concave profiles. Enroaching gullies coarse material are generally poor agricultural soils. Liang
appear suddenly, often at the bottom of almost vertical (1964) states that such areas are often devoted to "swidden"
headwalls, tens of feet in height. Liang (1964) notes the agriculture, in which an area is cleared, farmed for a few years,
danger of confusing the LC pattern with that of basalt or and then abandoned to native vegetation for some tens of
sandstone, which often form a caprock over a base of less years. The resulting land pattern is often quite apparent in
permeable, more easily eroded material. aerial photography which shows scattered small fields, some of
which are under cultivation and others in various stages of the
Another clue to the identification of terrain areas with reestablishment of native vegetation. In contrast, nearby areas
possible laterite crusts is provided by the fact that in a given with soils having coarse lateritic materials are uncultivated and
peneplain surfaces have formed only at a few stages in geologic those areas of less weathered clay or alluvial soils present
time. Often the approximate elevations of these surfaces may patterns of intensive perennial agricultural use.
be known or can be inferred from elevations of the sur- Although sources of sound gravel are rare near the ground
rounding areas. The Voltaian basin of east-central Ghana surface in the tropics, materials other than laterite may
presents several surfaces, usually very discontinuous, but sometimes be found. Deep residual weathering of rocks
similar enough in elevation that correlation can be made. The containing veins or inclusions of quartz can result in a terrain
most noteworthy occurs on all of the higher hills at an where a layer of quartz fragments is concentrated at or near
elevation of 1400 to 2500 feet above sea level (Brash, 1962). the ground surface. In spite of its convenient location, it is
Ferricrete is a laterite-like material formed by the cementa- unfortunately, often too thin and scattered t o be of much use.
tion of the sands and gravels in stream banks and terraces. The
formation of ferricrete is a continuing process which takes If a fairly thick layer is present it can sometimes be identified
place in areas close to the level of present day streams as iron in aerial photography through tonal lightness and lack of
oxide is precipitated from emerging groundwater. Ferricrete continuous vegetative cover. Identification is easier on more
may also be associated with systems of terraces at levels above sloping terrains where sheet erosion has washed the finer
the elevations of present streams where the stream regimen material from the surface or where it is known that the
previously was stable for a long period of time. In the Voltaian underlying country rock is one which contains large veins or
basin terraces are frequently found at 60-75 feet, 100-120 feet lenses of quartz. Occurrences of this type are more likely to be
and 250-300 feet above present base levels. The two lower present on youthful terrains.
terraces often contain well-cemented ferricrete deposits. Surface drainage conditions are often of considerable
Laterite gravel LC, like the laterite crust LC, may occur as a engineering importance. Natural color aerial photography
result of weathering in situ or by other processes which provides one means for drainage evaluation that can be
produce accumulations of particles high in iron oxide. LG is usefully combined with considerations of topographic posi-
often found on high-lying ancient erosion surfaces where the tion. Variations in the state of oxidation and hydration of iron
formation of LC is not complete. Such surfaces tend to be compounds are strongly reflected in soil color. Soil colors may
flat-topped and to present such evidence of high surface be observed where land has been cleared for agricultural use or
permeability as local aridity and infertility. In contrast to LC presents occasional bare spots for other reasons. Brammer
areas the sharply defined break at the heads of encroaching (1962) reports a change in color in typical forest soils of
gullies is usually subdued or missing as is the concave profile of Ghana from red for the best drained soils in which the iron is
the slopes. completely oxidized and not hydrated, to brown and yellow in
In northern Ghana some concretionary gravel is usually imperfectly drained soils in which the iron becomes increas-
ingly hydrated and finally to gray or white in poorly drained ground view of the features, indicates that it is composed of
soils which are perennially waterlogged so that the iron has thin-bedded sandstone. Plates 1 and 2 also illustrate the type
been reduced to the ferrous state or completely leached from of farming carried on in areas of lateritic forest soils. Several
the soil. borings made in the area failed to reveal any concretionary
Tropical black clays rich in montmorillonite have attracted materials. The size and number of fields seem to indicate that
considerable interest because of the special engineering pro- the soil is at least fairly fertile.
blems which they present. They are discussed in detail in In the vicinity of the Afram River crossing relatively clear
Chapters 15 and 16. Where black clays occur as residual soils, photography (Plates 4 and 5) was available for terrain having
they are nearly always associated with basic bedrock materials. concretionary laterite. A comparison of Plate 4 with Plate 1
Topographically, occurrences are generally located in positions shows a distinct contrast in the color of both soil and
where soil drainage tends to be impeded. Thus, black clays are vegetation, as well as some apparent differences in the
found in areas of gently rolling topography or in flat intensity of agricultural use. Swidden type agriculture seems to
depressions, but seldom in areas of mountainous relief. Natural predominate in the area shown in Plate 4 whereas the fields
vegetative cover is usually savannah and not forest, even if the shown in Plate 1 appear to be more permanent. The river can
soil is located in an area where much forest cover is present. In be seen in only a few locations but it and its tributaries are
many tropical areas these soils are farmed intensively. Even well-marked by the heavier tree growth. The large borrow pit
though shrinkage cracks may attain considerable size under at B contained clayey laterite gravel below 3 ft as shown in the
dry conditions, Liang reports that they are not apparent in ground photo (Plate 6). Laterite gravels were also found at the
photographs at a scale of 1:20,000 of areas in Colombia and surface at C and below 16 inches at D.
Burma. In areas of Africa with longer dry seasons and sparser The study of color photos of the two types suggests several
vegetation photo identification may be easier. Certainly the preliminary conclusions. Open water can be identified more
use of larger-scale color photography should make possible readily on infrared film, but in the area studied the presence of
their identification. water courses can be identified about equally well in both
In most tropical areas unweathered country rock is seldom types by the density and darker color of the vegetation.
found within many feet of the surface, but there are Inselbergs are more prominent in infrared color because of
exceptional occurrences of isolated rock outcrops or insel- blacker shadows, but positive identification depends on
bergs. Such features frequently appear as isolated "rock stereoscope coverage so that the terrain can be seen in
castles" which rise some tens of feet above the surrounding three-dimensions. Black and white stereo photos would pos-
terrain and present a relatively unweathered surface of bare sibly permit as ready identification of such features. Stereo
bedrock. In most cases, these features seem to bear little coverage is essential if shapes of slopes, presence of stream
relation to the surrounding deeply weathered terrain. They are terraces, and locations of peneplain remnants are to be
useful sources of aggregates and coarse material if they can be identified positively. Both natural and infrared color have
found. Most inselbergs can be easily identified in aerial advantages over black and white in revealing subtle differences
photography by their light appearance and contrast with their in the condition of the vegetation. Such differences are often
shadows. soil-related.
Infrared color is preferable to natural color in the identifi-
COLOR PHOTOGRAPHY FOR PRESENT STUDY cation of vegetation which is diseased or starved for lack of
nutrients or water. On the other hand, natural color photos are
Color aerial photography, both natural color and infrared, more useful for observing tonal changes in the surface soil. As
along the highway from Mampong to Ejura in Ghana was is the case with all airphoto interpretation, the amount of
taken primarily to determine whether differences between information which can be extracted from photographs in-
areas of ferrisols and of ferruginous soils could be detected. creases rapidly as photo to ground studies provide the data for
Much of the area was obscured by clouds at the time of the more meaningful correlations. Undoubtedly, much can be
flights, and as the length of traverse was apparently too short accomplished with monochrome photos alone since with
to encompass significant variations, the experiment was stereo coverage the pattern of terrain shapes can be seen as
inconclusive. On the other hand, features of special interest well in gray tones as in color. Only further comparative studies
were noted in some of the clearer photos and in a few cases it of all three types of photography can reveal the relative
was possible to compare the findings with field site examina- advantages of each.
tions.
Plates 1 and 2 illustrate the appearance of an inselberg (at Because of the scarcity of detailed geologic and pedologic
A) in natural color and color infrared photography. The data throughout much of Africa it seems obvious that terrain
feature is more prominent in the color infrared photography analysis techniques based primarily on photo identifiable
because of the intense blackness of the shadow. Shadow patterns are likely to provide the most logical basis for
intensity is less in the natural color photograph because of the correlating soil engineering data from one location and for
response of the film to diffused blue light. Plate 3, which is a extrapolating it to another. It is anticipated that the use of
such techniques will increase with increasing number of side of the large borrow pit located at B on the airphotos.
engineering projects in a given area. Material to a depth of 3 feet is a brown clayey silt. Below this
depth is a reddish brown clay gravel with lateritic nodules up
Description of Plates 1, 2 and 3 t o 112" in size. The large rock by the shovel is sandstone.
Isolated rock outcrops or inselbergs are recognized in aerial A coppery tone, which may be seen in several parts of the
photography by their stark white appearance (A). Shadows photograph between the river and the boundary of the Forest
cast are somewhat more noticeable in the infrared photo- Reserve, was thought to be associated with lateritic gravel and
graphy. The ground view indicates a rather thin bedded ground investigation was made at C. This area was found to be
sandstone which is thought to be the lowest member of the covered by a rank growth of 5 foot grass. Lateritic pea gravel
Voltaian formation. It is thought that this material forms the was found in the first 16 inches of the soil profile, but below
uppermost bedrock under most, if not all, of the area this was a chocolate brown silty clay t o 72 inches.
photographed. While this material, if crushed, may have some Investigation of the light-toned area at D again revealed tall
value as aggregate, it would possibly be inferior t o well grass growth. Boring revealed a brown sandy silty clay to 16
indurated lateritic soil. In any event, outcrops of this material inches underlain by brownish lateritic iron-pan fragments t o
are infrequent and do not appear t o occur even on steep slopes 60 inches.
near streams in the area photographed between Mampong and The Afram River is the o n b stream in the area photo-
Ejura. graphed in which exposed water surfaces can easily be seen.
The somewhat lighter appearance of the bare soil in the While most of the stream in the infrared photograph is
community of Nkwanta compared to that in Bobin and the obscured by clouds, an occasional view is available (A) and it
somewhat stippled and arid appearance of terrain along the may be noted that the blue of the water contrasts strongly
road and in the vicinity of B suggests that the village and its with the pink-magenta of the surrounding vegetation. The
immediate surroundings are located on an unusually sandy soil small pond at E could easily go unnoticed in Plate 4 but shows
to some depth. Field examination of the surface indicated a as a distinct blue patch in Plate 5 even though light cloud
fine sand, although individual grains appeared to be of a rather cover.
soft material which might break down easily to silt. A boring
alongside the highway at the village revealed a dark brown REFERENCES
sandy silty clay to 12 inches, underlain by 3 6 inches of reddish
Dowling, J. W. F., 1968. The classification of terrain for road
brown sandy clay which was unusually moist and contained
engineering purposes. Paper 16, Conf. on Civil Eng. Prob.
bits of sandstone. Thus the lighter appearance may be due to
Overseas, Inst. of Civil Eng., London: 58 p.
somewhat poorer drainage in the locality.
Dowling, J. W. F., and Beaven, P. J., 1969. Terrain evaluation
The extensively cleared and farmed agricultural lands
for road engineers in developing countries. Jour. of the lnst.
around Nkwanta represent a more or less typical lateritic
of High. Eng., June: 5-15.
forest soil. About one foot of dark brown sandy silty clay was
Liang, T., 1964. Tropical soils: characteristics and airphoto
found to be underlain by reddish or reddish brown sandy clay.
interpretation. Cornell Univ., Rept. AFCRL-64-937: 163 p.
No evidence of concretionary or indurated lateritic soil was
Wills, J. Brian (editor), 1962. Agriculture and land use in
found in the area.
Ghana. Oxford Press, London.
Description of Plates 4, 5 and 6 Reference was made to the following chapters:
The photographs taken at the Afram River Crossing show Bates, D.A.; Geology; Chapter 3: 5 1-61.
an area where extensive deposits of concretionary lateritic Brash, Helen T.; Geomorphology ;Chapter 5: 77-87.
material are present. The ground photograph illustrates the Brammer, H.; Soils; Chapter 6: 88-126.
 CHAPTER 8

SUITABILITY OF CURRENT ENGINEERING CLASSIFICA TION

SYSTEMS FOR TROPICALL Y WEATHER ED SOILS

INTRODUCTION The distribution of the AASHO classification is more

Two of the soil classification systems most widely used uniform. Twenty-three percent fall in the A-2-6, 19 percent in
today were developed and based on soils from temperate A-2-7, 17 percent in A-7-6, 13 percent in A-6, and 11 percent
zones. The AASHO classification system had its origins in in A-7-5.
1929 (Hogentogler and Terzaghi) when it was known as the More than three-quarters of the samples are concentrated
Public Roads (PR) classification. This system has undergone into 2 groups in the USC system. The SC and SM groups are
many revisions; the latest modification (AASHO M 145-661) each spread into 7 AASHO Classifications. Fifty-two percent
involved changes in the Group Index which was originally of the samples fall into the A-2 group and 41 percent into the
added t o the system in 1945 (HRB). The Unified Soil A-6 and A-7 groups.
Classification System was developed in 1942 and was then Generally, if lateritic soils contain gravel, the amount
called the Airfield Classification (AC) System or the Casa- passing the No. 200 sieve is less than 35 percent and the
grande System after its author. In both of these systems, the classification is usually in the A-2 group. If there is little or no
grading and Atterberg limits are the basis for classification. gravel, more than 35 percent passes the No. 200 sieve, and the
The AASHO system classified soils into 12 major groups and, soil is classified as either A-6 or A-7. The majority (93%) of
within each, the group index further classified the fine-grained lateritic soils tested fall within these groups. Soil derived from
materials on the basis of variations of grading and Atterberg sandy parent materials fall into the A-1 or A-4 groups. No
limits. The Unified System classifies soils into 15 major lateritic soils were found in the A-3 or A-5 groups.
groups. Lateritic soils are known t o possess high strengths in
The D'Hoore (1964) pedological system, which reflects the relation t o the plasticity, higher than those of soils in
temperate regions. The CBR strengths found in t h s study were
climate, drainage, topography and parent material, differen-
much higher than typically reported values for temperate
tiates among the three major lateritic soil groups. This has
zones. The difference may be due t o determination of the
been discussed in Chapter 3. However the range in properties
CBR values at 100 percent modified compaction and the
within one group is too great for an engineering classification
values at lesser compactive efforts. The engineering
and a more detailed system is required. The D'Hoore system
s of a material are the best indication of the value of
provides at best a meaningful terminology and suggests a
general range of properties in each group. erial when used in road construction. Chapter 12
The AASHO classification system is convenient as a basis shows that the CBR value is a good indication of the expected
performance of a material in a road. Figures 55 and 5 6 show
for classifying the tropically weathered soils of Africa because
it is understood and used by all the African Public Works that there is a better relationship of the CBR with the AASHO
Departments. The basic structure of the system is closely tied system than with the USC system.
t o highway requirements. The Unified System possesses more It has been suggested by Vallerga et al. (1969) that the
major groupings, but the AASHO system with its group index Unified system be modified by including durability (Chapter
allows relative evaluation within a given group. 6) as a physical characteristic. They propose adding one
character t o the G (gravel) or S (sand) series of symbols. The
APPLICABILITY OF EXISTING proposed symbols would indicate high durability as "1";
CLASS1FICAI'ION SYSTEMS medium as "2"; and low as "3". The symbol "1" indicates a
durability more than 35; "2" a durability between 20 and 35;
A total of 221 lateritic soils were classified in accordance and "3" less than 20. They found for Thailand that durability
with the Unified and AASHO systems. The soils came from was a good measure of performance. The results of the present
Ghana and other African countries and were distributed study on soils in Ghana and other African locations did not
among the three soil groups. One-thud of the soils were show any relationship between durability (either with G or S
sampled during the roadway condition survey and the remain- soils) and performance (Chapter 12) or with CBR.
ing in connection with the physical properties testing program. The AASHO system rates a soil within a group by the
The results of these classifications are shown in Table 15. It Group Index. The new Group Index (AASHO M-145-661) is
can be seen that according t o the Unified Classification determined from the same formula originally proposed, but
System, 77 percent of the samples are classified as SC or SM; with no upper limit for LL, PI, and percentage minus 200.
60 percent are designated as SC. Almost all of the SM soils
were so classified because they plotted below the "A" line of Group
the Plasticity Chart. With a small increase of plasticity, the SM Index = (F-35) [0.2 + 0.005 (LL-40)] + 0.0 1 (F-1 5)(PI-10) in
soils would be classified as SC. which:
 F = percentage passing No. 200 sieve, expressed as a whole This is a multiple regression equation, with 0.2, 0.005 and
number. This percentage is based only on the materiai 0.0 1 as correlation coefficients.
passing the 3-in. sieve. These same variables, with the above lower limits for PI, LL
LL = liquid limit and minus 200, were re-examined for new coefficients using a
PI = plasticity index computer for the multiple regression. These new coefficients
The group index was calculated for all soils and compared were then applied in calculating new Group Index values for
to the CBR (Fig. 57). The correlation coefficient between the each soil. The new Group Index values were then compared
two factors was about 0.40. with CBR. There was no significant improvement in correla-
The pavement condition survey indicated (Chapter 12) that tion.
higher Atterberg limits and a larger percentage passing the No. Further attempts at improving the correlation were no
200 sieve could be allowed for lateritic than for temperate more successful. The lower limits of LL, PI and minus 200
zone soils. The AASHO classification system was examined to were varied (from LL-40; PI-10; F-35; and F-15). The best
determine if a more descriptive classification would result by correlation occurred when the lower limits for LL and PI
modifying these limits. approached 0 , and as the minus 200 approached 20, producing
It was believed that the coefficients in the Group Index the equation:
equation might not be fully applicable to lateritic soils. GI = (F-20) [0.2 + 0.005 (LL)] + (F-20)(PI).
The Group Index equation can be restated: However, the correlation with CBR was still low, not
GI = 0.2a + 0.005 ac + 0.0 1 bd justifying any change in either coefficients or lower limits.
where: a = F-35 in the equation above The AASHO system was also examined to determine if the
b = PI-40 limitations for each group should be altered; that is, if the
c = LL-40 minus No. 200 sieve value of 35 percent, the LL of 40 or the
d = F-15 in the equation above PI of 10 could be relaxed to reflect the acceptable higher

TABLE 15

Breakdown of the Unified Soil Classification

and AASHO systems for Lateritic Soils

- - - - - - - --- - - -- - -

AASHO % of
USC - -
A- 1 A-2-4 A-2-6 A-2-7 A-4 A-6 A-7-5 A-7-6 Total Total

GC

GM

SW-SC

SC

SM

ML

CL

MH

CH

Total 5 23 51 42 10 28 25 37 22 1

% of
Total 2% 10% 23% 19% 5% '1 3% 11% 17%
 LEGEND 4
LEGENDE 4
A FERRUGINOUS W
FERRUGINEUX LL
0
0 FERRALLlTlC
FERRALLITIQUE 0 2
0 FERRISOLS $2 ::
FERRISOLS a2 2
w2
"a! t
E w

8"
e
2
I

c!!
WU)
Z
a
0
ax g
$5 k
w- 0
5t I

I5 P
gz 2
a
$2 t
(DO

43
a e
b-

"g a
3'E

AASHO CLASSIFICATION
CLASSIFICATION A ASH0
FIG.55 C B R VS. AASHO CLASSIFICATION
C B R EN FONCTION DE L A
CLASSIFICATION AASHO GROUP INDEX

FlG.57 C B R VS. GROUP INDEX

C B R E N FONCTION DE L'INDICE DE GROUPE

limits for lateritic soil. The results of the pavement condition

survey were employed (Chapters 12 and 13). These limits were
A
LEGEND
LEGENDE I1 increased t o values as high as F = 40; L L = 50; and PI = 20.
1.0 1I l A @

1
1

I
A
•
FERRUGINOUS
FERRUGINEUX
EeR~LLITlc
FERRALLITIQUES
Again no significant improvement in correlation with CBR
occurred.
0 FERRISOLS
FERRISOLS
These results do not nullify the findings in Chapter 5 with
respect to the CBR and the AASHO classification. It was
stated there that a broad relationship exists between CBR and
the AASHO groups and that a minimum value of CBR can be
predicted with reasonable accuracy. The present investigation
was to determine if a more precise relationshp could be
established which would result in a more descriptive
classification.
Any classification system is only a guide to the range in
properties and general behavior of a soil. Variations in
properties in lateritic soils, as with temperate-zone soils, are
such that completely definitive classifications are not possible.

REFERENCES
UNlFlE D SOlL CLASSlFlCATl ON American Association of St ate Highway Officials (AASHO) ,
SYSTEME DE CLASSIFICATION UNlFl E E DES
SOLS 1966. Interim recommended practice for the classification
FIG.56 CB R VS. UNIFIED SOlL CLASSIFICATION of soils and soil-aggregate mixtures for highway construct-
C B R EN FONCTlON DU SYSTEME DE CLASSIFICATION
UNlFlEE DES SOLS ion purposes. Desig. M145-661.
Casagrande, A., 1948. Classification and identification of soils. Hogentogler, C.A., and Terzaghi, K., 1929. Interrelationship of
Trans. Amer. Soc. of Civil Eng., vol. 1 13: 9 0 1-930. load, road and subgrade. Public Roads, May.
D'Hoore, J. L., 1964. Soil map of Africa scale 1 t o 5,000,000, Vallerga, B. A., Shuster, J. A,, Love, A. L., and Van Til, C. J.,
explanatory monograph. Comm. for Tech. Co-op. in Africa, 1969. Engineering study of laterite and lateritic soils in
Publ. 9 3 : 205 p. connection with construction of roads, highways and
Highway Research Board, 1945. Classification of highway airfields. Phase I - Southeast Asia (Thailand). Final Report
subgrade materials. Proc. High. Res. Board, vol. 25: for U.S. Agency for International Development, Contract
376-392. AID/csd-18 10. Soil and Pavement Consultants, Oakland,
California: 165 p. and Appendices.
 CHAPTER 9

STABILIZA TION OF SELECTED AFRICAN LA TERITIC

GRAVELS AND SOILS

INTRODUCTION MATERIALS
In African countries where economic development is Three soils from each of the lateritic soil groups were
based primarily on agriculture, a network of major highways, selected from various borrow pits in Ghana. The locations
secondary roads, and low-cost feeder roads is of utmost are shown on the map, Fig. 58. The two Congo soils were
importance. Economy requires that local materials be used in selected from sandy materials being assessed in a current
the construction of these roads. Yet, untreated lateritic design study for use as base course for a highway. A clayey
materials have presented many problems in road construction sample from Ivory Coast was selected as a subgrade-type
and maintenance. Consequently, for the past three decades material. All other materials were selected from convenient
stabilization of lateritic materials by admixtures has been borrow pits in their respective countries. The physical and
practiced and studied in Africa. chemical properties of the tested materials are listed in Table
The aim of stabilizing soil is t o alter its physical prop- 16.
erties, increase its strength and durability and thus provide a The lime and cement used in the study were all obtained
satisfactory foundation material. The admixtures most com- from local sources. The cement is Type I Portland cement,
monly used today in road construction throughout the world made locally in Ghana, which met ASTM specifications for
are cement, lime asphalt and sand. Cement appears t o be the mortar strength and setting time. Lime is not produced
most common additive used in Africa; its use has been locally in Ghana so an imported brand bras used for the lime
reported in Ghana, Gambia, Sierra Leone, Ivory Coast, stabilization. A typical coastal sand from Ghana was used for
Nigeria, Kenya, Uganda, Tanzania, Zambia, Southern the admixture in the sand stabilization studies.
Rhodesia and South Africa. The use of lime has been
reported in Uganda, Kenya, Zambia, Nigeria, Basutoland and PROCEDURES FOR LIME AND CEMENT
South Africa. The use of asphalt in Africa has been limited
t o lateritic soils, of a sandy nature. Stabilization by ad- &r dried soils were used in all of the stabilization tests.
mixture of sand has been investigated and used in ~ i ~ ~ ~ flow
The i ~chart
, for the lime and cement tests is presented on
but because of the limited availability of sand in Africa has Fig. 59.
not been widely used.
Compaction
SCOPE Moisture-density relationships were investigated in ac-
To evaluate the use of lime, cement, asphalt and sand as cordance with ASTM D l 557-70 Method B. Recent studies
stabilizers for lateritic materials from Africa, a series of (Arman and Saifon, 1967), (Mitchell and Hooper, 1961)
studies was carried out. A total of 20 representative soil have shown that the density and strength show a con-
materials were selected for this purpose, nine from Ghana, siderable decrease if there is a delay between mixing and
two from the Congo, and one each from Ivory Coast, Kenya, compaction. This decrease may occur if the delay is longer
Uganda, Dahomey and Niger. The materials from Sierra than 2 hours for soil-cement mixtures and 24 hours for
Leone were tested at Fourah Bay College. Four samples, one lime-soil mixtures. It was decided t o compact specimens
from Niger and Dahomey and two from the Congo, were containing either stabilizer immediately after mixing.
sent t o the Asphalt Institute Laboratories in College Park, The required amount of stabilizer was blended with the
Maryland for asphalt stabilization testing. soil in a mechanical mixer. Water was then added and
The materials from Ghana, Congo and Sierra Leone were thoroughly mixed in the mechanical mixer. Immediately
stabilized both with lime and with cement, whereas the soils after mixing, the soil was compacted in the mold. Any
from Ivory Coast, Kenya, Uganda and Dahomey were sta- excess material was sealed until needed for compaction.
bilized with lime only. The material from Niger was stabi- After being weighed, the sample was extracted from the
lized with cement only. One clayey gravel from Ghana was mold and placed on a pallet in a sealed plastic bag in the
stabilized with sand. curing room. Water was added t o each bag t o maintain
A quick and simple method has been devised b y Eades approximately 100 percent relative humidity.
and Grim (1966) for determining lime requirements for soil
stabilization. The method can also be used for determining CBR
cement requirements. It was adopted for this study t o After determination of the optimum moisture content
investigate its applicability for lateritic soils. (OMC) for each stabilizer content, two CBR samples were
 .. Hurnic
,H. ..Humique

prepared in accordance with ASTM D1557-70 Method D at from the curing room and soaked in water for 4 days under
the optimum moisture content. The samples were placed in a a 10 pound surcharge weight. Readings were taken before
storage box on a layer of sand that was kept moist and and after soaking to determine swell. After the soaking
moist sand was placed in the collar extension of the molds. period the. CBR's were determined in an automatic testing
One sample was cured for 7 days and the other for 28 days. machine. On account of the limitation of the testing a p
After their respective curing periods, the molds were taken paratus, .the maximum tested CBR was 500%.
 TABLE 16

Properties of Soils Investigated

Torma- Free- Kaba- Bat-

PROPERTIES 65-1 G 1-14 bum town la kanu
D'Hoore BcJd JdBc LC Lm LC LcNa
Gravel60mmto4.5mm 45 30 0 16 39 50
Sand 4.75 mm to .074 mm 35 60 76 35 33 29
Silt .074 mm to .002 mm 12 5 10 16 14 8
Clay 0.002 mm 8 5 14 33 14 13
Physical Properties
Liquid Limit % 29 23
Plastic Limit % 18 12
00
\O Plastic l ndex 11 11
Specific Gravitya 2.87 2.90
Max Dry Densitybpcf 136 140
Opt. Moist Contentb% 8.9 7.2
Portuguese Swell % 13.3 8.8
Classification
Unified SC SM-SC SM-SC SM SM SM
AASHO A-2-4 A-2-4
(0) (0)
Chemical Properties
PH 6.6 6.5
Cation Exc. Capacity
MegI100 gm 7.23 4.37

a. On minus No. 10 Fraction b. AASHO Modified

 ATTERBERG L I M I T S L l M l T E S D' A T T E R B E R G
CHEMICAL ANALYSES CHlMlQUES
MINERALOGICAL ANALYSES MINERALOGIQUES
AIR DRIED
+
2 % DE A D D 2 Oo/ SAMPLE
STABILISATEUR STABILIZER

SSCHE A L'AIR
CBR 4 DAY SOAK C B R IMBIBITION DE 4 JOURS
+41 ,
'0 6 Oo/ , 8 % ADD 4 % , 6 % , 8 %
STABILISATEUR STABILIZER

RELATION TENEUR COMPRESSlON

EN E A U - DENSITE RELATIONSHIP COMPRESSION SIMPLE
ASTM 1 5 5 7 - 7 0 ( 8 ) ASTM D 2166 - 6 6

1
12 E C H A N T I L L O N S
2 SAMPLES 4 SAMPLES 2 SEALED 12 S A M P L E S COMPACTAGE STATIQUE
2 ECHANTILLONS
ASTM 1 5 5 7 - 7 0 ( D l
ASTM 1 5 5 7 - 7 0 ( 0 )
@ 100% DENSITY
MOISTENED
SAMPLES
I
1 STATIC COMPACTlON
@ 1 0 0 % DENSITY B 0 M C I A DENSITE 1 0 0 % E T TENEUR
E N E A U OPTIMUM

d DENSITE 100% + 2 ECHANTILLONS

HUMIDIFIES
SCELLES
I 1--
7 1 JOURS 28' 28 7 r,---- 2 8 l 28 7 ry- 28'28
1 I I 1 I
DAY
t
JOURS
+ DAY DAY JOURS JOURS DAY DAY JOURS JOURS DAY

I /
I
WET-DRY T E S T 3 SAMPLES
12 C Y C L E S LIMITS SPLIT - TENSILE
COMPRESSION
ASTM D 5 5 9 - 6 5 ( 8 ) 8 S. 1 3 7 7 TEST

E S S A I HUMIDIFICATION LlMlTES 3 ENCHANTILLONS 3 ENCHANTILLONS

SECHAGE - I 2 C Y C L E S D' A T T E R B E R G COMPRESSION S I M P L E ESSAl DE T R A C T I O N
ASTM D 5 5 9 - 6 5 ( 0 ) 8,s. 1377 ASTM D 2 1 6 6 - 6 6

CBR T E S T
ASTM D1194-66

FIG 59 F L O W CHART FOR L I M E AND C E M E N T S T A B I L I Z A T I O N

SUITE DES ESSAIS DE STABILISATION A L A CHAUX E T AU CIMENT
Wet-Dry Tests SAND STABILIZATION
Samples for wet-dry tests were prepared and tested in Compaction
accordance with ASTM D l 559-65 Method B. Two samples To compare the results of lime and cement with sand
were prepared for the 7-day and for the 28-day curing stabilization, compaction samples with sand admixtures were
periods. After extraction from the molds, each sample was prepared in the same manner as with lime and cement. The
laced on a pallet in a ~ealedplastic bag and then placed in mixtures contained 10, 20, 30, 40 and 50% sand by weight.
the curing room. Water was added t o each bag. Both samples
of a pair were subjected to 12 cycles of wetting and drying; CBR
one sample was used for the brushing test and the other to
measure volume change. Samples were molded at the optimum moisture content
and 100% modified AASHO compaction for each sand
content. CBR's were performed on the samples after 4 day
Unconfined Compression and Split-Tensile Tests soaking with a 10 pound surcharge.
Past research (Thompson, 1965) has shown that the
ASPHALT STAB1 LlZATlON
split-tensile strength is closely correlated with unconfined
com~ressivestrength" for lime-soil mixtures. To see if this All samples were prepared and tested at the Asphalt
correlation would be similar for lateritic stabilized soils, Institute. Results of these tests are presented in the ap-
split-tensile tests were performed along with unconfined pendix.
compressive tests.
All speciments were prepared in 2 by 4 inch molds by RESULTS AND DlSCUSSlON
static compaction at the modified AASHO maximum density Lime and Cement Stabilization
and optimum moisture content for each stabilizer content.
COMPACT1 O N
Twelve samples were prepared for each stabilizer content:
three specimens each for unconfined and split -tensile tests at The changes of maximum density and OMC for those
7 and 28 day curing periods. The specimens were placed on soils stabilized with lime or cement are similar to those
a pallet in sealed plastic bags in the curing room for their reported for other soils, including those in temperate zones
respective curing periods. Water was added to each bag. and other lateritic soils. The maximum density decreased
Loading strips 0.25 inches wide, which gave a width to with an increase of lime content. With the addition of 2%
diameter ratio of 118, were used for the split-tensile test. lime, the maximum density decreased from about 1 to 6
pounds per cubic foot. With additional lime the maximum
densities decreased further, up to 8 pounds per cubic foot
Atterberg Limits less than the natural maximum density. The OMC increased
Two samples were prepared in sealed plastic bags at each from 1 to 2 percent with 2% lime, but further addition of
stabilizer content, one each for 7 and 28 day curing periods. lime produced little increase in most soils.
The samples were prepared at about the plastic limit (PL) The changes in maximum density and OMC for soil-
and water contents were checked during the curing periods. cement were smaller than for lime. The most significant
Water was added when necessary to allow proper action change took place with the addition of 2% cement, but with
between the stabilizer and soil. A minimum of three plastic increased cement contents the changes were small. Except
limit determinations were made for each sample. Liquid for three soils (G9-3, G12-2, Kabala), the maximum densities
limits (LL) were determined in accordance with B.S. 1377 and OMC did not vary by more than 3 pounds per cubic
(see Test Procedures). foot and 2 percent moisture.
All samples were cured at approximately 100% relative The maximum density-optimum moisture relationships for
humidity at temperatures varying between 2 1" and 23°C. All all of the soils tested at the BRRI laboratory are illustrated
CBR, unconfined compression and split-tensile tests were on Figs. 60 and 61, and tabulated in the Appendix.
performed at a strain rate of 0.05 inches per minute.
ATTERBERG LIMITS
The results of the Atterberg limit tests are shown in
pH METHOD tables in the Appendix and the change in the Plasticity
The pH determination for lime and cement requirements Indexes are illustrated on Figs. 62 to 79.
was performed at the University of Illinois. The test mea- Almost all research on lime stabilization has indicated the
sures the amount of cement or lime consumed by a soil after ability of the lime to alter the plasticity of the soil. Addition
one hour. The pH of cementllime-slurries is measured to of lime reduces the plasticity index. The amount of re-
determine the minimum percent of stabilizer required to duction varies directly with the quantity of lime added and
maintain a pH of 12.40 for lime and 12.20 for cement. with the type of clay mineral and organic matter present in
 F E R R A L L I T I Q U E F E R R U G I N E U X F E R R A L L I T I Q U E
F E R R A L L I T I C F E R R U G I N O U S F E R R A L L l T l C

F E R R A L L I T I Q U E F E R R A L L I T I Q U E
F E R R A L L I T I C F E R R A L L l T l C
L E G E N D E L E G E N D

N A T U R E L -N A T U R A L

% C I M E N T 2----- O/ o CEMENT

% CHA U X 2 . . .......... % LIME

F E R R A L L I T I Q U E F E R R U G I N O U X
F E R R A L L I T I C F E R R U G I N O U S

100
I
1206
~
12 14 16 18 19 8 10 12 14

MOLDING WATER C O N T E NT - PERCENT T ENE UR EN EAUX DE MOULAGE - POUR C E N T (%)

FIG 6 0 R E L A T I O N S DENSITE MAXIMUM-TENEUR EN EAU OPTIMUM

MAXIMUM DENSITY - OPTIMUM MOISTURE RELATIONSHIPS
 F E R R U G I N E U X

F E R R U G I N O U S

F E R R I S O L

F E R R A L L I T I Q U E

F E R R A L L I T I C
130

126

122

118 -- P
Ap
-
-
118

~
I
\ \I

- i I
114 - -- ..
.-. - -- - 114 .
.. -- - -

--
I
i --
114

I
I

110 110 I 110 1

12 14 16 18 10 12 14 16 9 II 13 5

MOLDING WATER CONTENT - PERCENT T E NE UR EN EAUX DE MOULAGE - POUR C E N T (%)

L E G E N D E L E G E N D

NATUREL -N A T U R A L
O/o C I M E N T 2---- % CEMENT

% C H A U X 2 - % L I M E

FIG 61 RELATIONS DENSITE MAXIMUM-TENEUR E N EAU OPTIMUM

MAXIMUM DENSITY OPTIMUM -
MOISTURE RELATIONSHIPS
 F E R R A L LlTlQUE
FERRALLlTlC
100.
LEGEND
CBR-CALIFORNIA BEARING
\ I I
RATIO
P I - P L A S T I C I T Y INDEX
q u 4 " - UNCONFINED COMPRES-
SION STRENGTH - 4
INCH DIAMETER SAMPLE
qu 2"-UNCONFINED COMPRES-
SION STRENGTH - 2 INCH
DIAMETER S A M P L E
ST - S P L I T T E N S I L E STRE-
NGTH

WO.LOSS-LOSS OF WEIGHT 60
AFTER 12 CYCLES OF
WET-DRY TEST j
--------- 7 DAY CURE L 45,

2 8 DAY CURE ' '01

LEGENDE 750
t
CBR -CALIFORNIA BEARING
RATIO
PI - I NDICE DE PLASTICITE
2; 450
qu4"-RESISTANCE A L A COM- 2
PRESSION SIMPLE
DAMETRE DE LEPROU- 3091
VETTE : 4 POUCES

qu2"-RESISTANCE A L A COM-
PRESSION S l MPLE
DIAMETRE DE LIEPROU-
- . - -

VETTE : 2 POUCES 1 I

S T -RESISTANCE A LA
TRACTION .- 400 -
Y)
W.DLOSS-PERTE DE POI DS (5
APRES 12 CYCLES DE I -
MOUILLAGE -SECHAGE 300
P
--------CURE 7 JOURS 2
C U R E 2 8 JOURS u 200-

i'----

LlME
CHAUX
STABILIZER O/o

STABlLl SATEUR %

FIG 63 STABILIZATION TEST RESULTS

LlME
CHAUX
STABILIZER %
STABl L l SATEUR %

FIG 6 2 STABl LlZATlON TEST RESULTS

 L lME
CHAUX STABILIZER % EzK LIME
STABILIZER %
STAB1L l SATEUR %
CEMENT
C~MENT

STABILISATEUR '10

FIG 64 STABILIZATION TEST RESULTS FIG 65 STABILIZATION TEST RESULTS

 LIME CEMENT
CHAUX STAB lLlZ ER OIO CIMENT
LIME CEMENT STABILISATEUR %
CHAUX STABILIZER % CIMENT
STAB1 Ll SATEUR %
FIG 67 STAB1 I-IZAT I ON TEST RESULTS
FIG 66 STABILIZATION TEST RESULTS
 CON- , FERRALLITIQUE
FERRALLlTlC C-2
FERRA LLlTlQUE
FERRALLlTlC

LIME CEMENT LIME CEMENT

CHAUX STABILIZER % CIMENT CtiAux STABlLiZER % ~~MENT
STABlLl SATEUR % STAB1 L l SATEUR %

FIG 68 S T A B l L l Z A T l ON TEST RESULTS FIG 69 STABILIZATION TEST RESULTS

 L l ME
c~nux STABILIZER % C',rJX STABILIZER % ggg
STABl L l SATEUR % STABl LISATEUR Ofo

FIG 7 0 STABl L l Z A T ION TEST RESULTS FIG 71 STABl LIZAT ION TEST RESULTS
 LIME
c~aux STABILIZER %
CEMENT
C~MENT kAth STABILIZER O/o
CEMENT
C~MENT
STABILISATEUR % STABILISATEUR

FIG 72 STAB1 LlZATl ON TEST RESULTS FIG 73 STABILIZATION TEST RESULTS

 LIME r.E ME NT
CHAUX STABILIZER % C~MENT
STABl LISATEUR %

FIG 74 STABl Ll ZAT l ON 'TEST RESULTS FIG 75 STABlLlZAT l ON TEST RESULTS

 FERRUGINEUX
K-
FERRUGINOUS
75
5

60 -
4

-,
0

I-I
45- -3 N

5
V) 30 - -2 Y
~3

750 50
- 45
g600 - - 40
a - 35
N 450 -
3
(J

300 -

I I

I
- 30
- 25
,3001 - 2oN5
-.
--__ z
d
rn
3 200- C- ,,*+--- -15

__---
C-
v

-__---
_i-- 10

too - 5

LIME
CHAUX
STABILIZER O/o
STABILISATEUR %

FIG 76 STABILIZATION TEST RESULTS

CHdUX
L lME STABILIZEROh E,":,"!
STABILISATEUR %

FIG 77 STABILIZATION TEST RESULTS

 CEMENT
LIME CEMENT STABILIZER '10 C~MENT
CHAUX STAB'LiZER 9b ClMENT STABILISATEUR '%
STABILISATEUR %

FIG 79 STAB1 LlZAT l ON 'TEST RESULTS

FIG 78 STABILIZATION 'TEST RESULTS
the soil. The liquid limit normally decreases with increasing lime. However, higher precentages of lime did not reduce the
amounts of lime (Herrin & Mitchell, 1961), although the plasticity index t o less than 9 .
converse has also been reported for some soils (Diamond &
Kinter, 1965). WET-DRY DURABl LlTY
In the tests performed for this study, almost all the In African countries where the seasonal variations of
lateritic samples tested experienced a decrease in liquid limit rainfall are severe, weathering is the major process which
with the addition of lime. The only two soils that showed an causes deterioration. In most tropical countries the durability
increase of the liquid limit were K-4 and C-2, soils with requirements of soil-cement mixtures are assumed t o be
sandy parent material. However, the increase of the liquid satisfactory if the stabilized soil is in conformance with their
limit was not as great as the accompanying increase in plastic minimum strength requirements. Moreover, Moh et al.
limit, and the plasticity index decreased accordingly. In- (1967) discussed some soils in Thailand, stabilized with
creases in the liquid limit occurred in the majority of soils cement, for which the 7-day compressive strength was 150
stabilized with cement. pounds per square inch, which possessed adequate stability
Clare and Cruchley (1 957) reported that an increase in the t o withstand the wet-dry durability test. The assessment of
liquid limit of a soil stabilized with lime may be due to the wet-dry durability properties of lime-soil mixtures by
calcium saturation of the clay particles. This explanation means of laboratory tests is difficult. Results from standard
could apply to the lateritic soils studied. The phenomenon wet-dry tests using brushed samples, when compared t o
may be caused by flocculation and cementing of the indi- criteria developed for soil-cement mixtures, indicate that
vidual particles, whereupon the soil becomes silty and stif- lime-soil mixtures have little resistance t o weathering. How-
fens so that it flows with greater difficulty in the liquid limit ever, field research has shown that lime-soil mixtures when
cup. protected from weathering by a bituminous wearing surface
There was very little difference in liquid limit between are more durable than indicated by most laboratory
the 7 and 28 day curing periods for lime or cement- durability tests (Herrin and Mitchell, 1961).
stabilized soils. To investigate the extent of durability, nevertheless, all of
The plasticity indexes of lateritic soils are higher than the 9 Ghana gravels and IV-5 were subjected to the wet-dry
normal soils with similar strength characteristics. In almost test. The results of the study are tabulated in the Appendix
all cases where comparitive data were obtained, a greater and illustrated in Figs. 62 to 79. Using the Portland Cement
reduction in the plasticity index occurred with the addition Association (PCA) criteria of 14 percent, 20 percent or 7
of lime than with cement. This finding is in agreement with percent loss of weight depending upon the AASHO soil
previous research with both types of stabilizers. Lime as classification, the experimental data show that lateritic
compared with an equal weight of cement has more calcium gravels can be made durable with the addition of 4% by
available for cationic exchange. This reaction will alter the weight of cement. One soil, G8-2, would require 6% cement
water holding capacity of the clay to form a less plastic soil. t o meet the criterion. There was greater loss of weight for
The plasticity index of the natural soils ranged from 10 the lime-soil mixtures than for the soil-cement mixtures.
t o 34. The Kabala soil from Sierra Leone. was the only one Even with 8 percent lime, most of the soils did not meet the
t o attain a non-plastic state. All the other materials tested PCA criteria.
showed a decrease of the plasticity index to less than 15, The volumetric-change specimens that were molded for
except soil IV-5 which exhibited a reduction from 31 to each test condition and stabilizer did not experience any
only 20 with 8 percent lime. extensive volume or length changes. These specimens were
The ferruginous soils and soils derived from sandy parent subjected t o the 1 2 cycles of wetting and drying with-
materials all showed a great reduction in the plasticity index out brushing.
with the addition of 2% lime. Higher percentages of lime STRENGTH
resulted in smaller reductions. The other soils showed a
steady reduction in the plasticity index with increasing lime The California Bearing Ratio (CBR) test and the uncon-
content. All cement treated materials showed steady re- fined compression tests are the most common procedures
duction. used t o assess strength gains in stabilized soils. They have
been extensively reported in research on temperate and
In previous research, Cartmell (1958) showed for a Zam- tropical soils and are used as design parameters in most
bian "laterite" the steady reduction from an original plas- African countries. The tensile strength characteristics of base
ticity index of 19 with increasing quantities of lime to zero courses, subbases and subgrades are also important consi-
at 5 percent lime. Hayter and Cairns (1966) showed that the derations in pavement design. Recent literature (Thompson
plasticity index of a Nigerian lateritic soil with an original 1965, Kennedy and Hudson, 1968) has shown that the
value of 29 did not decrease t o less than 14 with the split-tensile test can be used t o evaluate stabilized materials.
addition of up t o 8% lime. Wiggington (1967) reported a The results of these strength tests are summarized in
reduction of the plasticity index from 20 t o 10 with 2% tables in the Appendix and illustrated on Figs. 62 to 79.
CBR was a ferrallitic sandy clay from Freetown which had a CBR
For the lime-treated soils there did not seem to be any value of only 137 percent at 4 percent cement content. The
distinction in the behavior of samples from the different 3 gravels from the ferruginous group all had CBR values of
lateritic soil groups. There was some distinction with respect over 400 percent at 4 percent cement content. Some of the
to cement-treated soils, as will be discussed subsequently. soils from the ferrisol and ferrallitic groups also attained very
The CBR values for cement treated soil were much higher high CBR values at the higher cement contents. Generally
than for lime treated soil. In most cases the highest CBR most soils were not tested above a CBR of 500.
value with lime treatment was less than the 28 day CBR for Most of the countries in Africa use as design criteria a
2% cement content. In no case was the highest soil-lime CBR minimum CBR value of 180 percent for cement and lime
value more than the 28 day CBR value for 4% cement by treated soils when tested at 100% modified AASHO com-
weight. paction. O'Reilly and Millard (1969) recommend a CBR
The CBR value for most of the lime treated soils in- value of 80-100 percent at the density to which the soil will
creased linearly with lime at low lime contents, but there be compacted in the field. This requirement is for 7 day test
was a tendency toward a smaller increase or even a slight results, comprising 3 days moist curing followed by 4 days
decrease in strength with higher lime contents. Peaks in CBR immersion in water.
values occurred at about 4 to 6 percent lime by weight. On the basis of the CBR test results it appears that most
There were small increases in strength between the 7 and 28 lateritic soils can be successfully stabilized as base course
day curing periods. material with 3 to 4 percent cement. Many of the lime-treated
The smallness of the gain in strength may be due t o the soils did not reach a CBR value of 180 even with the higher
lack of pozzolanic material in the natural soil. Usually percentages of lime, but could be considered for subbase. Lime
aluminous and siliceous materials in soil react with the lime treatment could also be considered for improvement of
t o cement the soil particles. Lateritic soils are usually low in subgrade soils.
siliceous materials, which might explain the lack of increase
in strength with longer curing periods. Herrin and Mitchell U N C O N F I N E D COMPRESSION
(1961) state that neither base exchange capacity, the pH of
Unconfined compression tests were performed on the
the soil, nor the plasticity characteristics seem to be satis-
extracted 4 by 4.5 inch compaction specimens. The 4 inch
factory indicators as t o when a soil will give large strength
compressive strengths of each compaction point were plotted
gains with lime. Further research may be necessary to
against the molded moisture content and the strength at the
determine whether the addition of a pozzolan, such as fly
OMC selected for consideration. Figure 8 0 shows a typical
ash, might be helpful to aid the cementing action in lateritic
compaction curve with corresponding unconfined compressive
soils. Irwin (1958) found that after 26 weeks of curing,
strengths. It will be observed that a small variation in the OMC
lateritic gravels from Uganda stabilized with 4 percent of
could change the compressive strength considerably.
lime were approaching the 7 day strength of the same gravels
The 2 inch diameter unconfined compressive strengths for
stabilized with 4 percent of cement. Hayter and Cairns
cement-treated soils were much higher than those for lime-
(1966) reported a very marked gain in strength between 7
treated soils.
and 28 days for a lime treated lateritic soil. They did not Most of the cement-treated soils had strength gains in
give any explanation for this reaction. The long time re- proportion to the cement content. Except for the Batkanu soil
quired for the gain in strength is due to the slow cementing
from Sierra Leone, all soils attained a 28-day strength of 250
reaction. The time may be accelerated with the addition of pounds per square inch at 4 to 6 percent cement.
pozzolans. Most of the lime-treated soils did not attain 28-day
Most of the cement-treated soils in the present study strengths of 250 pounds per square inch even at the higher
developed high CBR values with low percantages of cement. lime contents. There was very little or no gain in strength from
The values generally increased linearly with increase of 7 to 28 days. The one exception was IV-5 which showed gains
cement content, although some soils exhibited smaller in- from 50 to over 100 percent. The strengths generally increased
creases at 4 to 6 percent. Sherwood and Sullivan (1960) linearly with an increase in lime content but some soils
observed smaller measures when testing a Gambian lateritic exhibited peak values at 4 to 5 percent and then decreased in
gravel at optimum moisture content. They showed that the
strength at the higher lime contents.
decreased rate was due to insufficient water being available
A decrease in compressive strength at higher lime contents
for the larger quantities of cement to hydrate fully. When
as well as complete failures with high percentages of lime were
the gravel was tested at higher moisture contents, the
reported for lateritic soils by Winterkorn and Chandrasekharan
unconfined compressive strength increased linearly with an
(195 1). Hayter and Cairns also reported little or no further
increase in cement content.
gain in strength beyond a certain point, usually in the range of
With all cement treated soils except one, the CBR values 4 to 7 percent.
in the present study were over 200 percent. The exception No distinction could be made for any of the D'Hoore soil
 RESISTANCE MAXIMUM
PEAK STRENGTH
---------
TYPICAL
TYPIOUE
CHAUX
7AND 28 DAY RESULTS
~RESULTATSAWES 7ET 28 JOURS
L'ML
A FERRUGINOUS
FERRUGINEUX
1
STRENGTH AT OMC

RESISTANCE A 700 FERRALLITIQUES

TENEUR ENEAU
OPTIMUM
I
1
I
I
I
MAXIMUM DRY DENSITY I

MOISTURE CONTENT
TENEUR EN EAU

0 10 20 30 40 50 60
FIG. 8 0 DRY DENSITY AND UNCONFINED COMPRESSIVE STRENGTH S T = SPLIT-TENSILE STRENGTH - p s i
DENSITE SECHE E T RESISTANCE A LA COMPRESSION SIMPLE RESISTANCE A L A T R A C T I O N L I V R E S I P O U C E ~

FIG.82 COMPARISON BETWEEN SPLIT- TENSILE AND UNCONFINED

COMPRESSIVE STRENGTHS
COMPARAISON E N T R E RESISTANCE A L A T R A C T I O N E T
RESISTANCE A L A COMPRESSION S I M P L E .

groups with the unconfined compression test results.

Research by Eades (personal communication) has shown
that the chemical reaction between lime and soil is time and
temperature dependent. Therefore, there is an optimum lime
percent for a given curing period.

SPLIT-TEIUSILE STRENGTH
The split-tensile test procedure (ASTM C496-64T) has been
primarily used to evaluate the tensile strength of concrete.
However, it appears to be a practical method t o evaluate the
tensile strength of lime-, cement- and asphalt-treated materials.

Thompson (1 965) reported that the ratio of split-tensile

strength to compressive strength (ST/qu) showed small but
significant differences for different soil types, and that the
best estimate of the ST/qu ratio for lime soil mixture was an
overall average of 0.13, with a range from 0.1 1 3 to 0.155.
From the test results of the present study, Figs. 8 1 and 82, it
is seen that for lime-treated soils the ST/qu ratio has an
overall average of 0.096 with a range of 0.041 to 0.153. The
cement treated soils showed an average of 0.097 with a range
0
0 10 20 30 40 50 60 of 0.040 to 0.164. The ST/qu ratios were similar for all soil
ST = SPLIT -TENSILE STRENGTH - PS i groups. There were no major differences in the ST/qu ratio
RESISTANCE A L A TRACTION LIVRES / P O U C E ~
between the lime and cement treated soils. It is apparent the
FIG.81 COMPARISON BETWEEN SPLIT- TENSILE AND UNCONFINED changes in compressive and split-tensile strengths are related
COMPRESSIVE STRENGTHS
COMPARAISON ENTRE RESISTANCE A L A TRACTION E T
within the range of stabilizer contents and curing periods used
RESISTANCE A LA COMPRESSION SIMPLE. in this program.
 On the basis of the results of this study it appears that the
split-tensile test may have application in the stabilization of
lateritic soils. An ST/qu ratio of 0.10 for lime- and cement-
treated mixtures seems appropriate. Further research would be
desirable to verify this ratio.

S T R E N G T H RELATIONSHIPS
The Road Research Laboratory (O'Reilly and Millard,
1969) established for five soil types a correlation between the
CBR and unconfined compressive strength for soil-cement
mixtures. A similar correlation has been investigated in the
present study for both lime- and cement-treated samples. This
is shown on Fig. 83 for soil-cement mixtures and on Fig. 84
for lime-soil mixtures. De Medina (1964) reported the same
relationship for laterite gravels from Ghana as the Road
Research Laboratory did for concretionary gravels. He believes
that the slopes of the lines differ from each other because of
the absence of lateral pressure during the unconfined compres-
sion test whereas such pressure exists in the CBR test.

Sand Stabilization
The strength and plasticity of a lateritic gravel are improved
with the addition of sand.
C B R

FIG.84 COMPARISON BETWEEN COMPRESSIVE STRENGTH AND C BR

COMPARAISON ENTRE RESISTANCE A LA COMPRESSION ET CBR.

I CEMENT
CLMENT
I 7AND 28 DAY RESULTS
RESULTATS APRES 7 E T 28 JOUR
900 A FERRUGINOUS
FERRUGINEUX

With the addition of sand the maximum dry density

increased up t o a percent of sand by weight equal t o 4 0 and
then decreased with higher percentages of sand. The maximum
dry density was higher for all percentages of sand, than for the
natural soil. The maximum increase in the maximum dry
density was 6 pounds per cubic foot. The optimum moisture
content decreased linearly by about 3 percent from the natural
soil t o 5 0 percent sand content, Fig. 85. In performing a
similar study on a lateritic soil from Thailand, Muktabhant and
Ongskul (1969) found that the highest density and lowest
OMC occurred with about 50 percent sand.

CBR
Generally the CBR value of the soil can be improved by
adding sand. However, the 28 day strength for the same soil
treated with 2 percent lime is about 25 percent higher than the
maximum value obtained with sand. The maximum soaked
0 100 200 300 400 500 600 700 CBR value was attained with 20 percent sand, then decreased
CBR with an increase of sand content. The CBR values for all
percentages of sand were higher than the CBR value of the
FIG.83 COMPARISON BETWEEN COMPRESSIVE STRENGTH AND C B R
COMPARAISON ENTRE RESISTANCE A LA COMPRESSION
natural soil. Muktabhant and Ongskul found the maximum
ET CBR soaked CBR at about 60 percent sand by weight.
 sion test method. Most soils had a pH lime design content of 4
GI2-2
NO SAND - PAS D E S A B L E
percent, while the test method showed from 4 to 6 percent.
10% SAND ....... 10% DE SABLE
2 0 % SAND -- 2O%DE SABLE One soil, IV-5, had a pH-method content of 3 percent
3 0 % SAND - - -- 30%DE SABLE
compared t o a test method content of 5 percent. Four soils
4 0 % D E SABLE
5 0 % DE SABLE showed a design content of 4 percent for both materials.
For soil-cement mixtures, the pH method gave higher
design contents than the unconfined compression test method.
The pH method indicated values mostly at 6 percent with 2
soils at 9 percent and 1 at 10 percent. The test method
indicated design contents from 3 to 6 percent with most soils
at 3 to 4 percent. The two Congo sandy soils had pH-method
design contents of 9 and 10 percent compared to 3 and 4
i4OISTURE CONTENT, PERCENT
percent respectively for the test method.
rENEUR EN EAU , POURCENT Test results for 2 soils, DAH-3 and NIA-1, stabilized, with
lime and cement respectively, at the pH design content as
FIG.85 EFFECT OF SANG ON THE MOISTURE-DENSITY RELATIONSHIP
INFLUENCE DU SABLE SUR LA RELATION TENEUR E N EAU- shown in Table 18. The strength increases of NIA-1 were h g h .
DENSITE The CBR increased from 78 to 500 in 7 days with 4 percent
cement. The unconfined compressive strength increased from
AT'TERBERG L I M I T S 8 0 to 360 pounds per square inch in 7 days. There was a slight
reduction of the PI and maximum dry density, with a small
The results of the data show that both the liquid limit and increase in the OMC.
plasticity index decreased linearly with an increase of sand The increase of strength was smaller for DAH-3, stabilized
content. It is interesting to note that at a sand content of 50% with 5 percent lime. The CBR increased from 16 to 129; the
the plasticity index was only 8 in contrast to a value of 22 for unconfined compressive strength increased from 1 10 to 231
the natural soil. Charernnit (1965) found that for lateritic soils pounds per square inch at 28 days. The plasticity index
in Thailand the reduction in the plasticity index is nearly reduced from 24 t o 16, while there was a slight decrease in the
proportional t o the increase in percentage of sand. He maximum density and a slight increase in the OMC.
suggested the following equation for the plasticity index of the
lateritic soil-sand mixture: ASPHALT STAB1LIZATION
(PI of mixture) = (% of lateritic soil*) (PI of lateritic soil) Asphalt stabilization studies on four sandy lateritic soils
Muktabhant and Ongskul found their test results to be in good were conducted by The Asphalt Institute, College Park,
agreement with this formula. The dashed line in Fig. 8 6 Maryland. Three ferrallitic soils were submitted, two from the
represents the plasticity index according to this formula. The Congo Basin and one from Dahomey. One sandy ferruginous
solid line represents the results obtained from the tests in the
present study. It can be seen that the actual decrease of
plasticity index was greater than the decrease found by using
the formula.

pH Method
G 12-2
BY T E S T
EXPERIENCE
BY FORMULA-
-I
--
CALCUL
The minimum stabilizer contents needed to maintain the
required pH for the lime or cement mixtures are listed in Table
17. Also listed are the design contents determined from test
results.
The design contents of the soil-cement mixtures were based
on attaining a minimum CBR value of 180 and an unconfined
compressive strength of 250 pounds per square inch. Most of
the lime-soil mixtures exhibited peak values for the range of
lime contents studied. The design lime content was considered
to correspond to the peak value even if this value was lower
than the minimum strength required for base materials. PERCENTAGE OF SAND
The pH method for lime-soil mixtures indicated design POURCENTAGE DE SABLE

contents lower than those found by the unconfined compres-

FIG.86 VARIATION OF ATTERBERG LIMITS WITH PERCENTAGE OF SAND
VARIATION DES LIMITES D'ATTERBERG AVEC L E POURCENTAGE
*percent in the mixture passing No. 40 sieve. DE SABLE
soil from Niamey, Niger was also submitted. differently t o stabilization. From the results of this study, it
The data indicates that all four soils can be successfully appears that some differences occur in the reaction of the
stabilized and waterproofed with bituminous binders. The soils different soil groups to stabilizations. The ferruginous gravels
with appreciable content of fines required special studies, when stabilized with cement showed large gains in strength at
however. As with all types of stabilization, the type and small percentages of cement, higher than the materials in any
amount of binder have to be established by means of of the other soil groups. The ferruginous and ferrallitic soils
laboratory tests involving actual project soils. from sandy parent materials showed a great reduction in the
A complete report on the full asphalt stabilization study, plasticity index at a small percentage of lime, but with further
which was prepared by The Asphalt Institute, has been increases in the lime content the decrease in the plasticity
reproduced in the Appendix. index was slight. The other soils showed a lineal decrease with
increasing lime content.
In almost all cases the strength and durability of lime,
CONCLUSIONS cement and sand treated materials increased with increased
One of the goals of this stabilization investigation was t o stabilizer content. For each soil tested, the soil-cement
determine if soils of the three major groups would react strength and durability were greater than the lime-soil strength

TABLE 17

Design Content of Lime and Cement

Stabilized Materials

Lime Cement
Design Content Design Content
Sample By pH Method By Test Method By pH Method By Test Method
 TABLE 18

Stabilization Test Data

Stabilizer Content Determined from pH Test

NIA- 1 DAH-3
WO 4% Cement Wo 5% Lime
7 day 28 day 7 day 28 day

MDD 134 1.33 - 113 11 2.5 -

OMC 8.7 9.5 - 16.6 17.0 -

CBR 77.5 500 500t 16.2 98.7 129 .O

Swell 0.18 0 0 0.7 0 0

and durability. The sand-soil mixture showed CBR values not reach a non-plastic state. In the lime-soil mixtures the
lower than the same soil treated with lime. plasticity index was reduced considerably more than in the
For the cement treated soils, the recommended minimum soil-cement mixtures. The plasticity indexes of all the lime-
values of unconfined compressive and CBR strengths were treated gravels were reduced to below 15.
attained with 3 t o 6 percent cement by weight. Because the In the soil-cement mixtures there were large increases in
experimental test data included a wide variety of lateritic strength between 7 and 28 days. With lime-soil mixtures there
gravels, it can be concluded that lateritic gravels can generally was only a slight increase in strength between 7 and 28 days.
be stabilized with from 3 to 6 percent cement for use as base The split tensile test appears to be a good method to
course. evaluate the tensile strength properties of lime- and cement-
From the same experimental data it can be seen that lime treated laterite materials. The split-tensile strength increases
stabilized gravels are likely to have limited use as base course proportionately with the unconfined compressive strength.
materials. Most lime treated soils showed maximum strength The lime-soil and soil-cement mixtures both showed an overall
values when the percent lime by weight ranged from 4 to 6. average for the ST/qu ratio of 0.10. Further studies should be
Higher percentages of lime caused reduction of strength. Lime made on the application of the split-tensile test for stabiliza-
may have some value as an admixture for subbase material or tion of lateritic soils.
t o improve subgrade soils where lower CBR's are acceptable. The CBR test appears to be the best means of evaluating
Additional research would be required to see if the addition of the effectiveness of sand, lime- and cement-stabilized ma-
a pozzolan to a lime-soil mixture would increase its strength terials. The performance of roads and highways has been
and durability. correlated with the CBR test more than with any other test. It
The recommended PCA values for the loss of weight after is used by most African countries as a criterion for pavement
12 cycles of the wet-dry durability test was attained with 4 design.
percent cement in all soil-cement mixtures. The recommended The unconfined compressive test appears to have limited
PCA values were generally not attained with the lime-soil application as the basis for design of stabilized materials. There
mixtures even at the higher lime contents. This test may not does not appear t o be any clear correlation between .the
be a true indication of the actual durability of lime treated unconfined compressive strength and the CBR values. There is
soils with a surface covering. very little agreement upon design criteria based on the
In almost all cases the lime- or cement-treated materials did unconfined compressive strength of stabilized materials.
 Although Portland Cement and lime, Ca(OH), , were used REFERENCES
t o stabilize the twenty selected soil samples, it must be
remembered that this project was a laboratory study. Regard-
less of the plasticity and water content, laboratory size soil Arman, A., and Saifan, F., 1967. The effect of delayed
samples can be dried and disaggregated to the desired size compaction on stabilized soil-cement. High. Res. Board
fraction. Dry soil irrespective of its grain size and plasticity can Record 198: 30-38.
be blended with any dry admixture before adding water. The Cartmell, H. S., and Berch, A. O., 1958. Lime stabilization of
mixture can be compacted immediately before harmful chemi- soils for use as road foundations in Northern Rhodesia.
cal reactions take place. Dept. of Sci. and Ind. Res. (U.K.), Road Res. Lab. Overseas
Under field conditions it is very hard and expensive to dry Bull. No. 9 .
and break down a plastic soil mechanically. In a high rainfall Charernnit, W. Swelling characteristics of plastic soil with fine
area it is impossible. It has been recommended that cement be sand stabilization. Thesis No. 1 12, SEAT0 Graduate School
used t o stabilize those soils having a plasticity index of 15 or of Engineering, Bangkok, Thailand.
less and not more than 15% minus 0.074 mm. Soils with these Clare, K. A., and Cruchley, A. E. Laboratory experiments in
physical characteristics can be mixed with the crudest type of stabilization of clays with hydrated lime. Geotechnique,
mixing equipment. Lime should be used where the plasticity V O ~ .7: 97-1 11.
index is greater than 15 and the minus 0.074 mm. exceeds De Medina J., 1964. Les laterites et leur application a la
15%. Soils with a plasticity index greater than 15 tend to "ball technique routiere. Rev. Generale des Routes, No.
up" when a pulvermixer is used. To insure that a soil will be 382: 81-94.
properly stabilized at least 60% of the soil binder must pass a Diamond, S., and k n t e r , E. B., 1965. Mechanisms of soil-lime
No. 4 sieve. Cement treated soils must be mixed and stabilization. High. Res. Board Record 92: 83-95.
compacted w i t h n a two hour period after the cement is added Eades, J . L., Grim, R. E., 1966. A quick test t o determine lime
t o the soil. requirements for lime stabilization. High. Res. Board
Reactions w h c h take place between lime and soil proceed Record 139: 61-72.
slowly and, therefore, the final mixing can be delayed for Hayter, M. A., and Cairns, H., 1966. A field and laboratory
considerable time. If the soil is very plastic, the lime can be investigation into the strength of a trail stretch of lime-
mixed very crudely at first and compacted only enough to stabilized soil pavement in Eastern Nigeria. Dept. of Sci.
protect the lime from carbonation and the subgrade from and Ind. Res. (U.K.), Road Res. Overseas Bull. No. 20.
becoming saturated in case of a heavy rain until the lime has Herrin, M., and Mitchell, H., 1961. Lime-soil mixtures. High.
time to react with the clays. Final mixing should proceed as Res. Board Bull. 304: 99-138.
soon as 60% of the soil binder passes a No. 4 sieve. At the time Irwin, M. J., 1958. A laboratory investigation of the stabiliza-
of final mixing the pH of the soil should be determined and if tion of five tropical gravels from Uganda using Portland
needed enough lime added t o bring the pH t o 12.4 befpre cement or hydrated lime. Dept. of Sci. and Ind. Res.
compaction. (U.K.), Road Res. Lab. Note No. RB/3327/MJI
If the material t o be stabilized has a plasticity index greater (unpublished).
than 15 and laboratory strength tests show that the lime Kennedy, T. W., and Hudson, R. W., 1968. Application of the
treated soil will not give a bearing value which will meet the indirect tensile test t o stabilize materials. High. Res. Board
design standards, lime and cement should be used in combina- Record No. 235: 36-48.
tion. The soil or base material should be mixed with enough Laboratoria Nacional de Engenharia Civil et al., 1969. Portu-
*
lime t o give a pH of 11.0 0.2. The lime should be added first guese studies on engineering properties of lateritic soils.
t o disaggregate the soil by reducing the plasticity. At the time Proc. of the Spec. Session on Eng. Prop. of Lateritic Soils,
of final mixing the cement should be added. The mixing and Seventh Int. Conf. on Soil Mech. and Found. Eng., Mexico
compaction of the cement and the lime treated soil can be City, vol. 1: 85-96.
accomplished in the two hour time limit and adequate strength Mitchell, J. K., and Hooper, D. R., 1961. Influence of time
will develop. between mixing and compaction on properties of a lime-
Sand stabilization slightly improves the strength and plasti- stabilized expansive clay. High. Res. Board Bull.
city characteristics of lateritic materials. However, because of 304: 32-50.
the limited availability of sand in most of tropical Africa and Muktabhant, C., and Ongskul, 1969. Stabilization of lateritic
the relatively large amounts of sand required for successful soil with sand. Proc. of the Spec. Session on Eng. Prop. of
stabilization, sand-soil mixtures are not widely used. Lateritic Soils, Seventh Int. Conf. on Soil Mech. and
The results of the asphalt stabilization tests indicate that Found. Eng. Mexico City, vol. 1: 153-164.
the 2 sandy soils from the Congo and particularly CON-2 Moh, Z. C., and Mazhar, F. M., 1969. Effects of method of
could be successfully stabilized and waterproofed with bitumi- preparation on index properties of lateritic soils. Proc. of
nous binders. CON-2 had less fines and was less plastic than the Spec. Session on Eng. Prop. of Lateritic Soils, Seventh
CON-1. Int. Conf. on Soil Mech. vol. 1: 23-35.
O'Reilly , M.P. and Millard, R.S:, 1969. Roadmaking materials gravel. Dept. of Sci. and Ind. Res. (U.K.), Road Res. Lab.
and pavement designs in tropical and sub-tropical countries. Note No. RN/3798/PTS/GAS (unpublished).
Dept. of Sci. and Ind. Res. (U.K.), Road Res. Lab. Report Thompson, M. R., 1965. The split-tensile strength of lime-
LR/279: 31p. stabilized soils. High. Res. Board Record 92: 11-23.
Sherwood, P. T., and Sullivan, G. A., 1960. A study of the Winterkorn, H. F., and Chandrasekharen, E. C., 1951. Laterite
factors affecting the strength of a cement stabilized lateritic soils and stabilization. High. Res. Board Bull. 44: 10-29.
 CHAPTER 10

PROPOSED PAVEMENT TEST SECTIONS

INTRODUCTION STRENGTH
The use of stabilized soil for the bases of flexible pavements It is important to evaluate the strength requirements, as
in Ghana is becoming increasingly important because of well as rate of development of strength in stabilized lateritic
increased traffic and the rather poor quality of the lateritic soils under field conditions. Different requirements of strength
gravels available. The practice, though not new in this country, for stabilized lateritic soils to be used as road bases have been
still leaves a number of questions unanswered about the suggested. Maclean (1956) has suggested a minimum uncon-
performance of roads constructed with stabilized lateritic fined compressive strength of 250 psi for the base. Clare and
gravels. Mohan (1947) used 300 psi for assessing stabilized laterites
Cement stabilization, being costly, warrants careful investi- from Nigeria. In the laboratory studies conducted in Ghana
gat ion, design, and construction to achieve optimum results. (Road Research Laboratory, 1958) an unconfined compressive
The considerations prompt the suggestion that profitable use strength of 250 psi at 7 days was used as the criterion for
of cement stabilization can hardly be anticipated until there assessing the strength of soil-cement samples. This is the
has been a quantitative assessment of the performance of criterion used in the United Kingdom. It is, however, uncertain
cement-stabilized materials in the field. Previously, studies whether a lower strength criterion would be adequate in view
conducted on soil stabilization in Ghana have been restricted of the absence of frost. The major point to be established in
t o laboratory samples. Studies of this type (Road Research this experiment is the level of strength at which t o aim in
Laboratory, 1958) have indicated that most of the soils in the order to obtain the most satisfactory performance.
country are amenable to stabilization with cement, and that On the basis of the previous laboratory study (Road
between 3 and 7 percent cement should usually be adequate. Research Laboratory, 1958) in which between 3 and 7 percent
It has also been noted that some regional effects are cement were found to be adequate for Ghanian soils under
apparent; the soils from the forest area (ferrallitic) show laboratory conditions, five levels of strength should be assessed
smaller gain in strength with increasing cement content .than corresponding t o 070, 2%, 4%, 6%, 8% cement (Fig. 87). Once
soils from the savannah area (ferruginous). However, strength the subgrade strength has been evaluated, the thickness of base
gains are unpredictable. It is therefore necessary that field should be selected according to the proposed procedure
evidence be obtained t o supplement knowledge from labora- described in Chapter 13.
tory studies.
Information is required on the necessary thickness of BASE THlCKRlESS
cement-stablized materials in the road structure t o give a 'The effect of base thickness should be studied at one level
pavement satisfactory for various traffic and soil conditions. of strength only. This will reduce the number of sections and
Information is also required on the optimum mixes which will so reduce the cost of the experiment. Three levels of thickness
give satisfactory performance, will experience minimum crack- should be studied at a strength level corresponding to 4%
ing, and will withstand the effect of weathering in the field. cement. The subgrade strength is not known but, for purposes
The adequacy of pavement design methods for cement- of illustration, we assume a subgrade CBR value of 6 and
stabilized lateritic materials and the rate of gain of strength in subbase material with a CBR of 20. For a traffic class of 450
the field under various regional factors are some of the t o 1500 commercial vehicles per day, a total thickness of 15
questions for which answers are sought. inches will be required. The subbase will require a cover of 8
T h s chapter describes proposed designs of test sections for inches. A uniform subbase .thckness of 7 inches should be
construction, when funds are available, as experimental pave- used. The base thicknesses should then be 6 , 8 and 10 inches
ment structures utilizing soil cement. These test sections will
supplement the present study on laterite and lateritic soils. - respectively. This is illustrated in Fig. 88.

LOAD FACTORS
SELECTlORl OF FACTORS
A preliminary traffic volume study should be performed to
In genera', a number of may be in the ens,re that the test sections will experience adequate traffic
design of a test section. In this experiment three factors are of loads. Traffic counts with measurements of axle loads should
interest. These are strength requirements, thckness
continue after construction to enable the accumulated axle
requirements, and influence of regional factors. Different
loads t o be evaluated.
levels are assigned to these variables but a complete factorial
analysis (Irick and Hudson, 1964) cannot be considered REGIONAL FACTORS
because the cost would be prohibitive. There are three distinct rainfall areas with corresponding
 4
A . 1( 1000~-04~)
-----I
A.Z (1000'-0") 40' -
A.3 (1000' 0") b&4 ~ . 11000'-0")
4 ~ . 5 ( 1 0 0 0-'0 " )
MIN. MIN MIN. MIN.

COMPACTED SUBGRADE - ASSUMED C.B.R.=6%

SOUS-SOL COMPACTE-VALEUR ADMISE POUR LE CBR=6O/o

(I) SPECIFIED THICKNESS IS TENTATIVE BASED ON ASSUMED SUB-GRADE C.B.R.OF 6 %

LEPAJSSEUR DONNEE EST PROVISOIRE E T BASEE SUR UNE VALEUR DU CBR
SUPPOSEE EGALE A 6 %
(2) TRANSITIONS SHOULD BE 4 0 ' MINIMUM
LES RACORDEMENTS DEVIENT E T R E 4 0 PIED AU MINIMUM

FIG. 87 TEST SECTION A , VARIABLE STRENGTH, UNIFORM THICKNESS

TRONCON D ESSAl A, RESISTANCE VARIABLE, EPAISSEUR UNIFORME

Racordement Racordernent
(;ansition

4
B. I ( 1 0 0 0 ' - 0 " ) -
B. 2 ( 1 0 0 0 ~ 0 " ) -
r-
I

-1
B. 3 (1000~ 0") - b

- . . .
Subgrade C .B. R . = 6%
C B R du sous - sol = 6 OO/

SECTION No. 8.1 8.2 8.3 TRONCON

Thickness o f pavement 17" 15" 13" Epaisseur du revgtemenl
Thickness of sub 7" 7" 7" Epalsseur de la fondation
Thickness of base. 10" 8' 6" Epalsseur de la couche de base

FIG.88 TEST SECTION 6 , UNIFORM STRENGTH

TRONCON D'ESSAI 6, RESISTANCE UNIFORME
soil groups, namely: ferruginous, ferralitic and ferrisol. Though extending the stabilized base about 18 inches beyond the edge
the results of the pavement condition survey have shown that of the bituminous surfacing.
these soil types are not appreciably distinguishable with
respect to pavement performance, it will be advisable to take CONSTRUCTION CONTROL
these regional variations into account. It is proposed that one It is essential that construction be carefully controlled and
experimental section be constructed in soils in each of the that continual checks be made during construction of the
groups. experimental lengths. Mixing and compaction should be
thorough but must be limited to 3 hours or less to ensure
NUMBER OF TEST LENGTHS WITHIN adequate compaction before the mixture sets. The base course
ONE REGION should be compacted to 95% Ghana Standard.
Each experimental section should be composed of eight test
lengths. These are denoted A.1; A.2; A.3; A.4; A.5 ; B.l; B.2; Construction control should also include:
and B.3. A.l through A.5 should be of u n i f ~ r mthickness and 1. The condition of the subbase on which the soil cement
should be laid out according to laboratory subgrade strengths is to be compacted. The subbase material for the whole
of: experimental length should preferably be similar and of
1. 0% Cement in A.l uniform strength. The upper 6 inches should be
2. 2% Cement in A.2 compacted to a minimum of 95%. The uniformity of
3 . 4% Cement in A.3 material coming from the borrow pit (for stabilization)
4. 6% Cement in A.4 will require constant checking.
5. 8% Cement in A.5 2. The cement content should be controlled by observa-
B.1 to B.3 should be of the same strength (4% cement) but of tions of the quantities used and the method of
varying tllicknesses as explained above. application.
3. The moisture content at the time of compaction and
LENGTHS OF SECTIONS throughout the curing period should be carefully
Each test length should be 1000 feet in length, 600 feet of controlled. This is perhaps the most important factor
which should be used for observation and the 200 feet at each since the maximum attainable density, hydration of the
end reserved for destructive tests (see Fig. 89). Each of the cement, and the degree of loss of strength on soaking
eight test lengths in a region should be separated by a all depend on the initial mixing moisture content.
minimum transition of forty feet in which there should be no 4. The uniformity of dispersion of the cement and
tests. This requirement is in line with the experience of the moisture should be controlled visually. Occasionally,
AASHO Road Test (Irick and Hudson, 1964). Thus each 30-40 pound samples should be taken for chemical
experimental section should be at least 8,400 feet or 1.6 miles. analyses.
No test section should contain transitions between cuts and 5. Compacted densities should be checked to ensure that
fills. It would also be desirable to avoid curves and drainage the specified compaction has been obtained.
structures in test sections. 6. The final compacted depth of construction should be
It is recognized that this type of construction, like most measured to ensure that the design tlucknesses are
others in Ghana, will provide a base material with a very t h n maintained.
bituminous surfacing (surface treatment). To ensure that 7. Finishing should be achieved by blade grading to
runoff water from the roadway will be carried onto the remove minor irregularities.
shoulders, the subbase and subgrade should be protected by 8. Curing is required to ensure the hydration of the
cement, and for this the material must be kept damp.
Until the final grading is performed, curing should be
effected by spraying the surface with water. The
material should be allowed to dry between successive
sprays and before the surface is sealed.

TESTING PROGRAM
Note:
Four classes of variables should be measured. These are
-
Shaded area IS for sampling and destructive testing structural variables, load variables, climatic and regional
-
Note:
variables, and performance variables.
La zone hachuree est destinge aux e'chantillonnages et
essois destructifs 1 . Structural Variables:
The composite strength of the section should be
FIG.89 T E S T SECTION P L A N VIEW measured twice a year with a Benkelman deflection
V U E E N P L A N D U T R O N C O N D'ESSAI beam under an 18,000 pound axle load. All deflection .
 measurements should be made within the central 600 (c) Depth to water table.
feet of each section. (d) General character of drainage and surrounding
At pre-determined intervals of time, cored samples land.
for laboratory studies should be taken from the (e) Type and condition of shoulders.
reserved 200-feet ends of the section. Results should be 4. Performance Variables:
compared with those from similar mixes prepared at The pavement sections should be rated by visual
the time of construction. The following tests should be inspection at ten levels of performance depending on
performed: the present ability of the pavement to serve traffic.
(a) Atterberg limits This is the method used in the pavement condition
(b) Moisture content survey. Observations should include surface cracking,
(c) Field density of the specimens rutting, deterioration and riding comfort. The perform-
(dl pH ance should be rated each time the composite strength
(e) Unconfined compressive strength is measured by the Benkelman Beam.
(f) Cement content
2. Load Variables:
Two major load variables should be evaluated REFERENCES
during the study. These are: Clare, K. E., and Mohan, D., 1947. An examination of five
(a) The total number of equivalent 18,000 pound soils and an aggrekate from Nigeria. Dept. Sci. and Ind.
single axle load applications accumulated by the Res., (U.K.) Road Res. Lab. Note RN/882/KEC.DM
section. (unpublished).
(b) The years of service over which these load Irick, P. E., and Hudson, W. R., 1964. Guidelines for satellite
applications are accumulated. studies of pavement performance. National Cooperative
The evaluation of load variables will necessitate classifi- High. Res. Program, High Res. Board Report 2A.
cation counts with axle load measurements of traffic Maclean, D. J., 1956. Considerations affecting the design and
over the section. construction of stabilized-soil road bases. Jour. Inst. High.
3. Climatic and Regional Variables: Eng., vol. 111, 9 : 16-33.
Over the test sections the following information Millard, R. S., 1961. The need for soil stabilization. Roads and
should be collected : Road Construction, vol: 39 (468): 355-362.
(a) Yearly rainfall and monthly distribution of the Road Research Laboratory, 1958. Notes on the cement
rainfall. treatment of Ghana soils. Dept. Sci. and Ind. Res., Overseas
(b) Daily high, low and average temperatures. Bull. No. 8.
 CHAPTER 1 1

A SURVEY OF PAVEMENT DESIGN PROCEDURES

IN AFRICA

INTRODUCTION subgrade moisture content. A sample of the soil, compacted t o

The pavement design procedures now used by the indi- the specified density and at field moisture content, is allowed
vidual African countries were either established by the t o attain moisture equilibrium under the same effective depth
respective colonial administrators or have been developed or t o water table as prevails at the site. An undisturbed soil
modified since the colonial period. A survey of these pro- sample is used if the subgrade is not t o be compacted. The
cedures was conducted in association with field and soil procedure and apparatus are given in RRL Road Note 3 1.
samvline trivs into the various countries. The vrimarv source When equipment for making a suction test is not available,
I U I

of information in this canvass was interviews with locil public an estimate of the ultimate moisture content can be made
Works officials. A secondary source was a questionnaire from measurements below the zone affected by seasonal
circulated t o the various countries in Africa by the Building moisture changes.
and Road Research Institute. In designs for category (2) the RRL recommends a
Generally, the two design methods most cornrno-nly used conservative estimate of the subgrade strength by the CBR test
are the one developed by the Road Research Laboratory, compacted at British Standard Compaction. It is pointed out
United Kingdom, and the one adopted by Centre Experi- that within a given climatic zone the ratio:
mental de Recherche et d'Etudes du Batiment et des Travaux field moisture content
Public, Paris, France. plastic limit
for road subgrades tends t o be constant. Thus an investigation
Design Procedures of moisture condition under existing pavements in the climatic
The Road Research Laboratory procedure, which is widely zone concerned may prove to be a useful guide t o the moisture
used in the English speaking countries of Africa, was de- content for design. When an appreciable proportion of
veloped through experience gained by the RKL, Tropical material is retained on the No. 36 B.S. sieve, the following
Division. It is published in RRL Road Note 3 1, second edition relationship is suggested:
(1966). The two major considerations are: (1) The influence field moisture content
of different tropical climates on moisture conditions under (P x X) + (S x Y)
sealed surfaces and their effect on the strength of subgrades, where :
sub-bases and bases, and (2) the rapidly increasing traffic X is the proportion of soil passing the No. 36 B.S. sieve
which is a general feature of roads in developing countries of Y is the proportion of soil retained on the No. 3 6 B.S.
tropical and sub-tropical regions. sieve
Subgrades are classified into three main categcjries ac- P is the plastic limit of the fines
cording to the moisture conditions: S is the saturated surface-dry moisture content of the
(1) Subgrades in areas where the water table is suffi- material retained on the No. 36 B.S. sieve
ciently close to the surface t o control the subgrade In category (3) the normal road camber or any necessary
moisture content. The depth at which the water surface-water drainage measures are sufficient to prevent
table will influence moisture conditions depends on .accumulation of water and a corresponding high moisture-
the soil type: a) non-plastic soils where the water content beneath the pavement.
table is within 3 feet of the surface; b) sandy clay The design charts are shown in Fig. 90.
where the water table is within 10 feet of the The Road Research Laboratory has since published another
surface; and c) heavy clays where the water table is report entitled "Road Making Materials and Pavement Design
within 20 feet of the surface. in Tropical and Sub-tropical Countries." As this report was not
(2) Subgrades in areas where the water table is too deep published until 1969, it has not yet come into use. Recom-
t o affect the subgrade adversely and where the mendations are made for the selection of pavement thckness
seasonal rainfall exceeds the moisture loss by evapo- in terms of traffic loading and strength of the soil subgrade.
transpiration over at least two months of the year. The report also considers stage construction for developing
The annual rainfall in such areas is usually greater countries. This possibility has also been emphasized in papers
than I 0 inches. presented by Miller (1969) and Burrow (1969) at the First
(3) Subgrades in areas where climate is arid throughout African Highway Conference in Addis Ababa.
the year. The CEBTP design procedure, most widely used in the
In designs for category (1) the conditions expected in the French speaking countries of Africa, was developed at the
field are simulated t o obtain an estimate of the ultimate Laboratoire de Ponts et Chaussees of Algiers. Peltier has been
 REVETEMENT SUPERFlClEL REVETEMENT ENROBE

ET
POUR LA CIRCULATION A APPLIQUER LORSQUE
- IMMEDIAT
SURFACE DRESSING
LA CIRCULATION L' EXlGE
PREMIXED SURFACE TO BE-
To
TRAFFIC
CARRY IMMEDIATE LAID WHEN REQUIRED BY
INCREASING TRAFFIC
COUCHE DE FOND ET REVETEMENT SUPERFlClEL

LIc P%%kCS~ fl
A BASE DE LIANT HYDROCARBONE
BITUMINOUS PRIME AND SURFACE DRESSING
4
d
RE2'5'NEH"ENhEEMNI?foBE BNIPFMI"N'ou2S
LL BITUMINOUS PRIME AND SURFACE DRESSING
COUCHE DE FOND ET REVETEMENT SUPERFlClEL
fi HEAVY A BASE DE LIANT HYDROCARBONE
IC 6-INCH THICKNESS OF BASE TRAFFIC 6-INCH THICKNESS OF BASE
EGERE EPAISSEUR DE COUCHE DE BASE 6 POUCE INTENSE EPAISSEUR DE COUCHE DE BASE 6 POUCES

--------
CBR MINIMUM

rz 20
2% UJ--
cn
W
2 3 4 5678910 152025

p: C.B. R OF SUBGRADE - (Per Cent)

Z
Y
u C.B. R . OF SUBGRADE- (Per Cent)
a2 C.B. R. DU SOUS -SOL Oo
'/ C. B. R . DU SOUS - SOL
I11 I-
WI-

F I G . 9 0 FLEXIBLE PAVEMENT DESIGN CURVES ( R R L )

ABAQUES DE CHAUSSEES SOUPLES

given credit for the development of the equation (Liautaud MATE Rl AL SPEC1 FICATIONS
1966) To initiate the review of material specifications, the BRRI
100 + 150 P sent t o a number of African countries a questionnaire on the
e =
CBR + 5 uses of lateritic soils and particularly their success and failures.
where The answers received were as varied as the number of
countries to which the questionnaire was directed. The general
e = thickness of pavement (cm) trend, however, was to use lateritic soils for road construction
P = wheel load in metric tons except where such soils were uncommon. A few countries
In experiments conducted by the Laboratoire des Ponts et mentioned increasing use of crushed rock for base con-
Chaussees the best correlation with performance was obtained struction on account of increasing traffic and axle loadings.
by using the CBR value corresponding to a dry density equal From the answers received, it would appear that more than
t o 95% of the modified AASHO maximum density and at a 90% of road construction in tropical Africa involves the use of
moisture content near the optimum after a 4 day soaking lateritic soils.
period. Standards for the acceptance or rejection of material are
Peltier later modified his equation to: invariably derived from particle-size distribution, Att erberp,
l o o + P ( 7 5 + 50 LogN/10) limits and the California Bearing Ratio. Not surprisjngly the
e = CBR + 5 limits for acceptance vary from country to country.
which introduced the repetition factor N representing the Specifications employed by the various countries are shown
number of commercial vehicleslday on the pavement. A in Table 19. The data shown were obtained from the
commercial vehicle was defined as one having a gross weight questionnaire sent by the BRRI and through interviews with
exceeding 3 tons, N is the number of vehicles in both local Public Works Departments of the countries visited.
directions. Thls procedure was used extensively from 1950 to The following ranges of requirements are observed. In the
about 1965. The trend since 1965 has been to use the design base course, the minimum soaked CBR varies from 5 0 to 120.
procedures of the Asphalt Institute. Plasticity requirements put the upper limits of plasticity index
Design procedures based on the CBR are employed when and liqujd limits from 6 to 25 and 25 to 4 8 respectively;
the pavement structure is considered almost perfectly flexible whereas, the upper limit of the percentage passing No. 200
(E<5000 bars and CBR<100). If the layers are more rigid, as varies froin 13 to 38. The period of soaking for the CBR test
by stabilization, the elastic theory is utilized. Liautaud (1966) ranges from 24 to 9 6 hours. The CBR requirements for
outlined the development of the theory of two and multi- subbases range from 12 to 30, and the upper limits of
layered systems and its application in pavement design. plasticity index from 6 t o 20.
 TABLE 19

Material Specifications

GRADATION SOAK
COUNTRY TYPE MATERIAL 1% 314 318 3/16 No. 10 No. 40 No. 200 CBR LL PI

Republic of Wearing Gravel 100 35 25 20 12-20

Malawi Course
Base Gravel 100 60-90 45-75 30-60 20-50 10-30 5-15 85 >30 >6
Course
s
Base
Stab. 150 a 0 a 0
Course
Gravel
c 2
Base
Republique Wearing Gravel >75 <25 >60 a 4
Du Niger Course
Base Gravel 100 a 5 >so <22
Course
Base Stab. 100 <25 >so 4 0
Course Gravel
Sub Sable >20 <16
Base Argile
Ethiopia Wearing Crushed 100 70-90 45-75 30-60 15-40 0-30 0-20 >SO <35 4-9
Course Rock
Base Crushed 68-88 35-55 29-49 17-37 8-28 0-18 0-13 30 a 5 <6
Course Rock
Sub 75-85 45-65 0-10 >2 5 <25 <6
Base
Republic of Wearing Gravel 100 85-100 50-100 35-92 25-80 14-50 10-38 18 GO
Kenya Course
Base Gravel PIX % PASSING NO. 36 a50 80
Course
Sub Gravel PIX % PASSING NO. 36 <75 0
Base
Federation Wearing Gravel 39-54 24-30
of Nigeria Course
Base Gravel 100 80-100 55-80 40-60 30-50 5-15 80 <25 G O
Course
Base Sand 100 80-100 10-25 80 <25 GO
Course Clay
! Sub 30 <45 15
Base
1Mali Wearing Gravel
Course
Base Gravel AASHO SPECIFICATIONS 50 6-16
Course
Sub Gravel 12 20
- Base
I v r ~ yCoast Base Gravel 30-65 16-30 60 12
Course
Base Sand 26-40 60 5-20
Course
Base Sand 16-32 60 5-15 0-1
Course Clay
Sub 30 20
Base
Senegal Base Gravel <60 PASSING NO. 60 20-35 80 10-25
Course
Sub Gravel 30
Base

BEST AVAILABLE COPY

 TABLE 19 (Cont.)

Material Specifications

- -

GRADATION SOAK
COUNTRY TYPE MATERIAL 1% 314 318 3/16 No.10 No.40 No.200 CBR LL PI

Cameroon Base Gravel

Course
Base Gravel 80 10-25
Course
Sub Sand 30
Base
Gabon Wearing
Base >60 <20
Subbase
Gambia Wearing Gravel 80-100 60-100 35-83 28-62 25-50 2244 13-28 95% AA 20-37 13-22
SHO MOD
Base Gravel 80 60-100 35-83 28-62 25-50 22-14 13-28 95%AA 20-37 13-22
SHO MOD
Subbase GRADING NOT SPECIFIED
Zambia Wearing Gravel
Base Gravel 100 15-20 120 30 6
Subbase 25
Uganda Base Gravel 100 85-100 68-100 54-100 43-90 30-57 19-38 3748 16-25
Subbase 25

CONCLUSION REFERENCES
The review indicates that two principal methods of design Burrow, J.C., 1969. Design standards for stage construction of
have been in use in African countries; namely, the procedure highways in developing countries. First African High. Conf.,
developed by the U.K. Road Research Laboratory, and that Addis Ababa.
advocated by the French Cent re Experimental de Recherches Liautaud, G. A., 1966. Flexible Pavement Design-analysis of
et d'Etudes du Batiment et des Travaux Publics. Both methods the traditional methods, review of the modern methods
make use of CBR as a measure of strength. The review also with recommendations on their condition of application.
shows that lateritic gravel has been widely used in tropical CEBTP Tela. Note 006.
Africa and continues t o give reasonably good service both as Miller, D. F., 1969. Stage construction of highways in Africa.
base material and running course, especially where traffic First African High. Conf., Addis Ababa.
volumes are low. Road Research Laboratory, 1966. A guide t o the structural
design of bituminous-surfaced roads in tropical and sub- (
tropical countries, 2nd Ed. Ministry of Transport, Road
Note 31.
!
f'
 CHAPTER 12

PA VEMENT CONDITION SUR VE Y

INTRODUCTION during the past 25 years in many parts of the world; virtually
In recent years considerable effort has been devoted t o the all the African hghway testing experience has been with this
development of rational procedures of pavement design. It is test. The contention that the R-value determination is superior
well recognized, however, that such procedures can only be for pavement design is still debatable, furthermore its adoption
developed on the basis of knowledge of the performance of as a standard, would mean abandoning much of the experience
pavements under local environmental and traffic conditions. that has been accumulated in African countries.
Thus the first step toward rational design requires an evalu- The results of the pavement condition survey have provided
ation of the behavior of pavement components under loads of the basic information for establishing suitable design criteria
known magnitude in the locality where the design procedures and material specifications for lateritic soils in highway
are t o be employed. construct ion under various climatic and traffic conditions.
Most regional African pavements are built primarily from
local materials and may be described as surface treatment SCOPE
pavements. The structural strength of the pavement is there- Test sections were selected with the objective of evaluating
fore derived principally from the available geologic materials the pertinent design and performance variables through as
which constitute the base, subbase and the subgrade. An wide a range as possible. Table 2 0 lists these variables and
investigation of the relationships between the condition of Table 21 gives the relationships t o be investigated as outlined
these structural pavement components and the corresponding by Irick and Hudson (1964). Thus, site selection was based on
pavement performance ratings was therefore undertaken as a the observed pavement drainage conditions, soil group, climat-
part of this project. It was anticipated that the results of the ic conditions, and traffic characteristics. Each site was then
investigation should provide a basis for a rational procedure evaluated and assigned a pavement condition rating. The
for design of flexible pavements in tropical Africa. pavement condition was rated by observations which included
The condition survey included the determination of the longitudinal deformation, extent of cracking, rutting depths,
deflection characteristics and the structural strength of several rideability and judgment of the pavement's present ability to
test sections. A relationship was then developed between the serve traffic.
deflection characteristics of the pavement and the behavior or Pavement deflection characteristics were established for
performance of the pavement section relative t o the traffic each test section. The deflection characteristics were used as a
loading. A second relationship was developed between the measure of uniformity within the test section and as a measure
structural strength of the pavement and the deflection of composite strength. Sections for detailed analysis were
characteristics. The deflection characteristics were considered selected considering uniformity, and test pits were then
t o be represented by either of two parameters, (1) the excavated. In-situ tests and laboratory tests were conducted on
maximum rebound deflection of the pavement (Benkelman et each component material, to establish the strength character-
al. 1962) and (2) the slopes of the deflection basin (Kung, istics of the pavement and the subgrade.
1967). The second parameter is similar t o the surface A statistical analysis was then performed on the data t o
curvature index (Schrivner et al. 1968). The structural strength establish relationships between the design variables (strength
of the pavement section was evaluated by measuring the characteristics) and the performance variables. This analysis
thickness and in-situ CBR of each layer within the pavement provided the data necessary to establish suitable design criteria
structure. The structural depth of the pavement was consider- and also provide an evaluation of the adequacy of tropical soils
ed to be thirty-six inches because at this depth the induced as base and subbase materials under various trafik loading
stress is less than three percent of the applied stress at the conditions.
surface, as shown by means of Newmark's charts (Martin and
Wallace, 1958). PROCEDURE
The CBR test was retained for this study for two important
reasons. First, all African countries visited during this study Selection and Rating of Test Sections
use the CBR test in pavement design and all laboratories are Only sites with good surface drainage were used in the
equipped t o perform the test. On the other hand, very few study; it was considered axiomatic that poor ditching and
laboratories have the equipment necessary to determine the R other drainage provisions lead t o unsatisfactory performance.
value; the standard test of California, and the test utilized After this basic condition was satisfied, sites were selected on
during Phase I of this study in Thailand. Secondly, there has the basis of performance, traffic loading, soil type and climatic
been considerable experience accumulated with the CBR test conditions. Fig. 91 shows the location of test sections included
 TABLE 20 Pavement deflection characteristics were determined for
nine ty-six individual pavement sections. All of the sections
General Basis For Studies of One Pavement Type were evaluated relative t o uniformity o f the deflection
characteristics and fifty-seven selected for the structural
(after lrick and Hudson, 1964)
condition survey. The distribution of sections for each soil
group is as follows: ferruginous, fifteen sections; ferrisols,
twenty-five sections; ferrallitic, fourteen sections; and black
DESIGN VARIABLES
clay, two sections. The higher percentage of sections in the
ferrisol soil group was a consequence of the limited number of
Structural
paved roads in the other two major soil groups. The sections in
the Ivory Coast were in the ferrallitic soil group. These
Pavement Structure Strength Characteristics
sections compensated for the limited sites available in Ghana.
Since the soil type reflects the climatic conditions, the
Surface Courses Thickness of Pavement
distribution among soil groups also represents the distribution
Courses
of climatic variables. The distribution within the soil group
Base Courses (if any)
Other Design Features provided for the necessary variations in traffic as well.
Subbase Courses (if any)
Composite Strength Determi nation of Deflection Characteristics
of Pavement Sections
Roadbed Material Deilection characteristics were established for each of the
pavement sections. Two parameters were investigated; 1) the
Load Accumulated Axle Loads conventional maximum rebound deflection and 2) the slope of
the deflection curve as defined by Kung (1 967) or by Scrivner
Years of Service (1968) as the surface curvature index. Ail illustrated expla-
nation of the two deflection parameters is given in Fig. 9 2 and
Rate of Accumulation in the Appendix. Deflection measurements were taken at each
100 ft. station throughout a 500 ft. test section in both inner
Climatic and Regional Conditions of Precipitation (IWP) and outer (OWP) wheel paths. The deflection charact er-
istics were considered t o be a measure of the composite
Moisture, Temperature and strength of the pavement system. The maximum rebound
Frost deflection (MRD) is a measure of strain relative to the applied
stress whereas the slope of the deflection curve (tan Q),
Topography although also indicative of the strain associated with the
applied stress, is more particularly an indicator of the
Relative Strength in concentration of the applied stress. This distinction is in
different climates accordance with a method of analysis suggested by Lee
(1968). Lee suggests a curvature deflection number (N), which
Regional Factors he considers t o be a pavement strength index, and considers
the radius of curvature and maximum deflections t o be
PERFORMANCE VARIABLES indexes of stress and strain respectively. He shows that the-
curvature deflection number (radius of curvature divided by
Surface Behavior Deformation and Deteriora- maximum deflection) is inversely proportional t o the square of
t ion tan @.
Both deflection parameters were correlated with composite
Present Serviceability strength and performance. Both are considered t o be inversely
proportional t o the potential of the pavement t o perform well.
Performance Detailed Investigation
FIELD INVESTIGATION

in the preliminary survey and those that were included in the The field investigation included excavation of test pits,
structural or detailed condition survey. measurement of component thicknesses, in-situ density deter-
The pavement sections were rated by visual inspection at minations, cone-penetration tests and recovering laboratory
ten levels of performance depending on the present ability of samples of each soil component including the subgrade
the pavement t o serve traffic. material.
 .DABAKALA

IVORY COAST
COTE D'IVOIRE

LEGEND
LEGENDE

PAVEMENT DEFL ECTlON STUDY

ETUDE DE DEFLECTIONS DE LA
CHAUSSEE
Q PAVEMENT D E F L E C T I O N AND
DETAILED STUDY
ETUDE DE DEFLECTIONS DE LA
CHAUSSEE

FIG 91 PAVEMENT CONDITION SURVEY LOCATIONS

ETUDE DE L A VlABlLlTE DES CHAUSEES, SITUATION
 TABLE 21

General Basis For Relationships Among Variables

For One Pavement Type
(after l rick and Hudson, 1964)

Performance Variables Design Variables

Measures of Surface Deformation Structural

and Deterioration
Thicknesses hl , h2, h3
XI, X2, X3, etc.
Strength Characteristics Sl , s2, s3, s4
Present Serviceability Index
Other Design Features Cl , c2, c3, etc.

Composite Strength S
Performance Index
Load and Time EL, ADL, Y

Climatic or Regional
Note: Laterite Study p = 5.0
Climatic Variables

Note: Variables outside parentheses Relative Strength

are defined by variables
inside the parentheses Regional Factor

/
-- \
/' \ DIRECTION OF MOVEMENT
/ \ +
0 DIRECTION DU MOVEMENT
\ I
t I
I
1
. '.
\
\
\
L J / - --
D E F L E C T I O N BASIN AT CREEP SPEED
BASSIN DE DEFLECTION A VITESSE TRES LENTE
POINT OF INFLECTION D = MAXIMUM REBOUND DEFLECTION
POINT D'INFLECTION D = DEFLECTION MAXIMUM t

X = VARIABLE
X =VARIABLE

A. MEASUREMENT OF MAXIMUM REBOUND DEFLECTION

MESURE DE L A D E F L E C T I O N M A X I M U M

tan 0 = ( S L O P E O F DEFLECTION)
( P E N T E DE L A DEFLECTION)
POINT D'INFLECTION

B. MEASUREMENT OF THE SLOPE OF THE DEFLECTION CURVE ( a f t e r Kung, 1 9 6 7 )

M E S U R E D E L A P E N T E D E L A COURBE DE D E F L E C T I O N (d'aprgs KUng, 1 9 6 7 )

FIG 92 MEASUREMENT OF PAVEMENT DEFLECTION PARAMETERS

MESURE DES PARAMETRES DE DEFLECTION DE LA CHAUSSEE
 All test pits were excavated and tested to a depth of 36 LABORATORY INVESTIGATION
inches t o insure that the examination of the pavement The samples that were obtained during the field pavement
structure had been carried out to a sufficient depth. With condition survey were transported t o the Public Works
many of the CBR-thickness design curves, 36 inches would Department Soil Laboratory in Kumasi for testing. The tests
provide a sufficient cover for a subgrade with a CBR of two. conducted on the samples are shown in Table 22.
The procedure in the field was as follows. The test pit was CBR-Density-Moisture relationships were established for
laid out with a width of 3 0 inches and a length, perpendicular each material encountered during excavation of each test pit.
t o the center line of the highway, equal to one-half the The compactive efforts included AASHO Modified, Ghana
pavement width. The asphaltic concrete or surface treatment Standard and AASHO Standard compaction; a comparison of
was removed. Density tests were conducted in each wheel path compactive efforts is shown in Table 22. CBR tests were
of the base course material along with a sufficient number of conducted on the top (as compacted) of the sample prior to
cone-penetration tests to obtain an average value. If the values the four-day soak period. After submerged soaking for four
beneath the inner and outer wheel paths differed, they were days, CBR tests were conducted on the bottom (as compact-
recorded as individual average values for each wheel path. A ed) of the sample. The in-situ CBR value as well as the soaked
representative sample of approximately 150 pounds was CBR value of the material were estimated from these curves
obtained for laboratory testing. The base course material was using the in-situ density and moisture content, as determined
removed; if the pavement system was not a composite in the structural survey. Typical compaction curves are shown
structure, ten inches of material was removed. Testing was in the Appendix.
repeated on the subbase course or on the level after the ten
inches had been removed. Excavation and testing continued
until the subgrade (or 36 inch depth) could be tested and Traffic Analysis
sampled. After the excavation was completed, the thickness of In order t o arrive at the performance index for the sections
each component of the pavement structure was measured. of pavement tested, it was necessary t o have some knowledge

TABLE 22

Laboratory Testing Schedule For Pavement Condition Survey

Field Samples

Borrow Surface
Test Area Gravel Base Subbase Subgrade

Gradation X X X

Atterberg Limits X X X

. Specific Gravity X X X

2
,, Sand Equivalent
' California Durability

Aggregate Crushing Strength

CBR (as molded14 day soak)

AASHO Modified 56750 ft. lbs./cu. ft.

Ghana Standard 22500 ft. Ibs./cu. ft.

AASHO Standard 12375 ft. lbs./cu. ft.

Portuguese Swell Test

of the accumulated equivalent 18,000 1b.-single-axle load The growth factors for all classes of vehicles were estimated
applications at the time of the pavement condition rating. The from data on vehicle registration. Since vehicle registration
purpose of the traffic analysis was t o estimate, from available does not necessarily reflect the growth of vehicles on each
traffic counts, the equivalent 18,000 lb. single-axle load road, this assumption was bound to introduce some error.
applications which the pavements had experienced since However, since previous classification counts were lacking,
construction or resurfacing. there seemed no other way of estimating the growth factor for
Traffic counts up t o 1968 gave total vehicles per day. different classes of vehicles on different sections of the roads.
However, information was not available as t o the traffic The growth rates used in the analysis are as follows:
classification. The only data available for this purpose was the
Growth Rate %
Report on Ghana Sector Studies prepared for the Ministry of
Vehicle type per annum
Finance and Economic Planning by Nathan Consortium in
1970. The analysis of traffic was .therefore based on that
Cars 6%
report. To obtain necessary data relating t o construction and
Light Buses 10%
resurfacing a questionnaire was sent t o the Road Authorities in
Heavy Buses 10%
each district. Personal interviews of villagers and construction
Mammy Wagons 2 .O%
workers supplemented the data.
diminishing
Traffice charts showing average daily traffic were available
All Goods Vehicles 1.5%
for most of the main roads in Ghana. The composition of
Special Purpose Vehicles 4 .O%
traffic for each section was based on a traffic survey carried
out by the Building and Road Research Institute in 1969
under contract for the Ghana Sector Studies. The traffic was The traffic classification count for each test section was
classified into eight categories as follows: taken from the 1969 Ghana Sector Studies. By using the
1. Cars: Private cars, hired cars and taxis. growth rates, the number of vehicles per day in each class was
2. Light buses: Buses with seats for not more than 22 calculated from the date of construction or resurfacing. The
people. vehicles per day so obtained were summed for each class and
3. Heavy buses: Buses with seats for more than 22 people. multiplied by the appropriate equivalence factor. Having
4. Mammy Wagons: Lorries with wooden bodies and converted the daily totals into equivalent 18,000 lb .-single-axle
covered tops for transport of passengers and goods. load applications, a grand total of the equivalencies in all
5. Light Goods Vehicles: Goods vehicles with 2 axles and classes was taken. This total multiplied by the total number of
single rear tires including pick ups and vans. days gave the accumulated equivalent 18,000 lb .-single-axle
6. Medium Goods Vehicles: Goods vehicles with 2 axles load applications in both directions from the date of construc-
and dual rear tires, single unit without trailer. tion or resurfacing t o the time when the pavement condition
7. Heavy Goods Vehicles: Goods vehicles with 3 or more rating was assigned.
axles, vehicles with trailers and articulated vehicles, e.g. Sections IV-1 t o IV-6 were tested in cooperation with the
authorities in the Ivory Coast. The traffic data available listed
timber trucks, low loaders.
essentially cars and commercial vehicles. Observations of axle
8 . Special Purpose Vehicles: Tractors, bulldozers, graders,
loads on typical roads in five countries by Ellis (1968) have
rollers, other motorized agricultural or construction
shown that, in countries where the maximum allowable axle "
machinery.
load is limited t o 10,000 kg. or 1 0 tons, 100 commercial ,
For purposes of analysis, axle loads were assigned t o each
vehicles are equivalent t o fifty 18,000 lb. axle loads. Since this
traffic category and 18,000 1b.-single-axle-equivalent factors
loading is used in the Ivory Coast, the above assumption was'.
were assigned as follows:
Equivalent made in the analysis of the Ivory Coast traffic. A factor of
Estimated Factor 0.0002 was used for cars as in Ghana.
Axle Load 18,000 1b.-
in single R ESU LTS
Vehicle Type 1000 lbs. axle Deflection Study
Cars 2 0.0002
Light Buses 5.5 0.01 1 The results of the pavement deflection studies are shown in
Heavy Buses 10 0.090 Figs. 9 3 and 94. Fig. 93 shows the relationship between inner
Mammy Wagons 7 0.035 and outer wheel paths for both deflection parameters. Fig. 9 4
Light Goods Vehicles 4 0.002 shows the relationship between the deflection parameters for
Medium Goods Vehicles 18 1.OO both wlzeel paths.
Heavy Goods Vehicles
(Tandem axle) Detailed Study
Special Purpose Vehicles 1.OO The results of the structural evaluation of the pavement
 I 1
I
0 00 I I
000 001 0 02 0 03 0 04 0 05 0 06
MAXIMUM REBOUND DEFLECTION (IN) I.W. P. SLOPE OF DEFLECTION CUHVE(tan ~ x I o - ~1.W.P
)
PENTE DE LA COURBE DE DEFLECTION (tan 0 x INTERIEURE
I
00 0 25 0.50 0 75 100 125 1.50
DEFLECTION MAXIMUM INTERIEURE (MM)

FIG.^^ DEFLECTION CHARACTERISTICS OF INSIDE AND OUTSIDE WHEEL PATHS

CARACTERISTIQUES DES DEFLECTIONS A L'INTERIEUR ET A LIEXTERIEUR
DES TRAJECTOIRES DES ROUES.

e#ty tan 0 x lo-3. +

38.58 MRD 0.013
tan 0x 39.83 MRD-
r= 0.888
6. *
0.283

V I I I I I I I I
0 0 0 01 0.02 0.03 0.04 0.05 0-06 0.07 0.08 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
MAXIMUM REBOUND DEFLECTION (IN) I.W. I? MAXIMUM REBOUND DEFLECTION (IN) O.WF!
l l l l l l l l l I I I I I I I I 1
.OO 025 0.50 075 to0 125 150 175 2.00 0.00 0.25 0 5 0 075 1.00 1.25 1.50 175 2.00
DEFLECTION MAXIMUM INTERIEURE (MM) DEFLECTION MAXIMUM EXTERIEURE (MM)

FIG 94 RELATIONSHIP OF DEFLECTION PARAMETERS FOR INSIDE AND OUTSIDE WHEEL WTHS
RELATIONS ENTRE PARAMETRES DE DEFLECTION A L'INTERIER ET A LIEXTERIEUR
DES TRAJECTOIRES DES ROUES
 sections are given in Table 23. The general in-situ conditions of The regression equation is as follows:
base, subbase and subgrade are given in Table 24. MRD = 0.047 - 0.0024 Rating
The ferruginous base course materials on the average Standard deviation, (o) = k0.0 1 1
exhibit higher CBR than the ferrisols or ferrallitic soils. The Coefficient of correlation, (r) = -0.476
average compaction of the base course in all groups was close Standard error, (oo) = k0.130
t o 90% AASHO Modified Compaction. The higher CBR of the The relationship between the slope of the deflection curve
ferruginous soils indicates that these soils are generally better (tan $) and the pavement condition rating is shown in Fig. 96.
base course material or that better base course material is more The regression equation is as follows:
readily available. Moisture conditions which exist in the field t a n $ x l o 3 = 1.97 - 0 . 1 17 Rating
are somewhat lower than initially expected. If adequate Standard deviation, (o) = k0.500
drainage is provided one should consider the as-molded CBR Coefficient of correlation, (r) = -0.530
for design as opposed t o the soaked CBR value. This is Standard error, (oo) = -+ 0.130
especially true for the lower rainfall areas of the ferruginous Although the relationship is represented by a low coeffi-
soil group. cient of correlation, the coefficient of correlation is approx-
The compaction and CBR are higher for subbase materials imately four times the standard error (ao) at zero association
in the ferrisol and ferrallitic soil groups. However, the moisture (Tippett), 1952. The standard error (0,) at zero association is
content is lower in the ferruginous soil group. dependent upon the number of data used in the analysis. It
The subgrade CBR values were generally low in all groups. can be concluded that there is a real, although somewhat
The degree of compaction of the subgrades was also low. The weak, association be tween the deflection characteristics and
CBR of the subgrades could be improved considerably through pavement condition rating. Similar results have been reported
compaction alone. The moisture contents of the subgrade by Lee and Chow (1968) in a study of deflection character-
material of the ferrisols and ferrallitic soil groups are 3-6 istics and pavement condition ratings in the State of Maryland.
percent higher than those of the subgrades of the ferruginous They report that the pavement condition rating is quite
soils. consistent relative t o the curvature-deflection number, but not
t o the average deflection. The curvature-deflection number as
Traffic Study described earlier is related t o the slope of the deflection curve.
The results of the traffic analysis are shown in the Performance
Appendix.
A traffic analysis was conducted for each test section of the
detailed survey t o correlate deflection characteristics with
performance. All traffic was converted t o its equivalent 18,000
ANALYSIS A N D DISCUSSION lb. single-axle dual-tire loading. The accumulated axle loadings
Deflection studies conducted in the United States (Lee and for each section were determined from the date of
Chow, 1968; Benkelman, 195 1) have led t o the conclusion construction or from the date of last resurfacing t o the date of
that the inner wheel path (IWP) experiences lower deflections the pavement rating. The data (rating, deflection and accu-
than the outer wheel path (OWP).The hlgher strength of the mulative 18,000 1b.-axle loadings) were analyzed through a
inner wheel path is attributed to the additional reinforcement discriminant analysis computer program to establish the
of improved material due t o the increased distance from the relationship between allowable deflection at various levels of
shoulder. As shown in Fig. 9 3 a large number, approximately traffic which maintain a serviceable pavement condition rating.
+, one-half, of the sections evaluated in this study have shown In the system used in this study, a rating of below 4 would
d
higher deflections in the inner wheel path. These higher indicate termination of the life of the pavement, at which time
deflections are due t o a lower CBR and a thinner section as resurfacing or reconstruction would be desirable. Using this
compared t o the outside wheel path. rating as the discriminatory factor, the deflection equation
yielded the maximum permissible deflection t o maintain a
Deflection Characteristics given pavement life. The pavement life is defined as the total
F ~ g4~ shows
. the relationship of the maximum rebound accumulated 18,000 lb. equivalent axle loads (EAL) that the
deflection and corresponding slope of deflection curve for Pavement must accommodate prior t o re-construction.
each wheel path. A good correlation exists between the two The results of the analysis are shown in Figs. 97 and 98.
deflection parameters with a coefficient of correlation of The sections which it would be desirable to resurface or
0.903 IWP and 0.888 OWP. reconstruct are shown as open circles. The allowable deflection
characteristics at a given repetition of loading can be deter-
Rating mined by the following equations:

The relationship between maximum rebound deflection 12.4 - Log, 0 2 18000 EAL
MRD =
(MRD) and the pavement condition rating is shown in Fig. 95. 204
 SUBJECTIVE PAVEMENT CONDITION RATING-COTATION SUBJECTIVE DE L'ETAT DES CHAUSSEES
FIG 95 RELATIONSHIP BETWEEN MRD AND SUBJECTIVE PAVEMENT RATING
RELATION ENTRE DEFLECTION MAXIMUM ET COTATION DE L'ETAT DES CHAUSSEES

PAVEMENT CONDITION RATING

COTATION DE CETAT DES CHAUSSEES

FIG. 96 RELATIONSHIP BETWEEN TAN0 AND SLIBJECTIVE PAVEMENT RATING

RELATION ENTRE TAN $8 ET COTATION DE L'ETAT DES CHAUSSEES
 0 PAVEMENT RATING > 4
COTATION DE LA CHAUSSEE > 4
0 PAVEMENT RATING < 4
COTATION DE LA CHAUSSEE > 4

--- - -- 0
--
h

-* d MRDz12.4-Log 5 E A L I 8 0 0 0 LB
204

(rn .. . .
0.. -- -- -- U
-*

000 100000 1000000 1000(

L E A L 1 8 0 0 0 L B LOADING

FIG 97 RELATIONSHIP BETWEEN MAXIMUM REBOUND DEFLECTION AND RATING AT

VARIOUS TRAFFIC LOADINGS
RELATION ENTRE DEFLECTION MAXIMUM E T COTATION PGUR DIFFERENTES
CHARGES DE CIRCULATION

PAVEMENT RATING > 4

0 COTATION DE LA CHAUSSEE >4
0 PAVEMENT RATING < 4
COTATION DE LA CHAUSSEE >4

a
0 0

0
0 0 TAN 0 =18.0-Log 1 EAL I8000 LB
o a
-- ------- go ;a
a0 / 9390

a -
a 8 "Laa
------a

a
ar
=,8
a 0 a.
a 8@
a
a
a a.
a

100000 1000000
1 E A L I8000 L B LOADING

FIG 98 RELATIONSHIP BETWEEN SLOPE OF DEFLECTION CURVE AND RATING A T

VARIOUS T R A F F I C LOADINGS
RELATION E N T R E P E N T E D E L A COURBE DE DEFLECTION E T COTATION POUR
DIFFERENTES CHARGES DE CIRCULATION
 18 - Logl 0 L: 18000 EAL Po = initial serviceability index
and tan @ =
9390 P1 = present serviceability index

The two equations provide a relationship between the It is necessary to assume a numerical value for B in order to
deflection characteristics of the pavement and the perform- use this formula. Experience with the AASHO Road Test
ance of the pavement relative to traffic. The two test sites performance data (Irick and Hudson, 1964) has indicated that
which fall below the regression line may be the result of initial suitable approximations for B are:
low compaction of the base course which resulted in deform- B = 1 for flexible pavements and
ations due to additional compaction by traffic carried to a B = 0.5 for rigid pavements
point of eventual self-stabilization. The pavement can be Most of the pavements in this study consisted of only
considered in a failed condition but exhbiting low deflection surface treatments. Consequently, the approximation of B
characteristics. Such a situation in practice would indicate a from the AASHO experience would be unreliable. Since a
pavement section that could be resurfaced as opposed t o a relationship has has been shown for deflection characteristic
section with higher deflections which would have t o be and performance, an analysis was conducted to determine
reconstructed. what value of B gave the best correlation. The best correlation
The two equations presented are based on the average between deflection and performance index was obtained when
deflection characteristics of the section. Since fifty percent of B was assumed to be 2. The equations were:
the deflections are higher, the limiting deflection curves are MRD = 0.0898 - 0.00947P
conservative. A more liberal approach t o the analysis is to Standard deviation, (o) = 20.009
consider the characteristic deflection of the section. The Coefficient of Correlation, (r) = -0.67 1 and
characteristic deflection is defined as: tan @ x l o 3 = 3.81 - 0.427P
Standard deviation, (o) = k0.350
Coefficient of Correlation, (r) = -0.714
where Xc = characteristic deflection
The results appear t o be reasonable for pavements lacking a
X = average deflection of section surfacing with structural strength.
o = standard deviation
The resulting allowable deflection equations are : DETAI LED INVESTIGATION
21.3 - Loglo L: EWL 18000 Axle The primary purpose of the detailed structural investigation
MRDc = was to relate the engineering properties of the materials in the
355
pavement structure to performance. The deflection character-
12.7 - Logl L: EWL 18000 Axle istics have been related t o performance by using average
and tan@, =
4000 deflection values for the test sections. Although the majority
These two equations permit higher deflections relative to of the pits were excavated at the average deflection station (a
performance. few test pits were located in areas of high deflections), it was
A relationship between deflection and the performance of decided to relate the material properties to the deflection
the pavement section has been shown on the basis of the characteristics which in turn were a measure of performance.
pavement rating at various applications of 18,000 lb. axle
loadings. Another approach is to relate the deflection param- Physical and Engineering Properties
. eters to the Performance Index of the pavement section. Irick The purpose of this analysis was to determine the effect of
and Hudson (1964) have defined the performance index of the the various engineering properties on the behavior of the
pavement as the logarithm of the number of applications of an pavement structure relative t o the deflection characteristics.
"'
18,000 lb. axle loading prior to the pavement's deteriorating The main analysis, which included soil type, thickness, CBR
t o a serviceability index of 2.5 (equivalent t o a pavement and compaction was analyzed through a factorial experiment.
rating of 4 in this study). The equation suggested for Secondary experiments included index properties, gradation,
estimating the performance index is: swell, durability, and aggregate crushing values; they were
analyzed by multiple regression analysis. The range in values of
the properties was not sufficient to allow assigning the various
(estimated) P = Log L: L1 + B [Log Log (Po12.5) - levels of these properties which are required in a factorial
Log Log (PoIP1 )I experiment. For example, only samples from pits showing
where P = Performance Index satisfactory compaction could be used so that this variable
could be removed. Compaction was very low in many of the
L1 = accumulated 18000 lb. axle loads
pits tested, and consequently these could not be included in a
B = numerical constant defining the slope factorial experiment considering the other properties. Those
of the performance curve remaining displayed such narrow ranges of index properties,
swell, durability, etc. that the effect of these properties on assessing the potential of the pavement to perform well. Since
performance could not be evaluated except by multiple soil type did not, as an independent variable, influence the
regression. structural performance it can be concluded that subgrades
The results of factorial experiment 002A-1 (Thick- having a low in-situ CBR provide the same support value
ness-CBR-Soil Type) are as follows: regardless of soil type. However, the ferruginous soils have
The analysis included twenty-seven sections at three levels hgher CBR values and lower in-situ moisture contents than do
of pavement thickness, three levels of base CBR, and the three both the ferrisols and ferrallitic soils and therefore the type
lateritic soil types (ferruginous, ferrallitic and ferrisols). Since has an indirect effect on performance. The range of values and
the data were obtained from a pavement condition survey of mean values of these and other properties were discussed
existing roads the complete factorial block could not be above.
satisfied (See Fig. 99). Six of the required sections were The results of factorial experiment 002A-2 (Thickness-
occupied by averaged sections. This factorial experiment compaction) are as follows:
considered the effect of thickness, CBR and soil type on the The analysis, shown in Fig. 100, included sixteen sections
performance of the pavement structure. The fixed variables at two levels of thickness and two levels of compaction. Four
were subgrade CBR and base compaction. Soil type, as a replications of each level were used to strengthen the analysis
variable, also represents a climatic variable. The results: since the sections provided only the two levels in each variable.
1. Maximum rebound deflection. The thickness of the This factorial experiment considered the effect of thickness
pavement structure has a pronounced effect on MRD at the 95 and compaction on the performance of the pavement struc-
percent confidence level. The CBR has a pronounced effect at ture, the fixed variables being the subgrade CBR and base
the 98 percent confidence level. Soil type has no effect on course CBR. The results:
performance as an independent variable. 1. Maximum Rebound Deflection. The thickness and
2. Slope of the deflection curve. Both thickness and CBR compaction have a pronounced effect on MRD at the 98
have a pronounced effect on tan r) at the 98 percent percent confidence level.
confidence level. Soil type (whether ferruginous, ferrallitic or 2. Slope of the Deflection Curve. The thickness and
ferrisol) does not influence the tan r$ as an independent compaction have a pronounced effect on tan @ at the 98
variable. percent confidence level.
Regardless of soil type (ferruginous, ferrisols or ferrallitic) When the CBR of the pavement structure is relatively fixed,
the thickness and in-situ CBR are significant criteria for both thickness and compaction of the material have a
pronounced effect on the performance of the pavement
structure.
Several attempts were made to establish a relationship
THICKNESS
EPAISSEUR
between the physical properties (liquid limit, plasticity index,
I-
aa dW minus 200, durability, swell, and aggregate crushing strength)
8: 3: I 2 3 and performance of the pavement section. A form of analysis
' (013) (0.22)
TP 23 AVERAGE
MOYENNE
(017) (0.85) (030) ( \ . o n
TP 05
could not be developed in which the physical properties as
independent variables could be related to pavement perform-
ance. The CBR and thickness of the pavement section control
the performance of a section. The physical properties have
been examined with respect t o their effect on the CBR value
in Chapter 5. Consideration will be given t o these relationships,
in Chapter 13, on standards and specifications.
TP IV06 TP lVO3
(027) (0.96) (037) (b3O)
b%FFN"WE 1 S-tructural Number
TP27 AVERAGE TP 0 6
(033) (1.16) (038) (1.41
MOYENNE
(038) (1.66) The structural number (SN) express an empirical relation-
ship between the thickness of a component layer in a
pavement structure and the type of material in the layer.
Although usually not given units, SN actually has units of
inches. The relationship is expressed by the general equation:

FIG 99 FACTORIAL EXPERIMENT - THICKNESS, SOIL TYPE

(FERRUGINOUS, FERRISOL AND FERRALLITIC) AND where Dl = thickness of asphalt surface (inches)
CALIFORNIA BEARING RATIO VS DEFLECTION
D, = base thickness (inches)
EXP~RIENCEFACTORIELLE- EPAISSEUR, TYPE ESOL
(FERRUGINEUX, FERRISOL E T FERRALITIQUE) E T
D, = subbase thickness (inches)
CBR E N FONCTION DE LA DEFLECTION
 zW
FO F REPETITIONS

$p REPETITIONS

00
0 I 2 3 4
0

TP21 TPlO TP 14 AVERAGE

MOYENNE
2
(026) (1.00) (021) (0.83) (035) (1.41) 1027) (1.08)

AVERAGE
MOYENNE

I TP09 I TPOl I -

TP19 1

FIG 100 -
FACTORIAL EXPERIMENT THICKNESS AND COMPACTION VS DEFLECTION
-
EXPERIENCE FACTORIELLE EPAISSEUR-COMPACTAGE EN FONCTION DE
LA DEFLECTION

Values of a , , a, and a3 are established from performance view, it is considered that an accurate assessment can be made
data but should be considered as coefficients without units. As of the pavement potential for reducing deflection on the basis
a means of comparison, the coefficients suggested by Chastain of CBR and thickness regardless of the gravel type.
and Schwartz (1965) for granular base and subbase courses Fig. 103 shows the relationship between the slope of the
were used initially in this study. These coefficients are shown deflection curve and the structural number of the pavement
in Fig. 101. The coefficients for the gravel base courses system. Three relationships are shown: 1) base material is
(ferruginous, ferrallitic, ferrisols, and quartz gravels) were classified as fine-grained lateritic soil; 2) base material is
initially assumed t o be equal, and the structural number was classified as quartz gravel, and 3) base material is classified as
computed on the basis of CBR alone. lateritic gravel. A single relationship for all gravels results in an
It is generally accepted that base course material should average error of approximately k2.2 inches in thickness. The
have a minimum CBR of approximately 75. The initial analysis individual relationship shows that an average additional 4.5
of structural number was limited t o sections with base course inches of quartz gravel is required t o have the same potential
materials having an in-situ CBR greater than 75 and having for reducing the slope of the deflection curve as lateritic gravel.
compaction greater than 90% Modified AASHO. The subgrade Unlike deflection, the potential for reducing the slope of the
in all sections had an in-situ CBR near zero as determined from deflection curve appears to be dependent upon the com-
laboratory tests; moreover the average compaction was 77% position of the gravel or gravel matrix as well as CBR and
Modified AASHO. thickness.
Fig. 102 shows the relationship between maxiumum re- The data are limited, but suggest that the potential of the
bound deflection and the structural number of the pavement base course material t o reduce .the maximum rebound deflec-
system. The lower scale is the equivalent thckness assuming a tion is dependent upon thickness and CBR, independent of .the
CBR of 75. Two relationships are shown: I ) base material is nature of the material, providing the compaction is greater
classified as fine-grained and 2) base material is classified as than 90% AASHO Modified and the CBR is greater than 75.
gravel. A single relationship for all gravels results in an average However, the potential of the base course material t o reduce
error of approximately ? 1.2 inches in thickness. If individual the slope of the deflection curve is dependent upon thickness
relationships are shown for lateritic gravel and quartz gravel, and CBR as qualified above, and apparently upon the nature
the results would indicate that an average additional 2.5 inches of the base course material. It cannot be determined from the
of quartz gravel would be required to have the same potential data available if the dependency is within the base course
for reducing deflections as lateritic gravels. However, the range material itself or if the quartz gravel and soft subgrade are
in CBR (75-1 15) considered would in itself incorporate an incompatible. The subgrade in all sections had an in-situ CBR
error of approximately 21 inch. From a practical point of near zero.
 0,150 0,150-

0.125 - 0.125 -

Pom0.100-
i-i- /
/ /' i
! i-
0.100-
0
/
/ 0/ /

!:
zz /
0.075- G0-075-
0

UL!
L LL
/
d 0 0
0

h 0.050 - / -
180.050
00 / 0
V U 1 0

0.025 -
/
I O0.025 - 0
/
/
/

/ 0
I I I l l 1 I I I I I l l I I I l l
O / I 0
10 20 40
60 80 100 200 0 2 4 6810 20 40 60 80 100
C B R VALUE C B R VALUE
VALEUR DU CBR VALEUR DU CBR

FIG 101 COEFFICIENTS O F R E L A T I V E STRENGTH O F BASE ( a 2 ) AND SUBBASE(a3)

(after Chastoin and Schwortz, 1 9 6 5 1
COEFFICIENT DE L A RESISTANCE RELATIVE DES COUCHES DE BASE (a2)
E T D E FONDATION ( a 3 ) (d'oprbs Chastain and Schwartz, 1 9 6 5 1

The deflection characteristics relative t o thickness and and conclusions are valid only if a double layer surface
in-situ CBR values were analyzed t o determine suitable treatment is applied t o the pavement structure.
coefficients (a,), (a3) and (a4). Since all of the pavement To preclude any extraneous variable such as compaction,
sections were paved with a relatively thin surface treatment the data used in the analysis were limited t o test sections
the component of the surface layer was considered as part of which exhibited sufficient compaction in base and subbase
the base course thickness. The subgrade was considered as layers so that the CBR of the materials rather than compaction
layer 4 with an appropriate coefficient a4. In this case, results controlled the deflection characteristics. The minimum com-
paction accepted for base and subbase materials was 88
percent Modified AASHO compaction. The variation within a
given test section, as measured by variations in deflection
AVEC tRAVlER A FERRUGINEUX
LEGENDE

0 FERRALLITIQUE
0 FERRISOLS
V WARTZ
FINES A

c m w SWS-S(X~0
0 FERRISOLS
V QUARTZ
WBGRADE CBR=O
LEGEND
WlTH GRAVEL A FERRUGINOUS FlNE A
0 FERRALLITIC . between stations, was limited t o 35 percent. The sections
meeting these requirements are shown in Table 2 5 .
1 6-2-4

Fourteen sections were analyzed in which the structural

5 A-7-6
COUCnE DE BASE
t W W E DEBASE
C B R 75-115
COYPACTAGE * 9 0 %
BASE CBR 75-115
0ASE COUPACTION + 9 0 % components of the pavement consisted of a base course and a

AVEC tRAVlER A FERRUGINEUX

LEGENDE

O FERRALLITIQUE
0 FERRISOLS
V QUARTZ
FINES A

CBR W m - S m - 0
. LEGEND
WITH GRAVEL A FERRUGINOUS F l N E A
0 FERRALLITIC
0 FERRISOLS
n OUARTZ
~uE&ADE~CBR=O
.
1 A-2-4

5 A-7-6
W W E DE BASE C B R 75-115 BASE C B R 7 5 - 1 1 5
COUCHE DE BASE COMPACTAGE t90% BASE COMPACTION + 9 0 X

I
0 00- 000
0 0 0 5 1 0 15 2 0 25 I 0
I
I S
STRUCTURAL NUMBER ISN)
INDICE DE STRUCTURE
-
1 I I I I I I I 0.0 I I I I I I I
0.0 0.5 10 15 2 0 2.5 3.0 3.5
(CBR75lO 40 80 120 160 200 240 280
STRVCTURAL NUMBER ( S N I
THICKNESS (INCHES) I N D I C E D E STRUCTURE
EPAISSEUR (POUCES)
b i do I$O 2010 2do 220
THlCl(NESS(INCHES1 - E P A I S S E U R IPOUCESI
FIG 102 MAXIMUM REBOUND DEFLECTION VS. PAVEMENT STRUCTURAL
NUMBER (SN) FIG 103 SLOPE OF DEFLECTION CURVE VS. PAVEMENT STRUCTURE (SN)
DEFLECTION MAXIMUM EN FONCTION DE L'INDICE DE LA PENTE DE LA COURBE DE DEFLECTION EN FDNCTION DE
STRUCTURE LIINDICE DE STRUCTURE
 TABLE 25

Coefficients of Base, Subbase and Subgrade Material

BASE -

-
1
TPN" D2 CBR MRD TAN@ MRD a2

* Base and Subbase with same CBR

SUBBASE

TPN" D2 CBR a2 D3 CBR MRD TAN@ MRD Base Residual a3

SUBGRADE
Re-
1 Base & si-
TPN" D2 CBR a2 D3 CBR a3 D4 CBR MRD TAN@ MRD Subbase dual a4
subgrade. The subgrade did not contribute to reducing
deflection since in each case the laboratory testing indicated
that the in-situ CBR was near zero. The coefficient of -thebase
course is considered t o be a function of the CBR value. The Good correlation is shown for CBR values between 2 0 and
coefficient is given as: 100. Values below a CBR of 2 0 do not contribute to reducing
deflection, suggesting that this should be the minimum CBR
1
a2 = for the subbase material.
MRD x Base course thickness Eight sections were analyzed in which the structural
Fig. 104 shows a plot of the coefficients vs the in-situ CBR components of the pavement consisted of a base course and
values of the base courses. The relationship is shown as: subgrade. The subgrade here contributed t o the reduction of
a2 = CBR/30 deflection since the CBR values ranged from 4 t o 24. The
Good correlation exists between CBR values of 4 0 and 120. coefficient established for base and subbase were used and the
There appears to be no correlation below a CBR of 4 0 coefficient of the subgrade determined from the residual
suggesting that base course materials should have a minimum deflection after accounting for base and subbase. The co-
CBR of 40. If the CBR of the base course is below 4 0 the efficient is given as:
stresses transmitted to the base course are greater than .the 1
shear resistance of the material. When this occurs deformation a4 =
MRD (residual x Subgrade Thickness
is independent of thickness of the base material.
Eight sections were analyzed in which the structural The pavement structure was considered t o extend t o a depth
components of the pavement consisted of a base course, of 3 6 inches. The subgrade thickness then is equal t o 36 inches
subbase and subgrade. As above, the subgrade did not minus base and subbase thickness. Fig. 106 shows a plot of the
contribute t o the reduction in deflection because of its low coefficient versus the in-situ CBR of the subgrade. The
CBR value. The coefficients established for the base courses relationship is shown as:
were used, and the coefficients of the subbase determined CBR-4
a4 = -
from .the residual deflection often accounted for the base 8
course. The coefficient is given: Good correlation is shown for CBR values between 5 and 24.
1 Values below a CBR of 5 do not contribute to reducing
a3 = deflection. This suggests that, where possible, the subgrade
MRD (residual) x Subbase Thickness
should be compacted t o provide an effective CBR of 5 .
Figure 105 shows a plot of the coefficient vs the in-situ CBR The structural number has been computed for each of the
value of the subbase. The relationship is shown as: sections using the limitations established for base, subbase and

8.0

HYPOTHESE DM = ~ ( C B RXEPA-
ISSEUR)

6.0
-I -. - ---
4 - d
-
II

a
0
I'
0
4.0
ASSUMPTION' MR.D = ~ ( c $ Rx THICKNESS)

I . I
ONE O F NO CORREL,4TlON
ONE SA'NS CORRELATION
I
I
ONE O( C O R R E L A T I P N
2.0
ZONE A V E C CORRELATION

0.0
0 20 4 0 60 80 100 120 140 160 180

CALIFORNIA BEARING RATIO

FIG 104 BASE COURSE COEFFICIENT VS BASE COURSE CBR

COEFFICIENT DE LA COUCHE DE BASE EN FONCTION DU C B R DE L A COUCHE DE
BASE
 subgrade materials and the coefficients based on the in-situ
CBR values. The maximum rebound deflection has been
plotted against the structural number. Fig. 107 shows this
relationship. The relationship that is shown emphasizes the
significance of CBR and thickness with respect t o the
deflection characteristics. When deflection is used as a measure
of performance (Figs. 97 and 98) it is difficult t o relate any
variable t o performance other than the structural component
of the pavement section as measured by CBR and thickness.

Performance
A relationship has been shown between the deflection of a
pavement and performance. The deflection has also been
CALIFORNIA BEARING RATIO related t o the pavement structure through the use of coef-
FIG. 105 SUBBASE COEFFICIENT VS. SUBBASE CBR
ficients and CBR. A general relationship can therefore be
- COEFFICIENT DE LA FONDATION EN FONCTION DU CBR shown between the structural number of a pavement and the
b - DE L A FONDATION. expected performance of the pavement as measured in terms
of accumulated 18000 lb. single axle loads before failure. The
permissible deflections as obtained in the deflection study are
shown in Fig. 108. The equations involve the average
maximum rebound deflection and the characteristic maximum
rebound deflection. The deflection values have been converted
t o a structural number by using the relationship shown in Fig.
107. The relationship between structural number and pave-
ment rating at various traffic loadings is shown in Fig. 109.
Two relationships are shown between the structural number
and performance. One is related to the average deflection and
the second to the characteristic deflection. The development
CALIFORNIA BEARING RATIO
of design criteria, Chapter 13, will be based on an average
FIG. 106 SUBGRADE COEFFICIENT VS. SUBGRADE CBR requirement with respect t o the two relationships.
COEFFICIENT DU SOUS-SOL EN FONCTION DU CBR DU
SOUS-SOL

0-050

-
W
V)

5 0,040
0 TWO LAYER, SECOND LAYER CBR ( 4

-f 2 COUCHES,CBR DEUXlEME COUCHE < 4

8 THREE LAYER. THIRD LAYER ,CBR ( 4
Z 3 COUCHES CBR TROlSlEME COUCHE ( 4
0
0.030
W
1
LL
n
W

g 0.020
0

u
W

I
5- 0.0'0

X
2

aooJ 0
0 10 20 30 40 50 60 70 80 90

STRUCTURAL NUMBER (SN)

INDICE DE STRUCTURE (CS)

FIG 107 PAVEMENT DEFLECTION VS. STRUCTURAL NUMBER

INDICE DE STRUCTURE EN FONCTION DE LA DEFLECTION
DE LA CHAUSSEE
 DEFLECTION MAXIMUM (MM)
COEFFICIENT DE STRUCTURE DE TRONCON DE CHAUSSEE P P P - -
0 h)
0 (I!
z
0 UI 0
0 iv
-in
STRUCTURAL NUMBER OF PAVEMENT SECTION 0

MAXIMUM REBOUND DEFLECTION (1/1000 INCHES)

9
0
-
0
w

0oV)
oo-l
z m a
2-c
m 7 o
EBZ
-'?
0
miE +-iw
6 $:a" D-l
2s Z r qs
0 ,"Em
-' D Z
0 M
$
m E8 5 r
D T D D m
, ~ E ZD
7g 1;gg
0
s
D
2;mr =
0
m m
T r CIH g g
m
r m w
0 sz g
2E ;!z 5
'?m mc
D 0
wr 8
0 -' m 7

V)
s sZ - l L
v,
P -rl
"8
Bs
5 5
f-'
,"
V)
%0 i
4
;D
%
T Q
5 5
V)
CONCLUSIONS AND FINDINGS Conf. on AASHO Road Test, High. Res. Board Spec.
The major conclusions of the pavement condition survey Report 73: 102-125.
are as follows: Benkelman, A. C., 1958. Analysis of flexible pavement
(1) A relationship has been established between the deflection and data. High. Res. Board Bull. 210: 39-48.
maximum rebound deflection and the slope of the deflection Chastain, W. E., and Schwartz, D. R., 1965. AASHO road test
basin as shown in Fig. 94. equations applied to the design of bituminous pavements in
(2) Moisture conditions which exist in the field were Illinois. High. Res. Board Record No. 90: 3-25.
found to be somewhat lower than one would expect. If Ellis, C. I., 1968. Axle-load distribution on roads overseas,
adequate drainage can be provided the use of the as-molded survey on roads in West Malaysia 1967. Ministry of
CBR as opposed to the soaked CBR should be considered for Transport (U.K.), Road Res. Lab. Report LR 187: 34p.
design purposes. This is especially true for the lower rainfall Irick, P. E., and Hudson, W. R., 1964. Guidelines for satellite
areas such as in the ferruginous soil group. studies of pavement performance. High. Res. Board, Nat.
Cooperative High. Res. Program, Report 2A.
(3) Base courses with CBR as low as 50 have performed
satisfactorily. The specifications of Atterberg limits and Kung, K. Y., 1967. A new method in correlation study of
gradation which reflect the higher CBR (+ 80) should be pavement deflection and cracking. Proc. of Second Int.
examined and consideration given to the relaxation of these Conf. on Struct. Design of Asph. Pave., Univ. of Mich.:
specifications provided the CBR values are maintained as 1037-1046.
recommended (Chapter 13). Lee, A., and Chow, C. H., 1968. Statewide investigation of
flexible pavements. Maryland State Road Comm. Res.
(4) Allowable deflection parameters have been presented
in various ranges of traffic loadings. Report (Reproduced by Clearinghouse, U. S. Dept. Comm.
as PB 180 264): 72p.
(5) A relationship has been shown between CBR and
thickness and the deflection of the section. Martin, J. R., and Wallace, H. A., 1958. Design and construc-
tion of asphalt pavements. McGraw-Hill: 305p.
(6) A relationship has been shown beween the structural
number of a pavement section and the allowable deflection for Scrivner, F. H., Moore, W. M., McFarland, W. F., and Carey, G.
a given traffic loading. Appropriate design criteria can be R., 1968. A systems approach to the flexible pavement
developed from this relationship. design problem. Texas Transp. Inst., Res. Report 32-1 1.
(Reproduced by Clearinghouse, U.S. Dept. Comm. as PB 18
REFERENCES 3 149): 99p.
Benkelman, A. C., Kingham, R.I., and Fang, H. Y., 1962. Tippett, L. H. C., 1952. The methods of statistics. 4th ed.,
Special deflection studies on flexible pavement. Proc. of Williams and Norgate, London: 395p.
 CHAPTER 13

PA VEMENT DESIGN PR OCED URES AND SPECIFICA TiONS

FOR TROPICAL SOILS

INTRODUCTION obtained the present design can be used only with similar
The design procedures used in tropical Africa have been for roads where al Dl = 0. Furthermore, only two values of base
the most part methods of developed countries and generally thickness, 6 inches and 8 inches were used, thus eliminating
associated with temperate climates. These methods have been D2 as a variable.
more or less extrapolated to tropical environments without the For a selected range of traffic, the allowable maximum
benefit of correlation of the performance of pavements rebound deflection is obtained from Fig. 108. Entering Fig.
designed by these methods with the behavior of tropical soils 109 with this value of maximum rebound deflection, the
within the pavement structure. The reason has been the lack of required value of the structural number of the pavement (SNr)
information with regard to roads with bituminous surfaces. is read. Knowing the CBR of the base course permits obtaining
An evaluation of performance requires a minimum of ten its structural coefficient (a2). With this and the specified
years unless the traffic is accelerated as during the AASHO thickness of base (6 inches or 8 inches) the structural number
road test. Until recently such an evaluation would not have provided by the base course is obtained as (SNb). The
been possible. Ten years ago only a small percentage of the difference has to be supplied by subbase and subgrade
roads in Africa had bituminou? surfaces; the remainder were strength. The solution is obtained by solving for D3 in the
either gravel or earth roads. Today, however, it is possible to equations:
evaluate the performance of relatively low cost bituminous a3 D3 + a4 D4 = SNr - SNb
surface pavements (surface treatment), such as in Ghana,
where a number of these pavements have been in service for a D3 + D4 = 36 - D2
period of years. Such an evaluation has been conducted as
reported in Chapter 12. DESIGN CHARTS
This chapter applies the results of .the pavement condition Following the above procedure, charts were drawn for six
survey to the design of surface treated pavements. Design inch and eight inch thicknesses of the three classes of base
curves are presented which are believed to be realistic for described in the specifications below. These are Class I, with a
lateritic soils as the curves have been developed from measured minimum CBR of 100; Class 11, with a minimum CBR of 70;
performance studies. Additional investigations in each country and Class I11 with a minimum CBR of 50. The charts are given
are recommended to verify the design procedure. Nevertheless, in Figs. 110 and 11 1 , each containing 3 sets of curves. The
it is felt that this method can be applied immediately and that curves are drawn for 3 levels of traffic for each base thickness.
doing so would result in satisfactory and economic roads. These are 100,000 accumulated 18,000 lb. axle loads for light
traffic; 1,000,000 accumulated 18,000 lb. axle loads for
DESIGN PROCEDURE medium traffic; and 10,000,000 accumulated axle loads for
It was pointed out in Chapter 12 that a relationship exists heavy traffic.
between the maximum rebound deflection and the structural The first step in using the design charts is to select the
number of the pavement. Also, in the analysis of the suitable chart for the design traffic and base thickness. The
deflection characteristics, allowable values of deflection for chart selected is entered from the appropriate subgrade CBR
varying traffic were established for satisfactory performance. value. A line is extended vertically t o intersect the appropriate
Relationships were also established between the structural base course class, and then horizontally to the left hand
coefficients a, , a2 , a 3 , a4 (for pavement, base, subbase, and column to read the necessary subbase .thickness. A minimum
subgrade respectively) and CBR. These relationships are shown CBR of 20 is required for the subbase material.
in Figs. 107, 108 and 109. The design is based on a maximum An example is provided for light traffic. The compacted
thickness of 36 inches. subgrade CBR is 6 and a 6 inch base course is adopted. A
Two simultaneous equations were used: vertical line from a subgrade value of 6 intersects the Class 111
a, D l + a 2 D2 + a 3 D3 + a 4 D4 = S N base course material at a subbase thickness of 22 inches. The
Dl + D2 + D3 + D4 = 36 same line intersects the Class I1 base course material at a
where D l , D, , D3 and D4 are thicknesses of pavement surface subbase thickness of 11 inches. The line also intersects the
course, base course, subbase and subgrade respectively. Most CBR = 85 line (midway between Class I and Class 11) at a
of the pavements included in this study consisted only of subbase thickness of 2 inches. For the given subgrade CBR
surface treatment so that a, could not be calculated since Dl value of 6, and a subbase CBR value of 30, three base
was essentially zero. Until additional informat ion can be thicknesses are possible, depending on the quality of the
CIRCULATION LEGERE CIRCULATION MOYENNE
LIGHT TRAFFIC MEDIUM TRAFFIC

BASE=&
COUCHE DE BAsE=6'

SUBBASE C B R = 3 0
CBR DE L A FONDATION-30

1
6 7 8

SUBGRADE CBR SUBGRADE CBR

CBR W SOUS-SOL CBR DU SOUS-SOL

CIRCULATION INTENSE
HEAVY TRAFFIC

30
4 5 6 7 8 9 10
SUBGRADE CBR
CBR DU SOUS-SOL

FIG 110 DESIGN CURVES FOR 3 LEVELS OF TRAFFIC AND

6" BASE COURSE
 EPAISSEUR DE L A FONDATION, POUCES
SUBBASE THICKNESS. INCHES
h) h) - 0
VI VI
-0 rn
P

EPAISSEUR DE L A FONDATION, POUCES

SUBBASE THICKNESS. INCHES

1
0
-
mw
=m
CDcrC"
gt
m
0
0 f=
oz
C<
gcn
c
z m w
-D
"a 25 <? EPAISSEUR DE LA FONDATION, POUCES
"1 0 SUBBASE THICKNESS. INCHES
0 cn" -4z N h) - -
VI 0 VI 0 VI 0
= E P
7;
r g a
m r
<
m
r
V)

%
+
z
%
2
0
b
material. A CBR value of 100 for the base course is not tests of tropical soils from all over Africa, and on the basis of
necessary and a value of over 8 5 does not require a subbase. If the performance of various tropical soils as determined by the
the base course material has a CBR of 70, 11 inches of subbase pavement condition survey in Ghana and the Ivory Coast, it
must be added, and if the base course has only a CBR value of was concluded that the material specifications should be
5 0 , 2 2 inches of subbase is required. modified.
In the design charts presented, base course material of CBR It is recognized that there is need for low cost roads in
as low as 5 0 can be used for road construction. This is Africa. For this reason, lateritic base course materials are
particularly applicable where traffic volume is low. These classified into three classes, namely, Class I, Class I1 and Class
charts may be used in low cost road construction. 111. Class 111 base course is t o be used for low-cost, low-traffic
Determinations of equivalent axle loads are explained in roads. Class I1 base course is t o be used for intermediate traffic
Chapter 12. Utilizing accumulated 18,000 Ib. axle loads may and Class I for heavy traffic. The following specifications are
appear awkward at first, but once the traffic patterns of the recommended:
country and region are known, the actual applied loads can be
more closely approximated. Scope
Comparison of the proposed procedure with the current 1. These specifications cover the quality and grading of
methods used in Africa is difficult because the current lateritic materials for use in construction of subbase, base and
procedures do not consider the variation in both surface courses. The requirements are intended to cover only
CBR and base CBR. 26 and 27 give some materials which can be considered as indurated laterite or
Where traffic volume is intermediate, Class I1 base course is lateritic gravels.
required, and in the case of heavy traffic, Class I base course
needs to be used.
The procedure can be extended t o cover a range o f subbase General Requirements
materials varying in strength from CBR of 25 to 50 percent. 2. (a) Coarse aggregate retained on the No. 10 sieve shall
consist of hard, durable particles. Material that crushes under
SPEC1 FICATIONS the pressure of thumb and finger shall not be used.
The current standard specifications of the Public Works (b) Coarse aggregate shall have an aggregate crushing
Department of Ghana as well as those of several tropical value conformil,g to the following:
African Countries were ascertained through questionnaires
sent t o various road authorities and through interviews Class I < 35
conducted during visits t o some of the countries. The Class I1 35-40
specifications were reviewed with reference t o those parts Class I11 40-50
which relate to the use of lateritic material as subbase, base (c) All material shall be free from vegetable matter. The
course and running course. It was found that the material soil-aggregate material shall conform t o the grading require-
specifications varied a great deal. On the basis of laboratory ments shown in Table 28.

TABLE 26 TABLE 27

Selection of thickness of subbase (CBR = 30). Selection of thickness of subbase (CBR = 30)
Low traffic, base thickness = 6". Medium traffic, base thickness = 6".
Subgrade CBR = 5. Subgrade CBR = 5.

Thickness of subbase inches Thickness of subbase inches

Base Course Base Course
CBR percent Proposed R.R.L. CBR percent Proposed R.R.L.
Procedure Road Note 3 1 CEBTP Procedure Road Note 31 CEBTP

50 24 50 Not applicable
 TABLE 28 TABLE 30

Grading Requirements For Laterite and Lateritic Recommended Criteria For Base Material
Gravels

Criteria Class I Class I1 Class 111

- --

Base Courses Surface

(PI) (% passing 200) 200 max. 400 max. 600 max.
Sieve Subbase I I1 111 Course
(LL) (% passing 200) 600 max. 900 max. 1250 max.
2" 100 100 100
Design CBR 100 min. 70 min. 50 min.
1%" 100 82-100 82-100 90-100
EAL 18000 lb Heavy Med. Light

Per day 15 year design

- - - pp ---

3/16" 55-76 19-73 19-73 36-59 60-85 compaction at maximum dry density and the fineness index
(see Chapter 5 Fig. 23) should conform to the materials
No. 8 40-57 8-51 8-51 24-43 40-70 exhibiting the relationship above the B-B' line.

NO. 30 34-48 4-31 4-31 20-35 20-45

- -
DISCUSSION
No. 200 25-42 0-15 0-15 15-25 15-30 The grading and Atterberg limit requirements can be
modified provided the CBR requirements are fulfilled. Figs.
112 and 113 illustrate the effect of liquid limit, plasticity
Subbase Materials index and percent passing the No. 200 sieve on CBR. Higher
liauid limits and ~lasticitvindexes can be a c c e ~ t e d~rovided
1 1

3. The materials for subbase shall conform t o the require- the percentage 200 sieve is decreased.
ments of Section 2 a and c. Modification of gradation The grading requirements for base and subbase courses in
requirements can be accepted provided a CBR of 20 Table 28 are based on the classification of gravels presented in
is obtained in the field. Additional recommendations for Chapter 5. The grading requirements for surface course
subbase material are given in Table 29. materials are based on the shaded area shown in Fig. 114. It
was found during the course of this study in a number of
Base Course Materials African countries that road failures due t o instability are
4. The materials for base course shall conform t o the caused by materials finer than the indicated band, and
requirements of Section 2 a and c. Modification of grading corrugations are more prevalent in materials coarser than the
requirements can be accepted provided the minimum CBR as indicated band.
recommended in Table 30 can be obtained in the field. Results from the pavement condition survey have shown
the importance of compaction in reducing the deflection of
Surface Course Materials the pavement. Compaction specifications in Ghana have been
5. Materials for surface course shall conform t o the reduced to prevent the breakdown of the lateritic gravels. A
requirement of Section 2 a and c. When possible, the special study was conducted on three materials which have
been used as base materials in Ghana. Three compactive efforts
were applied in compaction of the three samples. Post
TABLE 29 compaction gradations were conducted to determine the
percentage of fines generated during compaction. The results
Recommended Criteria For Subbase Material are shown in Table 31.
The results of this limited study indicate that the break-
down due to compaction is not appreciably different under
AASHO the Modified AASHO compactive effort than under the Ghana
Classification CBR Gradation PI LL -200 standard. The loss in gravel size is reflected by an increase in
sand-sized material, rather than fines. This may actually
A-1-a t o A-6 >20 Table 28 25 40 40 improve the gradation since many la teritic gravels are deficient
in sand-sized material. It is, therefore, recommended that
FIG 112 RELATIONSHIP OF LIQUID LIMIT AND C B R AT VARIOUS PERCENTAGES OF MATERIAL PASSING
RELATION ENTRE LlMlTE DE LlQUlDlTE ET CBR POUR DIFFERENTS POURCENTAGES PASSANT
LE TAMlS 2 0 0

PERCENTAGE PASSING NO. 2 0 0 SIEVE

POURCENTAGE PASSANT LE TAMlS 2 0 0
FIG. 1 13 RELATIONSHIP OF PLASTICITY INDEX AND CBR AT VARIOUS PERCENTAGES
OF MATERIAL PASSING THE NO. 2 0 0 SIEVE
RELATION ENTRE L'INDICE DE PLASTlClTE ET L E CBR POUR DIFFERENTS
POURCENTAGES PASSANT L E TAMlS 2 0 0
 t I I I FREWCU AFWORX- 11- 501
30 31 26 24

FIG 114 GRADATION LIMITS FOR SURFACE COURSE

LlMlTES GRANULOMETRIQUES DU REVETEMENT

TABLE 31

Results of Compaction Study

Sample Before After Sample Before After

No Compaction Compaction No Compaction Compaction

Gravel 33A 36.5 M 34.0 Silt & Clay 38A 4 M 4

G 34.0 G 3
> No. 10 S 36.0 S 3
- -

Sand 33A 58.1 M 60.0 Gravel 34A 75 M 65

G 60.0 G 73
< No.10, > No.200 S 58.0 S 73

Silt & Clay 33A 5.4 M 6.0 Sand 34A 23 M 33

G 6.0 G 25
< No. 200 S 6.0 S 24

Gravel 38A 45 M 37 Silt & Clay 34A 2 M 2

G 39 G 2
S 37 S 3

Sand 38A 51 M 59
G 58 M = AASHO Modified Compaction
S 60 G = Ghana Standard
S = AASHO Standard
Modified AASHO compactive efforts be applied t o lateritic Moisture conditions of subgrade materials examined during
soils, and that the following compaction requirements be the pavement condition survey showed the moisture t o be
adopted : reasonably close to AASHO Standard optimum moisture. In
this case one could not expect compaction greater than 90%
Base Course I, I1
AASHO Modified compact ion which is essentially 100%
and I11 95% AASHO Modified compaction
AASHO standard compaction.
Subbase 95% AASHO Modified compaction
Subgrade 9W0 AASHO Modified compaction
 CHAPTER 14

APPLICA TIONS OF TROPICAL SOILS IN CONSTR UCTION

OTHER THAN IN ROADS OR HIGHWA YS

BUILDING MATERIALS made from the same soil as the blocks. Block walls are
The use of earth as a building material must date back to generally used in areas of low rainfall, such as the savannah
the time men first emerged from the shelter of their caves. In and semi-desert areas of Africa. In tropical rain forests, the
many areas of extreme climate, earth was abandoned in favor short dry season does not allow sufficient time for drying and
of more durable materials. However, in the tropics earth construction. There is also extensive erosion and loss of
remained widely used. Since lateritic soils are found through- strength from wetting. When blocks are made in Ghana they
out the tropical belt of the world, these materials have been are stabilized with cement.
and are still used in the construction of dwellings and other Most of the unstabilized lateritic soils used in housing in
structures. Ghana are silty or clayey gravels, and clays of low t o medium
Handbooks on building homes of earth have been prepared plasticity. Fine grained clayey soils have higher dry strength
by the U.S. Agency for International Development and the than coarse grained or gravelly soils. Small amounts of coarse
Ghana Ministry of Construction. grained material reduce shrinkage cracks; a balance between
Since ancient times, many types of natural fibrous and shrinkage and strength should be maintained. If the soil is too
cementitous materials, such as straw, wood ashes, rotted fine-grained, shrinkage is too high. If the coarse-grained
plantain leaves, cow dung, cattle urine and coconut oil have content of the soil is high, shrinkage is reduced but so is
been mixed with the soil t o improve its strength, durability strength.
and waterproofing ability. In the last few decades much Swish block samples tested in Ghana (Bawa, 1967) showed
research has been carried out using bitumen, lime and cement that soils having the following characteristics gave the highest
to stabilize and improve the strength and durability of the soil. performance.
Gradation
Natural Materials Gravel (314" t o 2mm) 15 - 25%
Buildings of soil generally consist of swish or mud wall, or Sand (2mm - 0.074mm) 30 - 50%
of block construction. Silt and Clay (-0.074) 20 - 40%
Swish is a widely used term applied to lateritic soils that are
Atterberg Limits
not too sensitive t o moisture change, that swell and shrink
very little when wetted and dried, and that can remain fairly Liquid Limit
stable for many years. Swish buildings over 50 years old, for Plastic Limit
example, have been observed in Ghana. Swish walls are Plasticity Index
constructed by molding wet soil into a wall. They have the Linear Shrinkage
disadvantages of susceptibility to extensive erosion, loss of 4% Maximum
strength during the rainy season, and shrinkage cracks from
drying. Bamboo reinforcement in walls, commonly used in Strength
Ghana, reduces cracks and strengthens the wall. However, the A number of swish blocks were sampled from various
reinforcement is usually eaten by termites within 5 years and villages in Ghana. The compressive strengths of these blocks,
voids are then left in the wall. tested flat, ranged between 9 0 and 150 pounds per square
Blocks are made in forms or are cut from natural soil. The inch; whereas, for a one-story structure with light roofing a
term laterite was first used by Buchanan in 1807 t o describe a strength of only about 30 pounds per square inch is required.
material "soft enough t o be readily cut into blocks by an iron However, with unstabilized blocks, strength is not the only
instrument, but which upon exposure t o air quickly becomes criterion. The best indications of performance are the size and
as hard as brick and is reasonably resistent t o the action of air number of shrinkage cracks and the loss of strength on
and water". Many temples and buildings in Southeast Asia wetting. The native builder without a knowledge of soil
were constructed with this natural block. Although substantial mechanics develops a "feel" to determine the required amount
structures of this type have been built in Africa, for example of water in the molded soil.
in Sierra Leone, it appears that the formed block is more Surface coatings are usually applied t o earth walling to
common and conveniently used in construction. The blocks reduce erosion and cracks and t o give a more uniform
are hand or machine rammed into molds and allowed to dry in appearance. The use of ordinary cement-sand plaster is not
the sun. Clay block walls are smoother than swish walls and generally recommended as large slabs of plaster break off from
experience fewer shrinkage cracks. The mortar is generally the mud walls. Slurries of cement and lime are frequently used
but require extensive maintenance. Alcock (1956) recom- square inch is generally accepted. Fitzmaurice suggests--
re-
mends a wash made from one part of powdered red laterite ducing the 28 day wet strength to 100 pounds per square inch
mixed with 2 parts of a liquid made by boiling banana or when the gravel and sand content is on the order of 70%.
plantain stems in water. Bituminous cutbacks are sometimes In an earlier study of Ghana soils (Bawa and Gidigasu,
used and produce a durable waterproof plaster. Generally, 1967) it was found that almost all soils could be stabilized
about 5 t o 6 percent bitumen by weight is used. Roof with cement although some required uneconomically high
overhangs are also advisable to reduce erosion on the walls. percentages to achieve desirable strength and durability. The
amount of cement content was determined by the durability
test with a minimum loss of weight of 10% after 12 cycles.
Stabilized Materials
Less than 10% of these soils required more than 10% cement,
Soil stabilizers for construction of earth buildings appear to while the other 90% required from 4 to 9 percent. Generally, 6
be limited to use in blocks only, since block construction is t o 8 percent of cement should be suitable to obtain adequate
used more extensively than rammed walls. Of all the materials strength and durability; with proper selection of soils 4 to 5
used for stabilization, cement has so far proved to be the most percent should give satisfactory results. Fitzmaurice (1 95 8)
reliable (Sperling 1962). Lime stabilization of building soils is has recommended criteria for the selection of soil for
not in widespread use in Africa, possibly because some soils do soil-cement blocks. These are shown in Table 32.
not react sufficiently with lime to give adequate strength for The Indian Standard, IS 1725, states that satisfactory
block manufacture. In Chapter 9 it is noted'that by using results have been obtained with soils having a plasticity index
equal amounts of lime or cement on the same soil, much between 7 and i 2 . Webb, Colliers and Stullerheim (1950)
higher strengths and better durability are obtained for the recommend a maximum liquid limit of between 30 and 50
cement treated soil. The main benefit of the use of lime is to percent, 2 maximum plasticity index of between 10 and 20
reduce the plasticity and sensitivity to changes of moisture. percent and a maximum clay content of between 10 and 20
There is a slight pozzolonic cementing action, but with smaller percent.
strength increases than with cement. Where lime has been It appears that the required amount of cement cannot be
used, about 5 percent by weight seems to be the optimum satisfactorily predicted from any of the index tests. Sperling
with most lateritic soils (Bawa and Gidigasu 1965). (1962) found that durability cannot be predicted from the
Cement is by far the most widely used stabilizer for compressive strength results,that the wet compressive strength
building construction in Africa. The best method of building provides an indication of the load bearing capacity, but that
with cement and swish is to construct blocks of the mixed this bearing capacity will decrease on weathering. Durability
material and cure them before using them in building. These seems to be the most important requirement that determines
blocks are commonly referred to as "swishcrete". The blocks the overall performance of soil-cement blocks.
are rammed into molds by hand or machine.
By using cement with the soil, shrinkage is reduced and FOUNDATIONS
strength and durability are increased. Because of the wide
Lateritic soils, like any other group of soils, present certain
variations in soil properties, various amounts of cement are
problems as foundation materials, though there are no specific
used in swishcrete. Chapter 9 shows that there is no substantial
increase in the durability of a soil-cement-mixture if more than
8% of cement (1 part cement to 12% parts soil is used). Many TABLE 32
of the soils did not show a marked increase with more than 4%
Selection of Soil For Soil-Cement Blocks
of cement. Consideration should be given to the reduction of
disintegration of a cement-soil-mixture on exposure to rain. Type of For Permanent Urban For rural housing
The durability test (ASTM D559-69, Wetting and Dry Test on Location buildings with low rainfall
Compacted Soil-Cement Mixtures) has been generally used as a
For areas with
basis for selecting the cement content. The recommended limit Physical rainfall exceeding
for minimum durability is not more than 10 per cent loss of Properties 30 inches
weight after 12 cycles of wetting and drying. Lower minimum
values of around 5% are recommended for permanent building Clay Content Between 5 and 20% Between 5 and 30%
developments and for urban areas with more than 20 inches of
Sand Content 33% Minimum 40% Minimum
annual rainfall (Fitzmaurice 1958).
It has always been found that cement stabilized soils which Liquid Limit 40% Maximum 50% Maximum
satisfy the durability test have ample strength. Different
minimum requirements of compressive strengths for soil- Plasticity Index Between 2.5 to 22% Between 2.5 to 30%
cement blocks have been recommended. They vary from 180
Optimum Moisture Between 10 to 14% Between 7 to 16%
t o 235 pounds per square inch, when tested wet at 28 days. A (Proctor Content)
minimum average compressive strength of 200 pounds per
stability or deformation problems associated particularly with building within limits.
the process of laterization. The cementing action of iron oxide The raft is usually placed at depths between 6 8 ft. from
tends to improve the strength and deformation characteristics the ground surface, unless greater depths are desired for
of lateritic formations. provision of basement in the building. Such rafts usually have
The general physical and engineering characteristics of the allowable bearing capacity values higher than 4 tonslsq. ft.,
soils of typical profiles over various geological formations are though the structural loads seldom exceed 2.5 - 3.0 tonslsq. ft.
shown in Table 33. It is seen from Table 33 that most of the for the type of medium to heavy structures built in Ghana (5 -
profiles in lateritic soils have satisfactory shear parameters. 12 story buildings).
This has also been confirmed by a large number of subsoil
investigations carried out on various sites throughout Ghana. Pile Foundations
Allowable Bearing Capacity Piles are used on a number of projects where the upper
strata, up to 15-20 ft. from the surface, are weak or have a
The bearing capacity of a soil, as is known, depends on the tendency to undergo high deformations when loaded. This is
type of foundation, depth of foundation, position of water often the case in lateritic soils formed over rocks having high
table and shear parameters of the soil. Although it is difficult percentage of mica, such as mica schist, muscovite granite,
t o suggest any specific figures of bearing capacity in different etc., or in soft clays produced as a result of weathering of clay
types of formations, it is possible on the basis of experience to shales. In addition, areas covered by river sediments have also
make the following general remarks on the bearing capacity shown very unpredictable properties of the soil and therefore
values in relation to various types of foundations in lateritic the need for ~ i l efoundations. Little use has been made of
soils. bored piles in West Africa; pre-cast piles are used on most of
the projects.
Spread Foundations The bearing capacity of a single pile depends on many
Foundations of this type include strip footings, individual factors and cannot safely be predicted on the basis of
column footing, combined column footings, etc. These are the generalizations. Because of variations in the depth of weather-
most economical foundations for small to medium sized ing, neighboring piles may often reach their working capacities
structures. In laterite soils generally such footings are placed at at widely different depths.
5-8 ft. from the ground surface. Shallower depths are generally
not recommended because of the frequent wetting and Settlements
desiccation of soils near the ground surface. As most lateritic The compression indices for non-micaceous lateritic soils
soils are at least slightly expansive, structural foundations are low with the result that consolidation settlements of
placed at shallow depths tend to cause cracks in the wall ordinary structures are likely to be small. For micaceous soils
panels on the buildings a few years after their erection. This is and clay shales, the compression indices are rather on the
a serious problem in small buildings with foundations laid at higher side, and therefore, on many sites the settlements of
shallow depths. structures present some problems. In a few instances, especial-
Generally, spread footings placed at a depth of about 5.0 ly in limestone terrains, highly compressible soils may be
ft., or so, and having widths between 4-6 ft., have bearing encountered just above sound rock.
capacity values even in weak lateritic soils of more than 2.0
tonslsq. ft. In profiles where laterization and desiccation have
produced a certain amount of cementing, the values of bearing SLOPES I N LATER lTlC Sol LS
capacity can be as high as 6-8 tonslsq. ft. There are, however, Stability and erosion are two factors that should be taken
local zones at certain sites where the bearing capacity values into consideration when designing slopes. The stability of any
are rather low. The weakness is associated with pockets of very slope is derived from its shear strength which, in turn, is
loose and weak soils,of transported soils or with the presence affected by the moisture and density of the slope material.
of a high proportion of certain minerals in the bedrock, such The statement is as valid with respect to the stability of slopes
as feldspar or biotite and their accumulation in the profile on tropical or lateritic soils as it is with respect to the soils of
during the process of weathering. In such exceptional cases the temperate zones. However, the properties of lateritic soils may
bearing capacity values at times are as low as 112 - 314 tonslsft. change with time and these changes generally improve the
stability of slopes, thus in some instances possibly allowing
Raft Foundations steeper slope angles and more economical road construction.
Beneath medium to heavy structures (generally more than 5 The change of properties is due to dehydration of iron
stories) raft or reinforced concrete mat foundations are often hydroxide on the surface of the slope. The iron oxide, which is
used on lateritic soils. The provision of raft foundations is very carried by ground water permeating through the slope, forms a
seldom dictated by the lack of bearing capacity in lateritic hard layer of iron-cemented materials which is more resistant
soils. It is generally necessitated to keep the settlement of the t o erosion and also increases the stability of the slope. During
 TABLE 33

Engineering Properties of
Lateritic Soils

- -

Dry Density
Soil Depth PCF LL PI C' PSF @' Cc

AIRPORT ROAD KUMASI

Hard reddish
brown micaceous (3-9'6"
lateritic clay
with lateritic
gravels.

PADMORE ROAD ACCRA

- 26 9 -

KUMASI - MAMPONG ROAD M I LE 22

112 43 24 750

Fine Sand

Grey silty
Clay

Micaceous
silty clay

U.S.T. KLlMASl
Highly plastic
lateritic 103 47 18 1640
residuum
 TABLE 33 (Cont.)

Engineering Properties of
Lateritic Soils

- - - - - - -

Dry Density
Soil Depth PCF LL PI C' PSF 4' Cc

U.S.T.KUMASI (Cont.)

Plastic to
silty
laterite
residuum

ACCRA DODOWA ROAD MI LE 35

Sandy clay 115 20 6 430
Red
Mottled sandy
clay with
pea gravel

BLACK STAR SQUARE - ACCRA

97 62 39 1870

Grey mottled
clay with high
plasticity

the dry season desiccation causes an irreversible induration of 1970). It was found that the existing slopes had reached their
the lateritic soil which also improves the resistance to erosion long term stability and observations and testing of their
and increases the shear strength. This process has been present state would determine the stable slope angles. The
previously discussed in Chapters 3 and 5 of this report. slopes were divided into two groups on the basis of the
geologic formation over which the soil was formed, whether
Ghana granitic rocks or schistose or phyllitic rocks. The areas were
In Ghana, during a highway study for the Accra-Kumasi located in tropical rain forests of the ferrallitic and ferrisol soil
route, a study of existing soil slopes was made (Liautaud groups. A total of 25 slopes were investigated, 12 in the schist
or phyllite areas and 13 in the granite areas. Soil samples were The side slopes that have recently been recommended for
taken from the existing slopes and the following average test 500 miles of road in the southwest sector of the Ivory Coast
results were obtained. are presented below and also plotted on Figure 115. It is
interesting to note that, for the same height, slopes with
Granite Schist or Phyllite schistose material are steeper than those for granitic materials.
% gravel 15 5 In schistose material for a maximum height of 20 feet a slope
% sand 45 15 angle of 81 degrees is recommended, while for the same height
% silt 25 50 in granitic materials a slope angle of only 72 degrees is
% clay 15 30 recommended. The maximum heights of granitic slopes are
Liquid Limit 50 65 below the critical lines found for slopes in Ghana, whereas the
Plastic Limit 32 38 maximum heights for schistose materials are above the line. It
Plastic Index 18 27 may be that the granitic soils are sandy and generally less
Moisture Content 9% 21% stable than phyllite slopes.
Dry Density 100 pcf 90 pcf
Unconfined Compressive 2 1 to 80 psi 10 to 18 psi Materials in Cutting Height Slope (H: V.)
Strength Schistose or Clayey 0-20 feet 1:5 to 1:6
Slope Between 1:1 & Between 1:2 & material >20 feet 1:3 or
1:6 (H:V) 1:9 special study
Height Between 6 & Between 7 and Granitic material 0-10 feet 1:5
33 feet 25 feet 10-20 feet 1:3
>20 special study
When the test results of samples from beneath the slopes
were compared with those of boreholes in similar geologic Liberia
formations, it was found that the slope materials exhibited
unconfined compressive strengths more than twice those from Considerable difficulties were encountered with slope fail-
the boreholes in spite of higher moisture content. It was also ures during construction of 270 kilometers of railway line in
determined that .the plasticity indices were less. The plasticity Liberia in 1962. The soils found along the railway are mainly
indices or materials taken from boreholes and pits plotted
above the "A" line. Those of materials from the existing slopes
plotted below. This change indicates that an irreversible
process had taken place during induration. The following cut 0 GHANA -GRANITIC
6HA1A - GRANITIOUE
slopes were recommended for this study:
CRITICAL HEIGHT AND
SLOPE ANGLE FOR
GRANITIC SOIL SLOPE XL LIBERIA

Materials in Cutting Height Slope (H: K )

O IVORY COAST- SCHISTOSE AND
Highly weathered schists <20 feet 1:4 PHYLLlTlC
COTE D'IvURE- SCHISTOSE E T R F I U A D E
XMlC MALAYA-IGNEOUS COARSE
Shales, phyllites-mainly
argillaceous >20 feet 1:3
Highly weathered granites <20 feet 1:3 41 MALAISIE- SEDIMENTAIRE- FERME

The records of 20 other soil slopes constructed in Ghana

from 7 to 10 years ago were also studied. These slopes were
also located in the ferrallitic and ferrisol soil groups and were
over granitic bedrock or over schistose or phyllitic bedrock. A
graph showing the angle and height of all the slopes studied in
Ghana is shown in Fig. 115. Slopes that had failed are shown
shaded and connected by a line to the present stable slope.
The lines of critical height and slope angle are shown for
granitic soils and for schistose or phyllitic soils. From these
lines it appears that eight types of soil can stand vertically with
a maximum height of about 15 and 10 feet respectively. The
soils derived from schists and phyllites should be stable with a SLOPE ANGLE, DEGREES
ANGLE DU TALUS - DEGRES
height of about 35 feet on a slope of 1:l and the granitic soils FIG.115 CRI'TICAL HEIGHT AND SLOPE ANGLE FOR LATERlTlC SOlL
with a maximum height of 50 feet for the same slope angle. HAUTEURS E T ANGLES CRITIQUES POUR SOLS LATERlTlQUES
residual over granitic bedrock. The soil group is ferrallitic. content in the "soft" rocks. Bulman found that 30% of the
Some cuts up to 92 feet deep were made without reaching coarse grained igneous rock slopes had slipped, while slips had
sound bedrock, although at many other places sound rock was occurred in only 15% of the fine grained rock slopes. In the
found at shallower depths. The design slope for the cuts was "firm" sedimentary slopes, 30 to 44% had slips, while there
1:2.1 (65") (H:V.) up to 30 feet. Cuts deeper than 30 feet were slips in 92% of the "soft" slopes.
were provided with a 7-foot wide bench at 23 feet; other In analyzing the results of his study, Bulman found that
benches were added at 46 and 69 feet if the cuts exceeded height did not affect the mean slope angle in any one category.
these depths. Below the bottom bench the slope was 1:1.85 The recommendations for stable slope angle are presented in
and above the bench the slope was 1:2.1. These specified Table 34. The maximum recommended slope angles are
slopes are plotted on Fig. -115. plotted on the graph in Fig. 115 at a slope height of 25 feet.
Twenty eight slopes, or about 2Wo of all the cuts deeper The igneous rocks, mainly granites, plot below the recom-
than 3 3 feet, failed by sliding during construction. One slope mended maximum heights for the granitic soils of Africa. The
72 feet high was initially cut at 62 degrees (1:1.7). An coarse-grained igneous Malaysian soils would be stable at 25
extensive slide occurred and the slope was cut back to 39 feet with a slope of 70 degrees or at about 3 3 feet at the
degrees (1.25: 1) where it has remained stable. recommended slope angle of 60 degrees. Since these values are
In many cases fallouts occurred along the planes where based on average results they are conservative and fall below
bedrock had been exposed or along the joints and fissures the critical curves. It is usual practice in Malaysia to slope
inherited from the parent rock during the process of decompo- shallow cuttings in lateritic soil at 70" or 80" and the majority
sition. In some cases the smaller fallouts initiated larger slides, of these are stable (Millard, personal communication). In
especially where the fallen materials blocked the drainage deeper cuttings, the top part is usually stable and slipping
ditches. Many of the slides occurred where the orientation of occurs in the weathered zone beneath the soil. Slipping seems
the joints had an unfavorable attitude. An inspection of the to be more common during the first years of the cutting, then
cut slopes along the railway line was made by one of the Lyon the slopes become more stable.
team engineers in the early spring of 1970. It was observed
that the overall condition of the slopes was generally good. Embankments
During heavy rains in 1966 some major slides had occurred. At Fills or embankments composed of lateritic soils can often
least one of these slides was caused by seepage directed toward be placed at steeper slopes than most temperate-zone soils
the slope by a stream deviation located at the top of the slope. because of the relatively greater strength and permeability,
It was also observed during this inspection tour that where other properties being equal.
the slopes were generally steep, 1:1 or steeper, the erosion was The embankment slopes recommended in the highway
small. However, on the slopes that were laid back, the erosion study in Ghana were 1.5: 1 (H:V.) for all fills over 3 feet in
was extensive. In a report on the construction difficulties on height in lateritic soils, and 3: 1 for fills less than 3 feet. In the
this project, Hayes (1963) studied the effect of erosion. He Ivory Coast the recommended slopes are 1.5: 1 for sandy
found that sandy silts eroded 5 to 10 times more rapidly than
gravelly lateritic soils. The resistance of the lateritic gravels is
probably due to surface desiccation. The other factors TABLE 34
affecting erosion of lateritic soils are the same as for other
soils, mainly slope angle, density and grading of the surficial Recommendations for Stable Slope Angles of Western
materials, and extent and nature of ground cover. Malaysia (After Bulman)

Malaysia Average Recommended

Bulman (1967) studied over 400 rock and residual soil cut Slope Slope
slopes in Western Malaysia. All of the cuts were over 25 feet in Angle- Angle-
height and the majority were over 10 years old. He classified Slope Category degrees degrees
the slopes into 4 main categories: (I) hard rocks; (2)
weathered igneous rocks; (3) weathered sedimentary rocks; Weathered Igneous Rock
and (4) volcanic rocks and conglomerates. The weathered Coarse Grained
igneous rocks were divided into fine-grained and coarse-grained Fine Grained
soils, with 0.6 mm. as the dividing point. The weathered
sedimentary rocks, which consisted mainly of shales, schists Weathered Sedimentary Rock
and phyllites, were divided into groups with 20% more or less Firm with 20% or more 44 40-50
of interbedded sandstone or quartzite (firm), and those with Firm with less than 20% 45 4045
60% or more of silt (soft). Clare and Newill (1965) have Soft 37 35-40
indicated a correlation between the slope stability and silt
materials and 1:1 for cohesive soils. In Liberia on the railway The design of adequate drainage facilities can greatly
line the slopes were 1.5: 1. At this slope angle the embankment improve the stability of a slope. However, a poorly designed
slopes appeared to have functioned satisfactorily, although drainage system can be harmful. Many of the deeper cuts
considerable local instability near the edges was experienced investigated in Malaysia were benched. In Liberia, benches
during construction where compaction with heavy equipment were recommended for cuts deeper than 30 feet. In Ghana,
was not practicable. No major fallouts or slides were observed some of the cuts deeper than 20 feet were benched. The
on the inspection tour. Erosion was slight and a moderate purpose of benches is to intercept and prevent water from
growth of vegetation had covered most of the embankments. running down the entire slope face, and to catch fallouts.
Generally, embankments present fewer problems than cuts Benches must be regularly maintained and the flat surfaces
because their composition is more homogenous and there are protected against erosion if they are to be effective. In none of
no water problems other than erosion. For this reason it may the slopes investigated had these measures been carried out
be possible with some soil types to achieve stability at steeper fully. Because fallouts are not effectively cleared from
slope angles in fills than in cuts (Bulman, personal communi- benches, surface water may accumulate and damage the slopes.
cation). The factor of safety of embankments may be easier to Unless a regular maintenance program is established for
predict by normal soil tests and theories. benches, it might be advisable to omit them in cuttings up to
Because of the unusually high rainfall intensities associated 50 feet deep (O'Reilly, 1967). "Cut-off' drains at the top of
with tropical forest areas, extensive slope protection measures cut slopes would be effective in reducing the flow of water
should be undertaken to reduce erosion. Hayes (1963) showed onto the slope. They generally do not require the extensive
that reasonably good slope protection could be obtained by maintenance as do mid-slope benches.
covering fills composed of compacted silts and sands by a layer Lumb (1962) observed that the majority of slides in the
of indurated lateritic soil. Hong Kong area iook place in decomposed volcanic rock,
while decomposed granite slopes were usually more stable. He
found that the effect of the degree of saturation on strength
Discussion was more pronounced for the decomposed volcanics than for
Theoretical stability analyses of cut slopes, based on shear the decomposed granitics. Since the stability is dependent
parameters from lateritic soils obtained prior to construction, upon the strength and the rate of infiltration of water into the
generally fail to give reasonable slope angles on account of the slope, an impermeable layer on the slope and beyond the top
influence of the relict structure of the rock in the weathered of the slope would be of great value. The effect of saturation
soils, and because the hardening effect cannot be accurately on volcanic soils is evident in a comparison of the similar red
predicted or reproduced in the laboratory. Only an empirical volcanic soils of Malaya and Kenya. Bryan and Ibrahim (1948)
design method of studying existing slopes composed of the regarded the soils in Malaya as the worst type for slopes, while
same material is at present capable of giving a reasonable the volcanic clays from Kenya are considerably stronger. This
indication of the slope performance (Bulman 1967). Slopes difference in performance may be due to the higher rainfall in
that are initially too steep subsequently slip to a flatter slope Malaya (Clare and Newill 1965).
angle. At any particular site the selection of slope angle should The growth of vegetation on a slope face greatly increases
be based on local conditions and economics. the stability by reducing the runoff of water on the slope.
The classification of safe cut slopes on the basis of geology Because of the sterility of most lateritic soils, difficulties have
could be developed into a more comprehensive classification been encountered in attempting to provide a cover of
that would include landform, drainage characteristics and vegetation. It may take one or two years before a dense
more geological divisions. For instance, mica occurs in Ghana covercrop takes hold (Hayes, 1963). On cut slopes, steeper
and highly rnicaceous soils may be included as a separate angles minimize the erosion problem. Natural vegetation
category. The critical lines presented on Figure 115 should covers the slope only after several years. This natural vegeta-
serve as a useful working guide for cuttings in granitic and tive cover was observed on many of the stable older cut slopes
schistose soils. in Ghana.
Although many methods are available to analyze the Bulman (1967) found that bitumen emulsion or cut-back
stability of cut slopes, no analytical method is capable of bitumen surface treatments were most successful when applied
including the influence of the detailed geological environment. to the stiffer granitic soils. A single application has a life on
When the soil undergoes the final stage of decomposition and the order of 3 to 5 years. On the other hand, such treatment is
becomes lateritic, the natural joints and fissures are inherited vulnerable to rapid deterioration if a small defect develops;
from the parent rock and effect the stability of slopes. Hayes test covers of this type on the railway cuts in Liberia failed
(1963) concluded that all the slides along the Liberian railway during the first rainy season. The surface stabilization of slopes
had taken place along the joints and fissures inherited from the with lime or cement has been successfully used in Malaysia.
parent rock. Clare and Newill (1965) observed that a high However, this treatment is not usually economically justified
proportion of slips occurred in laminated and fissured for general application. It is used mostly for the repair of local
material. failures.
EARTH DAMS mined to a great extent by the moisture content in the borrow
Earth dams have been constructed since earliest times. In areas, as will be discussed subsequently. The maximum dry
the last 40 years, with the introduction of soil mechanics and density at optimum moisture content may be as low as 70
the knowledge gained through experience, the heights of earth pounds per cubic foot for volcanic clays to as high as 125
dams of residual and transported soils have increased. Dams of pounds per cubic foot for granitic sandy clays. Shearing
residual soil have performed as well as those of sedimentary strengths for clays of both types may be similar. These
soil even though the liquid limits may be higher and the dry properties can be used to good advantage in dam construction
densities at optimum moisture content lower for residual than by placing the heavier materials at lower levels and the lighter
sedimentary soils with similar strengths and compressibilities. materials at higher levels. This arrangement will increase the
Terzaghl (1958) noted that earth dams constructed with overall stability of the slopes and dam. The compaction of
weathered volcanic clays and similar materials had performed soils into the dam is one of the most important aspects of
satisfactorily for many years. He listed several such dams, in construction. It influences the important embankment proper-
Java, Australia, the Philippines and Kenya. In contrast, he ties, such as compressibility, shear strength and permeability.
mentioned that another Javan dam, built of sedimentary clay In most of the dams studied, 10W0 of standard Proctor
fill, failed during construction even though its Atterberg limits maximum dry density has been specified as the minimum
were lower and it had a flatter up-stream slope than its lateritic compactive requirement. Terzaghl (1958) showed that the
counterpart. The factors that cause the unusual yet favorable process of compaction may alter the physical characteristics of
properties of lateritic soils were discussed in previous chapters. some tropical soils. He found that the liquid limit of the
The pertinent features, along with the physical and engi- Sasumua fill was roughly equal to that of samples from the
neering properties, of several dams constructed with lateritic natural ground, but that the values of plasticity index were
soils are given in Table 35. much higher for the fill. This breakdown phenomenon has
The initial selection of embankment slopes depends on the been previously discussed in Chapter 5. Studies are needed to
site geology, embankment materials and experience with dams investigate the change in strength characteristics with the
of similar types. The initial design may be modified after alteration of physical properties.
detailed analysis. The upstreams slopes of the dams listed in In order to achieve maximum density with minimum effort,
Table 35, are generally between 2.5: 1 and 4: 1 (horizontal to a soil should be compacted close to optimum moisture
vertical). Terzaghl (1958) was of the opinion that the original content. Opinions vary widely as to the most desirable
design of the Sasumua dam with an upstream slope of 3 : l placement water content (ASCE-USCOLD, 1967). Some re-
(upper part) and 3.5:l (lower part) would have functioned quire fill to be placed on the dry side; others on the wet side;
satisfactorily. The downstream slopes of the listed dams are and some at optimum moisture content. Placement of the clay
either 2:l or 2.5:l. Sherard et al. (1963) state that, as a fill on the dry side of optimum usually results in higher shear
general rule for homogeneous dams of fine grained soil, the strength and permeability, but may cause a susceptibility to
higher the dam the flatter must be the slopes; however, for cracking or excessive settlement upon inundation. Placement
thincore dams in which the bulk of the embankment consists on the wet side results in lower permeability and a greater
of p e ~ o u sgranular soil or rock, the allowable slopes are ability to deform without cracking. The low permeability is
almost independent of the height. They also suggest that associated with a slow dissipation of pore water pressures. De
variable slopes should be considered for all earth dams higher Graft-Johnson et al. (1968) suggest that low lateritic earth
than 100 feet. These considerations may not apply to dams dams should preferably be compacted wet of optimum
constructed with lateritic materials. Lumb (1962) showed that moisture content whereas high dams should be compacted
the shear strength of a highly weathered granitic sandy silty slightly on the dry side.
clay containing from 40 to 50% fines (passing No. 200 sieve) In most tropical wet climates, it may be difficult or
was hardly less than that of the coarse constituents (retained impracticable to reduce the natural water content to near
on the No. 200) alone. The effective-stress friction angle was optimum moisture content. If conventional construction
30 degrees and the cohesion zero. Similar strengths may even methods cannot be used to reduce the natural water content,
be displayed by weathered volcanics containing hgher propor- the dam must be constructed at moisture contents above
tions of fines. The Sasumua clays exhibited effective-stress optimum and the design must be based on the corresponding
friction angles between 30 and 36 degrees and cohesion strengths. It was shown in Chapter 5 that the compaction
between 540 to 630 pounds per square foot. Terza& (1958) characteristics of lateritic clays, especially those containing
felt it justifiable to assume that the shearing resistance of the halloysite, are different when compacted from a dry state or
Sasumua clays would not decrease in the course of time. As from the natural state. The laboratory compaction tests should
more experience is gained with lateritic soils and dams, it is simulate the field conditions which the soil will be
quite probable that the upstream and downstream slopes of experiencing.
future dams may become steeper. Settlements of lateritic earth dams do not appear to be a
The compaction characteristics of lateritic clays are deter- problem. During the design stages of the Sasumua Dam,
 TABLE 35

Properties of Some Earth Dams

in Tropical Soils

- - -- --- -

Year Embankment
of Slopes Proctor
Comple- UP Down Founda- Clay pcf CM/SEC C
Name tion Height Stream Stream tion <
Soil
-- LL PI -2u Max WOIT k psi 4'

Akosombo 1965 370' inorga- 33-45 17-24 26-46 105-122 13-19 1.0 to -9 15-24 7-12
main dam nic clay 8.8 x 10
core ' and
weathered
saddle Da-m- - L -- - -
shales
ljo' - -
1965 sand- silty & 25 8 12 125 11 - - -
Core ~ h a n a ' stone sandy
---- - - -- - -- -- - - - - - - -- - - -- -- clays
Barekese 1970 6!? 2.5:l 2:l 53 29 30 105 22 4.1 10-8 10 psi 37"
Dam -
- -- -
~ h a n- -a- ~- - --- 3.5:l 3:l
Lirnoeiro - 35' 2.5:l 2:l Gneiss clay 32-55 10-25 8.40 97 21 4.0 to 16 6 psi 15"
~razil~
Bariri - 29' 3.5:l 2.5:l Basalt clay 28-45 12-28 10-40 107 19 1 10-7 4 psi 29O
CI -
~f!z?!~- -
Q\ Chavantes 91' 35:l 2:l Basalt clay& 28-58 9-29 19-47 99 23 -
o 1 psi 29O
~razil~ sand- sandy 16-48 0-25 6-30 112 12 - 33O
- - - . -- - - - -
-- - - 142' 3:-1-- stone- -
---- - clay - -
Silvan 1931 2: 1 mudst- basaltic 45-86 9-36 50 69 49 - - -
~ustralia' w/5 one & clay
Berms shales -
Sasumua 1955 110' 4:l 2:l Basalt clay 87 33 40-50 70 49 0.8 to 3.8 to 3Ck36O
~en~a' w/2 Tuff 3x10-7 4.4
-- - -- -- - -- -- ----p-p
Berms
Lumot 1950 88' 3:l 2.5:l weathered Red 106 32 - 69-90 30-45 6 to - -
~hilli~ines' 5.5:l 3 Berms Tuff & Clay 14 x
- - -- - - - -. - Lavas
Guma - 222' 2: 1 2: 1 Gabbro Clay 45-90 32-58 - 83 41 3.1 x - -
Sierra h o n e 4 3: 1 2: 1 to
3.5:l w/2 6.5 x
Berms
Hills Creek 1960 340 2:l 2.25: 1 Basalt coarse 30-50 - - - 24-28 6.8xl0-' - -

Oregon, 2.25:l clayey

USA' 2.5:l gravel
Tiipandenjang 1929 108' 2.5:l 3:l Anderite red 115 58 6-33 72 48 - - -

~ava~ w/4 and clay

Berms Tuff

'Peck - (personal communication) 2 ~ . ~ . ~ .3 ~ a.r g aet

s al., 1965 ' ~ i x o n (1963)
calculations based on soil properties determined from conven- Fitzmaurice, R., 1958. Manual of stabilized soil construct ion
tional tests predicted extremely high settlements in the dam for housing. Tech. Assoc., New York: 125p.
and foundation material. After the dam was completed the Ghana Ministry of Construction, 1962. Handbook on rural
maximum settlement on the crest of the dam reached 1 inch housing, village and urban development in Ghana.
and then ceased to increase. Terza& (1958) noted that the Highway Research Board, Stability of rock slopes, five reports,
settlement of the crest of volcanic clay earth dams hardly Highway Res. Board No. 17,1963.
exceeds 112% of the height of the dams. He also found that Indian Standards, I. S. 1725, 1960. Specification for soil
the soils of the Sasumua dam with liquid limits about twice cement block use in general building construction. India
those of normal clays showed nearly the same compressibilities Stand. Inst., New Delhi, 1961.
as did the normal clays. Liautaud, Gerrard, 1970. Side slopes. Investigation of safe side
slopes in Ghana (unpublished).
REFERENCES Lumb, D. W. Decomposed Granite as Fill Material with
Alcock, A. E. S., 1956. Better earth building in the tropics. particular reference to Earth Dam Construction. Sympo-
Corona, London: 9p. sium Hong Kong Soils, Hong Kong Joint Group, 1962
ASCE-USCOLD. Current United States Practice in the Design pp. 57.
and Construction of Arch Dams, Embankment Dams. O'Reilly, M. P., 1967. Investigation into road building practice
Concrete Gravity Dams, 1967. in the tropics. Third Simposi. Sabre Pesquisas Rodoviorias,
Bawa, N. S., 1967. Manual on building materials in Ghana. Rio de Janeiro.
Build. and Road Res. Inst., Ghana: 20p. Sherard, J. L., et al. Earth and Earth Rock Dams. John Wiley
Bawa, N. S., and Gidigasu, M. D., 1965. Use of stabilized soil & Sons, New York, 1963.
as building material. Build. Res. Inst., Ghana: Report No. Sperling, R., 1962. Soil-cement blocks, effect of soil composi-
9: 18p. tion on durability and other properties. West Afr. Build.
Bulman, J. N., 1968. A survey of road cuttings in western Res. Inst., Note No. 14, Accra:
Malaysia. Proc. of the Southeast Asian Reg. Conf. on Soil, Terzaghi, K., 1958. Design and performance of the Sasumua
Bangkok: 289-300. Dam. Proc. Inst. Civil Eng., vol. 9: 369-395.
Bryan, J. J., and Ibrahim, Y. B. H., 1948. Malayan roads and United Nations, Land and Water Survey in the Upper and
some soils which support them. Proc. Conf. on Civil Eng. Northern Regions of Ghana, FOA, vol. I11 & IV, Rome,
Problems. Overseas, London, Inst. of Civil Eng.: 137-156 1967.
Clare, K. E., Newill, D., 1965. Some characteristics of road U. S. Agency for International Development. Handbook for
cuttings in Malaya. Road Res. Lab. Note No. building homes of earth. Department of Housing and Urban
LN/785/KECDN, Ministry of Transport, Harmondowarth, Development. Washington: 159p.
(unpublished). Vargas, M. Silva, F.P., and Tubio, Mr., 1965. Residual clay
De Graft-Johnson, J. W. S., Bhatia, H. S., Gidigasu, M. D. The dams in the State of Sao Paulo, Brazil. Sixth Int. Conf. on
Engineering Characteristics of Lateritic Residual Clay of Soil Mech. and Found. Eng., SMFE, vol. 578.
Ghana for Earth Dam Construction. Symposium on Rock Webb, T. L., Collier, T. F., and Stutterheim, N., 1950.
and Earth Fill Dams, New Delhi, India 1968, vol. 1 pp. 94. Properties of compacted soil and soil-cement mixtures for
Dixon, H. H., 1963. A review of test results on halloysitic soils use in buildings. So. Africa Coun. of Sci. and Ind. Res., Nat.
and their performance in the field. Third Reg. Conf. for Build. Res. Inst .: 62p.
Africa on Soil Mech. and Found. Eng., vol. 1: 183-185.
 CHAPTER 15

A REVIEW OF THE LITERATURE ON TROPICAL BLACK CLAYS

AND BLACK COTTON SOILS

INTRODUCTION occurrence in the literature and because it is widely used in

Tropical black clays, often called black cotton soils, are the many parts of Africa.
major problem soils of Africa. They are poor materials to An appropriate engineering definition for tropical black
employ for highway or airfield construction because they clays is "dark gray to black soils with a high content of clay,
contain a large percentage of plastic clay and are often usually over SO%, in which montmorillonite is the principal
expansive, swelling as water is absorbed. clay mineral and which are commonly expansive". The main
In many areas where these soils occur, there are no suitable characteristics common among most tropical black clays are
natural gravels or aggregates. Most deposits cover sufficiently high clay content, dark color, tendency t o expand and shrink
large areas that avoiding or by-passing them is not feasible. with changes in moisture, and appreciable plasticity in the clay
Few roads constructed through these soils have proved fraction. It should be emphasized that expansion occurs only
satisfactory. Many have failed completely. when the black clays are subjected to an increase in moisture
Most of the tropical black clays are formed residually by content. If they are" maintained in either a desiccated or a
weathering of basic rocks such as basalts. Alluvial deposits also saturated condition, expansion cannot occur. Unfortunately,
occur but, except in Sudan, are much more restricted in size. construction activities often at least modify the moisture
The Lake Chad Basin is the only extensive lacustrine deposit content. In such modification lies the problem with the black
of black clay soils in Africa. clay soils, as with all expansive soils. Designs must allow for
expansion due to increasing moisture contents, or else the
DEFINITIONS changes in moisture must be controlled in some manner.
The term "black cotton" soils is believed to have originated
in India where the locations of these soils are favorable for OCCURRENCES
growing cotton. Many other terms have been applied locally, Indian occurrences of black cotton soils have been
such as "regur" soils in India, "margalitic" soils in Indonesia, described by a number of authors. Tamhane and Sen (1954)
"black turfs" in Africa, and "tirs" in Morocco. found that these soils were formed mainly over basic volcanics,
Bal (1935, p.261) offers the following definition: "soils part of the Deccan basalt series. They state that the soils are
derived from the weathering of trap rocks, in particular, which not produced from acidic rocks in India. Aganval and Mukerji
are black, heavy, and climatologically suited to the growth of (1949); Tamhane and Namjoshi (1959); and Basu and Sirur
cotton, are known as black cotton, or regur soils." (1 954) also agree that the soils are derived from basalts of the
Mohr and Van Baren (1959, p.411) proposed the term Deccan series. Bal (1935) found, however, that they are
"margalitic" soils w h c h they define as "black or grayish-black, occasionally located over gneissic rocks. Uppal (1965) con-
gray, or in the eroded phase grayish-white, heavy l o a m or siders "black cotton" soils one of the major soil groups in
clays: they crack when dry and swell when moistened; they India, covering an area of 200,000 square miles.
are mostly rich in alkaline earths; horizons of calcium Australian occurrences have been described by Hosking
concretions develop sometimes or lime concretions are found (1935, 1940) and Hallsworth (1 952), among others. Hosking
scattered through the profile; they are characterized by found .that climate was important in development of soil
montmorillonite or other minerals of the montmorillonite types, and that in areas where red soils could be expected
group as the clay compounds." under the prevailing climate, black clays would develop if the
Oakes and Thorp (1950) proposed the term grumusol (from drainage was poor. Hallsworth concluded that soil color is
the Latin grumus, little heap or hillocks) for dark clay soils associated with leaching; dark soils are less subjected to the
which have formed under varying climatic conditions but with leaching process.
distinct wet and dry seasons. They include in this Group Black clay soils in Africa have received considerable
rendzina (shallow gray to black soils derived from limestone); attention. Black soils over basalts and dolerites in Basutoland
black cotton or regur; and tirs of Morocco. have been examined by Pollard (1964). Kenyan black cotton
More recently, Dudal (1 965) has adopted the more descrip- soils have been investigated by Dumbleton (1 963); Strongman
tive term "dark clay soils". This seems preferable to most (1953); and Williams (1961). The black clays of Ethiopia,
other pedological terms. The terms "tropical black clay soils", formed over basalts of the Ethiopian ,and Somalian Plateaus,
or "black clay soils", and "black cotton soils" are used more have been described by Morin and Parry. Stephen (1953)
or less interchangeably in this report. Although the first term believes that the soils derived from pre-Cambrian hornblende-
is preferred, black cotton is retained because of its extensive garnet gneiss in Ghana are typical black cotton soils. Aubert
(1941) describes black clay soils over basalt in Madagascar in South Africa has been shown to produce black clays by D.
(now Malagasay). Pedological descriptions include deposits in H. Van der Merwe (1964, 1962), Weinert (1965) and D. H.
Angola (Azevedo and Da Costa, 1954), Cameroon (Laplante, Weinert (1 964).
1954), Togo (Leneuf, 1954); and the Congo (Sys, 1959). The distribution of dark clay soils in Africa is given in Table
Occurrences in southern Africa have been well documented. 36 taken from Dudal (1965). This table lists a number of
Van der Merwe (1962) has described the black clays in South fluvial and alluvial deposits, made up of heterogeneous soil
Africa. Maure (1928) has discussed the formation of black mixtures and usually from limited areas. Exceptions are the
clays in Rhodesia. The weathering of the basic intrusive rocks Lake Chad Basin and Sudan. The distribution of the major

TABLE 36

Distribution of Dark Clay Soils in Africa

(after Dudal, 1965)

Estimated coverage
Country Area (million acres)

Angola Valleys o f the Cunene and Cubango,

region of Calete and southwestern
part of the country

Botswana West of Livingstone and Tuli; large

areas around the Okovang and Makarikari
Swamps

Cameroon Logone Chari Basin and part of

the Chad Basin; peneplain of Kaele
and Marona region

Chad Chad Basin and scattered areas

Congo Ruzizi Plain, valleys of the

Semliki and Lufira

Dahomey Northern part of the country and in the

region of Divo

Ethiopia Rift Valley and Ethiopian Plateau

Ghana Mainly Accra, Ho-Keta and Winneba

Plains; scattered patches near Kpandu,
Kwamen and Kwesi

Ivory Coast In northern part of the country and

in the region of Divo

Kenya Athi Plains near Nairobi and other

areas

Lesotho Drakensberg

Malagasay Some valley depressions in the

western part of the country and in
uplands in the western and northern
part
 TABLE 36 (Cont.)

Distribution of Dark Clay Soils in Africa

(after Dudal, 1965)

Estimated coverage
Country Area (million acres)
Malawi Chyre valley and areas around Malawi
and Chilwa Lakes

Morocco Mainly in the northwest Gharb and in

the Doukkalas (south of Casablanca);
scattered areas

Mozambique Alluvial plains of the Limpopo and

Inkornati Rivers and surrounding
uplands; Zarnbezi valley upstream of
Tete; coastal area north of
Mozambique city

Niger Central Niger valley and several

scattered areas

Nigeria Northeast Bornu and Benoue River

Basin

Senegal Senegal valley, lowland of the north-

west; on the Cape Verde peninsula

Somalia In the plains extending between the

Giuba and Scebeli Rivers

South Africa Bushveld and Springbok Flats (Trans-

vaal); Highveld (Orange Free State
and Transvaal)

Southwest Probable occurrence in the Caprivi

Africa Strip along the Cubango River and
around the "Etosha Pan"

Rhodesia The area around Salisbury; part of

the valley of the Zarnbesi; Livingstone
area and southern part of the
country

Sudan Region between the White and Blue Nile

extending east of the Blue Nile into
Ethiopia and covering an area west of
the White Nile; widespread in South
Sudan, Bahr el Gasal, Upper Nile and
Fquatoria Province

Swaziland Mainly in the Middleveld, eastern

Lowveld and Lebombo areas
 TABLE 36 (Cont.)

Distribution of Dark Clay Soils in Africa

(after Dudal, 1965)
Estimated coverage
Country Area (million acres)

Tanzania Valleys of the Mayowasi and Malagarasi,

Great Ruaha valley areas near Tendigo
Swamps and Rukwa Lake; scattered
upland areas

Togo In the northern part of the country

and in the Mono valley

Uganda Valley of the Sernliki, areas near

George, Albert and Edward Lakes, areas
in the east and northeast of the
country

Upper Volta Souron valley; poorly-drained basin

deposits spread over the country

Zambia Valleys of the Luangwa, Lukushashi

and Zambesi; Kafue flats

TOTAL

areas containing black clay soils is shown in Figure 116. iron content relatively low. The silica-sesquioxide ratios are
above 2.50, considerably higher than in the red soils.
The mineralogical composition of the clay fraction in-
MORPHOLOGICAL, CHEMICAL variably includes montmorillonite (Morin and Parry; Mohr and
AND MINERALOGICAL CHARACTERISTICS Van Baren, 1959; Tarnhane and Namjoshi, 1959; McAleese
The black color was at first explained as being due to the and Mitchell, 1958; Clare, 1957; Van der Merwe and Heysteck,
organic content. It was later found that the quantity of 1955; Tarnhane and Sen, 1954; Nagelschrnidt, et al., 1940; and
organic matter was neglible in many soils which had the dark Hardon and Favejee, 1939). Kaolinite is occasionally reported
gray to black coloring. The presence of titanium has also been but always in minor quantities. Morin and Parry found
proposed as a reason for the black color. However, titanium halloysite as well as kaolinite in Ethiopian black clay soils.
oxidizes very readily and the oxide is white. Theron and Van The mineralogical composition of the fine and coarse
Niekerk (1934), working with the black turfs of South Africa, fraction of various black cotton soils of India was determined
believed that the black color was due to the high percentage of by Tamhane and Sen (1954). They found that the sand-size
silica in the clay fraction which allows a small quantity of derived from basalt is made up principally of augite with
organic matter t o impose the black color on the entire soil appreciable magnetite and limonite.
mass. Dudal (1965) states, "the color results more from the The black clay soils rarely develop thick profiles. Van der
formation of a complex between clay and organic matter Merwe and Heysteck (1955) describe South Africa occurrences
rather than from high amounts.of organic matter; indeed the that vary from 14 inches t o 6 feet. Agarwal and Muke rji
content of the latter in dark clay soils is relatively low." (1 949) report a depth of 6 feet in India.
Chemical compositions of the clay fraction of some black Profiles in tropical black clays are even less developed than
soils are given in Tables 37 and 38. Analyses of selected in the red soils. Bal (1935) found few marked horizons in the
margalitic soils from Indonesia are shown in Table 37. Partial black cotton soil profiles in India. Similarly Muir (1951) found
chemical analyses of South Africa black clays are listed in little evidence of profile development over the basalts in Syria.
Table 38. The chemical composition differs considerably from Tamhane and Sen (1954) found that the chemical properties
the tropical red clays. The silica content remains high and the change very little with depth; Mohr and Van Baren (1959)
FIG. 116 DISTRIBUTION OF BLACK CLAY SOIL IN AFRICA

DISTRIBUTION DES SOLS ARGILEUX NOlRS EN AFRIQUE

 TABLE 37 cobbles of unweathered basalt, covered with a thin shell of
unweathered material, lying in black clay. If the basalt is
Chemical Composition of the Clay highly fractured or joined, a horizon of saprolite or weathered
Fraction of some Margalitic Soils, rock can develop. A partial chemical analyses of the clay
lndonesia fraction of a profile in India is provided by Agarwal and
Mukerji (1949) and given in Table 39. It can be seen that the
(after Mohr and Van Baren, 1959)
composit ion varies only slightly with depth.

ENVl RONMENT AND FORMATION

OF BLACK TROPICAL SOILS
Sample
12843 12461 33193 30277 The formation of the oxides and hydroxides of iron and
alumina has been discussed with regard t o lateritic soil. The
SiOz environment under which these and black clay soils are formed
will now be considered. However, the formation of the clay
A1203
minerals will be discussed first as these are generally believed
to be the precursors of the silica-free oxides and hydroxides.
0 3 The clay minerals that form are principally of four groups:
kaolinite, halloysit e, mont morillonite and illit e. Each group
Ti02 has a number of mineralogical variations reflecting differences
in mineral structure or chemical composition. Most of these
MnO are relatively unimportant insofar as engineering behavior is
concerned. One exception is metahalloysite.
CaO The clay minerals are essentially hydrous aluminum sili-
cates. Hydrous silicates of magnesium and iron are less
MgO common.
The parent rock determines in a negative sense which clay
K2 0 minerals form initially. Kaolinite cannot form, for example,
unless alumina is present. The parent soils may be a dominant
Na2 0 factor or only of transitory significance (Keller , 1956).
Weathering in a warm humid climate common in the tropics is
p2 05 rapid and the influence of the parent material is usually
shortlived. However, as Keller has pointed out, the parent
co2 material must provide the constituent elements of which the
clay minerals are formed.
H20 (-105") Ross and Hendricks (1945) and Kerr (1955) have con-
cluded that montmorillonite forms in an alkaline environment
H 2 0 (+105") where there is an adequate supply of magnesium, silica and
alumina. However, Clauss (1967) concluded that montmoril-
Organic lonite seems to form only where an initially alkaline envi-
matter ronment changes gradually into a slightly acid environment
because such a condition favors the mobilization of iron,
100.10 99.29 100.33 100.36 magnesium, silica and aluminum and the removal of iron.
Kaolinite forms under oxidizing conditions. Extremely wet,
Si02-R2 o3 warm clirnat es, "by insuring rapid weat hering and oxidizing
ratio 2.96 2.5 1 2.5 2 3.12 conditions, probably favor formation of kaolinitic and lateritic
types of soils from parent material, that, under other
Si02-A1203 conditions, might have given montmorillonite-type clays"
ratio 3.82 3.66 3.40 3 -74 (Ross and Hendricks, 1945, p.60).
state that a sharp boundary usually is found between basic Hosking (1940) found that montmorillonite forms from
rocks and soil, total transformation takes place within a layer basic rocks in an alkaline environment. Granite and acidic
of a few millimeters or even tenths of a millimeter. The basalts rocks do not form montmorillonite or black clay soils in
in Ethiopia pass very abruptly into black clay with only a thin Australia. Hardon and Favejee (1 939) found similar envi-
zone of weathered rock, usually only a fraction of an inch. ronmental conditions required for montmorillonite formation
Spherodial weathering is also common with boulders and in Java.
 TABLE 38

Partial Chemical Analyses

of South African Black Clays
(after Van der Merwe and Heysteck, 1955)

Clay
Parent Depth Fraction Si02 Si02
Material (in -1 (%) Si02 Fe2 0 3 Alz03 R2 0 3 A1203

basalt

dolerite 18-28
34-54

The kind of alkali or alkaline earth is also considered and Van der Merwe (1962) report that montmorillonite always
important. Rocks that do not contain alkalies and alkaline forms first over the South Africa basalt, but it may alter to
earths can produce only kaolin-type clay mineral or lateritic Kaolin in well-drained areas. Morin and Parry found a similar
weathering products (Grim, 1968). transformation presently occurring in Ethiopia. Ferguson
The transformation of montmorillonite to kaolinite has (1954) believes that the black montmorillonite soils of
been well documented. Van der Merwe and Heysteck (1955) Australia alter to red kaolinitic soils through loss of silica and
alkali, alkaline earths, and iron. This change t o kaolinite has
been noted by others: Altschuler, et al. (1963); Correns
TABLE 39 (1962); McAleese and Mitchell (1958); Allen (1 948, 1952);
and Nagleschmidt et al. (1940).
Partial Chemical Analysis Bates (1962), in his study of halloysite and gibbsite
of the Clay Fraction of a formation in Hawaii, distinguished three broad mineral groups:
Soil Profile at Ata, India (1) montmorillonite-group clay minerals in dry areas where
(after Agarwal and Mukerji, 1949) leaching has only removed part of the bases; (2) halloysite and
allophane in areas with sufficient rainfall t o remove bases but
Depth (inches) not all the silica; and (3) hydrated alumina-laterite-gibbsite in
0-12 12-28 28-42 42-56 56-60 60-72 continually wet areas or alternately wet and dry for long
enough periods of time to remove the silica. Considerable
overlapping of areas 1 and 2, and areas 2 and 3 occurs.
The Ethiopian soils include similar broad areas (Morin and
Parry). The black clays form in dryer areas where the rainfall is
under 40 inches per year or in poorly-drained areas wi.th over
40 inches per year. Red soils occur in areas with greater than
4 0 inches. The corresponding figure for South Africa was
found to be 32 inches. However, the black clays in Ethiopia all
contain kaolinite as well as montmorillonite, which in most
cases is subordinate to kaolinite and halloysite.
Sherman (1952) concluded that not only -the amount of
rainfall but also the intensity and seasonal d.istribution of that forms. Kaolinite and illite are common in subordinate
rainfall controlled the intensity of the weat hering process. amounts. Halloysite has been reported only in Ethiopia. The
Also, the length of time that the materials have been subjected principal exchangeable cations are calcium and magnesium
to weathering is important. The weathering cycle he derived is with potassium and sodium usually in minor amounts. Sodium
reproduced in Figure 1 17. The kaolinite shown is actually is the principal ion in the Burma soil. Cation exchange
reported as kaolin-like clays and may therefore include capacities are within the normal ranges of montmorillonitic
halloysite. Sherman also determined the progressive formation soils.
of clay development given in Fig. 1 18 and 1 19. The range in The gradation of the soils varies considerably. There is little
rainfall in which montmorillonite, kaolinite and hydroxide or no gravel size in most soils and those which show
form is shown against the percentage of clay that has formed. appreciable percentages, such as those in 'India and Ethiopia,
Montmorillonite forms first and persists only where rainfall is may include rock fragments not yet decomposed. The amount
relatively light. of sand varies from 0 to 39% and silt from 2 to 47%. The clay
Bates (1962) lists five factors which affect the rate and size reported (the upper limit for the clay size is rarely stated)
amount of weathering in Hawaii: (1) age, or length of ranges from 13 to 100%.
exposure, (2) composition and (3) texture of the rock; (4) Almost all liquid limits are over 40% with most over 50%;
amount and seasonal distribution of rainfall; and (5) the most plasticity indexes fall between 20 and 60. Shrinkage
nature of the slope and the resulting drainage pattern. limits range from 7% to 28%. The AASHO classification for
The following -points summarize the general findings most of the soils is A-7-5 or A-7-6. Organic contents are
reviewed: usually fairly low.
Compaction and CBR data are limited. CBR values are low
1. basic rocks + montmorillonite (black clays) + kao-
when reported after the normal 4 day soaking period.
linite-type minerals (red soils) + gibbsite, etc. or
A suggested classification system was proposed by the
geothite, etc. or both (lateritic soils).
Indian Roads Congress (1953) in an attempt to differentiate
2. acid rocks + kaolinite-type mineral + as in 1. the varieties of black cotton soil. All 21 0 soils tested under
3. Fe-rich or Al-rich (and in combination) soils may form this study fell within A-7-5 or A-7-6 groups; the group index
indurated crusts as -they approach the end products of values were 20 for almost all of the soils. A Table 41 was
weathering and silica is largely removed. Harder crusts proposed embodying the division classification system, A-7;
are produced by the more iron-rich soils. A-7a; A-7b and A-7c. The group index was expanded by
raising .the maximum value from 20 to 50.
4. Degree of weathering is controlled by time; by char- This classification system was statistically examined by
acteristics of the parent rocks; by xnount and Pattern Uppal and Sehgal (1958). They found that the variation
of rainfall; and by drainage conditions. between the adjusted groups was significantly greater than the
variation within each group. Only 6% of 210 samples could be
GEOTECHNICAL PROPERTIES O F considered as misclassified. However, since all are poor
TROPICAL BLACK CLAYS material for subgrades, the breakdown into four groups does
not appear to be particularly useful or meaningful. The
The literature of the geotechnical properties of expanded group index adopted by AASHO M145-661 is more
black clays is limited. Most of the data are from Africa and descriptive.
India. Much of this data has been tabulated by Morin and
Parry; a similar tabulation is given in Table 40. Included are
descriptions from Australia (Hosking, 1935); Burma (Wool- ENGINEERING PROPERTIES
torton, 1951); India (Katti et. al, 1966); Mehra and Uppal, Most of the engineering properties reported in the literature
1964; Mohan, 1957, Clare, 1957); Palestine (Road Research are from Indian sources, not only because of the widespread
Laboratory, 1947); Nigeria (Evans, 1957, Grainger, 1951); occurrences of,black cotton soil in India but also because the
Ghana (Clare, 1957; Stephen, 1953); Ethiopia (Morin and term "black cotton" is so widely used in that country.
Parry); Kenya (Williams, 196 1); Rhodesia (Grainger , 1952); Tropical clays from other parts of the world are not always
Uganda (Evans, 1957); Basutoland (Pollard, 1964); South described as either "tropical black clays", or "black cotton
Africa, D.H. Van der Merwe, 1967); Morocco (Moussaoui, soils", and appropriate papers in the literature may have been
1967); and British Honduras (Evans, 1957). missed.
The data in Table 40 show that the development of black Very little information is available on the shear strength of
clay soils follows a fairly characteristic pattern. Parent rocks tropical black clays or black cotton soils. Mohan (1 955b,
for the residual soils are usually basic, most commonly basalt. 1957) investigated the shear strength of a number of Indian
Rainfall is generally under 50 inches and drainage conditions black cotton soils and found a relationship between the
are usually poor. All pH values reported are alkaline. Under remolded strength and the liquidity index (W-PL)/(LL-PL).
these conditions, montmorillonite is the principal clay mineral Barai (1953) studied the stress-strain and shear characteristics
 BASALTIC LAVA V O L C A N I C ASH
L A V E BASALTIQUE CENDRES VOLCANIQUES

GEOLOGICAL WEATHERING
GEOLOGIOUE

?! PRIMARY MINERALS

IL '
ALTERNATING
W E T A N D DRY SEASON
SAISONS HUMIDES E T CONTINUOUS W E T C L I M A T E
SECHES , A L T E R N E E S C L I M A T HUMIDE CONTINU

OPEN FOREST TROPICAL RAIN FOREST

F 0 R E T DECOUVERTE
- F'REE OXIDE FORMATION-
FORMATION D'OXYDES L I B R E S
FORET TROPICALE HUM l D E

HYDRATED IRON +TITANIUM OXIDE HYDRATED ALUMINUM OXIDE

F E R HYDRATE + OXYDE DE T I T A N €

T-
I-
FERRUOINOUS TlTANlFEROUS LATERITE CRUST
CROGTE LATERITIOUE FERRUGINEUSE TITANIFERREUSE
D E H Y DRATlON
DE HYDRATATION
-
BAUXITE LATERITE
BAUXITE LATERITIQUE

I\ /
EROSION
\ DEORADATION
/ DEGRADATION /
/
NEW WEATHERING SURFACE /
NOUVELLE SURFACE DE LESSIVAOE R E S I L 1 C AT10 N ?

7l-7
PEAT
TOURBE
SWAMP
MARAIS
CERAMIC CLAY
AROILE CERAMIOUE

FIG. 1 1 7 O U T L I N E OF WEATHERING I N T H E TROPICAL S O I L S OF THE H A W A I I A N I S L A N D S SHOWING

PRIMARY, I N T E R M E D I A T E , AND END PRODUCTS OF PEDOLOGICAL WEATHERING
( a f t e r Sherman, 1952)
SCHEMA DE LESSIVAGE DES SOLS T R O P I C A U X DES I L E S HAWAII M O N T R A N T DES MATERIAUX
P R I M I T I FS, l WTERMEDlAlRES E T F l NAUX PRODUITS PAR LESSIVAGE PEDOLOGIQUE
 /
/
0

.~GEAVY M I N E R A L S
,'/ MINERAUX LOURDS

0
/
.0 HEMATITE
ANATASE
/
0
/ MAGNETITE

HYDRATED F R E E OXIDE

30 60 90 120
ANNUAL RAINFALL IN INCHES
PLUVlOSlTE ANNUELLE EN POUCES
FIG 118 PROGRESSIVE TYPES OF CLAY DEVELOPMENT IN HAWAIIAN
SOILS UNDER A CLIMATE HAVING ALTERNATING WET AND
DRY SEASONS ( a f t e r SHERMAN, 1952)

DEVELOPPEMENT D'ARGILES DE TYPES PROGRESSIFS DANS

DES SOLS HAWAIIENS SOUS UN CLIMAT A SAISONS
HUMIDES E T SECHES ALTERNEES ( d'apres SHERMAN, 1952)

/ F R E E O X I D E (HYDRATED)
o x y o E s L l s R E s (HYDRATES)

I K-1 DlASP0RE
BOEHMITE
LIMONITE
BAUXITE
)
K AOLlNlTE

ONTMORILLONITIQUES- KAOLINIQUES

ANNUAL RAINFALL IN INCHES

FIG. 119 PROGRESSIVE TYPES OF DEVELOPMENT IN HAWAl lAN SOILS

UNDER CONTINUOUSLY WET CLIMATE ( a f t e r SHERMAN, 1952)

DEVELOPPEMENT DE TYPES PROGRESSIFS DANS DES

SOLS HAWAII ENS SOUS UN C L l MAT CONSTAMMENT
HUMIDE ( d'apres SHERMAN, 1952)
 TABLE 40

Properties of Tropical Black Clays

Author's Des- Drainage Cation
cription of Depth Parent Rainfall Character- Clay Miner* Ex. capacity
Location Material (ft) Material --in./yr istics pH _. logy m.e/100 gm Cation

Darling Downs black earth 3+ basalt <50 8.3 Ca; Mg;

~ustralia' Na

black cotton basalt illite; montmoril- Na; Mg

lonite
India black cotton trap, rnontmorillonite; 40-50 Ca; Na;
3
(various) basalt some illite K
India Black cotton - 44-72
(various)4
India black cotton 3-15 mostly mainly montmo-
(various)' trap rillonite
Hyderabad, black clay poor montrnorillonite;
1ndia6 tr. of kaolinite

Nagpur, black clay poor rnontmorillonite;

1ndia6 some illite; tr.
o f kaolinite
black cotton probably 8.2
chalk
black clay basalt, poor t o 7.2-8.4 montmorillonite; Ca; Na;
trachyte good kaolinite; K
halloysite
black cotton basic
extrusive
black clay

black cotton poor montmoril lonite;

with kaolinite;
illite
Basutoland ' black clay basalt,
dolerite
black clay norite rnontmoril lonite

black cotton poor

black clay-
cotton
black clay 3-5' hornblende Poor 7.9' rnontmorillonite; Ca; Mg; ~ a '
garnet- tr. o f kaolinite
gneiss
alluvial rnontmorillonite;
chlorite; tr. of
kaolinite
British black clay chalk or poor montmorillonite
~onduras'O marl
 A n n -
, " = = z
A
0
N
w - - w w

? T O ' ? ?
? ? ? ? ?
a a a a a
 TABLE 41

Classification of l ndian Black

Cotton Soils

(after l ndian Roads Congress, 1953)

Passing No. 200 sieves, % up to 55 5 5 -95 80-95 85-100

Liquid Limit, % 55 <65 >65 >65

Plasticity Index 25 <42 <42 >42

Group Index 20 max. 20-30 30-40 40-50

Rated as subgrade Fair to poor poor poor poor

of a black cotton soil from India. Mohan and Jain (1957) pounds per square foot in Burma. In an earlier paper on Burma
concluded that field vane tests give more reliable values than (1936), Wooltorton suggested a minimum loading of one ton
the unconfined compression test. per square foot. Annual vertical movements of over 4 inches
Mohan (1957) also found a close relation between the were measured in the black turf soils in Central Transvaal,
compression index and the liquid limit. The points from eight South Africa (de Bruijn, 1967).
tests fell close t o Skempton's relationship C, = 0.007 A number of other Indian workers have studied the swelling
(LL - 10). pressures of Indian black cotton soils. Deb (1961) reviewed
Ranganatham (1961) determined that a change in the laboratory and field studies of swelling pressure versus bearing
electrolytic environment causes significant changes in consoli- capacity. Mohan and Goel (1960) determined the swelling
dation characteristics. Lime increases the coefficient of con- pressures and volume expansions of 20 soils. Nagarajan and
solidation; black cotton soils therefore should consolidate Sundaraman (1 960) evaluated different methods of laboratory
more quickly during rolling after being treated with lime. A measurements of swelling pressures. Palit (1953) describes a
dispersant has the opposite effect. Ranganatham and Henry method of measuring swelling pressures with a simple loading
(1963) give results of consolidation tests on black cotton soils frame and proving ring. Ranganatham and Satyanarayana
under different chemical environments. (1965a) also describe a procedure for measuring swelling
The expansiveness of tropical black clays is notorious and pressures in a consolidation cell. They later (1 965b) advocated
has been studied by a number of investigators. Uppal and Palit the use of the shrinkage index (liquid limit minus shrinkage
(1959) found that swell pressures: (1) increase as the dry limit) t o predict the swell potential. Gupta et al. (1967) state
density increases; (2) increase with the height of the specimen that the montmorillonite content is mainly responsible for
being tested; (3) increase with increasing ranges of initial and swelling, but other clay minerals such as illite and beidellite
final moisture; (4) are not significant until moisture exceeds may also cause swelling. They believe that the base-exchange
the shrinkage limit; and (5) increase as the area of confinement capacity "is a good indicator". Katti (1965) and Katti and
increases (that is swelling pressures are greater under the Sadasivan (1 967) found that swelling pressure can be reduced
central part of the roadway or building, etc.). Deb and Joshi t o nearly zero with repeated cycles of loading. Katti and
(1960) found that the swelling pressure under a circular plate Kulkarni (1 967) state that a superimposed nonswelling soil
in the field reduced as the diameter of the plate increased. The layer is very effective in resisting the swelling pressure; the
maximum swelling pressure measured, for a 10 inch diameter shear characteristics of a nonswelling soil contribute t o a great
plate, was 3550 pounds per square foot. Mehra and Uppal extent in resisting the swelling pressure. Jennings and Knight
(1964) found that the swell pressure increases as the degree of (1957-58) devised the double oedometer test to determine the
compaction increases and as the difference between initial total heave that will occur. The test takes into account the
moisture content and moisture content at saturation increases. degree of desiccation of the soil, the overburden and applied
They measured swelling pressures of up t o 3000 pounds per load, and the activity of the soil.
square foot in India. Wooltorton (1 95 1) measured up t o 4000 The following recent work was summarized by Morin
(1969). Swelling clays and clay shales are found in many parts counteract swelling pressures may lead t o bearing capacity
of the world in temperate as well as tropical regions. The failures since swelling pressure is bound t o dissipate with time
causes are usually high clay content with montmorillonite as while the moisture content does not change. Bearing capacity
the principal clay mineral. Therefore, work on swelling clays in against shear failure is often considerably less than swelling
temperate areas can be applied, except for unusual cases, t o pressure.
tropical black clays. Gizienski and Lee (1965), determined "Under-reamed" pile foundations in black cotton soils are
that expansion decreases with increased surcharge loads, commonly employed in India (Chandra, 1962); Chandra and
decreasing density and increasing moisture content. However, Khepar, 1964; Deb and Chandra, 1962; Mohan and Chandra,
actual heave measured in the field was much less than the 1961a, 1961b; Mohan and Jain 1958a; 1958b). These are
amount indicated by the laboratory tests. Ho (1967) found bell-shaped piles that resist upward forces. Double under-
that swelling pressure can be reduced by allowing some reamed piles are also utilized for increased bearing capacity
expansion t o occur. (Subhash and Khepar, 1964). Mohan and Rao (1965) state
Kormornik and David (1969) statistically analyzed over that under-reamed pile foundations are satisfactory if carried
200 disturbed and undisturbed samples from Israel and found t o depths of 1 2 feet or more, at depths where moisture
that the logarithm of the swelling pressure should be based on: variations are not significant.
They believe a simple means of preventing moisture changes
log P = 2 1 3 2 + 0.0208 (LL) + 0.000665 (nd) is to provide an underground concrete apron.
A research program sponsored by the U.S. Air Force is
= 0.0269 (w) described by Gaffey (1967). The behavior of drilled caissons in
expansive soil is being observed under a variety of load and
where P = swelling pressure in Kg/cm2 soil-moisture conditions. Observations are continuing over a
two year period.
LL = liquid limit Strongman (1963) utilized black cotton soil successfully in
a 25 foot embankment in Kenya by placing a protective
nd = natural dry density of the soil in Kg/m3 blanket of weathered phonolite which prevented moisture loss
in the black cotton soil.
w = natural moisture content Ranganatham and Satyanarayana (1 967) have proposed an
approximate method of analysis for the determination of earth
pressures in expansive (black cotton) soils.
The correlation coefficient between log P and each of the
single variables was low, but collectively it was 0.60. This Road construction in black cotton soils has been described
suggests that swelling pressure depends not only o n the by Nageswara (1945) for India; by Wooltorton (1959) for
properties but also on initial placement conditions, or density, Kenya; by Moussaoui (1967) for Morocco and for South
water content and soil structure. Africa by Weinert (1 968); Williams (1 965, 1959a); Williams
and Simons (1 963); and Williams and Jennings (1 960).
Foundations for structures in black cotton soils have
Moussaoui states that granular material is placed as a buttress
received some attention by South African and Indian authors.
t o embankments of black clay soils or tirs; this prevents
Donaldson (1 963, 1962) described the foundation problems
shoulder settlement caused by alternating wetting and drying.
that arise in South Africa and design methods that have been
Uppal (1 965) found that the moisture content of black cotton
adopted. Changes in moisture content that bring about heaving
as subgrade soil under pavements varied from only 1 3 t o 16
in the black clays have been studied by Simons and Williams
percent from dry t o wet seasons, whereas, under the
(1 963) and Williams (1 959b). Deb and Chandra (1 963)
uncovered shoulders the variation was between 6 and 23
suggested that recurrent cracking in small buildings could be
percent.
prevented by keeping the moisture content beneath and
around the building as constant as possible. Mohan (1961) Black clay soils have been stabilized in the laboratory with
discussed various types of building foundations that might hydrated lime. Clare (1957) found that the addition of 4%
prove t o be successful in black cotton soils. He recommended lime eliminated all plasticity in a Southern Rhodesian black
a pier or pile-and-beam construction at a depth at which cotton soil. Mehra and Uppal (1964) state that 4%lime
moisture contents do not fluctuate. Earlier Mohan (1955a) improves the texture of an Indian black cotton soil sufficiently
reviewed the research carried out on ground movements and t o permit its use in construction. The soil properties were
damage to buildings founded on expansive clay in an attempt improved from liquid limit of 90% t o 55%; plasticity index of
t o apply methods developed elsewhere t o Indian black cotton 44 to 5; and CBR of 1.6 to 17. The increase in CBR occurred
soils. Treatment of foundations of buildings in black cotton with only 2% lime. Clauss (1967) describes the use of pH
soils is also described by Narain (1 943). Deb and Joshi (1 960) determinations of fresh and weathered dolerite as indicators of
describe various factors which influence the bearing capacity the stabilization requirements. Fossberg (1965) has inves-
of expansive clays. They believe designing bearing pressures t o tigated the engineering properties of a lime-stabilized clay.
 Soils with montmorillonite and degraded weathering soils of the United Provinces, 111: Pedochemical character-
products retard the hydration and hardening of cement; such istics of the black soils of the plains. Indian Jour. Agric.
soils are more suitable for stabilization with lime than cement Sci., vol. 19: 31-40.
(Croft, 1967). Cement stabilization of Indian black cotton Allen, V. T., 1952. Formation of minerals above or below
requires 17% cement (Mehra and Uppal, 1964). Stabilization the ground water table. In: Problems of clay and laterite
with special cement containing finely divided lime was genesis. Am. Inst. Min., Met. and Pet. Eng., New
suggested by Bhatia (1956). York: 204-205 (discussion).
Other research on stabilization has been performed in India. Allen, V. T., 1948. Formation of bauxite from basaltic rocks
Katti and Kulkarni (1962) studied the effect of trace inorganic of Oregon. Econ. Geol., vol. 43: 619-626.
chemicals on lime stabilization of an expansive soil from Altschuler, Z. S., Dwornik, E. J., and Kramer, H., 1963.
Poona, India. Very small quantities of calcium chloride, Transformation of montmorillonite to kaolinite during
sodium hydroxide, sodium silicate and potassium per- weathering. Sci., vol. 141: 148-152.
manganate were used to improve the early strength charac- Aubert, G., 1941. Les sols de la France d'0utre-Mer.
teristics of black cotton soil-lime mixtures. Katti and Sudhar Monogr. Lab. Rech. Agron., Paris: 90 p.
(1962) and Katti et al. (1966) investigated stabilization with Azevedo, A. L., and Da Costa, J. V. B., 1954. Les "terres
combinations of cement and lime. Mehra and Chadda (1957) noires" d'Angola. Trans. Fifth Int. Cong. Soil Sci.,
also studied stabilization with gypsum. Misty (1957) gives the Leopoldville (now Kinshasa); vol. 4 : 149-154.
details of an experimental road where black cotton soil was Bal, D. V., 1935. Some aspects of the black cotton soils of
mechanically stabilized. Cement stabilization is further central provinces. India Emp. Jour. Expt. Agric., vol.
described by Venkatesan and Bapaye (1 955). 3: 261-268.
Clare (1957) found that heating a Southern Rhodesian Barai, S. C., 1953. Study of the stress-strain and shear
black cotton soil to 500°C eliminated all plasticity, and that characteristics of soils by triaxial compression tests. Irri-
vitrified aggregate was produced between 500°C and 700°C. gation and Power (India), vol. 10: 92-101.
Similar findings are reported by Mehra and Uppal (1964). The Basu, J. K., and Sirur, S. S., 1954. Soils of the Deccan
Indian soil was made non-plastic by heating at 500°C for two canals, their classification, nomenclature and genesis. Natl.
hours. Grainger (1951) found that a Nigerian black clay Inst. of Sci. of India, Bull. No. 3: 83-88.
became nonplastic after heating for one hour above 500°C. Bates, T. F., 1962. Halloysite and gibbsite formation in
The heated soil broke down on impact to give a material Hawaii. Ninth National Conf. on Clays and Clay
whose particle-size distribution was similar to a sandy gravel. Minerals: 3 15-328.
Thermal stabilization was also attempted by Radhakrishnan, et Bhatia, H. S., 1956. Cement stabilized base in highway
al. (1967). It appears doubtful that such treatment would be construction. Indian Concr. Jour., vol. 30: 219-220.
practical or economic for hghways, but could prove advant- Chandra, S., 1962. Under-reamed pile foundation in black
ageous were localized stabilization is desirable. cotton soils. Indian Concr. Jour., vol. 36: 27-30,37.
The influence of soil mineralogical compost ion on stabili- Chandra, S., 1962, and Khepar, S. D., 1964. Double under-
zation has been studied by Croft (1967). His findings may reamed piles for foundations in black cotton soils. Indian
be summarized as follows: (1) kaolinitic and illitic soils are Concr. Jour., vol. 38: 50-62.
more suitable for cement stabilization than soils with ex- Clare, K. E., 1957. Airfield construction on overseas soils,
pansive clays; (2) heavy clay soils are difficult to stabilize Part 2: Tropical black clays. Proc. Inst. Civil Eng. vol.
with cement ; if expansive montmorillonite is present in 81, 223-231.
quantities sufficient to control the behavior of the soil, Clauss, K. A., 1967. The pH of fresh and weathered dolerite
cement will probably be uneconomical and lime should be as an indication of decomposition and of stabilization
used; (3) cement reduces the plasticity indexes and increases requirements. Fourth Reg. Conf. for Africa on Soil Mech.
the shrinkage limits of the clay minerals; the changes in and Found. Eng., Capetown, vol. 1 : 101-108.
expansive clay minerals are not enough to stabilize the clay Correns, C. W., 1962. Beobachtungen ueber die Bildung und
and are less than changes produced with lime; (4) one of the Umbildung von Tonmineralen bei der Zersetzung von
problems with cement stabilization is the difficulty in ob- Basalten. Colloq. Intern. Centre Natl. Rech. Sci., Paris, No.
taining an intimate mixture with the soil; cement does not 105: 109-121.
improve the soil workability as does lime and consequently Croft, J. B., 1967. The influence of soil mineralogical
must be effective with the lump-size material that is avail- composition of cement stabilization. Geotechnique, vol.
able, and (5) the addition of lime before cement will assist 17: 119-135.
stabilization by reducing the size of the clods. De Bruijn, C. M. A., 1967. Annual vertical movements and
soil moisture redistribution at the Onderstepoort test site.
REFERENCES Fourth Reg. Conf. for Africa on Soil Mech. and Found.
Agarwal, R. R., and Mukerji, P., 1949. Studies on Bundelshand Eng., Cape Town, vol. 1 : 235-242.
Deb. A. K., 1961. Swelling pressure vs. bearing capacity of Soil Mech. and Faund. Eng., Haifa, vol. 1 : 84-89
black cotton soil. Sym. on Found. Eng., Bangalore, paper Hallsworth, E. G., 1952. An interpretation of the soil
Al: 1-10. formation found on basalt in the Richmond-Tweed region
Deb. A. K., and Chandra, S., 1963. Remedial measures for of New South Wales. Austr. Jour. Agric. Res., vol.
the prevention of recurrent cracking in small buildings 2: 411-428.
founded in black cotton soil. Indian Concr. Jour., vol. Hardon, H. J., and Favejee, J. C. L., 1939. Mineralogische
37: 190-191. auder-zolkingen aan klein en kleimineralen 111: Qualitative
Deb, A. K., and Chandra, S., 1962. Under-reamed pile x-ray analysis of the clay fraction of the principal soil
foundation in black cotton soil. Cement and Concr. vol. types of Java. Medediel Landbouwhoogeschool, vol.
3: 45-50. 43: 55-59.
Deb, A. K., and Joshi, S. S., 1960. Basis for foundation Ho, M. M. K., 1967. The effect of swelling on swelling
design on expansive clays. First Asian Reg. Conf. on Soil pressures. Third Asian Reg. Conf. on Soil Mech. and Found.
Mech. and Found. Eng., New Delhi, Part 111, Subj. 11, Eng., Haifa, vol. 1: 90-98.
Paper No. 18: 12 p. Hosking, J. S., 1940. The soil clay mineralogy of some
Degens, E. T., 1965. Geochemistry of the sediments. Australian soils developed on granitic and basaltic parent
Prentice-Hall: 342 p. materials. Jour. Austr. Counc. Sci. and Ind. Res., vol.
Donaldson, G. W., 1963. Soil investigation and foundation 13: 206-216.
design in relation to municipal engineering. Proc. Ann. Hosking, J. S., 1935. A comparative study of the black earths
Conf. Mun. Eng., Bulawayo (Rhodesia). of Australia and the regur of India. Trans. and Proc. Royal
Donaldson, G. W., 1962. Foundation problems experienced Soc. of So. Austr., vol. 59: 168-200.
in Africa, C.C.T.A. Publ. No. 81: 17 p. Indian Roads Congress Journal, 1953. A suggested classifi-
Dudal, R., ed., 1965. Dark soils of tropical and subtropical cation system of the black cotton soils of India. vol.
regions. F.A.O. Agric. Develop. Paper No. 83: 161 p. 18: 281-290.
Dumbleton, M. J., 1963. The clay minerals of soils from Jennings, J. E. B., and Knight, K., 1957-58. The prediction
some roads in Kenya. Dept. Sci. and Ind. Res. (U.K.), of total heave from the double oedometer test. Sym. on
Road Res. Lab. Note LN/372/MJD (unpublished). Expansive Clays, So. African Inst. of Civil Eng.: 13-19.
Evans, E. A., 1957. Laboratory examination of the prop- Katti, R. K., 1965. Effect of preload and side surcharge on
erties of fifteen tropical soils. Dept. Sci. and Ind. Res. the swelling characteristics of black cotton soils. Behavior
(U.K.), Road Res. Lab. Note No. RN/3020/EAE of soil under stress, Proc. Fifth Sym. of Civil and
(unpublished). Hydraulic Eng. Dept., India Inst. of Sci., vol. 1.
Ferguson, J. A., 1954. Transformation of clay minerals in Katti, R. K., and Kulkarni, K. R., 1967. Studies on non-
black earths and red loams of basaltic origin. Austr. swelling soil layer as a means to resist swelling pressure of
Journ. Agric. Res., vol. 5: 98-108. expansive soil system. Southeast Asia Reg. Conf. on Soil
Fossberg, P. E., 1965. Some fundamental engineering prop- Eng., Bangkok: 147-157.
erties of lime-stabilized clay. Sixth Int. Conf. on Soil Katti, R. K., and Kulkarni, K. R., 1962. A study of the
Mech. and Found. Eng., Montreal, vol. 1: 221-225. effect of trace inorganic chemicals on lime stabilization of
Gaffey, J. T., 11, 1967. Foundations in expansive soils. an expansive soil from Poona. Indian Roads Cong. Jour.,
Military Engineer, vol. 59: 4 1 0 4 12. v01. 27: 215-227.
Gizienski, S. F., and Lee, L. J., 1965. Comparison of Katti, R. K., and Kulkarni, K. R., and Radhakrishnan, N.,
laboratory swell tests to small scale field tests. Proc. Int. 1966. Research on black cotton soils without and with
Res. and Eng. Conf. on Expansive Clay Soils, Texas A&M inorganic additives. Road Res. Bull. Indian Roads Cong.,
Press: 108-119. No. 10: 92m.
Grainger, G. D., 1952. The stabilization with lime and a Katti, R. K., and Sadasivan, S. K., 1967. Consolidation and
local wood tar of a black cotton soil from Southern rebound characteristics of an expansive soil under re-
Rhodesia. Dept. Sci. and Ind. Res. (U.K.), Road Res. Lab. peated cycles of loading and unloading. Third Asian Reg.
Note No. RN/ 1778/GDG (unpublished). Conf. on Soil Mech. and Found. Eng., Haifa, vol.
Grainger, G. D., 195 1. The production of low-grade ag- 1: 99-102.
gregate from black cotton soil by heat treatment. Dept. Katti, R. K., and Sudhar, S., 1962. Stabilization of black
Sci. and Ind. Res. (U.K.), Road Res. Lab. Note No. cotton soil sample with combinations of cement and lime.
RN/ 1482/GDG (unpublished). Jour. Indian Natl. Soc. of Soil Mech. and Found. Eng.,
Grim, R. E., 1968. Clay mineralogy, 2nd Ed., McGraw- vol. 1: 43-54.
Hill: 596 p.
Gupta, S. N., Gupta, B. N., and Shjukla, K. P., 1967. Keller, W. D., 1956. Clay minerals as influenced by en-
Physico-chemical properties of expansive clays in relation vironment of their formation. Bull. Am. Assoc. Pet. Geol.,
to their engineering behavior. Third Asian Reg. Conf. on vol. 40: 2689-2710.
Kerr, P. F., 1955. Formation and occurrence of clay rnin- Civil Eng. (London), vol. 52: 1387-1388.
erals. In: Clay and clay technology. Calif. Div. of Mines Mohan, D., and Rao, B. C., 1965. Moisture variation and
Bull. 169: 19-32. performance of foundations in black cotton soil in India.
Komornik, A., and David, D., 1969. Prediction of swelling In: Moisture equilibria and moisture changes in soils.
pressures in clays. Proc. Am. Soc. Civil Eng., Jour. of Soil Butterworths, Australia: 175-183.
Mech. and Found. Div., vol. 95: 209-226. Mohr, E. C. J., and Van Baren, F. A., 1959. Tropical Soils,
Laplante, A., 1954. Les sols fonces tropicaux d'origine W. van Heove, The Hague: 498 p.
basaltique au Cameroun. Trans. Fifth Int. Cong. Soil Sci., Morin, W. J., 1969. Design considerations for hghways over
Leopoldville (now Kinshasa); vol. 4: 144-158. tropical black clays. First African Highway Conf., Addis
Leneuf, N., 1954. Les terres noires du Togo. Trans. Fifth Ababa: 16 p.
Int. Cong. Soil Sci., Leopoldville (now Ki~shasa);vol. Morin, W. J., and Parry, W. T., in preparation. Geotechnical
4 : 131-136. Properties of Ethiopian volcanic soils.
Maufe, M. B., 1928. On the formation of red soil and black Moussaoui, M. M., 1967. Le comportement au Maroc des
vlei soil from dolerite at Salisbury, Southern Rhodesia. chaussees sur les "tirs." Thirteenth Int . Road Cong.,
So. Africa Jour. Sci., vol. 25: 156-167. Tokyo: 1 1/ 1-20
McAleese, D. M., and Mitchell, W. A., 1958. Studies on Muir, A., 1951. Notes on the soils of Syria. Jour. Soil Sci.,
basalt soils of northen Ireland. IV: Mineralogical study of vol. 2: 163-182.
the clay separates. Jour. Soil Sci., vol. 9: 76-80. Nagarajan, R., and Sundaraman, R., 1960. Swelling pressure
Mehra, S. R., and Chadda, L. R., 1957. Some laboratory of shrinkable soils. First Asian Reg. Conf. on Soil Mech.
investigations for improving black cotton .soil in road and Found. Eng., New Delhi, Part 111, Subj. 11, Paper
construction. Indian Roads Cong., Road Res. Bull. No. 22: 12 p.
4: 51-72. Nagelschmidt, G., Desai, A. D., and Muri, A., 1940. The
Mehra, S. R., and Uppal, H. L., 1964. Laboratory and field minerals in the clay fraction of a black cotton soil and a
experiences in the construction of roads in black cotton red earth from Hyderabad, Deccan State, India. J. Agric.
soil areas. Second Asian Reg. Conf. on Soil Mech. and Sci., vol. 30: 639-653.
Found. Eng., Tokyo, vol. 1 : 294-300. Nageswars, Ayyar, R. B. A., 1945. Road Construction in
Misty, J. F., 1957. Experimental soil stabilized road con- black cotton soils. Indian Roads Cong. Jour., vol.
struction in black cotton soil area. Indian Roads Congr. 10: 125-132,
Jour., vol. 22: 261-296. Narain, S., 1943. Treatment of foundations of buildings in
Mohan, D., 1961. Foundations on black cotton soils. Indian black cotton soil. Inst. of Eng. (India) Jour., vol.
Natl. Soc. of Soil Mech. and Found. Eng., Bull. No. 24: 51-54.
9: 102-105. Oakes, H., and Thorp, J., 1950. Dark clay soils of warm
Mohan, D., 1957. Consolidation and strength characteristics regions variously called rendzina, black cotton soil, regur
of Indian black cotton soils. Fourth Int. Conf. Soil Mech. and tirs. Proc. Soil Sci. Soc., vol. 15 : 347-354.
and Found. Eng., London, vol. 1 : 74-76. Palit, R. M., 1953. Determination of swelling pressure of
Mohan, D., 1955a. Foundations on expansive clays. India black cotton soil-a method. Third Int. Conf. on Soil
Central Build. Res. Inst., Bull. No. 3: 9-19. Mech. and Found. Eng., Zurich, vol. 1: 170-172.
Mohan, D., 1955b. Laboratory study of Indian black cotton Pollard, A. E., 1964. The road-making materials of Basuto-
soil. Indian Central Build. Res. Inst., Bull. No. 3: 20-27. land. Part 11. The engineering properties of the soils.
Mohan, D., and Chandra, D., 1961a. Frictional resistance of Dept. of Sci. and Ind. Res. (U.K.), Road Res. Lab. Note
bored piles in expansive clays. Geotechnique, vol. No. LN/689/AEP.
1 1 : 294-301. Radhakrishnan, N., Katti, R. K., and Hussain, M., 1967.
Mohan, D., and Chandra, D., 1961b. Short bored piles in Studies on thermal stabilization of black cotton soil.
shallow black cotton soils. Natl. Build. Org. (India) Jour., Third Asian Reg. Conf. on soil Mech. and Found. Eng.,
vol. 6: 197-202. Haifa, vol. 1: 111-115.
Mohan, D., and Goel, R. K., 1960. Swelling pressure and Ranganatham, B. V., 1961. Soil structure and consolidation
volume expansion in Indian black cotton soils. Inst. of characteristics of black cotton clay. Geotechnique, vol.
Eng. (India) Jour., vol. 40: 58-62. 11: 333-338.
Mohan, D., and Jain, G. S., 1958a. Pile loading and pull-out Ranganatham, B. V., and Henry, A. V., 1963. Permeability
tests on black cotton soil. Inst. of Eng. (India) Jour., vol. of black cotton clay as related to void ratio and absorbed
38: 409-421. water. Sols, vol. 2: 22-29.
Mohan, D., and Jain, G. S., 1958b. Under-reamed pile Ranganatham, B. V., and Satyanarayana, B., 1967. Earth
foundations in black cotton soil. Indian Concr. Jour., vol. Pressures in expansive backfills. Third Asian Reg. Conf. on
32: 203-207. Soil Mech. and Found. Eng., Haifa, vol. 1 : 116-119.
Mohan, D., and Jain, G. S., 1957. Field vane shear tests. Ranganatham, B. V., and Satyanarayana, B., 1965a. Labora-
 tory studies on swelling pressures of black cotton soils. Van der Merwe, D. H., 1967. The composition and physical
Fifth Sym. Civil and Hydr. Eng. Dept., India Inst. of Sci., properties of the clay soils on the mafic rocks of the
V O ~ .1: 13/1-15. Bushveld igneous complex. Fourth Reg. Conf. for Africa
Ranganatham, B. V., and Satyanarayana, B., 1965b. A on Soil Mech. and Found. Eng., Cape Town, vol.
rational method of predicting swelling potential for com- 1: 141-146.
pacted expansive clays. Sixth Int. Conf. on Soil Mech. Van der Merwe, D. H., 1964. The weathering of some basic
and Found. Eng., Montreal, vol. 1: 92-96. igneous rocks and their engineering properties. The Civil
Road Research Laboratory, 1947. An examination of a Eng. in Southern Africa, vol. 6: 213-222.
sample of cotton soil from Palestine. Dept. of Sci. and Van der Merwe, D. H., 1962. Weathering of various types of
I n d . Res. (U.K.), N o t e N o . RN/831/KEC.JS Karoo dolerite in the Middelbury Area, Cape Province.
(unpublished). Trans. Geol. Soc. So. Africa, vol. 65, part 2: 1 19-137.
Ross, C. S., and Hendricks, S. B., 1945. Minerals of mont- Van der Merwe, C. R., 1962. Soil groups and subgroups of
morillonite group; their origin and relation t o soils and South Africa. Dept. of Agric. Tech. Services, Chem. Series
clays. U.S. Geol. Survey Prof. Paper 205B: 23-77. No. 165: 355 p.
Sherman, G. D., 1952. The genesis and morphology of the Van der Merwe, C. R., and Heysteck, H., 1955. Clay
alumina-rich laterite clays. In: Problems of clay and minerals of South African soil groups. 11: Subtropical
laterite genesis. Am. Inst. Min., Met. and Pet. Eng., New black clays and related soils. Soil Sci., vol. 79: 147-158.
York: 154-161. Venkatesan, M. N., and Bapaye, D. R., 1955. Soil cement as
Simons, N. E., and Williams, A. A. B., 1963. Preliminary a construction material. Irrigation and Power (India), vol.
report on moisture studies at Ondersterpoort site near 12: 495-509.
Pretoria. Third Reg. Conf. for Africa on Soil Mech. and Weinert, D. H., 1964. Basic igneous rocks in road foun-
Found. Eng., Salisbury, vol. 1: 41-46. dations. So. ~ f r i c aCoun. Sci. and Ind. Res., Res. Rep.
Stephen, I., 1953. A petrographic study of a tropical black No. 218. (Natl. Inst. of Road Res. Bull. No. 5),
earth and grey earth from the Gold Coast. Jour. Soil Sci., Pretoria: 47 p.
v01. 4: 21 1-219. Weinert, H. H., 1968. Engineering petrology for roads in
Strongman, F. S., 1963. Black cotton soil road embankment. South Africa. Eng. Geol., vol. 2: 363-395.
Third Reg. Conf. for Africa on Soil Mech. and Found. Weinert, H. H., 1965. Climatic factors affecting the weath-
Eng., Salisbury, vol. 1: 89-91. ering of igneous rocks. African Meteorol., vol. 2 : 27-42.
Subhash, C., and Khepar, S. D., 1964. Double under-reamed Williams, A. A. B., 1965. The deformation of rocks resulting
piles for foundations in black cotton soils. Indian Concr. from moisture changes in expansive soils in South Africa.
Jour., vol. 38: 50. In: Moisture equilibria and moisture changes in soils.
Sys, C., 1959. Terres noires tropicales au Congo Belge. Butterworths, Australia: 143-155.
Trans. Troisiime Conf. Interafr. Sols, Dalaba, vol. Williams, A. A. B., 1959a. Some regional aspects of soil
1: 209-214. mechanics in South Africa. Second Reg. Conf. for Africa
Tamhane, R. V., and Manjosh, N. G., 1959. A comparative o n S o i l Mech. a n d Found. Eng., Lourenco
study of black soils formed from different parent rocks. Marques: 11-16.
Jour. Indian Soc. Soil Sci., vol. 9: 49-63. Williams, A. A. B., 1959b. Some observations of vertical
Tamhane, R. V., and Sen, N., 1954. The black cotton soils movement as an indication of moisture movement and
and the basic rocks of south India. Inst. of Sci. of India distribution under roads and buildings. Second Reg. Conf.
Bull.,No. 3: 177-184. for Africa on Soil Mech. and Found. Eng., Lourenco
Theron, J., and van Niekerk, P. Le R., 1934. The nature and Marques: 59-63.
origin of black turf soils. S. African Jour. Sci., vol. Williams, A. A. B., and Jennings, J. E., 1960. Problems on
31: 320-346. pavements and earthworks on heavy clays in tropical
Uppal, H. L., 1965. Field study on the movement of climates. Conf. Proc. on Civil Eng. Problems Overseas,
moisture in black cotton soils under road pavements. Inst. of Civil Eng.
In: Moisture equilibria and moisture changes in soils Williams, A. A. B., and Simons, N. E., 1963. The heaving of
beneath covered areas. Butterworths, Australia: 165-174. roads on expansive clays with suggested remedies based
Uppal, H. L., and Palit, R. M., 1959. A study of some on field moisture conditions. Third Reg. Conf. for Africa
factors influencing the swelling pressure of clays. Jour. on Soil Mech. and Found. Eng., Salisbury, vol. 1: 47-5 1.
Indian Roads Cong., vol. 24: 543-555. Williams, F. H. P., 1961. The laboratory testing of soils from .
Uppal, H. L., and Sehgal, T. R., 1958. A statistical ex- six sites in Kenya. Dept. Sci. and Ind. Res., (U.K.), Road
amination of the classification suggested for black cotton Res. Lab. Note No. RN/3939/FHPW (unpublished).
soils o f India. Jour. Indian Roads. Cong., vol. Wooltorton, F. L. D., 1959. Low cost design in Kenya.
23: 577-588. African Roads and Transport, vol. 17, No. 5 : 5 , 7, 9 , 12.
Wooltorton, F.L.D., 195 1. Movements in the dessicated the subject of foundations in the "Black Cotton" and
alkaline soils of Burma. Trans. Am. Soc. Civil Eng., vol. "Kyatti" soils of the Mandalay District, Burma. First Int.
1 16: 433-479. Conf. on Soil Mech. and Found. Eng., Cambridge, Mass.,
Wooltorton, F. L. D., 1936. A preliminary investigation into vol. 3: 242-257.
 CHAPTER 16

GEOGRAPHICAL DISTRIBUTION AND PHYSICAL

PROPERTIES OF AFRICAN BLACK
CLAY SOILS

INTRODUCTION over recent alluvial sediments.

The physical properties of the tropical black clays are A brief description of the major black clay occurrences
distinctly different from those of ferruginous soils but there given in Table 36. Chapter 15, permits evaluation of the
can be considerable overlap in properties with ferrisols and relative severity of the problems they impose on construc-
ferrallitic soils. This is particularly true over geologically'recent tional activities. The appraisal is necessarily cursory; many of
volcanic rocks. There has been insufficient time t o allow the locations were not examined in the field.
transformation of most of the clay minerals from montmoril-
lonite t o kaolinite t o gibbsite, except in areas of high rainfall GEOGRAPHICAL DISTRIBUTION
where erosion has not proceeded as rapidly as soil develop- West Africa
ment. Over other rock types, most of which are geologically Dark clay soils in West Africa occur in Dahomey, Ghana,
older, the index properties of the ferrisols or ferrallitic soils Ivory Coast, Nigeria, Senegal, Togo and Upper Volta (Dudal,
still overlap those of the black clays, but the differences in 1965). Most of these pose no serious problems, and samples
engineering properties are generally distinguishable. Over older were selected only from Ghana and Nigeria. The dark soils of
basic rocks, leaching has altered almost all of the montmoril- Ghana, Togo and Dahomey are all formed over Precambrian
lonite t o kaolinite and gibbsite and has resulted in a greater metamorphics-schists and gneisses. Their clay contents are
iron content as well. not h g h and the soils do not cause constructional problems.
The physical properties of selected tropical black clay soils Only a few index tests are reported.
are discussed in this Chapter with emphasis on swelling The deposits in Ivory Coast and Upper Volta are limited
properties. Chemical, mineralogical, and engineering properties areas of eutrophic brown soils. These are not considered
are also discussed. troublesome soils and consequently were not included in the
The soils investigated were selected from the major deposits study.
discussed in Chapter 15 and shown in Fig. 1 16. Minor residual A small occurrence of black clay is found on the Cape
and alluvial deposits of limited area are not included as other Verde peninsula in Senegal. The clay has formed residually
soils are generally available and their use may be more over a small basic intrusive body. No samples or test data were
economical. Moreover, the properties of these limited deposits collected.
can be expected t o fall within the range described in this Nigeria contains both residual and alluvial occurrences.
Chapter. The emphasis here is on larger deposits, where Volcanic bodies of Tertiary age or younger occur in north-
construction difficulties have occurred and where alternative eastern Nigeria in the vicinity of Biu and Meringa and in other
sources of material are restricted or absent. isolated erosional remnants of the basalt flows which once
capped the plateau. Other small, isolated patches of black
SCOPE clays have formed residually over calcareous rocks. Black
The objectives of the laboratory investigation of the alluvial soils occur along the Benoue River Basin, above its
tropical black clay were t o determine typical properties of junction with the Niger. However, the largest area of black
various deposits, t o explain these properties with regard to the clay, also in northeastern Nigeria, is part of the lacustrine and
chemical and mineralogical composition, and t o select tests alluvial sediments within the Chad Basin. These soils will be
that suitably describe the engineering behavior. discussed further below.
Sampling was limited because of the short time available One sample was forwarded t o Kumasi by the Nigeria Public
and because of transportation difficulties. Obviously, it was Works Department. It was collected from the Bauchi-Gombe-
not possible t o sample many of the occurrences of black clays Yola road in the northeastern part of the country. Between
shown on the Soils Map. Samples were shipped to Kumasi by Gombe and Yola, the road traverses a few small remnants of
air, except for one large sea shipment from East Africa. This basalt-capped plateaus and also passes some distance through
imposed limitations not only on the number of samples, but the alluvial river plain. However, large sections pass through
also on their size. "vertisols" or black clays derived from calcareous rocks. It is
The samples were collected from areas selected from the not known which of the three soil types were sampled, but the
Soils Map, after considering the geology of the area. Soils over chemical data in Table 42 suggest that the soil formed over
recent volcanic rocks were augmented by samples in poorly calcareous rocks.
drained areas over various rocks including Precambrian, and Black clays formed over calcareous rocks are not common
 TABLE 42

Chemical and Mineralogical Properties

Tropical Black Clays
Cation Exchange
Capacity in Exchangeable Organic
Location Sample No. Parent Material Clay Mineralogy M.E.1100 Grams Cations pH Content

Chad Basin CB-1 Alluvium 34.90 mostly Ca; 8.4 <1.0%

with Mg; minor
K and Na
Chad Basin CB-2 Alluvium 32.83 mostly Ca; 6.3 <2.0%
with Mg; minor
K and Na
Chad Basin CB-3 Alluvium Kaolinite, illite,
montmorillonite,
Chad Basin CB4 Alluvium quartz, kaolinite 6.3
montmorilloni te
Nigeria NBC-1 Calcareous montmorillonite, .O%
Rocks quartz, kaolinite with Mg; minor
K and Na
Ethiopia ETH-1 Basalt montmorillonite, 64.90 mostly Ca; 7.2 <I.O%
kaolinite, halloy- with Mg; minor
site K and Na
Kenya KEN-6 Basalt montmorillonite, 8.8
kaolinite
Kenya KEN-9 Basalt montmorillonite, 45.02 mostly Ca; 7.2
kaolinite with Mg; minor
K and Na
Zambia ZAM-1 Alluvium 32.16 mostly Ca; 1.O-1.5%
with Mg; minor
K, Na, Mn
Zambia ZAM-2 Basalt 33.34 mostly Ca; 6.4 <z .O%
with Mg; minor
K, Na, Mn
Uganda U-12 Basalt 56.74 mostly Ca; 7.4 1.O%
minor Mg, K
and Na
Tanzania TAN-X-1 Alluvium <z .o%
Tanzania TAN-3 A Alluvium 30.07 mostly Ca; 9.2
with Mg; minor
-- K and Na
Morocco MOR-2 Alluvium from montmorillonite,
basalt chlorite, kaolinite
Morocco MOR-3 Alluvium from montmorillonite,
shales chlorite, kaolinite
Sudan SU-1 Alluvium from montmorillonite, 8.9
basalt kaolinite, illite,
quartz, calcite
Sudan SU-2 Alluvium from montmorillonite,
basalt illite, kaolinite,
calcite
Sudan SU-3 Alluvium from montmorillonite,
basalt kaolinite, illite
quartz, calcite
Sudan SU-5 Alluvium from montmorillonite,
basalt kaolinite, illite
.- . , calcite
auartz.
Sudan SUD-D Unknown montmorillonite,
vermiculite,
misc.
West ~ a m e r o o n CAM-2
West Cameroon WC-2 Basalt vermiculite,
metahalloysite
in Africa but they have been reported in other parts of the origin. The volcanic rocks from the Ethiopian Plateau extend
world, such as the western United States and Israel. Such rocks into Sudan and undoubtedly some residual soils occur in the
are rich in the bases Ca and Mg, and with an alkaline pH and eastern parts of the country. However, the black clays of the
plentiful silica, montmorillonite will form as the weathering plateau have been transported down into the former lacustrine
product. basin and are being eroded now by the various streams,
principally the Blue Nile, that flow from the Plateau into the
Central Africa Sudan Basin.
The most important black clay area in Central Africa is the Most of the Ethiopian Plateau and part of the Somalian
Chad Basin, w h c h covers northeastern Nigeria, northern Plateau are covered with tropical black clays formed over
Cameroon and much of southern Chad. Most of the soils are basalt. The intervening Rift Valley has a variety of soils-
lacustrine or alluvial, but some residual soils over basic alluvial, lacustrine, and residual black clays over the more
extrusive rocks occur in northern Cameroon as well as in recent volcanics. Much of the k f t Valley is shown on the Soils
northeastern Nigeria. Basic rocks have been reported in Map as raw mineral soils, indicating that soil development is
Cameroon between Maroua and Mora. immature and has not overtaken erosion. Ferrisols also occur
Actually the soils of the Lake Chad Basin vary consider- on the plateaus, representing the initial stages of laterization.
ably. Although black clay soils are common, silty clays and Most of the tropical black clay on Kenya, Uganda, and t o a
sandy clays are also common. Pedological workers have lesser extent Tanzania were derived from basic volcanic rocks.
divided the basin soils into three types: ( I ) mostly clay, locally Low-lying areas in Uganda, associated with the Rift valley
called "firki" soils; (2) sandy clay; and (3) clayey sand. In lakes, contain alluvial black clays. Large areas of northwestern
Nigeria these soils are called the Ngala Series. Four small Tanzania, southeast of Lake Victoria, also have developed
samples were collected from the Chad Basin, two in Nigeria black clay soils. The alluvial soils in East Africa were sampled
and two in Cameroon. in Tanzania, west 'of Dar-es-Salaam. The samples from Uganda,
The volcanic mountains and rocks of West Cameroon and Kenya and Ethiopia were collected from residual soils over
the island of Fernando Po have developed brown t o black soils volcanics.
which, under the high rainfall of the area, are slowly being Alluvial black clays, associated with intermit tent streams,
altered t o ferrisols or ferralitic soils. The darker soils are occur in southern Somalia. These were not sampled during this
included here under black clay soils; the red soils were study. Ample granular materials are found along the coastal
included in the lateritic soils, Chapter 5. regions, so construction problems should be relatively minor.
The only other areas with tropical black clays in Central Further south, Malawi and Zambia both contain small areas
Africa are the Congo and Angola. Occurrences include clays of black clays. Most of these in Malawi are associated with the
formed over basic rocks in the southern part of Angola as well Rift Valley lakes and are believed t o be alluvial. The
as black clays formed over calcareous rocks around Luanda. occurrences in Zambia are both alluvial, along stream valleys,
The Congo deposits are mostly in or near the Rift Valley, in and residual, over volcanics, in the southern part of the
poorly drained areas. country, along the Zambesi River. One sample of alluvial black
clay was collected near Lusaka and one sample of residual
Eastern Africa black clay was collected near Livingstone.
Eastern African occurrences are by far the most widespread
in Africa. Large parts of Sudan and Ethiopia are covered with Southern Africa
black clay soils, as are smaller areas in Tanzania, Kenya, None of the various deposits in southern Africa or
Uganda, Malawi, Zambia and Somalia. The topography of the Malagasay was sampled during this study. However, the
Rift Valley system is the main cause of the black soil composition and properties of these soils have been well
formation in poorly drained areas of Tanzania, Malawi and established in South Africa, where the soils are largely
Zambia. However, the associated volcanic rocks are the main sub-tropical. Most of the soils in South Africa and Rhodesia
cause of the black clays forming jn Kenya, Ethiopia and were derived from basic intrusive or extrusive rocks. Van der
Sudan. Merwe (1962) states the black clays occur over norite
A large area of southern Sudan contains black clay. Much (gabbro), basalt, and diabase.
of the area contains partially black clay and partially eutrophic In Botswana, black clays are found in poorly-drained
brown clay. Also the flood plains of the White Nile and Blue alluvial deposits and swamplands, and also over basic rocks in
Nile include black clay among various silty and sandy alluvial the northern part of the country, near the Zambesi River and
deposits. The White Nile in southern Sudan broadens into the the Zambian border.
large swamplands of the Sudd with its accumulation of organic Lesotho is virtually completely covered with basalt which
matter and floating vegetation. has weathered t o form black clay soils. Swaziland has less areal
The soils of Sudan were only cursorily examined during this coverage of basalt but the soil development has been the same.
study, but it is believed that most are lacustrine or alluvial in Rhodesia also contains alluvial and residual black clays. The
residual clays have developed over basalts that are part of the There is considerable variation in the index properties
Karroo formations of Rhodesia. These are found in the shown in Table 43. The silt and clay content varies from 58 to
southeast adjoining Mozambique, and in the west adjoining 90%, while the clay fraction varies from 4 0 to 70%. Liquid
Zambia. The area around Salisbury contains black clays limits range from 22 t o 124% and plasticity indexes vary from
believed t o be alluvial. 11 to 74. The shrinkage limits are fairly high; in fact these may
Other deposits are found in southwest Africa, Mozambique be suspect since high shrinkage limits indicate little volume
and Malagasay. Southwest Africa contains a continuation of change in the material as it approaches the plastic limit.
the residually formed soils in Angola, as well as other black Activities are all below 1.70 and most are under 1.0. Thus,
clays over intrusive and extrusive bodies further south and most of the clay would be considered on this basis as
over alluvial deposits along the streams. Karroo volcanics "normal" by Skempton (1953). Almost all of the soils are
extending from Rhodesia into west ern Mozambique have either A-7-5 or A-7-6. The recent AASHO interim procedure
produced black clays; alluvial deposits occur in the northeast, (M145-661) for calculating the group indexes was adopted.
along the coast, north of Mozambique City. Black clays in With this expanded range, the group index values vary from 13
Malagasay have formed in the northern and western parts of to 89.
the country over basalts. The basalts are Cretaceous or Engineering properties include moisture-density relation-
Tertiary and Quaternary in age (Furon, 1963). Clays have also ships, California Bearing Ratios and swelling properties. The
developed in alluvial deposits in valley depressions. first two are listed in Table 43, while the various swelling tests
results are given in Table 44. Both Modified AASHO and
North Africa Standard AASHO densities and moisture contents are reported
Expansive black clay soils are found in northern Morocco. for most soils. The Modified AASHO maximum densities vary
These soils, locally called "tirs", are derived from fine-grained from 85.3 pcf t o 119.7 pcf w h l e the Standard AASHO
sedimentary rocks, such as shales, which are being eroded in densities range from 67.2 pcf to 101.8 pcf. The residual black
the highlands t o the east. The black clays are therefore alluvial clays over volcanic rocks display noticeably lower densities
and sub-tropical. Two small samples were collected from these and higher optimum moisture contents. In t h s respect they
soils for testing in Kumasi. The objective was to compare the resemble the ferrisols and ferrallitic soils formed over volcanic
index properties with those of the tropical black clays; rocks.
extensive swelling tests were not conducted. CBR values are all very low and reflect the instability of the
black clays. The range is from 0.5 to 0.8 when compacted at
TEST PROCEDURES AND RESULTS AASHO Standard density and from 0.8 to 1.6 when com-
pacted at AASHO Modified density. Relationships between
Test results for the black clays are given in Tables 42, 4 3 density, moisture content and CBR are shown in Fig. 120 for
and 44. Chemical and mineralogical data are given in Table 42. three soils. In all cases the CBR values increase with increasing
Index properties and engineering properties are shown in Table moisture contents, and swelling decreases.
4 3 and swelling tests are reported in Table 44. Swelling was determined by means of four different tests:
The samples are believed to be fairly representative of the CBR swell, free swell, expansion-pressure and P.V.C. (Potential
tropical black clays of Africa. Both residual and alluvial soils Volume Change). Expansion-pressure tests were performed in
are represented. The residual soils tested were mostly over accordance with AASHO T-190-61 with slight modifications as
volcanic rocks, but a few were over basic intrusives and adopted by the Utah State Department of Highways. The test
metamorphic rocks. The parent material for each soil tested is procedure is given in the Appendix. Expansion-pressures were
given in Table 42. determined at various densities and moisture contents. The
Samples were collected from Morocco, Ghana, Nigeria, figures reported in Table 4 4 are for tests conducted at both
Cameroon, Chad Basin, Sudan, Ethiopia, Kenya, Uganda, maximum densities and corresponding optimum moisture
Tanzania and Zambia. Results are included from unpublished contents. The results range from 180 t o 1460 psf for samples
reports concerning Ethiopia, Kenya and Sudan and from a compacted at Modified AASHO 'density and from 284 to
published report from Morocco (Moussaoui, 1967). 1150 psf for .those compacted at Standard AASHO density.
, The mineralogy of the samples is basically similar. Montmo- Relationships between expansion pressures and dry densi-
rillonite is the principal clay mineral reported, but kaolinite, ties at initial molding moisture contents are shown in Fig. 121
halloysite and illite are also common. Only alluvial soils for four soils. The increase in expansion pressure with
contain illite and only soils over volcanic rocks contain increasing density and decreasing initial moisture content is
halloysite. The cation exchange capacities range from 30.07 to apparent.
64.90 milli-equivalents per I 00 grams. The principal exchange- Free swell is reported for most soils. This is an extremely
able cation in all cases is calcium. Magnesium is present in simple test and it was desirable to compare the results with
significant quantities, but potassium, sodium and manganese those from other swelling tests. The free swell is determined
occur only in minor amounts. Most of the soils are basic, with by slowly pouring 10 cc of soil passing the No. 4 0 sieve into a
pH values varying from 6.30 t o 9.20. graduated cylinder containing 100 cc of water. After the soil
 TABLE 43

Physical Properties
Tropical Black Clays
Standard Mod~fied
Gradation: (%I Atterberg Limits: (%) AASHO AASHO
AASHO Max. Opt. Max. 0 P..'
Sample Class~fica- Group Speciflc Density Moist. Dens. Mols.
Location No. Gravel
--
Sand Silt Clay LL P L PI SL tlon Index Activlty Gravity pcf. %/BR pcf. % CBR
Chad Basin CB1 0 --
10 -
20 --
70 58 16 42 12 A-7-6 - -
40 0.61 -
107.3 20.2 1.6
-

Chad Basin CB2 0 19 28 -

-
53 44 14- 30 9 A-7-6 24 0.59 --- - -
119.7 13.0
Chad Basin CB3 0 14 34 52
- --
56 30 26 15 A-7-6 20 040 89.4 18.9 - - -
103.6
-
18.6
Chad Bas~n CB4 0 lo -
-
30 60 52 19 33 -
15 A-7-5 25 0.56 --
89.7 19.3--

Nigeria NBC-I 2 11 -
30 57 -
66
-
20 46 -- -
A-7-6 43 0.82 - - --
2.52
- - -
101.6-
- -
24.0
-
- -
105.5 20.0 1.5
Ethiopia ETH-1 2 38 56 109 28 81 14 A-7-6 89 1.36 75.4 32.5 0.5 92.7 23.0 1.0
Kenya KEN-6 - 8 22 8 62-
-
104 34 70 - --
A-7-5
- ---
52 1.15 _
--
2.28
--
67.2 37.2-
-
--
- --
85.3 32.2 1.O
Kenya KEN8 0 --
8 37 55 -- -72 24 48-- ---
A-7-6 - -
50 0.88 --
2.47 -
72.9 24.3 94.6 23.7 0.9
-

-I Zambia ZAM-1 0 14 35 51 49 16 33 l6 A-7-6 38 0.65 2.58 -- 95.5 18.8 112.5 16.0

0 0 - - -- - ---
Zambia ZAM-2 0 30 20 -
50 -- 50 17 33 15 A-7-6 21
-
0.67 2.6 1 - -
94.7
.-
17.8
Uganda U-12 6 6 32 56 - 87 24 63 l8 -
A-7-6
--
61 1.12 - -
-
2.33 82.9
--
31.4 0.8 90.8 24.0 0.8
Tanzania TAN-X-1 2 19 31 48 60 30 30 A-7-5 --
- -26
---
- -0.63 100.5 20.2 108.3 - 18.7
Tanzania TAN-3A -
4 6 50 40 58 16 42 15
--
A-7-5 - - 40 1.04 99.8 21.1 0.7 114.2 15.3
Morroco MOR-2 0 14 32 54 56 24 32 A-7-6 30 0.59 - -
101.8 24.8 - -

Morroco MOR-3 0 19 25 56 59 31 28 --
A-7-5 -
25 0.50 --
100.3 25.3 107.4 19.4
Sudan SU-I 70 29 41
Sudan SU-2 5 8 30 28 --
1.2
Sudan SU-3 -- 30 22 8 --
Sudan SU-5 40 17 23
Cameroon CAM-2 1 41 58 64 37 27 A-7-5 13 1.70
Cameroon WC-2 0 19 43 38 62 35 27 A-7-5 26 0.71
Cameroon WC-5 - -- 85 64 21 -- -
Ghana GH-4-7 47 18 29
Ghana GH-4-13 22 6 16 ---
Morocco
- MOR-M 5-10 25-30 60-70 - 60e 20e
- 40e 10 -.
- --- -
97-100 --
3-6
Sudan SUD-D 61 25 36 11 -
2.80 - -
Ethiopia ETH-T 96- 124 54-74 10-17 2.87-
- 81-85 20-31
Kenya KEN-T 75 94 38 55 11.3-16.0 2.63 83.0 32.5 94.2 26.6
 TABLE 44

Swelling Properties
Tropical Black Clays

- -- -
Volume Change CBR
Expansion Pressure, psf - - - Swell Index (PVC), psf % Swell, %
Load To At At
Sample At AASHO Mod. At AASHO Std. At AASHO Mod. At AASHO Std. Under Retain Std. Mod. Free Swell,
Location NO-.
--
MDD and OMC MDD and
.-
-
OMC - -
MDD and OMC
-- -- -- --- - -
MDD and OMC -
-
1 psi Vol., Ibs. Dens. Dens. %

Chad Basin CB 1 5850 3300 5.3 70

Chad Basin CB 2 895 55

Chad Basin CB 3 4450 3200 50

-
Chad Basin CB 4 6000 3350 50

-
00
00
Nigeria
-.-
NBC-1 1030 585 5275 3300 9 -0 7000 5.2 72

Ethiopia ETH-1 1460 1150 7000 4525 15.8 16000 11.4 12.4 88
-- -- - .- - -.

Kenya KEN-6 1125 646 7000 3450 11.5 92

Kenya KEN-9 1520 5487 25 00 13.4 87

Zambia ZAM-1 800 410 4975 2975 5.5 4000 57

- -

Zambia ZAM-2 180 4375 2025 58

Uganda U-12 1550 284 6150 4375 14.6 14.5 82

Tanzania TAN-X-1 650 385 4075 2350 5.7 65

Tanzania TAN-3A 1232 390 4975 2 175 5.2 4000 7.0 7 .O 62

Sudan SUD-D 100

MDD = Maximum Dry Density

OMC = Optimum Moisture Content
PVC = Potential Volume Change
 E T H I O P I A- l
E T H I O P IE- l

93 - AASHO MODIFIED
MODlFlE

2 6 2 8 3 0 3 2 34
MOISTURE CONTENT T A N Z A N I A-3A MOISTURE CONTENT
TENEUR EN EAU T A N Z A N I E-3A TENEUR EN EAU

I , AASHO STANDARD

19 20 21 22 23 13 15 17 19 21
MOISTURE CONTENT MOISTURE CONTENT
TENEUR EN EAU TENEUR EN EAU

UG A N D A-12

28 32 36 40 44 21 23 25 27 29
MOISTURE CONTENT MOISTURE CONTENT
TENEUR EN EAU TENEUR EN EAU

FIG. 120 MOISTURE -DENSITY -CBR RELATIONSHIPS (CBR SWELL SH OWN BELOW C BR CURVE )

RELATION TENEUR EN EAU-DENSITE-CBR (LA COURBE INFERIEURE EST L E GONFLEMENT AU

COURS DU CBR
 I LEGEND
0MC*25°/o
LEGENDE
TENEUR E N EAU MAXIMUM= OMC /a / OMC
LEGEND
* 10%
X 9 5 % OMC *I 0 %
TENEUR
LEGENDE
EN EAU MAXIMUM =OM C

0 105%OMC*l On

I SOIL ETHIOPIA -I / / TANZANIA - X-l

TANZANIE - X-l

01
DENSITE AASHO STAND*
I
I
50 60 X) 80 90 90 92 94 96 90 100 102 104 I06

DRY DENSITY DRY DENSITY

DENSITE SECHE DENSITE SECHE

2000
LEGEND LEGENDE 2000
OMC=IO% TENEUR EN EAU MAXIMUM = 4 M C LEGEND-- LEGENDE
OMC t 2 5% TENEUR EN EAU MAXIMUM=OMC
a 95%OMC *I 0 %
a 95%OMCt25Y.
0 105Yo OMC * 2O X N 0 1 0 5 % OMC 1 1 0 %
-
W
a
SOlL NIGERIA - N B C - I - x

I500
SOL NIGERIA - NBC- I i $ -
SOIL
!
KENYA-9
x -
SOL KENYA-9
)I

I500

::
Z

J
c

s5 0
0

1000 ax/ CZ g
W
o a

$
E
W

0
z:
:2
E:

500
50 0

MOO AASHO DENSITY

DENSITE AASHO M O D I F I ~ MOD AASHO DENSITY
DENSITE AASHO MOD IF^

95 100 105 110 I15 86 88 90 92 94 96 98 100

D R Y DENSITY
DENSITE SECHE DRY DENSITY
DENSITE SECHE
FIG. 121 RELATIONSHIPS BETWEEN EXPANSION-PRESSURB AND DRY DENSITIES AT VARIOUS INITIAL
MOISTURE CONTENT
RELATION ENTRE PRESSIONS DE GONFLEMENT ET DENSITES SECHES AVEC DIVERSES
TENEURS EN EAU INITIALES
 16,000
change test are shown in Fig. 122. Volume change is plotted
against surcharge load; a load of 10,400 lbs. was required to
restore the sample t o its initial volume.
Since the purpose of the expansion study was to evaluate
suitable tests t o adopt for highway design, no undisturbed
testing was performed. The double oedometer test, devised by
- I 4010.-----
10,400 LIVRES

LBS
POUR

FOR ZERO
GONFLEMENT

SWELL
ZERO
Jennings and Knight (1957-58), has been useful in estimating
heave under buildings, but undisturbed samples are required.

DISCUSSION O F PROPERTIES
The classifications of the black clays reported in Table 43
are similar t o those given in the literature, Chapter 15. A
plasticity chart, Fig. 123, shows similar properties and both
the correlation line and Casagrande's A-Line are shown. The
correlation equation is:
SOIL : NIGERIA - NBC - I
SOL NIGERIA - NBC - I

A correlation performed by Kassiff, et al. (1969) for Israeli

black clays produced a similar equation:

VOLUME CHANGE. %
VARIATION DE VOLUME '10 Most of the clays plot above the A-Line. Those that do not are
FIG 122 EFFECT OF SURCHARGE LOAD ON VOLUME CHANGE
from Cameroon and contain halloysite.
INFLUENCE DE L A SURCHARGE SUR L A VARIATION A few other significant correlations were established. Fig.
DU VOLUME 124 shows the relationship between liquid limit and both
comes to rest at the bottom of the cylinder, the expanded Modified and Standard AASHO maximum density. The
volume is measured. It is calculated as follows: relationship between Modified densities and the amount
passing the No. 200 sieve multiplied by the PI is shown in
Final Volume - Initial Volume x 100
Free swell, % = Fig. 125.
Initial Volume (1 0 cc) The relationship between optimum moisture contents and
plastic limit is shown in Fig. 126. Both optimum moisture
The results ranged from ~ Wt O o 100%. contents at Modified and Standard AASHO densities are
PVC (potential volume change) tests were conducted included. Most optimum moisture contents are either near or
following the procedure described by Lambe (1960), who above the plastic limit.
developed the apparatus for a rapid determination of the A modification t o Skempton's colloidal activity (1953) was
expansiveness of soils. As Lambe's report may not be readily proposed by Williams (1957-58) to classify the soils into very
accessible, .the test procedure is reproduced in the Appendix. high, high, medium and low degrees of potential expansive-
The apparatus is now available commercially from various ness. The black clays were plotted on the basis of this
suppliers of soils laboratory equipment. The swell indexes classification (Fig. 127). All fall within the very high to high
were determined at both maximum densities and corre- degrees of potential expansiveness.
sponding moisture contents. This also is a relatively simple test Swelling in soils depends on the initial moisture content
which can be easily incorporated in testing programs. Results and the subsequent amount of moisture the soil is allowed or
varied from 4075 psf to 7000 psf for soils compacted at able to absorb. This point is considered inqgreaterdetail in the
Modified AASHO densities, and 2025 psf to 4525 psf for soils next chapter. However, if a soil is completely desiccated it will
compacted at Standard AASHO densities. exert high expansion pressures and volume change as it
A few volume change tests were performed employing approaches saturation. When highways are constructed over
oedometers. This is a very time-consuming test if a series of expansive black clays, the amount of expansion and swelling
samples at varying initial moisture contents and surcharge pressure depends on the moisture content existing at the time
loads is tested. The tests were performed only at Standard of construction.
AASHO maximum density and optimum moisture content. The expansion-pressure and volume change tests in the
Expansion was allowed under the normal 1 psi load until present study were usually performed on soils at either
equilibrium was reached. The sample was then loaded in Modified or Standard densities and optimum moisture con-
increments until the original volume was restored. Volume tents. These are reasonable initial moisture contents at which
change varied from 5.2% to 15.8%. The results of one volume t o conduct swelling tests for highway purposes, as they should
 LIQUID LIMIT
LlMlTE DE LlQUlDlTE
FIG. 1 2 3 PLASTICITY CHART, AFRICAN BLACK CLAYS
DIAGRAMME DE PLASTICITE, ARGILES NOIRES
D'AFRIQUE

be the moisture-density condition that will be applicable specimen leads t o a considerably longer time for saturation
during construction. than the 0.75 inch PVC specimen. As a consequence the
Of the swelling tests investigated, free swell and PVC at expansion pressures are considerably lower than the PVC swell
Modified AASHO are considered the best indicators. Correla- indexes.
tions between free swell and Atterberg limits are shown in Fig. The effect of time on expansion pressure was investigated
128, and correlations between PVC at Modified AASHO and by Seed et al. (1962). They advocate a 7-day testing period to
Atterberg limits are shown in Fig. 129. Correlation coefficients develop complete expansion. Testing one sample, Kenya-9,
for each are also given. The relationships between PVC at indicated that 30 days were required t o develop complete
Modified AASHO and Group Index; and free swell and Group expansion, and the final expansion pressure was more than
Index are given in Fig. 130. The expansion pressure at double that developed in 24 hours. In fact, in 5 other samples
Modified AASHO also correlated well with liquid limit, tested, water did not percolate through the specimen in a
plasticity index, group index, activity, and plasticity index testing period of 7 days. The final expansion pressures in 4 of
multiplied by the amount passing the number 200 sieve. the 5 samples was approximately double the 24 hour value. In
However, correlations are not as significant as with the free one, it was 3 times the 24-hour value. From 21 t o 30 days
swell and PVC at Modified AASHO density. were required t o develop the full expansion pressure.
Neither expansion-pressure nor PVC at Standard density A single PVC test on the Nigerian black cotton soil
showed very high correlations with the index tests. Expansion indicated that 8 hours were required t o attain full expansion
pressures at lower densities are evidently more erratic. Com- and that the full swell pressure was over twice the 2 hour
pacting the soil in the PVC device at Standard AASHO is value. Nevertheless, the standard swell index test is preferred
accomplished with one layer and this may produce a non-uni- for providing a useful approximation of expansion pressures
form density and irregular expansion within the samples that are likely t o develop after construction.
during the two hour testing period. It has been well documented that volume change and
The expansion-pressure test is not considered as practical as expansion pressures in swelling soils can be reduced in a
the PVC. It requires a considerably longer time, 24 t o 48 hours number of ways. If a small amount of expansion is allowed, a
as opposed t o 2 hours. Also the 4 % inch height of the considerable reduction in expansion pressure and volume
DENSITE SECHE MAXIMUM A A S H O MODIFIE DENSITE SECHE MAXIMUM AASHO STANDARD
MODIFIED AASHO MAXIMUM DRY DENSITY, p c f STANDARD AASHO MAXIMUM DRY DENSITY
- - - - -
w
o
5
o 6 R) w UI
0
m 03
0
w i5 0
- I\)
0
TENEUR EN EAU O P T I M U M % DENSITE SECHE MAXIMUM AASHO M O D I F I ~
OPTIMUM MOISTURE CONTENT '10 MODIFIED AASHO MAXIMUM DRY DENSITY
N w P
0 0 0

INDICE DE PLASTlClTE
DE L' ECHANTILLON ENTIER
P I. OF WHOLE SAMPLE
 PLASTICITY INDEX - INDICE DE PLASTlClTE

50 60 70 80 90 100 11 0

LIQUID LIMIT
LlMlTE DE LlQUlDlTE

FIG 128 RELATIONSHIP BETWEEN FREE SWELL AND ATTERBERG LIMITS

GONFLEMENT LIBRE EN FONCTION DES LlMlTES D'ATTERBERG
 CAPACITE POTENTIELLE DE GONFLEMENT A DENSITE SECHE
AASHO MODlFlE ET TENEUR EN EAU OPTIMUM

PVC AT MODIFIED AASHO DRY DENSITY AND OPTIMUM

MOISTURE C O N T E N T

r
0
r 5-
s-'
E
-'
m
change result. Expansion is reduced considerably by surcharge
loading and by an amount out of proportion to the intensity
of the load. Expansion is also dependent on the initial density
and moisture content. These considerations indicate that
accepting lower densities and higher moisture contents may be
advisable as a means of reducing volume change and expansion
pressure.
No single test evaluates all these possibilities. It is felt that
the PVC at Modified AASHO density or Standard AASHO
density provides an approximation that can be useful for
design purposes. Densities and moisture contents can be varied
t o investigate these effects. The PVC device allows a small
amount of expansion to occur (as does the expansion-pressure
device) in deflecting the proving bar or proving ring. However,
the change in volume is probably considerably less than the
amount that would normally be tolerable in a highway
subgrade. Surcharge loading cannot be evaluated in the PVC
device. The volume change test in an oedometer is believed to
be the best means for evaluating this factor.
Holtz and Gibbs (1953) provide an estimate of probable
expansion based on the index tests, shrinkage limit, plasticity
index and the amount smaller than 1 micron. These estimates
are reproduced in Table 45.
For African black clays, the shrinkage limit was found to be
a difficult test to conduct properly and the results correlated
poorly with those of the swelling tests. The approximation
proposed by Williams (1957-58) and shown in Fig. 127 is
considered more applicable.
The free swell test is also recommended for soils suspected
of being expansive. Correlations with index tests are good and
the test is extremely simple to conduct. It provides no
quantitative measure of either expansion pressure or volume
change, but nevertheless is useful in identifying problem soils.

CONCLUSIONS
The major occurences of African black clay soils are in
northeastern Nigeria, Cameroon, Lake Chad Basin, Sudan,
Ethiopia, Kenya and other countries in eastern Africa. Those
soils causing the most severe problems have formed over basic
volcanic rocks which are of geologically relatively recent 20 30 40 50 60 70 80 90
origin. GROUP INDEX
The properties of the African black clays are similar to INDICE DE GROUPE
FIG. 130 RELATION BETWEEN FREE SWELL,P.V.C. AND
those reported in the literature on black cotton soils and also GROUP INDEX.
t o other expansive soils. Montmorillonite is the clay mineral RELATION ENTRE GONFLEMENT LIBRE,CAPACITE
POTENTIELLE DE GONFLEMENT ET INDICE DE GROUPE.
responsible for swelling, but swelling can occur only if the clay
is desiccated, or partially desiccated and allowed t o absorb
moisture.
African black clays all contain calcium as the principal by Williams (1957-58), Fig. 127 is considered valid for all
exchangeable cation. Clays with calcium cations display lower African black clays and is useful for estimating potential
volume change than clays with sodium cations, but they expansiveness. The free swell test is a good, simple test which
display greater expansion pressures (Donaldson, personal can be used to verify swelling tendencies. The PVC test is
communication). considered the best test for estimating the amount of
The expansion pressures vary considerably depending on expansion pressure which may be exerted by the compacted
the clay and on the test employed. The classification proposed clay. No single test can evaluate all the factors that affect
expansion pressure, namely, initial density and moisture Lambe, T. W., 1960. The character and identification of
content, the amount of volume change that is tolerable, and expansive soils (soil PVC Meter). Fed. Housing Adrnin.
the effect of surcharge loads. Report FHA-701.
Moussaoui, M. M., 1967. Le compartement au Maroc des
REFERENCES Chaussees sur les "tirs." Proc. Thirteenth Int. Road Cong.,
Dudal, R., ed., 1965. Dark clay soils of tropical and Tokyo: 11/1-20.
sub-tropical regions. F.A.O. Agric. Develop. Paper, No. 83: Seed, H. B., Mitchell, J. K., and Chan, C. K., 1962. Studies of
161p. swell and swell pressure characteristics of compacted clays.
Donaldson, G. W., 1970. Personal communication. High. Res. Board, Bull. 313: 12-39.
Furon, R., 1963. The geology of Africa. Oliver and Boyd, Skempton, A. W., 1953. The colloidal activity of clays. Third
Ltd., London: 377p. Int. Conf. on Soil Mech. and Found. Eng., vol. 1: 57-61.
Holtz, W. G., and Gibbs, H. J., 1956. Engineering properties of Van der Merwe, D. H., 1964. The prediction of heave from the
expansive clays. Trans. Am. Soc. Civil Eng., vol. 121 : plasticity index and percentage clay fraction of soils. Trans.
So. African Inst. Civil Eng., June (6): 103-107.
64 1-663.
Jennings, J. E. B., and Knight, K., 1957-58. The prediction of Williams, A. A. B., 1957-58. Discussion on: The prediction of
total heave from the double oedometer test. Sym. on total heave from the double oedometer test, by J. E. B.
Jennings and K. Knight. Sym. on Expansive clays. So.
Expansive Clays, So. African Inst. of Civil Eng.: 13-19.
African Inst. of Civil Eng.: 57-58.
Kassiff, G., Levneh, M., and Wiseman, G., 1969. Pavements on
expansive clays. Jerusalem Academic Press: 21 8p.
 CHAPTER 17

DESIGN CONSIDERA TIONS FOR ROADS 0 VER TR OPICAL

BLACK CLA Y S

INTRODUCTION RECENT DESIGN PRACTICES

The tropical black clays (or "black cotton" soils) have been Recent work on expansive soils, summarized in Chapter 1 5 ,
a persistent problem in hghway and airfield construction in has shown that swelling of expansive clays can be reduced
those parts of Africa where they are widespread. This is by: (1) accepting lower densities and hgher moisture con-
particularly true of those deposits formed residually over basic tents; (2) allowing some initial expansion t o occur; and (3) by
intrusive rocks such as gabbros or, more commonly, over basic surcharging. Obviously, preventing changes in moisture con-
extrusives, mainly basalts. Weathering of these rocks is rapid tent would also eliminate swelling.
because their constituents are among the minerals least stable Pavement design practices over expansive clays in the U.S.
under chemical attack. Without resistant mineral crystals to have been described by Sallberg and Smith (1965). There are
form sand-size material, the rocks tend to weather directly to four major approa~hes:
clay minerals. The predominant clay mineral formed, at least 1. Avoid expansive clays by either realignment or under-
initially, is montmorillonite. cutting and backfilling.
Three problems are associated with black clay soils in 2. Reduce expansive characteristics..
highway construction. First, the material is predominantly 3. Confine expansive clays under embankments.
clay-principally montmorillonite, which is potentially expan- 4. Minimize moisture changes in the expansive clay after
sive and may have a very low bearing capacity. Secondly, in paving.
the larger residual deposits there is a lack of a suitable natural The first, avoiding the clays, is not often possible in many
aggregate where these soils occur, although alluvial deposits of the locations in Africa where these soils occur. However, a
generally include layers of sandy materials within the profile. routine swell test, such as the PVC, would at least provide a
Finally, road construction alters the moisture pattern in the basis for selecting alignments through less expansive sections.
subgrade soils. Surface evaporation is prevented or reduced Reducing expansive characteristics has been attempted by
and the moisture content of the subgrade soil invariably rises. stabilization. In plastic clays, hydrated lime is usually more
In a montmorillonitic soil, this causes swelling. The swelling, in successful than cement (Clare, 1957; Mehra and Uppal, 1964;
turn, reduces the bearing capacity. A pavement structure that Croft, 1967). Lime improves the workability of the soil while
is adequately designed for the as-constructed subgrade density cement does not; moreover, montmorillonitic clays retard the
may fail as the density is reduced through expansion. hydration of cement. If the black clays have appreciable
There are no inexpensive means of utilizing these soils organic matter, about 5% or more, stabilization may not be
satisfactorily. At least, no easy way has been found for dealing possible. The limited stabilization study on African black clays
with these or any other expansive soils that occur throughout is reported further below.
the world. The obvious solution is to avoid them wherever The third point, confining clays under embankment
possible, or if shallow, to excavate and waste the material. This material, assumes that suitable embankment material is readily
is rarely possible in large deposits such as the black clays over available. On the Ethiopian Plateau, for example, suitable
basalts on the Ethiopian Plateau. material is scarce and the available embankment material may
A very great thickness of base and surface is usually itself be expansive.
required, as dictated by low CBR values even in the absence of Minimizing moisture changes is undoubtedly the most
swelling. Economizing on the thickness of the pavement important aspect. A recent investigation by the Road Research
structure invariably results in greater expenditures later. This Laboratory on pavement design over tropical soils (O'Reilly et
has been the experience all too often in many African al., 1968) showed the importance of eliminating any design
countries. Inadequate designs have resulted in extensive features which allow the accumulation of moisture in the
failures; large sections have had to be rebuilt within one or two pavement structure. Adequate provisions must be made for
years. surface drainage; cracks must be filled and the surface
The various means of dealing with expansive black clays are maintained.
discussed in this chapter. Recent design practices are summa- Ponding water over or pre-wetting the subgrade soil has
rized from the literature and from the interviews conducted been successful in Texas and South Africa. McDowe11 (1965)
during the course of this study. The applications of these found that ponding water for 3 0 days over an expansive Texas
practices are discussed for the tropical black clays of Africa. soil was sufficient t o reduce the heave t o under one inch. The
water did not penetrate deeper than 4 feet in 24 days but the structure, seasonal drying or partial desiccation will occur
moisture content at a depth of 20 feet began to increase after along the shoulders. Kassiff et al. (1969) recommend extend-
a few days and the zone of increase progressed upward to 4 ing the shoulders to a width equal to the depth of the active
feet within 24 days. Williams and Simms (1963) believe that zone and locating the surface drainage as far away from the
pre-wetting expansive clays in South Africa may provide a pavement as practicable.
solution t o prevent localized heaving around culverts. Blight The compactive effort applied t o expansive soils and the
and DeWit (1965) report that flooding in conjunction with a corresponding optimum moisture content should correspond
grid of vertical wells will induce most of the heaving in a few t o Standard AASHO (or British Standard) rather than to more
months. Pre-wetting has also been employed by the Bureau of severe standards. O'Reilly et al. (1965) state that "Field
Reclamation in California (Bara, 1969), where troublesome measurements indicated that either the optimum moisture
soils are identified by the relation between their water content content in the British Standard compaction test or the plastic
and liquid limit. limit would provide a useful guide t o the moisture conditions
Possible treatments for expansive subgrades have been under bituminous surfaced pavements when these conditions
described by Williams (1965). Clearing the vegetation and were not governed by the water-table or where the drainage
allowing the soil to "fallow" reduces water loss due to arrangements for the pavement layers were not defective."
evapotranspiration and allows moisture to accumulate from They recommend compaction to at least 90% B.S. maximum
rainfall. The moisture content of the subgrade is raised dry density and preferably 95% to 100%. Kassiff et al. (1965)
significantly and a large part of the swelling occurs prior to state that swelling can be minimized by compacting clay above
covering with structural courses of the highway. Pre-wetting the plastic limit to 90% Modified AASHO density.
for some months, with a grid of auger holes, is advisable for The final moisture content beneath pavements has been
culvert sites. Williams believes lime-stabilization is beneficial discussed by Kassiff et al. (1969). Soil conditions are divided
for a number of reasons: (1) it reduces potential swelling in into three categories:
the top layer of the subgrade which would contribute most 1. groundwater level close to ground level.
toward total heave; (2) more water has to be added than for 2. deep groundwater level and a yearly rainfall of over 250
the untreated soil, which helps to raise the moisture content mm.
nearer to final equilibrium moisture content; (3) the treated 3. arid climatic conditions and a yearly rainfall of less than
layer is more impermeable, so moisture variations are reduced; 250 mm.
and (4) the workability of the soil is improved, particularly In the first category are pavements over clay subgrades with a
during the wet seasons. Williams also notes the benefits of groundwater depth of 5 to 6 meters. In this case the water
surcharge. He estimates that heave can be reduced by 10% for table controls the moisture conditions in the subgrade. For
every foot of fill. clays in the second category, field measurements have shown
Fallowing to increase the subgrade moisture content and that the final moisture contents are close to the plastic limit,
induce heave is a simple and inexpensive procedure to or to the optimum moisture content at Standard AASHO
eliminate a major part of the problem of expansive soils. It is compaction, which is usually close to the plastic limit.
particularly adaptible to highways and stage construction. It has been stated that the CBR values for black clay soils
Placing a layer of filter material or a subbase course is also are always low; the results of tests in this study after the
beneficial. DeBruijn (1965) found that a 6-inch sand cover four-day soaking period (Chapter 16) are all under 2. It has
over black clay induced almost all the heave in one season by been found in South Africa that the in-situ CBR values
permitting the moisture content to increase. The sand pre- improve with time. Measurements on a number of black clay
vented evapotranspiration and evaporation and conserved the subgrade soils in roads up to 15 years old were always 6 or
moisture from precipitation. Sand was found to be more over (Williams, personal communication). An explanation of
effective than a plastic cover. Kassiff et al. (1965) also found a this gain in strength with time has been provided by Eades
sand filter to be effective in reducing drying at the edges of the (personal communication). The initial water content intro-
pavement by retaining rainwater and reducing evaporation duced during compaction partially fills voids and is adsorbed
through capillary rise. Jennings (personal communication) on mineral and soil particle surfaces. With time, some of the
believes 5 or 6 years are required for heave to reach adsorbed water becomes absorbed or structurally-held water
equilibrium under structures. However, he considers that within the clay mineral lattice. Also, adsorbed water, being a
allowing the soil t o fallow for one year, particularly with a dipole molecule, becomes electrically oriented with time and,
granular cover, can accelerate heave by conserving moisture, once oriented, is more firmly held on the clay mineral
and that it is quite likely that most of the heave will occur in 1 surfaces. In either event, the strength increases. The tendency
or 2 years. to gain strength appears to be another advantage to stage
Drainage is particularly important in highways over black construction, whereby subgrade CBR values can be expected
clays. If drainage is poor, seasonal variations in the subgrade to improve somewhat with time.
moisture are unavoidable and heaving results. However, if Design curves recently adopted by Israel for expansive soils
drainage ditches are too deep, or too close to the pavement are shown in Fig. 131. These curves are based on the Road
FIG.131 C B R DESIGN CURVES, REVISED BY ISRAEL P. W.D. FOR VARIOUS
TRAFFIC CATEGORIES AND A WHEEL LOAD OF 6 . 0 TONS ( a f t e r
Nedavia, 1 9 6 5 )
ABAQUES C B R REVISE PAR TRAVAUX PUBLICS ISRAELIENS POUR
DIVERSES CATEGORIES DE TRAFIC E T UNE CHARGE DE 6 TONNES
PAR ROUE (d1apr8s Nedavia, 1 9 6 5 )
Research Laboratory curves and Peltier's equation for a wheel
load of 6 tons. Kassiff et al. (1969) state that the curves lead
t o pavements 5 to 15 cm. -thicker than those of the R.R.L.;
this increase is required because of heavier wheel loads in
Israel.

STABILIZATION
A limited stabilization study was conducted on a few black
clay soils. Lime was the only additive investigated because
experiences reported in the literature (Chapter 15) indicate
that lime is far better for clay soils than is cement. The lime
content was determined by the quick pH test procedure
devised by Eades and Grim (1966). In all soils but one, 6%
lime was found to be the optimum content for stabilization.
One soil required 7%.
It can be seen from the results given in Table 46 that in all
cases the strengths improved considerably. The unconfined
compressive strength of one sample, NBC-1, decreased with
lime and with curing time, but the CBR value increased.
Decreases in unconfined compressive strength of treated clay
soils can be attributed to the loss of cohesion; this is not
uncommon. In these cases, the CBR test is a better repre-
sentation of the true gain in strength.
Expansiveness was reduced in all samples. CBR swell values
were 0 or near 0. Expansion pressures after stabilization were
10% or less than the comparable untreated values. Although
the expansion pressures were reported after only 14 days, all
samples had reached equilibrium.

RECOMMENDED DESIGN PROCEDURE

It is recommended that the following additions be made to
the normal soil survey procedures for highways over black clay
soils and t o the design considerations.
The depth of the troublesome clay layer should be
determined at sufficiently close intervals to provide a rea-
sonably accurate picture of its stratigraphic position. This may
be accomplished by normal hand or machine auger borings.
The expansiveness of the clay should be determined along the
length of alignment and with depth. The index properties can
be used for this purpose, using Williams' chart (Fig. 127) as an
approximation of the potential expansiveness. The expan-
siveness of a few selected samples can be checked by means of
the PVC test.
The natural moisture contents and water table depths
should be determined. These are important in all soil surveys
but particularly important in expansive soils. Investigations for
the water table should be carried to depths of at least 20 feet.
Low lying, poorly-drained areas with high water tables should
be avoided.
The potential total heave can be approximated in accord-
ance with the metl~od proposed by D. H. Van der Merwe
(1 964) which makes use of Williams' chart. The potential. unit
heave at ground surface is assumed to be 1 inch for each foot
of depth of very highly expansive soil; 112 inch for each foot
of highly expansive soil; 114 inch for each foot of medium TABLE 47
expansive soil; and 0 inch for low or non-expansive soil. The
decrease of the potential heave caused by the depth of cover is Value of factor F with depth D from the relation
given approximately by: D = 20 log F
D = k logF
where D = depth of soil in feet
F = factor indicating relative decrease in Depth in f t Mean value of F Depth in f t Mean value of F
heave at depth D
k = empirical coefficient, approximately 20
The values of F are given in Table 47.
The potential expansiveness of all the soil layers are then
added to determine the total potential heave. As an example,
if only depths 1 t o 4 feet are of concern and the soil is highly
expansive, the total heave could be estimated by:
1 - 2 ' , ~ ,= 0 . 8 4 2 x 1 / 2 = 0.421 in.
2-3', F3 = 0.750 x 112 = 0.375

Tlus approximation does not take into account the natural

moisture content or what the ultimate moisture content might
be. If the water table is below 20 feet and the natural moisture
content is near optimum, the approximation is too conser-
vative. Nevertheless, a total potential heave of over 1 inch
should be considered detrimental for highways and special
design procedures or relocation of the alignment should be
during construction. Paper presented at the Proc. Second
investigated .
Int. Res. and Eng. Conf. on Expansive Clay Soils. Texas
Stage construction and allowing the subgrade t o lie fallow
A&M University.
for one year should be adopted whenever possible. A granular
Blight, G. E., and De Wet, J. A., 1965. The acceleration of
cover should also be provided not only t o conserve moisture,
heave by flooding. In: Moisture equilibria and moisture
but also to act as a filter. Kassiff et al. (1 969) have found that
changes in soils beneath covered areas. Butterworths,
quarry material in the 318 inch t o No. 10 sieve size is the most
Australia: 89-92.
suitable filter size over cohesive soils.
Clare, K. E., 1957. Airfield construction on overseas soils. Part
Standard AASHO requirements for compaction and
2: Tropical black clays. Proc. Inst. Civil Eng., vol.
optimum moisture content, rather than Modified AASHO
8 : 223-231.
requirements, should be specified. The lower density and Croft, J. B., 1967. The influence of soil mineralogical
higher moisture content will reduce potential swelling. The composition on cement stabilization. Geotechnique, vol.
lower density does not appreciably effect bearing capacity.
17: 119-135.
The data listed in Chapter 1 6 indicates that CBR values at De Bruijn, C. M. A., 1965. Annual redistribution of soil
both compactive efforts are very low. Moisture contents above
moisture suction and soil moisture density beneath two
optimum do not lead t o lower CBR values. Therefore, there
different surface covers and the associated heaves at the
appears t o be no advantage in specifying greater compaction.
Ondersterpoort test site near Pretoria. In: Moisture equi-
The CBR curves provided in Chapter 1 3 can be applied to
libria and moisture changes in soils beneath covered areas.
black clay soils. The minimum CBR value in these curves is 4,
Butterworths, Australia: 122-1 34.
and this value can be used for soils having CBR's below this
Eades, J.L., and Grim, R.E., 1966. A quick test t o determine
minimum. From the pavement condition survey, Chapter 12,
lime requirements for lime stabilization. High. Res. Board
it was concluded that subgrades with CBR's less than 4
Record 139: 61 -72.
contribute no appreciable strength t o the pavement structure.
Kassiff, G., Livneh, M., and Wiseman, G., 1969. Pavements on
The new curves appear more conservative than the Israeli
expansive clays. Jerusalem Academic Press: 2 18 p.
curves, Fig. 131 ; however, they allow for variations in base
Kassiff, G., Komornik, A., Wiseman, G., and Zeitlen, J. G.,
course tluckness and CBR value. For the axle loads applicable 1965. Studies and criteria for structures on expansive clay.
in Israel, the new curves are not more conservative. In: Engineering aspects of moisture changes in soils. Texas
A&M Press.
REFERENCES McDowell, C., 1965. Remedial procedures in the reduction of
Bara, J. P., 1969. Controlling .the expansion of desiccated clays detrimented effects of swelling soils. In: Engineering
 aspects of moisture changes in soils. Texas A&M In: Engineering aspects of moisture changes in soils. Texas
Press: 239-254. A&M Press: 208-238.
Mehra, S. R., and Uppal, H. L., 1964. Laboratory and field Van der Merwe, D. H., 1964. The prediction of heave from the
experience in the construction of roads in black cotton soil plasticity index and percentage clay fraction of soils. Trans.
areas. Second Asian Reg. Conf. o n Soil Mech. and Found. So. African Inst. Civil Eng., vol. 6: 103-107.
Eng., vol. 1 : 294-300. Williams, A. A. B., 1965. The deformation of roads resulting
O'Reilly, M. P., Russam, K., and Williams, F. H. P., 1968. from moisture changes in expansive soils in South Africa.
Pavement design in the tropics. Ministry of Transport In: Moisture equilibria and moisture changes in soils
(U.K.), Road Res. Lab., Road Res. Tech. Paper No. beneath covered areas. But terworths, Australia: 143-155.
80: 124 p. Williams, A. A. B., and Simons, N. E., 1963.The heaving of
Sallberg, J. R., and Smith, P. C., 1965. Pavement design over roads on expansive clays with suggested remedies based on
expansive clays; current practice and research in the U.S. field moisture conditions. Proc. Third Reg. Conf. for Africa
on Soil Mech. and Found. Eng., vol. 1 : 47-5 1 .
 CHAPTER 18

SUMMAR Y AND CONCL USIONS

The terms "laterite" and "lateritic soil" have been used by found, due largely to the diversity of the materials reported
various people t o describe a wide variety of materials. While it and to variations in the test procedures employed.
is realized that the geological definition is perhaps the most The physical properties of many red, tropically weathered
appropriate, such a definition is virtually impossible t o apply African soils were investigated. The pedological terms com-
in the field and, therefore, has limited engineering applica- monly in use in Africa were found t o be appropriate and a
bility. The following definition for laterite appears in the discriminant analysis revealed significant differences among
Glossary of the American Geological Institute: "A red, three major groups. Ferruginous soils occur in the more arid
residually developed tropical soil formed in humid, well- extremes for lateritic soils, in areas of pronounced dry seasons
drained regions. It is leached of silica and bases and contains and less than 72 inches of rain per year. Ferrallitic soils occur
concentrations of iron and aluminum hydroxide." The degree in the more humid extremes for lateritic soils, in areas with
of leaching and relative accumulation of iron and alumina is over 60 inches of rain per year and supporting dense
usually determined by chemical and minerological analyses. It vegetation. Ferrisols occur in intermediate t o h g h rainfall
is for this reason that these criteria are not suitable for areas where erosion has kept pace with profile development.
engineering purposes. Facilities for such analyses are rarely Ferruginous soils are distinctly different from both ferrisols
available t o the engineer and few African laboratories have the and ferrallitic soils. They show lower clay contents and
necessary equipment and personnel to perform such analyses Atterberg limits, and higher compacted densities and CBR
routinely. values. Their natural moisture contents are lower, reflecting
the more arid climate under w h c h they were formed.
1. All tropically weathered red residual soils were included Ferrallitic soils and ferrisols display a small but distinguishable
in this study. Some of these were true lateritic soils difference in properties. Average values of the Atterberg limits,
under the chemical criterion. Others are presently
percentage of fines and Modified AASHO dry density and
undergoing or have undergone laterization, but leaching
CBR for each group are shown in Table 7.
of the silica and bases is incomplete. There is no
practical means of distinguishing between the two.
Further, the engineering properties of both groups are 2. The more descriptive terms ferrugirious soil, ferrallitic
quite similar. Most lateritic soils contain an appreciable soil, and ferrisol are recommended in lieu of lateritic
clay fraction which generally displays properties that are soil. The geographical distribution of each of these soil
similar t o those of other tropical soils which are only groups is shown on the Soils Map, Fig. 13.
partially laterized.
3. Variations in properties among the three groups should
The geological setting of Africa can be described as a large be established ill each country. Possible variations within
crystalline Precambrian sheld with a coastal fringe of sedi- each group (among the subgroups) sllould also be
ments. The continent has been unusually stable and most of investigated.
the tectonic movements have involved uplifting of large
Correlations were established among various properties of
sections of the continent t o form a number of interconnected
the red tropical soils. The most significant of these are the
plateaus with intervening basins. The basins have received
relationships between: (1) liquid limit and plasticity index: (2)
thick deposits of lacustrine and continental sediments. The
optimum moisture and Modified AASHO maximum density:
largest and most important structural feature is the Rift Valley
(3) CBR and AASHO classification; and (4) for four groups of
system in eastern Africa. Volcanic activity has been associated
lateritic gravels from Ghana CBR and maximum dry density
with the Rift Valley and large areas of basalt occur in
divided by PI. General, but useful, relationships were estab-
Ethiopia, Kenya and other areas in eastern Africa adjacent t o
lished between (1) CBR and PI multiplied by the percentage
the Rift Valley. Volcanic activity in western Africa, principally
passing the No. 200 sieve; and (2) CBR, maximum density and
in Cameroon, was more or less contemporaneous with that in
fineness index.
eastern Africa. Since the volcailics are the most recent rocks,
they have led to soils which differ from those generally found, 4. The following useful relationships were established:
most of which have formed over Precambrian rocks. Other a) PI = 0.71 LL - 8.50 for ferruginous soils
exceptions are the large alluvial and lacustrine sediments of the PI = 0.57 LL - 3.62 for ferrallitic soils
basins. PI = 0.50 LL - 1.50 for ferrisols
A review was made of the literature on lateritic soils which and
was confined, with few exceptions, t o the engineering proper- PI = 0.58 LL - 4.54 for all tropical red soils.
ties of African occurrences. A wide variation in properties was b) MDD= 1 6 0 - 2.78 OMC
 C) CBR = 129 - 7.5 (AASHO) both water absorption and specific gravity.
where: (AASHO) = The numerical value of the AASHO Two types of swelling tests were evaluated. Potential

-
classification system, i.e., 1 'A-1-a;
2 = A-1-b; 12 A-7-6
d) Fig. 20 may be used to estimate the range of CBR
volume change (PVC) and the Portuguese swell tests both
showed that excessive swell is not a common phenomenon
with lateritic soils. The PVC test has limited value in measuring
values from the PI and percentage of fines. expansion accurately, but can be used as a quick and easy
e) The suitability of material for use as base course can procedure to determine if a soil is critically expansive. The
be estimated from Fig. 23, which shows the range in swell index, determined in the PVC test, cannot be reasonably
CBR values from the maximum density and fineness predicted by the ordinary index properties of the soil. Soils
index. with swell not exceeding 10% in the Portuguese device have
Lateritic soils may change properties as they are air dried. been found to perform satisfactorily in the Portuguese
There are two modes: (1) aggregation of the individual soil colonies. In such cases, the specifications on the Atterberg
grains as the hydrated iron and alumina minerals dry or limits and the amount of fines can be relaxed. These findings
oxidize; and (2) loss of structurally-held water in certain clay could not be critically evaluated because .the percentage swell
in most of the soils tested was less than 10%.
minerals or clay materials. The first mode takes place if the
soils have never been naturally desiccated. Ferruginous soils, 6. The one-point liquid limit test can be used for tropical
which commonly occur in the arid extremes for lateritic soils, soils.
display changes only if they are sufficiently deep in the profile The equation:
to have remained permanently moist. Ferrisols similarly can
N)
LL=w(% tan B
display changes of properties, but the ferrallitic soils are the
most susceptible. The second mode, loss of structural or
Can be applied, where: W = moisture content
attached water, causes changes in properties in soils containing
halloysite or gibbsite. Halloysite, by far the most significant, is N = The number of blows
found only over basic rocks such as those which form the tan B = The slope of the flow curve
volcanic extrusives of eastern Africa and Cameroon, and the of blows versus the moisture
small basic intrusive bodies found in some of the countries in content on a semi-log plot.
Western Africa. Only the hydrated form of halloysite is 7. Each point on the compaction curve should be deter-
susceptible. Metahalloysite has lost part of its structural water mined with new material. This prevents excessive break-
and does not change properties upon drying. Gibbsite appar- down during compaction, w h c h occurs when the same
ently experiences minor changes with drying; some of these material is reused. The soil should be "cured" overnight
are significant, or over 10% of the as-received values. at its compaction moisture content.
8. A minimum of three points should be determined to
5. Special pretreatments and test procedures are required
establish the CBR accurately at the optimum moisture
when ferrallitic soils or soils over volcanic rocks are
content or any other particular moisture content.
tested. These procedures can be ignored only if it is
9. To convert the Harvard miniature density t o Modified
certain that the soil will be dried during construction. If
AASHO density. The equation:
drying is not feasible, the laboratory testing should be
MDD Modified = 16.0 + .926 MDD HM
performed on the samples in the as-received condition.
may be used. To convert Harvard optimum moisture to
Otherwise, the laboratory results will apply only to the
Modified AASHO OMC, the following equation may be
altered material and not t o the material that obtains in
used:
the field.
OMC Modified = 0.1 23 + 0.802 OMC HM
Various test procedures and pretreatments were investi-
An experimental study on the applicability of natural color
gated. The one-point liquid limit and Harvard miniature
and color infrared air photos was conducted. This was
compaction procedure can be applied t o tropical soils.
necessarily abbreviated and hampered by persistent cloud
The CBR value is highly dependent on the molding
cover. Nevertheless, a few preliminary conclusions are possible.
moisture content. Samples molded wet of optimum display
Both natural and infrared colors have advantages over black-
considerably lower values than those compacted at optimum
and-white in revealing subtle differences in the condition of
moisture content.
the vegetation and such differences are often soil-related.
The California durability is not applicable to African
Infrared color is preferable to natural color in the identifica-
lateritic soils and gravels. It does not correlate with either the
tion of vegetation w h c h is diseased or starved for lack of
CBR test or the aggregate impact value. The aggregate impact
nutrients or water. However, natural color photos are more
value (British Standard No. 812) indicates that most lateritic
useful for observing tonal changes in the surface soil.
gravels are acceptable for use as base course materials.
Relationships were found between aggregate impact value and 10. A more complete study of the applicability of color

206
 and color infrared air photos in identifying tropical 13. Lime is generally a better stabilizing additive for
soils is recommended. The possibilities of distinguish- improving lateritic gravel whose clay content and
ing among the soil groups and locating indurated plasticity index are greater than 15. Cement in combi-
horizons and gravel deposits should be investigated. nation with lime is required when lime alone will not
sufficiently increase the strength.
The suitability of current engineering classification systems
14. Asphalt can be a practical additive for stabilizing
was analyzed for tropically-weathered soils. The AASHO
sandy tropical soils.
system distinguishes a greater number of soil groups among
those tropical soils tested than does the Unified system. In the A design has been prepared for proposed test sections to
Unified system the majority of the soils are classified into only evaluate the effectiveness of cement stabilized base courses. In
two groups. There also appears t o be a better relationship one section, a uniform thickness of stabilized base with
between CBR and the AASHO system than with the Unified. cement contents varying from 0% t o 8% is suggested. Three
The Group Index in the AASHO classification system was thicknesses of base course should also be evaluated in a second
further analyzed through multiple regression techniques to section, all at one cement content of 4%. These sections
determine if a better group index equation could be de- should provide data on the necessary thicknesses for cement
veloped, or if different limiting values for the applicable stabilized materials to provide satisfactory pavements under
parameters could be applied. No improvements in correlation various traffic loads and soil conditions, and on the optimum
were obtained. percentage of cement.
11. The AASHO classification system can be applied as A survey of the pavement design procedures and material
well with tropical soils as it can with temperate-zone specifications presently in use in the various African tropical
soils. countries was conducted. The survey was accomplished by
personal interviews and by questionnaires. The Road Research
Stabilization with both lime and cement was investigated in Laboratory procedure (RRL Road Note 31) is the common
the laboratory for selected soils and lateritic gravels from procedure in the English-speaking countries, while the one
Ghana and other African countries. Three soils, representative advocated by the Centre Experimental de Recherches et
of the typical range of properties, were stabilized from each d'Etudes du Batiment et des Travaux Publics is the common
major soil group, i.e., ferruginous soils, ferrallitic soils and procedure in the French-speaking countries. Both methods
ferrisols. Additional ferrallitic soils were tested from the make use of the CBR test as a measure of strength. Lateritic
Democratic Republic of the Congo and Ivory Coast. Ferrallitic gravels have been widely used in tropical Africa and have given
soils from Sierra Leone were tested at Fourah Bay College in reasonably good service both as base course and as surface
Freetown. course materials. Material specifications, however, vary widely.
Soils stabilized with cement showed greater gains in A pavement condition survey was performed in Ghana and
strength than did the lime-stabilized soils. The generally Ivory Coast to evaluate lateritic and tropical materials as
accepted minimum CBR value and unconfined compressive structural components of existing roads. A relationship was
strength were attained with 3% to 6% cement. Ferruginous established between the maximum rebound deflection and the
gravels displayed larger gains in strength than either ferrallitic slope of the deflection basin, both measured by means of a
gravels or gravelly ferrisols and they also required less cement. Benkelman beam and an 18,000 pound axle load. Both
Lime stabilized gravels showed maximum strength gains deflection parameters provided reasonable correlations with
with 4% to 6% lime. Ferruginous soils and sandy ferrallitic pavement performance, evaluated subjectively and adjusted for
soils displayed large reductions in plasticity index with small
loading history. Allowable deflection parameters were pre-
percentages of lime. Lime used alone appears t o have limited
sented for various ranges of traffic loading. Test sections with
use as a stabilizing additive for lateritic gravels for base course
representative (average) deflections were examined in detail by
materials. However, lime has application in reducing the PI of
means of test pits and field and laboratory testing. Factorial
soils which would otherwise not be effectively stabilized with
analyses and regression analyses were performed t o evaluate
cement. When the PI is greater than 15 and the percentage of
the adequacy of the various soil types and their characteristics.
fines is greater than 15, mixing cement with the soil is difficult
CBR, density, index tests, swell, durability and aggregate
and lime should be used. If lime alone does not produce the
strength were all analyzed. Deflection was found t o be highly
required strength, then lime and cement should both be
dependent on CBR, thickness of the structural component and
applied.
degree of compaction. Base courses with CBR values as low as
Sand stabilization can be used to improve strength and
50 have performed satisfactorily if the degree of compaction is
decrease plasticity. Strength gains are relatively small, how-
90% or greater than Modified AASHO density. It was found
ever, and large percentages of sand are required.
that subbase courses with CBR values below 20 do not
12. Cement stabilization is a practical means of improving contribute significant strength to the pavement structure with
lateritic gravel unless the clay content and plasticity regard to reducing deflection. A relationship was also obtained
index are greater than 15. between the structural number of a pavement section and the
allowable deflection for a given traffic loading. This relation- should be verified in the individual countries, as
ship was utilized to develop rational design criteria applicable additional data and local conditions may warrant
to tropical soils. Moisture contents which existed in the field modifications to the design charts. Nevertheless, it is
were somewhat lower than one would expect. If adequate felt that these design procedures should be adopted
drainage can be provided, the use of the as-molded CBR as now and doing so would result in more satisfactory
opposed to the soaked CBR should be considered for design and more economical facilities.
purposes. As might be expected, ferruginous soils generally The specifications tabulated in Chapter 13 are also
have lower moisture contents. recommended. The values listed for CBR, degree of
compaction, Atterberg limits and gradation are satis-
15. There is a strong relationship between maximum
factory values for insuring adequate pavement per-
rebound deflection and the slope of the deflection
formance.
basin for both wheel paths.
22. Figs. 1 12 and 1 13 show that modifications in the
16. Both deflection parameters can be used to predict
Atterberg limits or gradation are possible provided
pavement performance or serviceable life (Figs. 97
that the minimum CBR value for the base course class
and 98).
(design charts, Figs. 1 10 and 1 1 1) can be attained. If
17. CBR and thickiness of the structural component, and
the percentage of fines is low, then the LL (and PI)
the degree of compaction are the most important
can be increased without causing a reduction in CBR.
factors in reducing deflection. Compaction below 90%
Similarly, if the LL (and PI) values are low, then a
Modified AASHO density results in excessive deflec-
larger percentage of fines can be allowed.
tion (within the structural course). That deflection is
23. Modified AASHO compaction should be specified.
highly dependent upon CBR and thickness of the
However, the possibility that excessive fines may be
structural component is illustrated in Fig. 107.
produced should be investigated. Further research is
18. Subbase courses must have CBR values of 20 or
recommended.
greater in order to be effective in reducing deflections
(Fig. 105). The application of tropical soils to construction other than
19- As-molded, rather than soaked, CBR values can be highways was reviewed briefly. These soils are utilized in
applied particularly in the more arid regions (ferrugi- building construction either as blocks or as formed walls. Most
nous soils) if adequate drainage can be provided. building blocks today are stabilized with cement. A wet-dry
Design charts were developed for base course thicknesses of durability test is usually applied to evaluate stabilized blocks
six and eight inches. A CBR value of 3 0 was adopted for the and a maximum allowable loss of 10% is generally selected. A
subbase, although other values greater than 20 can be readily cement content of 6% to 8% is usually adequate.
applied to devise other design charts by means of the Cut slopes in tropical residual soils can rarely be analyzed
coefficieizts developed in Chapter 12. Different curves are by the usual theoretical methods because most soils display
given for various values of base course CBR which, with light relict structure from the parent rock in the profile. Such
traffic, can be as low as 50. Three charts are provided for each features as joints and fractures often control stability. Existing
base thickness, covering three ranges of traffic loading. slopes can be evaluated in order to predict slope performance.
Material specifications that are appropriate to the traffic A geological classification of safe cut slopes can be developed
were determined from the resu.lts of the pavement condition and can be broadened by including such characteristics as
survey. The specifications apply to subbase, base and surface landform and drainage as well as parent material.
materials. Further modifications to the At terberg limits and Earth dams of red tropically weathered soils have per-
grading requirements are also provided which are pernzissible formed satisfactorily even with higher liquid limits and lower
as long as the CBR requirements are fulfilled. compacted densities than normally allowed for temperate
Degradation due to compaction was also studied on a soils. Such soils may display strengths and compressibilities
limited number of samples. It was found that the breakdown similar to temperate soils with lower limits and higher
was not appreciably greater under the Modified AASHO densities.
compaction than under the Ghana Standard compaction. The The literature on tropical black clay and black cotton soils
loss in gravel-size material was reflected by an increase in the was reviewed. Most of the black clays have formed residually
sand-size, than fines. This may actually improve the Over basic rocks, particularly basalts. These soils generally
contain large clay contents. Montmorillonite is the predomi-
gradation since many lateritic gravels are deficient in sand-size
nant clay mineral in all cases where mineralogical analyses
material.
were performed. When the black clays are desiccated, or
partially desiccated, and exposed t o water, swelling occurs.
20. It is recommended that the design charts, Figs. 110 Expansion pressures can be considerable.
and 11 1 be adopted for roadway design in tropical The black clay soils in Africa occur over basic intrusive
Africa. The findings of the pavement condition survey rocks in Ethiopia, Kenya, Uganda, and in other areas in eastern
Africa over volcanic rocks associated with the Rift Valley. In conserved. Increasing the moisture content t o the plastic limit
western Africa, the black clays have also formed over basic or the Standard AASHO optimum moisture content (approxi-
extrusives in Cameroon and Nigeria. Black clays over intrusive mately the same) should be the objective as this has been
basic rocks and calcareous rocks occur in southern Africa and found t o be very near the equilibrium moisture content under
in smaller areas in western Africa. Large deposits of alluvial pavements. Flooding or prewetting has also been attempted
and lacustrine black clays have accumulated in the Chad Basin and various lengths of time have been utilized. However, the
and in southern Sudan. Smaller areas of alluvial deposits also soils have very low permeabilities, so the rate of moisture
occur, usually i11 poorly drained areas. increase is difficult to forecast. Ponding water at culvert sites
The properties of the African black clays are similar to may be effective in inducing heave if auger holes are drilled t o
those reported in the literature on black cotton soils and also help introduce the water.
to other expansive soils. The expansion pressures vary con- Black clay soils should be compacted at Standard rather
siderably, depending on the clay and on the test employed. A than Modified AASHO density. The lower density and higher
general classification of potential expansiveness devised in moisture content reduce potential swelling but do not appreci-
South Africa has been found t o apply t o other African black ably reduce the bearing capacity.
clays and is useful in estimating potential expansiveness. The Lime stabilization was investigated in the laboratory for a
free swell test is a simple test which can be used t o verify limited number of black clays. A lime content of 6% was
swelling tendencies and the PVC test is considered the best test required in all the soils except one, in which 7% was required.
for estimating the amount of expansion-pressure which may be Expansion pressure in all cases was reduced t o tolerable limits
exerted by the compacted clay. No single test can evaluate all and bearing capacities increased considerably.
the factors that affect swelling, which include initial density The CBR charts that were devised from the pavement
and moisture content, allowable volume change and surcharge condition survey on red tropically weathered soils can also be
load. used for the black clay soils. The charts show a minimum CBR
of 4 which is usually higher than the CBR values for black
24. Potential expansiveness of a black clay soil can be clays. However, a CBR of 4 can also be used for all lower
estimated from its PI and clay content by utilizing
values; this only means that CBR values under 4 contribute no
Fig. 127.
strength t o the pavement structure.
25. The free swell test and PVC test are recommended as
routine tests when investigating tropical black clays or 26. Moisture changes in black clay subgrades can be
other expansive tropical soils. minimized by:
Current practice in designing roads over black clays involves a ) Stripping the vegetation and allowing the soil t o
one or more of the following: fallow for a year. Adding a granular cover is
particularly effective in increasing the moisture
a.) Avoiding expansive soil areas by realignment or by content of the subgrade.
excavation.
b ) Compacting the subgrade at Standard AASHO
b.) Reducing expansive characteristics by stabilization,
density and optimum moisture content.
usually with lime.
c ) Providing adequate drainage facilities, and by
c.) Confining the clays under embankments. locating drainage ditches well away from the
d.) Minimizing moisture changes after construction. shoulders.
Stripping vegetation, or allowing the subgrade soil to fallow 27. In all the black clays tested, increasing the compactive
for a year, increases the moisture content by reducing effort t o Modified AASHO density did not increase
evapotranspiration. Covering the exposed subgrade with a the bearing capacity. This should be verified for
layer of granular material is even more effective in that individual projects.
evaporation is reduced because of the barrier t o capillary 28. Lime stabilization is a practical means of increasing
movement and also because water from precipitation is bearing capacity and reducing expansion pressure.
 APPENDIX FOR CHAPTER 5

LABORATORY TEST RESULTS

SOILS DA TA

LAB SOILAA 0 A/M SK

NO. GRP SHOGI L L P L P I 1 3/4 1/2 j/8 4 10 4U 200 LM GM MC MUD C6H Ct3K SG PH

GO920 J B l 7 41 20 21 97 95 90 83 54 23 13 8 4
GO461 J C 1 4 22 11 5 100 100 98 96 93 90 73 24 5
GO150 J C 1 7 0 44 15 29 100 100 100 93 70 28 10 5
GO520 JC 1 3L 20 12 100 100 98 98 83 49
GO910 J C 1 7 4 57 16 40 100 100 99 93 74 49 42 35 16
GO940 J C 1 7 2 53 27 26 100 100 99 97 84 62 46 30 17
GO950 JC1 6 1 31 12 19 100 100 99 99 88 68 46 27 12
GO112 J C 1 6 0 4 0 1 8 22 100 100 99 94 72 42 20 5
G0114JC1 4 20 11 9 88 84 76 72 59 34 13 5
G116A J C 1 10 9 35 18 17 100 100 100 100 93 77 74 67 L3 15 113
G116B J C 1 10 2 29 16 13 100 100 100 100 90 67 63 45 18 12 126
G116C J C 1 10 4 27 14 13 100 10U 100 100 96 74 67 40 16 13 122
G11IA JC1 7 4 49 L l 28 100 1 0 ~9 8 93 65 47 30 34 LL 11 132
G11/B JC1 12 15 6 3 25 40 100 100 100 100 88 57 5 50 L8 13 115
GO511 BCJU 4 0 28 12 16 95 92 92 90 11 43 14 b 12 l i d
GO530 BCJD 36 17 19 100 100 lU0 80 64 30 13 134
GO550 BCJD 40 16 24 100 96 89 82 54 21 13 132
GO610 BCJD 6 0 34 17 17 95 90 84 77 52 33 22 15 6
G 0 6 2 0 BCJD 4 0 23 16 7 100 100 99 94 61 28 19 15 4
GO930 BCJD 7 2 42 23 19 103 100 99 97 72 44 34 26 11
G096A BCJD 6 1 34 14 20 96 94 89 80 50 37 27 18 8
GO966 BCJD 6 1 30 15 15 100 100 100 99 92 69 45 26 12

G l O l O KB2 12 15 45 26 1 9 100 100 100 1 0 0 100 100 98 76 59

GS100 KC2 12 b 4 5 19 25 96 93 94 92 84 64 6U 47 35
GOZlP KC2 7 0 48 2 L L b 3 63 5 51 44 24 1L 3
G023P KC2 12 18 58 2 8 30 100 103 190 100 100 95 75 65
GO315 KC2 7 2 57 22 3 9 93 8 82 77 54 34 24 23 15
G 0 3 2 L KC2 7 3 52 2 1 3 1 93 92 91 90 55 64 4L 32 29
G033L KC2 7 0 56 2 1 35 100 100 81 74 43 16 8 6 4
GO630 KC2 6 1 38 17 2 1 100 99 96 94 82 40 3u 20 15
6 0 7 0 0 KC2 11 16 58 3 1 27 100 100 100 96 93 80 75 53 40
GO810 KC2 7 0 4 8 2 4 24 99 9b 95 87 60 3.2 25 23 16
GlOZOKC2 7 3 4 / 2 1 26 (35 90 dO 70 40 21 15 12 7
G 1 0 3 0 KC2 9 3 46 36 1 0 100 100 1.00 1 0 0 1 0 0 1 0 0 67 50 27
GSlUO KC2 12 08 4 3 1 9 25 0 9 6 095 94 92 84 64 6u 47 3j
LAB SOILAA 0 A/M SK
NO. GRP SHOGI L L P L P I 1 3/4 1/2 3/8 4 10 4U 2 0 0 2M GM MC M d D CBH CBH SG PH

GlllB LX3
G112A LX3
G112B LX3
G113A LX3
G113B LX3
G113C LX3
(51130 LX3
G114A LX3
G114B LX3
G114C LX3

TP30A JC 06 00 31 19 12
TP30B JC 08 04 20 16 04
TP30C JC 08 05 22 18 04
TP30D JC 10 01 29 17 12
TPO5A BCJD 06 00 26 15 11
TPO5B BCJD 8 1 25 15 10
TP06A BCJD 4 0 26 16 10
TP06B BCJO 4 0 20 20 0
TP2UA BCJO 6 2 31 16 15
TP21A BCJD 10 3 3/ 20 17
TP21B BCJD 10 1 37 22 15
TP21C BCJD 12 01 51 25 26
TP21D BCJO 06 01 38 22 16
TP22A BCJD 2 0 23 18 4
TP22D BCJD 6 0 30 18 12
TP23A BCJD 4 0 25 17 8
T P ~ ~BCJD
B 4 0 20 15 5
TP24A BCJD 06 01 30 18 12
TP25A BCJD 04 00 21 18 04
TP26B BCJD 06 00 25 13 12
TP27A BCJD 06 01 33 18 15
IP278 BCJD 08 02 17 13 04
TP28A BCJD 02 00 24 19 05
TP28B BCJD 04 00 24 21 03
TP29A BCJD 02 00 23 17 06
TP29B BCJD 2 0 20 16 4
TP31A BCJD 10 03 32 19 13
TP31B BCJD 08 01 19 17 02

TPOlA KC2 12 16 54 24 30
TP02A KC2 7 0 50 22 28
TP02B KC 12 32 70 2 6 44
TP03A KC2 6 0 32 15 17
TP03B KC 11 0 3 5 3 3 4 19
TPO4A KC2 6 33 1 5 18
LAB S O I L AA 0 A/M 5K
NO. GKP SHOGI L L P L P I 1 3/4 1/2 3/8 4 10 40 2 0 0 2M GM M C MOD C d H C 6 R SG PH

TP04B KC
TP07A KC2
TP07C KC
TPOBA KC2
TPOt3B KC
TP09A KC2
TP09B KC
TP09C KC
TPlOB KC2
TPlOC KC
TPllC KC2
TP15B KC2
TPl8A KC2
TPl9C KC2
TP36A KC2
TP36C KC
TP36D KC
TP38A KC2
TP38B KC
TP40A KC2
TP40B KC
TP41A KC2
TP42A KC2
TP42B KC

TP32A LC
TP32B LC
TP33B LC3
TP34C LC3
TP34D LC
TIVlA LM
TIVlB LM
TIV2A LM
TIV2B LM
TIV3A LM
TIV3B LM
TIV3C LM
TIV3D LM
TIV4A LM
TIV4B LM
TIV4C LM
TIV6B LM
TIV6C LM
TP14B LX3
TP16B LX3
TP17A LX3
TP17B LX
TP35B LX3
 mm7r7r
nnnn
7r7rNN
nn
I- P P P I - P P I-PI-++' I- I-+PI-I- I-
N C ~ ' ~ ' N ~ N ~ ~ ~ O + C ~ ~ - C ~ O C ~ , I - P ~ N 0 '~N U ' C P I - O O ~ ~ '

N P N I-I- W I- I-
r h \ r N + m 4 G N C r : r C a r N O ~ ; ( ~ O U C C L . O \ I - ~ U ~ W ~ W I O- C
C o c o a
N N C C
cnw0'0:

PPPI-I-+I-PPI-I- I-I-I- I-PP I-I-+I-PI-I-PI-I-I-I-I-I-I- I-I-PI-

0 0 0 0 0 0 0 0 C C C 9 0 0 0 ~ ~ 0 0 0 ~ G C 0 C O O O O O O C C O 0 O 0
0 0 0
O C 0 0 3 0 0 0 C C C u 0 0 0 0 ' c r 0 0 C ~ C O O O C O O O O C O O O O C 0 C 0 0

I - I - I - I - P W I - P P + t - I- I- I-I-I- I-I-I-I-I-I-I-I-I-P PI-I-

0 0 0 C O O O C a a C m 0 a 0 a ~ 0 0 a c C c C O 4 a O O O O O O O O O 0 o o o a
C G O O O O ~ C U O C W O ~ O ~ N C ~C C C W
C ~
~ ~ O O O O C C ~ C ~ O cGc o
C c
I- I-I-PI-PP P P P I- P I- I-I-PP I-
~ a o ~ ~ ~ 0 ~ 0 : a ~ m a ~ o x ~ o ~ a a ~ e ~ ~ a 0 0 0 ~ a a a
C ( ~ C O O O O C C ~ C C ~ ' U C I - C C C ~ ' ~ ' C C . ~ W C O C O O ~ ~ ~ W P C

I- PPI- I- P P I-I-I-
c a c o c a a w u l m ~ ~ a ~ ~ ~ u a ~ o ' u a o ' ~ c h : ~ a ~ ~ ~ ~ o ' o ' u u a
C N G C O ~
~ ~ ~ o a ~ ~ ~ ~ u ~ ~ a c ~ o ~ a ~ - ~ ~

C. hi h h:W L L : ~ ~ ~ W k
P P
C. C' N hr* NI-OI-I-C 4

N C N
b:WN I-
PI-!-'!-' I- I-' I-' I-' r r r r r rr I-' * 0-D
w r N r r d r d r C C C m C P m N N C ~ C Q C O ~ N O ~ O @ ~ N O C D O ~ C IC DC ( ~ O
0
W ~ I - !-' N r CL-( 0
0 9 9~ 0 ~ 0
~0 0
4 N 01-99 ~ C ~ ~ C ~ ~ W ~ O N O ~ O -~ C W ~ O O C ~

W W N C W W W W r r r r r r N N r r r r N N N N N W r r r N r r r r N r P N ~ - F C
U W C N O ~ N W ~ @ r e 4 m C 4 C W w N ~ ~ r ~ r ~ r ~ W u a S ~ c m Or r r w r C
C W C N NI-NN h: N N N r r r r r N N r r r r r C r N r r r r r T
O ' W 9 C T ; C 4 W C ( T 0 3 W C N 0 2 0 9 4 C C r G Q C W N C r + m C w d r p g o ? C + C r -
w r r r ~
~~ ~ r ~I r - ~~ ~ r r
~ r ~ rr r r~r ~rr r ~
~r r I - + '+' ~ r
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 0 0 0 0 C O O C C O O O C 0 0 & 0 C o o c o 9 9 Q o o c
O C O 0 0 0 0 0 0 0 0 0 0 0 0 C O 0 0 0 C O O O C C C 0 0 0 O O d C O O O O O w O c 4 C 3 O r

r r r r r r -I-r
h~N &) w Cr' N r r O
O O C t3lW W 4 N W
 DISCRIMINANT ANA L YSIS

The method of discriminant analysis was used t o provide an with Zi = mean of ith group
unbiased estimate of differences between the major soil Zij = value of 2 in the ith group evaluated with
groups-ferrisol, ferrallitic and ferruginous-and their respec- jth set of parameters.
tive subgroups.
Basically, discriminant analysis assumes that various ma- In terms of the individual parameters, Zij can be written
terials can be classified into major groups according t o certain
dominant characteristics, and then tested to see if the same
classification is valid using secondary characteristics. In this
study, soil groups were determined by a pedological classifi- Equation (3) can be resolved into the form:
cation and then tested according to physical properties.

MATHEMATICAL BACKGROUND
The method of discriminant analysis as developed by R. A.
Fisher, defines the variable Z1
2 = Xlxl + X 2 x 2 + . . . . . . . . +XkXk (1)
where 2 is an index used t o differentiate between the two
groups of materials based on the secondary parameters X I , X2
with dp = - q2 the difference between the
mean value of the pth parameter
. . . . . . . Xk, and the coefficients XI, X2 . . . . . . , Xk. If there
for each group.
is any relationship between the primary and the secondary
characteristics, then with the proper choice of k coefficients,
one can reasonably expect the frequency distribution of the
variable 2 to have a bimodal characteristic.
The samples grouped about the group means El and Z 2 ,
separated on the basis of the secondary parameters, are Maximization of G occurs by differentiation with respect to
likewise correctly separated on the basis of the primary the coefficient A, and setting the equivalent expression equal
characteristics. That is, a sample of primary class 1 will be to zero.
found t o have a value of 2 , based on secondary characteristics,
near Z1 , while a sample from primary class 2 will have a value
of 2 near Z2.
The object of the method is t o choose the parameters X I ,
X 2 , . . . . . . . , Xk such that the discriminant nature of
2-distribution is at a maximum-that is, the distance between
the group means Z, and Z, is as great as possible, while the
6B - 1 6A
variation within each group is as small as possible. Simplifying - - - - . .
Since 2 is a defined rather than a measured variable, 6Xr G 6Xr
calculation of the coefficients XI, X2, . . . . . . . . , Xk cannot be For each coefficient the following relationships hold:
obtained through direct regression. To circumvent this prob-
lem, the ratio of the differences between groups, as given by
the group means, Z1 and Z 2 , to the differences within the
groups, as represented by the standard deviations, T t l and
TZ, is maximized. A
and - = 2 ( X l d l + h 2 d 2 +. . . . . . + Xkdk)dr (7)
Both the maximization of bimodal differences and solu- r
tions for the coefficients X1, . . . . Xk can be obtained by
differentiating the function: Inserting (6) and (7) into (5), and simplifying yields

where C is (Al d l+ X2d2 + . . . . . . + hkdk)/G and independent

of r. For convenience C is set equal t o one.
Since (8) defines a linear equation for each coefficient, k
sets of equations are evolved which can be written in matrix
form.
 percent level. In the absence of more data, the result
is inconclusive.
(9) (3) At the five percent level both groups can be dis-
tinguished and the hypothesis of difference is
accepted.
The terms Sij and di are numerical quantities and can be Of course, as the number of samples is increased and the
readily calculated- Once the matrix (S) and the vector (d) have number of parameters extended, the results become more
been ccjmputed, standard reduction techniques such as a conclusive.
Gaussian reduction can be used to solve for the individual
coefficients. PROGRAMNIING
Due to extensive calculations, discriminant analysis be-
TEST OF SIGNIFICANCE
comes inapplicable,
- - in large
- scale studies, without access to an
After computing the coefficients h i , X2, . - - , Xk the sum electronic computer. A program was written to process
of the squares "between" and "within" groups are determined. discriminant analysis data and its listing is included at the end
of this section.
For "between" groups Written in FORTRAN I1 programming language, the pro-
gram runs on an IBM 1620-11 computer having a 60-K storage,
DB = Xldl + X2d2 + . . . . . + Xkdk . . . (10) card reader and typewriter output. Neither external storage
nor an on-line printer are required.
while "withn" groups Programming features include:
(1) Internal storage for two groups of data with a
(1 1) maximum of one hundred data sets, five parameters
per data set.
(2) Internal cycling and select feature to compare two
where n l , n2 = the number of data sets in groups one groups based on from one to five parameters which
and two respectively. need not be sequential.
A simplified flow chart is shown below.
The number of degrees of freedom for "between" and
"within" groups is computed. APPLICATION AND EXAMPLE
DFB = k, the number of parameters Sample input, sample output and an input sheet are
included together with a program listing.
DFw = n, + n2 - k - 1 As an illustration of the use of the program two groups of
soils, ferruginous and ferrallitic, each with liquid limit (LL),
Dividing each sum of squares by its number of degrees of plastic index (PI) and maximum dry density (MDD) para-
freedom, yields its mean square. Subsequent division of the meters are to be tested. Initially, both groups are to be tested
mean square between groups by the n ~ e a n square within using all three parameters (LL, PI, MDD) and subsequently
groups gives the F value. tested using only liquid limit and plastic index (LL/PI)
Once the F ratio has been calculated, a standard set of Parameters. Although only twenty data sets for each group are
statistical tables are consulted to test the significance of the presented here, substantially more were used in the final
results. Three outcomes are possible. testing.
(1) The computed F value is less than the statistical value
at both the one and five percent significance levels. Input
(2) The computed F value is greater than the statistical The input cards are prepared in the following sequence.
value at the one percent but less than the value at the 1. Constant data card - This feature allows alphabetic
five percent level. representation of the characters LL, PI and MDD to
(3) The computed F value is greater than the statistical appear on the output. The initials, LL, PI and MDD
values at both the one and five percent levels. are entered in columns one, five and nine
Interpretation of the above situations is as follows: respectively .
(1) Neither group can be distinguished from the other, 2. comparison card - only one comparison, namely
based on the given information. that be tween ferruginous and ferrallitic soils, is being
(2) While the groups can be distinguished at the one made; hence a one is entered.
percent level, they can not be distinguished at the five 3. Header Card - A written description of the problem
 I R E A D C O N S T A N T DATA AND I
NO. O F COMPARISONS C A R D
I

READ HEADER CARD AND

PARAMETER INDEX CARD
I
R E A D A N D S T O R E DATA
I

t
READ PARAMETER CONTROL CARD

I
I
CALL IISORT" TO SELECT DATA WITH I
lTHSET OF PARAMETERS AND
CALCULATE PARAMETER M E A N S
FOR EACH GROUP

OUTPUT VARIABLE * ESTABLISH S MATRIX AND

MEANS D VECTOR

CALL "SOLVE" TO SOLVE

FOR COEFFICIENTS

CALCULATE
1. DEGREES OF FREEDOM
2. SUM OF SQUARES
3. MEANS SQUARES
4. F RATIO
NORMALIZE X'S BASED ON
SMALLEST COEFFICIENT

I OUTPUT COMPUTED INFORMATION

I
1
I =I+,! - NO IS I= NO. OF SETS
OF VARIABLES
I

YES

IS J= NO. OF SETS -
J=J+i
OF COMPARISONS
I
 Ferruginous-Ferrallitic Sample Test is entered start- phased. In this case, two phases occur. Since both phases have
ing in column one. the same output format, only one need be discussed.
4. Parameter index card - since comparison is being Initially, the number of sets in each group used in testing is
made on two sets of parameters - first, all three presented. Seven sets of ferruginous soil and seven sets of
parameters together, and secondly, only liquid limit ferrallitic soil were used. These numbers differ from twenty,
and plastic index - a two is entered. since some elements in each group did not have maximum dry
5. Data - both sets of data are entered with ferruginous density values. Only those sets possessing all three parameters
group preceding the ferrallitic group. Each set of data are allowable.
is proceeded by a card indicating the number of The mean values of each parameter, liquid limit, plastic
points within each group, which in tlus case is index and maximum dry density are listed with each group,
twenty, followed by the data. The data is specified for those groups which were defined earlier as and q 2 .
first by an identification code, of the user's choice,
Two sets of coefficients are given for the discriminant
entered in columns one through six, followed by the
equation. The first set, under the heading LAMBDA define the
parameters liquid limit, plastic index and maximum
discriminant equation:
dry density starting in columns seven, fifteen and
twenty-three. Although the input sheet has been 1 + - 3 . 0 7 ~ 1CT3 (MDD)
Z = - 4 . 8 3 ~ 1CT2 (LL) + 2 . 8 8 ~CT2
abridged t o show only the first and last cards, twenty After computing the various sum of squares required, the
cards are entered for each group. coefficients are normalized by dividing the equation by the
coefficient with the least absolute value. In this case,
6 . Parameter control cards - this card selects the
-3.07x10-~. This operation yields the normalized equation:
number and the individual parameter to be used in
testing. There mtist be as many parameter control 2' = -1 5.743(LL) + 9.376 PI ( M D D )
cards as the number entered on the parameter index Since Z is only an arbitrary measure of difference, the sense of
card. In this case there must be two parameter the equation is unaltered. The quantities DELTA and PRO-
control cards. On the first card, a three, indicating all DUCT refer t o the differences between .the mean values of
three parameters are t o be used, must be entered in each parameter (%I - q 2 ) and the product of this

column three followed by the numbers: one (LL) in difference and the "unnormalized" lambda (Aidi).
column six, two in column nine (PI) and three in Lastly, the sum of squares (SS), degrees of freedom (DF)
column twelve (MDD). For the second z comparison, and mean square (MS) are given for "between" and "within"
a second card is used with a two entered in column groups. Following these is the F value.
three, indicating that only two parameters are to be In both cases the computed F values of 12.7 and 29.5 are
used, followed by a one in column six (LL) and a two greater than the statistical values of 6.55 and 4.36 at the five
in column nine (PI). percent level of significance. The two groups can then be
distinguished, one from the other, according t o F criteria.
The test has been so defined that group one is ferruginous,
group two is ferrallitic, parameter one is liquid limit, para- BIBLIOGRAPHY
meter two is plastic index and parameter three is maximum 1. Fisher, R. A., 1947. Statistical methods for research
dry density. workers, 8 t h Ed. Oliver & Boyd. London: 249-289.
2. Goulden, C. H., 1952. Methods of statistic analysis,
Output 2nd Ed. Modern Asian Edition: 375-385.
Where more than one set of parameters is chosen as a basis 3. Hoel, P. G., 1947. Introduction to mathematical
for comparison of two groups, the output becomes multi- statistics. John Wiley & Sons, New York: 258 p.
 FERRUGINOUS SOILS

Sample Sample
Code LL PI MDD Code LL PI MDD

LA JA
LC
JA
LC
JA
LC
JA
LC
LA JA

LC JA
LC JA
LC JA
LC JA
LC
JA
LC
JA
LC
JA
LC
LC JA

LC JA
LC JA
LCNA JAJD
LCNA JAJD
LCNA
JAJD
JAJD
JAJD
DISCRIMINANT ANALYSIS PROGRAM
DIMENSION A%lO,l 1
DIMENSION PROD%lOn
DIMENSION PSUM%2,25n1 DELTA%5n, FLMDA%5n
DlMENSlON PBAR%2,5n, PARM%2,5,100n1 DSET%2,5,100n1 N%2n
DIMENSION NSET%2n, NSORT%5n
DIMENSION FLMDC%5n
DIMENSION ITITLE%20n
DlMENSlON NAME%5n
COMMON A, PARM, DSET, PBAR, NSORT, NSET, N
READ 11 11,NAME
READ 100,NCOMBS
DO 2000 JPM#1 ,NCOMBS
READ 1050,l-rlTLE
PRINT 1051,ITI-rLE
R EAD l00,NCOM BV
DO 1 J#1,2
READ 100,NSET%Jn
INDEX#NSET%Jn
DO 1 K#l,IhIDEX
READ 102,%DSET%J,NI,Kn,M#1 ,5n
DO 2000 KLM#l,IVCOMBV
READ lOO,NVAR,%NSORT%Mn,M#l, NVARn
CALL SORT%NVARn
PRINT 700,N%l n,N%2n
PRINT 702
DO 701 J#l ,NVAR
M#NSORT%Jn
PRINT 703,NAME%Mn1PBAR%1,Jn,PBAR%2,Jn
NVARSQ#NVAR*NVAR
DO 3 JKL#l,2
DO 3 J#l ,NVAR
INDEX#N%JKLn
DO 3 K#l,INDEX
PARM%JKL,J,Kn#PARM%JKL,J,Kn-PBAR%JKL1J~
DO 4 J#1,2
DO 4 K # l ,NVARSQ
PSUM%J,Kn#O
J#1
L#O
DO 5 JKL#l,NVARSQ
L#L&1
DO 6 LLL#1,2
INDEX#N%LLLn
DO 6 K#l,INDEX
PSUM%LLL,JKL~#PSUM%LL,JKL~&PARM%LLL,J,K~*PARM%LLL,L,K~
A%J,Ln#PSUM%l ,.lKLn&PSUM%2,JKLn
IF%NVAR-Ln5,50,5
L#o
J#J&1
CONTI NLI E
DO 8 J#l,NVAR
DELTA%Jn#PBAR%l,Jn-PBAR%2,Jn
A%J,NVAR&1 n#DE LTA%Jn
CALL SOLVE%NVARn
DO 220 J#l,NVAR
F LNIDA%Jn#A%J,IVVAR&l n
FMIN#ABS%FLMDA%l n n
DO 221 J#2,NVAR
IF%FMIIV-ABS%FLMDA%Jnnn221,221,222
FNIIIV$ABS%FLMDA%Jnn
CONTINUE
DELSW
DO 223 J#l ,NVAR
F LNIDC%Jn#FLMDA%Jn/FM IIV
PROD%J~#DELTA%J~*FLIMDA%J~
DELS#DELS&PROD%Jn
FONE#N%l
FTWO#N%2n
SSWG#DE LS
SSBG#FONE*FTWO/%FONE~FTWO~*DELS*DELS
NDFBG#NVAR
NDFWG#N%l n&N%2n-NVAR-1
FDFBG#NDFBG
FDFWG#NDFWG
FMSWG#SSWG/FDFWG
FMSBG#SSBG/FDFBG
FTLST#FMSBG/FMSWG
PRINT 600
DO 601 J#l,NVAR
M#NSORT%Jn
PRINT 602,NAME%Mn,FLMDA%Jn,FLMDC%Jn,DELTA%Jn,PROD%Jn
PRINT 603
PRINT 604,SSBG,NDFBG,FMSBG,FTEST
PRINT 605,SSWG,NDFWG,FMSWG
CONTINUE
FORMAT%/,9X18HVARIABLE,6X16HLAMBDA,5X,1OHNORMALIZED,5X,5HDE LTA,17X17HPRODUCTn
FORMATXI 1X,A4,5X1E1 1.4,Fl 1.3,F13.3,4X1E1 1.4n
FORMAT%/,32X,2HSS19X,2HDF,9X,2HMS,1 2X,1 H F n
FORMAT%9X114HBETWEENGROUPS,4X,E12.5,16,4X,El2.5,4X,F5.1~
FORMAT%9X114HWITHIN GROUPS,4X,E12.5,16,4X,E12.5n
FORMAT%///,25X19HGROUP 1 #, 13,6X19HGROUP 2 #,I 3n
FORMAT%16X18HVARIABLE,13X16HMEAN 1,14X16HlVlEAN2n
FORMAT%18X1A4,F21.3,F20.3~
FORMAT%613n
FORMAT%6X15F8.0n
FO RMAT%20A4n
FOR MAT%///,20A4n
FORMAT%5A4n
END

SUBROUTE SORT%NVARn
DlMElVSlON A%lO,l 1n
DlMElVSlOlV PBAR%2,5n, PARM%2,5,100n, DSET%2,5,100n, N%2n
DlMENSlOlV IVSET%2n, NSORT%5n
COMMON A, PARM, DSET, PBAR, NSORT, NSET, N
DO 1 K#1,2
N%KnW
DO 1 J#1,5
PBAR%K,Jn#O
DO 2 JPM#1,2
INDEX#NSET%JPMn
INC#1
DO 2 JKL#l,IlVDEX
NOTE#O
 DO 3 K#1 ,NVAR
NTVAR#NSORT%Kn
1 F%DSET%JPM,NTVAR,JKLn n4,3,4
4 NOTE#NOTE&l
3 CONTINUE
IF%NOTE-NVARn2,5,2
5 DO 6 J#1 ,NVAR
NTVAR#NSORT%Jn
PARM%JPM,J,INCn#DSET%JPM,NTVAR,JKLn
6 PBAR%JPM,Jn#fBAR%JPM,Jn&PARM%JPM,J,INCn
INC#I NC&1
N%JPMn#N%JPMn&I
2 CONTINUE
DO 7 J#1,2
DO 7 K # l ,NVAR
D IV#N%Jn
7 PBAR%J,Kn#PBAR%J,Kn/DIV
RETURN
END

* 1008
SUBROUTE SOLVE%NROWSn
DIMENSION A%lO,l l n
COMMON A
C
C SIMLILTANEOUS EQUATION SOLUTION
C
NEND#NROWS&l
DO 1 J#D,NROWS
NBEG#J-1
DO 1 L#J,I\IROWS
IF%A%L,J-1 nn4,1,4
4 cON#A%J-1 ,J-1 n/A%L,J-1 a
DO 1 K#NBEG,NEND
A%L,KJJKON *A%L,K~-A%J-I ,KU
1 CONTINUE
NEND#NROWS
NSTOP #IVROWS&1
DO 2 J#Q,NROWS
NEND#IVEND-1

DO 2 L#l,NEI\ID
IF%A%L,NEND&I nn5,2,5
5 CON#A%NEND&l,NEND&l n/A%L,NEIVD&l n
DO 2 K#L,NSTOP
A%L,K~#CON*A%L,K~-A%NEND&I,K~
2 CONTINUE
M#IV ROWS&1
DO 3 J#1 ,N ROWS
A%J,M n#A%J,M n/A%J,Jn
3 A%J,Jn#1
RETURN
ElV D
 DISCRIMINANT ANALYSIS
(F-3)
CONSTANT DATA CARD
I

COMPARISON CARD
I

m
HEADER CARD

I . ~ I ~ _ ~L ~1I1 -s IF^ 11-

+PEriiqqG -mm~
PARAMETER INDEX CARD
I

Dzl
CARDS IF1 DATA SET ONE (1-100)
1
[W]
DATA SET ONE
IDENTITY , PARAMETER1 PARAMETER 2 2J PARAMETER 3 PARAMETER 4 39 PARAMETER 5

IJIAI I I I I 3 l o l I I I I 1 I ~ l ~ I ~I Il tI l l ~ l z l r l ~ l Is lI I I I I I I I I I 1 1 I I I I I ]
I
I

CARDS IN DATA SET TWO ( I - 100)

I
[7JT]
DATA SET TWO
IDENTITY 7 PARAMETER I 1s PARAMETER 2 23 PARAMETER 3 31 PARAMETER 4 39 PARAMETER 5
- -
lsTkT1~llrlol~l1II1121~1~1IIIIIIIII1IIIIII!IlI11II IIII
. -.

PARAMETER CONTROL CARDS ( S A M E AS NUMBER OF CARDS IN PARAMETER INDEX)

 APPENDIX FOR CHAPTER 6

METHOD OF TEST FOR THE PREPARATION

OF SENSITIVE SOILS

Hawaii Test Method TC-5

SCOPE: 8 . Air dry the sample until it is wet without any free
The method of test is intended for the preparation of water.
sensitive soil mixtures for subsequent tests such as liquid 9 . Mix the wet sample thoroughly and test.
limit, plasticity index, sand equivalent, gradation, and 10. Test the sample in accordance with AASHO Desig-
Proctor tests. The use of this method is covered in nation: T 8 9 - Standard Methods of Determining the
sub-section 6.7 of the Standard Special Provision. THIS Liquid Limit of Soils, with the exception that the
METHOD SHALL BE USED ONLY WITH THE WRIT- test will be run from the wet to dry state. This is
TEN APPROVAL OF THE MATERIALS TESTING & accomplished by air-drying the sample in increments
RESEARCH ENGINEER OR IF ALLOWED IN THE and performing the shock determinations at different
SPECIAL PROVISIONS FOR THE PROJECT IN WHICH moisture contents as the soil slowly dries out.
THE SOIL WILL BE USED. 11. Using a portion of the sample in the liquid limit test,
This method is used on soils such as volcanic ash at a high test the soil for the plasticity index in accordance
natural moisture content w h c h will irreversibly change its with AASHO Designation: T 9 0 - Standard Methods
properties if dried before testing. of Determining the Plastic Limit and Plasticity Index
of Soils.
APPARATUS:
1. Balance, sensitive to 0.1 gram.
PREPARATION FOR SAND EQUIVALENT TEST:
2. Sieves, No. 4 , No. 10, No. 40 and No. 200 1. A sample shall be selected by the method of
3. Sample splitter. quartering or by use of a sample splitter.
2. Sieve the sample over the 4 sieve. Rub the sample
PREPARATION OF SAMPLE: through the sieve with fingers if necessary.
The sample as received from the field shall be thoroughly 3. Without drying the sample, test in accordance with
mixed and aggregations shall be thoroughly broken u p in AASHO Designation: T 176 - Standard Method for
such a way to avoid reducing the natural size of individual Plastic Fines in Graded Aggregates and Soils by Use
particles. DO NOT DRY THE SAMPLE. of the Sand Equivalent Test.

PREPARATION FOR LIQUID LIMIT PREPARATION FOR GRADATION TEST:

AND PLASTICITY INDEX TESTS: 1. Two samples shall be selected by the method of
1. A sample, weighing approximately 1000 grams, shall quartering or by use of a sample splitter-one for the
be selected by the method of quartering or by use of gradation test and the other for moisture content.
a sample splitter. 2. The gradation sample shall weigh, in its natural
2. Place the 40 sieve in a pan. condition, not less than the amount indicated in the
3. Place a portion of the sample on the sieve. following table:
4. Add water with a syringe w h l e agitating the sample Nominal Maximum Approximate Minimum
Sieve Size Weight o f Sample, kg
with the fingers. Continue washing until all the minus
40 material has been washed into the pan. No. 4 1 .o
5. Repeat this procedure until all the sample has been 112"
washed. 314"
1"
6. Set the container holding the wash water and fine
1- 112" or over
material aside without disturbing for several hours
until all soil particles have settled to the bottom of 3. The moisture sample shall weigh at least 30 percent
the container and the water is clear. of the gradation sample.
7. Siphon or pour off as much of the water as possible. 4. Weigh the gradation sample in its wet condition and
Be careful not to lose any of the soil particles. place in a container. Add sufficient water t o the
 sample in the container t o cover it. PREPARATION FOR PROCTOR TEST:
5. Prepare a nest of two sieves with a 200 sieve on the 1. No other preparation other than that noted above in
bottom and a 10 sieve on the top. PREPARATION OF SAMPLE is required.
6. Transfer the sample t o the nested sieves and wash 2. With one exception, the material wi'l be tested in
with running water. When the sample is larger than accordance with AASHO Designation: T 9 9 - Stan-
can be handled at one time on the nested sieves, wash dard Method of Test for the Moisture-Density Rela-
a portion of the sample and transfer it t o a container tions of Soils Using a 5.5-lb. Rammer and a 12-in.
in which it is to be dried. Drop or AASHO Designation: T 180 - Standard
NOTE: Tapping of the sieves has been found to Methods of Test for Moisture-Density Relations of
expedite the washing procedure. Soils Using a 10-lb. Rammer and an 18-in. Drop,
7. Dry the washed material retained on the nested sieves whichever is applicable. The exception is that the test
to a constant weight at a temperature not to exceed will be run from the wet to dry state starting from
230°F (1 10°C). .the natural moisture content. This is accomplished
8. Test the sample in accordance with AASHO Desig- by air-drying the soil in increments and performing a
nation: T 27 - Standard Method of Test for Sieve density determination at different moisture contents
Analysis of Fine and Coarse Aggregates. as the sample slowly dries out.
NOTE: The dry weight to be used in the computa- 3. Run as many points as possible.
tion for the gradation shall be computed using the 4. If the test takes more than a day t o run, be sure t o
initial wet weight and the moisture content of the place the sample in an air-tight container so as not to
sample. lose any moisture.

ERROR SIGNIFICANCE IN ONE-POINT LIQUID LIMIT TEST

The expression relating liquid limit (LL) to the water In carrying out the 1-point test, there are four sources of
content (W) of a soil sample and the number of blows (N) to error that would not occur in the multi-point method. First,
close .the groove in Atterberg's liquid limit test is found the single water content reading may be incorrect, whereas
experimentally t o be of the form: with several determinations in the multi-point method errone-
tan B ous readings are more likely detected and discarded. Second,
N
L L = W (25) . . . . . . ... (1) the slope differs from sample t o sample. Third, the constant
value of the slope assumed in the one-point test may differ
where tan B is the slope of the line W vs N on a semi-log plot. from the average slope for the soil under study. Fourth, the
Usually, N lies between 15-35 blows. The values for tan B value of N for the one-point test cannot be closely controlled;
reported from many locations throughout the world lie in the the greater the difference of the number of blows from 25, the
range 0.090-0.150. more the errors due to errors in slope are accentuated. The
Within these narrow limits, the liquid limit may be relative contribution of these errors is determined in the
adequately represented by the first two terms in the series following. P

expansion of (1), neglecting higher orders of tan B, Assuming an associated error with each measurement, we
may express (1) as:

N
tan B + Sb + Se
LL + LLe = (W + We) (-)2 5 ... (3 ;'

From (2) it is readily seen that the term containing tan B

normally comprises no more than a small correction term in
the determination of the liquid limit. This fact is the basis for where Sb is a constant equal to the difference between the
the development of the 1-point testing method. average slope and the constant assumed, whereas S, represents
Since the slope of the line W vs log N is usually constant for the variation about the average slope.
a given test, the value of tan B is taken to be a constant. Eden Assuming form (2) and dropping second order error terms,
(1959) assumes a reasonable average of reported values to be it is easily shown that,
0.1, and using this value, he presents a table of correction
factors, equivalent to the bracketed term in (2), for blows
from 15 t o 35.
This indicates how the four sources of error mentioned above To illustrate the relative magnitudes of these errors, the
are related to the overall percent error in the liquid limit following data have been obtained from 564 soil samples
measurement. tested in Ghana using the multi-point method:
Since We, N, and S, are independent random variables, (4)
Mean tan B = 0.150
cannot be used directly to calculate the extent of the overall
Standard deviation = 0.08 17
error LL,. This can be calculated, however, by deriving the
variance of the combined variables. For each test, one measurement between 20 to 30 blows
To handle the product in the second term of (4), the was chosen t o represent what would have been obtained had
following relation is employed: For independent random the 1-point method been employed and limited to that range.
variables x and y (with respective mean values (This did not cause undue bias, since there was usually only
-
x and y, the variance of the product x.y is given by, one such point in each test.) Table 4 8 shows the distribution
of these measurements, the average percent error in the
uncorrected values for water content at each blow level, and
the associated standard deviation. The liquid limit determined
from the multi-point test was assumed to be the correct value.
Recalling that the variance of a sum of random variables is Figure 132 shows a plot of these data superimposed on the
equal to the sum of the respective variances, and assuming that theoretical correction function for tan B = 0.1 50.
the mean values From (6), as, = 0.0817. If Eden's value of tan B = 0.1 is
N - 1 ) and Se are zero, while the variance of Sb is zero,
of ( - assumed, then Sb = 0.05. From Table 48, oN is calculated to be
25 0.09. In the Ghana tests, the error in measuring water content
then we may write the variance of LLe/LL as,
is found to be in the neighborhood of *2%, so that ow, =
0.02.
Substituting these values in (5),
where o LLe
= standard deviation of -'
LLe LL

N
ON = s.d. ( - 1 ), and Thus, in this case, it is seen that the error inherent in the
test itself is the most significant factor in the 1-point method,
far outweighing any practical deviations in the value of tan B
for the soil tested.
In fact, if no corrections at all had been included in the test
TABLE 48 (tan B = O), the resulting standard deviation in the percent

Liquid Level Test Error Measurements

NO. of oerr
Blows Sam~les

NUMBER OF BLOWS
NOMBRE DE COUPS

FIG.132 PERCENT DIFFERENCE BETWEEN LIQUID LIMIT AND I-POINT

WATER CONTENT DETERMINATION
Overall 564 -0.30 2.60 DIFFERENCE RELATIVE ENTRE LlMlTE DE LlQUlDlTE ET LA
DETERMINATION DE LA TENEUR EN EAU PAR MESURE UNIQUE
error would be: if the basic measurement of water content has an accuracy
within only 2% of the true value. It is even debatable whether
the test result should be corrected t o a standard 25 blows, as
This agrees well with the 2.6% error observed in the long as the permissible range is limited between 20-30 blows.
uncorrected values (Table 48). By inserting a full correction When the test accuracy is better than 1%, the minor
term (tan B = 0.1 50), the percent error was reduced only to corrections are important. It will be noted, however, that a
2.4% as indicated by the above example. minor mean deviation (k.02) in tan B from the "average" value
The effect of changes in the distribution of blows permitted of 0.1 has very little effect in any instance, and is insignificant
in the test is illustrated by Eden. He presents data showing a compared to the deviations between .the samples themselves.
rather narrow range of errors for blows between 20 and 30
with O N about 0.09 as above, a s e = .036, Sb = 0 , and OLL, =
CONCLUSION :
0.006. By inserting these values in (5), it is seen that the slope
characteristics have very little effect so that the water content The conclusions can be summarized in three guidelines for
measurements have a value of owe near 0.006. A second set of use of the one-point liquid limit test:
data, showing a much broader range of errors, was based o n 1. Limit the number of blows t o the range 20-30.
the same soil group and testing procedures, but the range of
blows was changed. The range varied between 15 t o 20 and 3 0 2. If the error in the water content measurement is 1% or
t o 35, resulting roughly in G N = 0.32. Inserting this value in less, adjust the value t o 25 blows using (2) with tan B =
(5) along with the previous error values, 0.1.
3. If the water content measurement error is greater than
1%, any corrections have an insignificant effect on the
or 1.476, on the same order as the 1.6% Eden obtained result.
experimentally.
It would thus appear that neither the characteristic slope REFERENCE:
value for the soil, nor the range of blows employed, plays a Eden, W. J., 1959. Use of a one-point liquid limit procedure.
significant role in determining the accuracy of the 1-point test ASTM Special Tech. Publ. No. 254: 168-176.

BRITISH STANDARD 812 : 1960

METHODS FOR SAMPLING AND TESTING OF

MINERAL
AGGREGATES
SANDS AND FILLERS
BRITISH STANDARDS INSTITUTION

SECTION SIX : MECHANICAL PROPERTIES

Determination Of Aggregate Impact Value

29. a. General: The aggregate impact value gives a relative be prevented from rocking either by fixing it to the block
measure of the resistance of an aggregate to sudden shock or or floor or by supporting it on a level and plane metal
impact, which in some aggregates differs from its resistance to plate cast into the surface of the block or floor.
a slow compressive load.
b. Apparatus: (i) An impact testing machine complying A cylindrical steel cup of internal dimensions:
with the following: diameter 4 in. (102 mm)
Total weight not more than 130 lb. (60 kg) nor less depth 2 in. (5 1 mm)
than 100 lb. (45 kg). and not less than 114 in. (6.5 mm) thick with its inner
The machine shall have a metal base weighing surface case-hardened, that can be rigidly fastened at the
between 50 and 65 lb. (22 and 30 kg), with a plane lower centre of the base and easily removed for emptying.
surface of not less than 12 in. (30 cm) diameter, and shall A metal tup or hammer weighing 30-31 lb.
be supported on a level and plane concrete or stone block (1 3.6-14.1 kg), -the lower end of which shall be cylindrical
or floor at least 18 in. (45 cm) thick. The machine shall in shape, 3 15/16 in. (100.0 mm) diameter and 2 in. (5
 cm) long, with a 1116 in. (2 mm) chamfer at the lower straightedge. The net weight of aggregate in the measure shall
edge, and case-hardened. The hammer shall slide freely be determined t o the nearest gram (Weight A) and this weight
between vertical guides so arranged that the lower of aggregate shall be used for the duplicate test on the same
(cylindrical) part of the hammer is above and concentric material.
with the cup. d. Test procedure. The impact machine shall rest
Means for raising the hammer and allowing it t o fall without wedging or packing upon the level plate, block or
freely between the vertical guides from a height of floor, so that it is rigid and the hammer guide columns are
15 + 114 in. (381 + 6.5 mm) on t o the test sample in the vertical.
cup, and means for adjusting the height of fall within 1/ 4 The cup shall be fixed firmly in position on the base of
in. (6.5 mm). the machine and the whole of the test sample placed in it and
Means for supporting the hammer whilst fastening or compacted by a single tamping of 25 strokes of the tamping
removing the cup. rod.
NOTE. Some means for automatically recording the The hammer shall be raised until its lower face is 15 in.
number of blows is desirable. (38 cm) above the upper surface of the aggregate in the cup,
(ii) B.S. test sieves of sizes 112 in., 318 in. and No. 7. and allowed to fall freely on t o the aggregate. The test sample
(iii) A cylindrical metal measure, tared t o the nearest shall be subjected t o a total of 15 such blows, each being
gram, of sufficient rigidity t o retain its form under rough delivered at an interval of not less than one second.
usage, and of the following internal dimensions: The crushed aggregate shall then be removed from the
cup and the whole of it sieved on the No. 7 B.S. (No. 10 U.S.)
diameter 3 in. (76 mm)
sieve until no further significant amount passes in one minute.
depth 2 in. (5 1 mm)
The fraction passing the sieve shall be weighed to an accuracy
(iv) A straight metal tamping rod of circular cross-section, of 0.1 (Weight B). The fraction retained on the sieve shall also
318 in. (10 mm) diameter, 9 in. (23 cm) long, rounded at be weighed (Weight C), and if the total weight B + C is less
one end. than the initial weight (Weight A) by more than 1 g. the result
(v) A balance of capacity not less than 500 g, readable shall be discarded and a fresh test made.
and accurate t o 0.1 g. Two tests shall be made.
(vi) A well-ventilated oven, thermostatically controlled to
maintain a temperature of 1 0 0 - 1 0 0 " ~ . e. Calculations. The ratio of .the weight of fines formed
t o the total sample weight in each test shall be expressed as a
c. Preparation of the test sample. The test sample shall
percentage, the result being recorded t o the first decimal
consist of aggregate, the whole of which passes a 112 in. B.S.
place :
test sieve and is retained on a 318 in. B.S. test sieve. The
aggregate comprising the test sample shall be dried in an oven Aggregate impact value B x 100
for a period of four hours at a temperature of 100-100"C, and
cooled. where A = weight of oven-dried sample (g)
The measure shall be filled about one-third full with the B = weight of fraction passing No. 7 B.S.
aggregate and tamped with 25 strokes of the rounded end of (No. 10 U.S.) sieve (g).
the tamping rod. A further similar quantity of aggregate shall
be added and a further tamping of 25 strokes given. The f. Reporting of results. The mean of the two results
measure shall finally be filled t o overflowing, tamped 25 times shall be reported t o the nearest whole number as the aggregate
and the surplus aggregate struck off, using the tamping rod as a impact value of the tested material.

SOIL SWELLING TEST (L.N.E.C. E 200-1967)

SCOPE made up of: a perforated brass base fitted with one stem for
This testing standard covers the determination of volume fastening the dial gauge and two others for fastening the mold
change expressed in percentage of the initial volume, under- and its extension piece; a mold, 15 mm in height and 60 mm
gone by a soil fraction passing a No. 40 (A.S.T.M.) sieve when internal diameter; a 20 mm high extension piece for the mold,
in standard compaction conditions, it absorbs water by with the same diameter; a 4 mm thick perspex transparent
capillarity through a porous plate. plate with a diameter of 59 mm; a piece for fastening the dial
gauge at the stem of the base; a very fine grained porous plate
APPARATUS with a diameter of 65 mm and a thickness of 3 mm.
a) Apparatus for the determination of swelling essentially b) A tamper fitted with a spring capable of exerting a
force of 5 kgf when pressed in a travel of about 25 mm. The placing the porous plate on the base and fastening the mold
spring travel can be adjusted by means of a set of variable and its extension piece on the plate.
thickness washers so that the applied force remains equal to 5
b) Compact the soil sample, in two layers of about the
kgf.
same thickness, which must be such that after compaction it
c) A 0.01 mm dial gauge exerting a force of 100 gf when slightly exceeds the height of the mold. The sample is
a displacement of 2 mm is applied at the stem at a rate of compacted by the tamper by pressing the handle until the
about 0.5 mm per min. When the dial gauge does not comply flange edge is in contact with the part. Only the required force
with this condition, the spring which operates the stem must must be applied, t o be removed as soon as the pieces are in
be removed and replaced by a tare contained in a suitable contact. Each layer is compacted with 5 0 uniformly distri-
vessel, calculated to fulfill the above condition." When the buted compressions applied, as described, at rate of 1
force applied by the spring is less than 100 gf, it suffices to use compression per second.
a suitable surcharge.
c) Once the soil is compacted, remove the extension piece
d) An oven capable of maintaining a temperature of and raze the excess of soil with the spatula so as to obtain a
60°C. perfectly smooth surface flush with the brim of the mold.
Carefully remove the soil excess, replace the extension piece
e) A mortar and its accessories. and set the perspex plate on the soil.
f) A No. 40 (ASTM) sieve. d) Place the apparatus inside the glass container and
g) A glass bottle with suitable dimensions fitted with a assemble the dial gauge (including the tare in the respective
sealing cover. vessel, if necessary), resting the top of the stem on the center
of cylindrical projection of the perspex plate. In order to
h) A spatula with a blade more than 7 cm long in its
ensure a close contact, an initial reading above zero is
straight edge.
recommended. Write down this reading (1 ,).
i) A flat-bottom vessel of glass or any other suitable e) Pour distilled water into the glass container up to the
material with dimensions capable of holding the upper face of the base (corresponding t o the upper face of the
apparatus for determination of swelling. porous place in contact with the soil) and start the stop watch.
j) Stop watch. f) Take periodic readings in the dial gauge until two equal
k) Brush. (or decreasing) values separated by a 2h interval are obtained.
Calculations
PREPARATION OF THE SAMPLE
Swelling is given by:
a) The sample must be taken from a thoroughly mixed
portion of material passing No. 40 ASTM sieve
(0.420 mrn) in accordance with the rules of preparation of
samples for testing. Quantity: about 100g. Where:
b) Dry the sample in the oven at 60 f ~ O Cduring about
16 h and let it cool in the sealed bottle. It is advisable to use a E - swelling
sample that entirely fills the bottle. Ah - change in height (mm)
ho - initial height (mm); (ho = 15 mm)
1, - initial reading at the dial gauge (mm)
TESTING PROCEDURE 1 - final reading at the dial gauge (mm)
a) Assemble the apparatus for determination of swelling, The result is reported to the nearest unit.

* In order to remove the force of the spring applied to the stem of the
dial gauge, it suffices to free the end of this. After removing the
spring the surcharge can be calculated as follows: in the first place the
strength of the hair is checked, an operation which for simplicity's
sake can be done with the vessel for the tare already in its place. For
that, the dial gauge is lowered at a constant rate of about 0.5 mm/min
with the stem end resting on the pan of a balance sensible t o O.5g.
Readings at each turn of the pointer are recorded. The difference for
100g. of the value obtained for a displacement of 2 mm (2nd turn) is
the surcharge sought. After applying this surcharge, the calibration
should be repeated and any deviation observed duly corrected.
 APPENDIX FOR CHAPTER 9

TABLE 49. STABILIZATION TEST DATA - BRRI LAB

STABILIZATION T E S T DATA
SAMPLE No. G!
TYPE OF STAB. L I E
O/o STAB. OO/o 2 O/o 4 O/o 6 O/o 8 O/o

7 28 7 28 7 28 7 28
-- -- - .-- --

LL 28.8 25.4 23.4 24.8 24.5 26.3 23.9 25.4 25.5

-.-. . .- ------ --
PL 18.4 19.8 16.8 20.7 18.7 22.5 19.6 22.0 22.5

PI 10.4 5.6 6.6 4.1 5.3 3.8 4.3 3.4 3.0

MDD 136.3 134.2 133 0 - 132.1 130.6
PC f

OMC-O/~ 8.9 9.1 9.3 9.7 10.2

4"6 qu a
psi 56 30 51 35 35
2118 qu 64 24.8 26.3 47.7 51.0 69.7 74.3 129.0 153.2
PS'

ST: 6.2 1.9 2.2 3 . 3.7 7.5 8.5 6.1 16.7

PSI

ST/qu 9.7 7.7 8.3 8.6 7.2 10.7 11.4 5.0 10.5

CBR-O/~ 30.2 64.8 55.0 77.8 89.0 98.5 122.0 91.1 152.0

SWELL-Ok 0.746
W.D.LOSS
100 100 100 31.4 51.5 24.0 22.0 16.4 18.1
O/o

STABILIZATION T E S T DATA

-
SAMPLE
-
No. 1 - - - -
G! -3
~ Y P EOF STAB. LIME CEMENT
O/o STAB. OO/o 2 O/o 4 O/o 6 O/o 8 O/o 2% I 4 O/O 6 O/O 8 O/O

7 28 7 28 7 28 7 28
LL 33.0 27.0 28.5 28.3- 28.4 26.0 30.0 29.1 29.7
PL 17.0 19.0 22.4 20.2 22.0 20.1 23.0 21.0 23.3

PI 16.0 8.0 6.1 8.1 6.4 7.9 7.0 8.1 6.4

!sD 134.0 132.6 130.6 130.7 129.5

I I I I I 1 I I
 STABILIZATION TEST DATA
SAMPLE No. I 01-14 1
I TYPE OF STAB. I LIME I CEMDTT 1
O/o STAB OO/o 2 O/o 4 O/o 6 O/o 8 O/o 2 O/o 4 O/o 6 O/o 8 O/o

7 28 7 28 7 28 7 28 7 28 7 28 7 28 7 28
- -- -. -- -- - -- ~- . ----- - -. --
L L 23-4 22.8 23.7 23.6 23.6 22.9 23.7 24.0 23.3 31.6 32.4 35.3 34.5 35.7 34.5 34.1 35.1
- -. --- - - ---- -. -- -- --- -

P L 12.0 17.4 17.8 19.3 19.1 18.8 20.8 20.5 21.6 19.9 20.6 25.3 26.6 27.8 26.9 27.1 29.2

PI 11.4 5.4 599 4-3 4.5 4.1 2.9 3.5 1.7 11.7 11.8 10.0 7.9 7.9 7.6 7.0 5.9

F2D 140.0 134.7 135.5 13 5.7 132.0 137.1 139.0 139.1 140.0

OMC-O/~ 7.2 8.6 8.3 8.4 9.0 7.6 7.4 7.5 7-3
4"0 qu
PSI' 23 119 23 2 la6 139 334 343 53 0 500
2"0 qu I
PSI
63-2 99.7 144.7 124.3 206.0 173.0 214.0 164.0 208.7 156.7 222.9 272.7 445.7 551.3 563.0 1510.0 785.3
ST 3.9 7.2 11.1 10.1 29.6 16.8 22.9 15.4 16.5 18.2 17.4 30.5 33.2 41.9 43.5

--I4
PSI
29.9 37.5

S T / ~ ~ 6-2 7-2 7.6 d.1 11.4 9.7 10.7 9.4 7.8 11.6 7.8 10.4 7.4 5.4 7.5 7.4 6.2

CBR-O/~ 40.9 164.0 171.0 241.0 265 250.0 276 220.0 313 430 387 500 500 500+ 500+ 500+ 500+

SWELL-Ok

WDLOSS
loo 26.8 38.4 21.1 3 8.7 8.7 5.0 9.5 15.2 6.7 5.5 9.3 3.0 4.2 1.7 1.7
O/o
I

-
SAMPLE No

TYPE OF STAB.

OO
/ STAB.1 O0/0

ST
psi
1 2.8
 STABILIZATION TEST DATA
I SAMPLE No ( 02-1

STABILIZATION TEST DATA

 STABILIZATION TEST DATA
-
SAMPLE No. 1 08-2
rYPE OF STAB. I LILlE I GEM ENT

I I I I I , - I

4"0 qu
pri 33 74 119 163

STABILIZATION TEST DATA

 STABILIZATION T E S T DATA
SAMPLE NO. (312-1
TYPE OF STAB. LIME CEhIRTT
Ye STAB. 0% 2 'Yo 4 O/o 6% 81
'0 2% 4 O/o 6% 8 Oo/

7 28 7 28 7 28 7 28 7 28 7 28 7 28 7 28
-- - - -. ~

LL 47.6 50.7 48.3 51.6 48.5 48.8 46.0 49.4 44.3 47.0 49.0 49.0 48.6 48.5 48.5 48.5 48.3
- - - - - -- - . ..-

P L 25.8 32.8 31.4 35.2 3 3 33.5 32.8 36.0 34.0 25.9 28.8 31.2 1 32.7 33.6 34.5 34.1

PI 21.8 17.9 16.9 16.4 15.1 15.3 13.2 13.4 10.3 21.1 20.2 17.7 17.3 15.8 14.9 14.0 14.2

FsD 127.5 124.0 124.5 125.0 122.6 129.0 127.5 128.5 129.0

OMC-% 11.4 12.9 12. 5 12.0 12.5 12.8 13.6 13.0 14.0
4Y@qu

*"'
psi

PSI
qu
.2a-5
185

194 203.0 317.0

23 2 175

321.5 303.0 252

181

328 320
223

139.5 280
186

156.5 325
204

271.0 367
162

368.4 437
ST 21.4 18.0 16.5 29.5 28.0 31.8 28.9 38.6 18.0
29.3 14.4 6.0 27.9 22.7 30.3 25.8 38.8
psi

~ ~ / q 10.0
u 9.3 8.2 9.3 9.1 8.9 12.6 8.8 12.1 13.0 5.2 10.2 8.6 8.4 8.3 7.0 8.9
-
CBR-'Yo 17.3 36.0 48.9 47.7 54.5 51.8 65.1 79.7 86.1 50.0 107.0 113.3 224-6187.3284.0 273.333O.C
-

SWELL-% 0.9 0.7 -03

L

w.D.LOSS 100 100.OLOO.O 57.7 44.5 46.3 49.0 33.9 21.8 100 100 12.4 3.9 9.2 1.9 3.9 3.2
%

STABILIZATION T E S T DATA
b

I SAMPLE No 1
I
IV- 5 1
TYPE OF STAB. I LIME 1
 STABILIZATION TEST DATA
SAMPLE No K-4
4
TYPE OF STAB LIME
O/e STAB OO/o 2 O/o 4 Oo/ 6 O/o 8 O/o 2 O/o 4 O/o 6 Oo/ 8 O/o
7 28 7 28 7 28 7 28 7 28 7 28 7 28 7 28
-- --

LL 3 41.5 38.4 44.0--&.5 46.3--49.2--47..149.5

- - - - - - -- --- - -- - -- ------
PL 15.1 21.2 21.7 26.3 29.0 29.7 35.0 32.2 37.5

PI 29.2 20.3 16.6 17.7 15.5 16.6 14.2 15.0 12.0

MDD
pcf 122.0 122.5 121.0 121.0 121.0
OMC-% 12.2 12.0 12.4 13.0 13.5
4-0 qu
Pa' 130 163 181 218 195
2-0 qu
PSI 99.5 165.5 253.0 217.3 237.3 239.0 353.3 393.3 451.5
ST 15.2 25.2 28.3 30.3 33.8' 25.8 381 72.3
PSI

ST/qu 15.3 15.2 11.2 13.9 114.2 10.8 10.8 16.c

STABILIZATION TEST DATA

1 TYPE OF STAB. I LIME I 1

SAMPLE No
TYPE OF STAB.

STABILIZATION TEST DATA

F-2
CEMENT
 STABILIZATION TEST DATA
I -

SAMPLE No TORMABUM
TYPE OF STAB. 1 CEMENT

STABILIZATION TEST DATA

- - - - -. -- --

1 SAMPLE No I FREETOW I
1
I

TYPE OF STAB. 1 LIME 1 CEMENT

 REPORT OF THE ASPHALT INSTITUTE

1.0 PREPARATION OF SOIL-ASPHALT 3.1 Curing and Testing Condition A - Effect of Binder
MIXTURES Type and Concentration on Properties of As-Compac-
1.1 Soil as received was premixed with a sufficient ted Specimens (Tables IVA, IVB, IVC, IVD, VIA,
amount of water to bring it t o AASHO T 180-A VIB, VIC and VID).
optimum water content. Mixing time of two minutes After compaction duplicate specimens were stored and
was needed t o obtain uniform soil-water mixture. cured for 2 4 hours in polyethylene bags at room
Hobart mixer equipped with a stiff paddle was used temperature. Plastic bags were used t o prevent losses of
for mixing. water and volatiles. Another set of duplicate specimens
1.2 Asphalt binder (RC-800 or MC-800) preheated t o was immersed in water at room temperature (75-80'~) for
approximately 2 10°F was added in a predetermined 24 hours.
amount t o soil-water mixture in a mixing bowl. 3.2 Curing and Testing Condition B - Effect of Partial
1.3 Mixture was mixed for two one-minute mixing Curing on Properties and on Water Sensitivity of
periods. After first one-minute mixing, the mixture Stabilized Soils (Tables VA, VB, VC, VD, VIA, VIB,
was scraped from mixing bowl and paddle. VIC and VID).
1.4 After mixing, soil-water-asphalt mixture in a mixing Four specimens (compacted at AASHO T 180-A optimum
bowl was weighed t o determine loss of water or water content) were cured (dried) in 1 4 0 " ~forced-draft
volatiles. Small samples of mixture were removed for oven t o somewhat"less than a half of the original volatiles
volatiles content determinations. content. After oven drying, two specimens were stored in
2.0 COMPACTION OF SPECIMENS plastic bags for 24 hours t o prevent further evaporation.
'Two other specimens were immersed in water for 24
2.1 The soil-water-asphalt mixture was placed in a four- hours at room temperature.
inch diameter cylindrical steel mold. Mixture was 3.3 Curing and Testing Condition C - Effect of Complete
rodded with a steel rod for uniform placement. Curing on Properties and On Water Sensitivity of
2.2 Specimens were compacted by means of Mechanical Stabilized Soils (Tables VA, VB, VC and VD).
Gyratory Compactor applying 3 0 gyrations at one Four specimens were dried in 1 4 0 " ~forced draft oven to
degree gyration angle and with the ram set at 100 psi a constant weight. Two specimens were tested after
pressure. storing in plastic bags for 2 4 hours. Two other specimens
2.3 Specimen height was measured continuously during were tested after 2 4 hours immersion in water bath at
the compaction process by means of a dial gage. room temperature.
Readings were taken after each five (5) gyrations.
Height measurements allow comparison between 3.4 Curing and Testing Condition D - Effect of Mixture
compaction characteristics of different mixtures. Aeration (Table VII).
2.4 After compaction the specimens were trimmed exact- After mixing and before compaction soil-water-asphalt
ly t o the height of 2.5 inches. Close fitting spacers mixture was placed in flat pans and aerated to somewhat
were used t o move compacted mixture in a mold for more than half of the original amount of volatiles. After
top or bottom trimming. Trimmings were used for gyratory compaction, two specimens were tested after
determination of volatiles content. storing for 24 hours in plastic bags and two other
2.5 Specimens were extruded from the mold by means of specimens were tested after immersion in water at room
a hydraulic press and after extrusion, specimen temperature for 2 4 hours.
dimensions were accurately measured. These measure- 3.5 Curing and Testing Condition E - Effect of Lower
ments were used t o calculate as-compacted specimen Than Optimum Mixing and Compaction Water Con-
density. tent (Tables IXA and IXB).
2.6 Four specimens were compacted for each mixture Soil at water content two percent below AASHO T 180-A
composition or each test series. optimum was mixed with heated asphalt cutback. Regular
compaction procedures described previously were used.
3.0 DESCRIPTION OF DIFFERENT CURING Two specimens were tested after 2 4 hours storage in
AND IMMERSION CONDITIONS plastic bags and two other specimens were tested after
For a thorough evaluation of various factors on stability immersion in water at room temperatures for 24 hours.
and on water sensitivity of asphalt stabilized mixtures, 3.6 Testing and Curing Condition F - Effect of Higher
laboratory compacted specimens were subjected to a Than Optimum Mixing Water Content (Table VIII).
number of different curing and exposure conditions. Soil at water content two percent above AASHO T 180-A
Descriptions of these conditions follows: optimum was mixed with heated liquid asphalt. After
 mixing, the mixture was aerated to approximately opti- as-immersed density determination.
mum water content. After aeration, regular compaction 4.3 Marshall stabilities and flow values were determined
procedures were used. Two specimens were tested after using normal testing procedures. Load was applied at
storing in plastic bags for 24 hours. Two other specimens a rate of two (2) inches per minute and load-deforma-
were immersed in 140°F water bath for 24 hours tion graph was obtained for each specimen.
and tested. 4.4 After testing, water and volatiles content were deter-
mined for each individual specimen. Water distribu-
4.0 TESTING OF COMPACTION SPECIMENS tion gradient was determined for some immersed in
4.1 After curing, storage or after water immersion, the water specimens.
specimens were weighed to determine losses or gains 4.5 Percentage of stability retained after immersion was
in volatiles or in water content. used as an indication of the water sensitivity of
4.2 Specimen dimensions were measured for as-cured or asphalt stabilized soil mixtures.

NOTE: All data presented in charts and tables

following are average of two tests.
FIG. AI-I PARTICLE SIZE DISTRIBUTION OF AFRICAN LATERITE SOILS
DISTRIBUTION DE GRANDEUR PARTICULE DES SOLS LATER l Q UES 0' AFRIQUES
 OCON-I A CON-2 DAH-3 V NIA-4

I I I I I
RATE OF LOADING : 0.05 inch/min.
PROPORTION DE DONNEE : 0.05 pouce/min

4 6 8 10 12 14 16 18 20
WATER CONTENT, PERCENT
TENEUR EN EAU, '10

I 1 I I
COMPACTION; AASHO T 180-A
COMPACTAGE: AASHO T 180-A

100
4 6 8 10 12 14 16 18 20
WATER CONTENT, PERCENT
TENEUR EN EAU,'/o
FIG. A l -3 DRY DENSITY AND UNCONFINED COMPRESSIVE
STRENGTH OF LATERITE SOILS
DENSITE SECHE ET RESISTANCE A L A
COMPRESSION DES SOLS LATERIQUES
 O CON-I A CON-2 DAH-3 V NIA-4

WATER CONTENT, PERCENT

TENEUR EN EAU,O/o

COMPACTION :
MECH. GYR. COMPACTOR
3 0 GYRATIONS, ONE DEGREE
ANGLE, 100 PSI RAM PRESSURE
I I
COMPACTAGE:
MECH. GY R. COMPACTOR
3 0 GYRATIONS, UN ANGLE2
DEGRE, 100 LIVRES/POUCE
RAM PRESSION
90
0 2 4 6 8 10 12 14 16 18 20 22
WATER CONTENT, PERCENT
TENEUR EN EAU,%

FIG. AI-4 DRY DENSITY AND MARSHALL STABILITY OF LATERITE

SOILS
DENSITE SECHE ET STABlLlTE MARSHALL DES SOLS
LATERIQUES 248
 TABLE l

Properties of African Laterite Soils

SOIL IDENTIFICATION CON- 1 NIA- 1

AASHO Classification

Specific Gravity

Moisture Content
As received, Percent

pH (50: 50 Water Slurry)

Liquid Limit
I
Plastic Limit

P.I.

Sieve Analysis (Washed)

Percent Passing
U.S. Std. Sieve 318"

No. 8
No. 16
No. 30
4
1 100.0

97.5
94.8
100 .O
99.8
98.1
88 -5
79.2
72.4
No. 50 92.9 95.5 75.6 62.9
No. 100 54.7 61.9 64.8 42 .O
No. 200 38.1 28.9 59.5 26.4

NOTES: CON, DAH and NIA identifies Congo, Dahomey and Nigeria laterite soils respectively.
Graphs in Figure Al-1 show complete particle size distribution of laterite soils.
 TABLE II

Compaction Characteristics and Strength of African Laterite Soils

Compaction Method: AASHO T 180-A

Volume Unconfined
Water Dry Air Void Air and Compressive
Soil Content Density Content Water Strength
percent lbslcu ft percent percent psi
CON-1 9.2 116.2 13.7 30.8
9.7 121.7 8.7 27.6
10.8 125.1 4 .O 25.5
1 1.4 126.6 1.6 24.7
13.2 123.2 0.6 26.7
14.3 120.2 0.9 28.4
16.4 112.9 3.1 32.8
CON-2

DAH-3

NIA- 1

Unconfined compressive strength measured at rate of loading of 0.05 incheslmin.

Optimum water contents: CON-1 = 11.4 percent; CON-2 = 9.7 percent; rnDAH-3 = 15.5 percent and NIA-1 = 9.2 percent.
 -
BgL a o e e m - m m m e m
m -
m m
m mm mm r rm-mz
s ,- g
E
+ C U C U M ~ +--mm - - + +

=e3.ZX v,mooo o o o m o o m m o ~
m o m 0 0
%
sga
Z ~
m m e d m mar-vie O \ O N \ D
~ ~ S j r -m
r -a
mwm o w m m a
* (U - rr-em
BSjmvln

Q)m *
-go msa sg& 's.?1v!n
> C a
%YO???
- ~ r - o o ma e m v l a
V!V!h!O?v!
m m - - m
M C U C U ~ C U N ~ C U N CeUm m m m
JIVv!Y!Y
mr-ar-m
C U C U C U ~ ~

z**
>.a0 P50 O$
g 0 \ 4 \ s y q o(?(?yq oq\s?\4Gq
m
*
m m m m zmviee o
\4qPj?T
CU r - N - - d-or-aa
6 0

*=I s!?\4c!9 ?v!v!* *c'!c!?"! " ? " ! 9 " !

e m m o m 2 a b m m a - m e - ir-ormvi
$'P$ * - m N N
- - - * *
PJNNNflm - - - -
- - * * * 22222
B5
* * & d

* *
$ *8 5
-+
r
+ S e:
g ~=~~~
q ~ q y q
? d q ~ q? \ q q q q
r - m m*o + *
*V*) *
\ O*b *
W
q q y y q
bmorz=

.-0
.
I
&

i
Fn 0
U
 TABLE IVA

Effect of Binder Type and Concentration on Properties of Laterite Soil

IDENTIFICATION OF SOIL CON-1

Compactive Effort Gyrator Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder -
-
RC-800 MC-800
Residual Asphalt Content,' percent 5.O 6.5 5.O 6.5
Total Binder Content , I percent 6.2 8 .O 6.2 8 .O
Curing, and Testing Condition A2 Compacted Immersed Compacted Immersed Compacted Immersed
- Compacted - Immersed
-
Mixing Water Content, percent 12.1 11.5 11.8 11.5 11.6 12.0 11.6 11.4
Volatile Content After Compaction, percent --
10.9 10.9 10.8 10.7 11.6 11.4 11.4 11.2
Wet Density, lbs per cu ft 130.0 130.0 126.4 126.1 128.4 128.4 125.9 1237-
Dry Density, lbs per cu ft 112.2 112.1 107.7 107.5 110.2 110.3 106.8 106.8
--
Air Voids, percent 4.7 4.7 6 -0 6.3 5.2 5.3 5.7 6.2
Marshall Stability, lbs 823+3 653t3 47 5 400+3 593 440+3 405 365+3
Marshall Flow, 0.01" 32 40+3 37 40+3
Volatile Content After Testing. uercent 11.6 12.4 11.4 12.3
~ b s o r b e dWater, percent 0.76 -
0.7
Retained Stability, percent 79.3 84.2 74.2 90.1

NOTES: Binder and residual asphalt contents based on weight of dry soil.
See text for description of curing and testing condition A.
3 ~ l a s t i cmixtures; Marshall stability taken at flow value of 40.
 TABLE IVB

Effect of Binder Type and Concentration on Properties of Laterite Soil

IDENTIFICATION OF SOIL CON-2

Compactive Effort
Type of Binder RC-800
-
Gyrator Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
MC-800
Residual Asphalt Content,' percent 5O
. 6.5 5.0 7 6.5
Total Binder Content,' percent 6.2 8 .O 6.2 8 .O
Curing and Testing Condition A2 Compacted Immersed Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 9.7 9.6 9.5 9.4 9.7 9.6 9.6 9.6
Volatile Content After Compaction, percent 9.4 9.3 9.3 9.3 9.6 9.6 9.4 9.4
Wet Density, lbs per cu ft 131.3 131.2 129.5 129.4 130.7 130.7 128.6 128.3
Dry Density, lbs per cu f t 114.8 114.8 111.8 111.7 114.0 114.1 110.7 110.7
A r Voids, percent 4.9 4.9 4.9 5 .O 5.2 5.2 5-6 5.7
Marshall Stability, lbs 563 568 48S3 443+3 423 408 320 36S3
Marshall Flow, 0.01 " 40 50 50t3 50t3 33 40 34 Sot3
Volatile Content After Testing, percent 9 -4 9.7 9.3 10.0 9.3 9.9 9.6 10.2
Absorbed Water, percent 0.4 0.4 0.4 0.4
Retained Stability, percent 100.9 9 1.3 96.5 114.1

NOTES: 'Binder and residual asphalt contents based on weight of dry soil.
See text for description of curing and testing condition A.
3 ~ l a ~ tmixtures;
ic Marshall stability taken at flow value of 50.
 TABLE IVC

Effect of Binder Type on Properties of Laterite Soil

IDENTIFICATION OF SOIL DAH-3

Compactive Effort Gyratory Compactor: 3 0 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800 MC-800
Residual Asphalt Content ,l percent 5 .O 5 .O
Total Binder Content,l percent 6.2 6.2
Curing and Testing Condition A2 Compacted Immersed Compacted Immersed
Mixing Water Content, percent 15.6 15.6 15.5 15.3
Volatile Content After Compaction, percent 14.7 14.6 14.7 14.7
Wet Density, lbs per cu ft 126.7 126.6 126.1 126.1
Dry Density, lbs per cu ft 105.8 105.8 105.4 105.4
Air Voids, percent 3.3 3.5 3.8 3.8
Marshall Stability, lbs 1213 700+3 1060 6303
Marshall Flow, 0.0 1" 30 50+3 28
Volatile Content After Testing, percent 14.4 15.5 14.2 15.3
Absorbed Water, percent 0.9 0-9
Retained Stability, percent 57.7 P
59.4

NOTES: ' Binder and residual asphalt contents based on weight of dry soil.
See text for description of curing and testing condition A.
plastic mixtures; Marshall stability taken at flow value of 50.
 TABLE IVD

Effect of Binder Type on Properties of Laterite Soil

--
IDENTIFICATION OF SOIL NIA-1
Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800 MC-800
Residual Asphalt Content,' percent 5 .O 5.O
,'
Total Binder Content percent 6.2 6.2
Curing and Testing Condition A2 Compacted Immersed Compacted Immersed
Mixing Water Content, percent 9.4 9.5 9.4 9.4
Volatile Content After Compaction, percent 9.1 9.2 9.7 9.6
Wet Density, lbs per cu ft 135.2 135.3 135.O 135.4
Dry Density, lbs per cu ft 118.6 118.5 117.7 117.9
Air Voids, percent 5.7 5.7 5.3 5.5
Marshall Stability, lbs 545 388 443 375
Marshall Flow, 0.01 " 36 39 32 41
Volatile Content After Testing, percent 9.2 10.1 9.1 9.9
Absorbed Water, percent 0.6 0.6
Retained Stability, percent 71.2 84.7

NOTES: 'Binder and residual asphalt contents based on weight of dry soil.
2See text for description of curing and testing condition A.
 TABLE VA

Effect of Curing on Properties of Laterite Soil

IDENTIFICATION O F SOIL CON-1

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder I
RC-800
Residual Asphalt Content ,I percent 6.5
Total Binder Content,' percent 8O
.
Curing Condition2 A B C
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 11.8 11.5 11.8 12.0 1 1.3 11.6
Volatile Content After Compaction, percent 10.8 10.7 10.9 10.9 10.8 10.8
Wet Density, lbs/cu ft 126.4 126.1 127.3 127.3 127.4 127.6
Dry Density, lbslcu ft 107.7 -
107.5
-
108.4 108.5 108.6 108.7
Air Voids, percent 6 .O 6.3 5.2 5.2 5.3 5.1
Marshall Stability, lbs 475 400+3 1793 1158 4565 3580
Marshall Flow, 0.01" 38 40+3 16 27 9 12
Volatile Content After Testing, percent 11.1 12.0 5.7 8.4 0.3 1.7
Absorbed Water, percent -
1 .O - 2.6 - 2.1
Retained Stability, percent - 84.2 - 64.6 - 78.4

NOTES: ' Binder and residual asphalt content based on weight of dry soil.
For test condition A specimens mixed, compacted and immersed a t approximately optimum (AASHO T 180-A)water
content. Conditions B and C involve specimen drying before immersion. See text for detailed description of testing
conditions.
3 ~ l a s t i cMixtures: Marshall stability taken at flow value of 40.
 TABLE VB

Effect of Curing on Properties of Laterite Soil

IDENTIFICATION OF SOIL CON-2

Compactive Effort Gyratory Compactor: 3 0 gyrations, one degree angle, 100 psi ram pressure
Type of Binder --
RC-800
Residual Asphalt Content ,l percent 6.5
Total Binder Content,' percent
Curing Condition2 C
Testing Condition2 Compacted Immersed
Mixing Water Content, percent 9.7 9.5
Volatile Content After Compaction, percent 8.9 9 .O
Wet Density, lbs/cu ft 130.4 130.5
Dry Density, lbs/cu ft 113.1 112.9
Air Voids, percent 4.7 4.5
Marshall Stability, lbs 401 3 3630 -
Marshall Flow, 0.0 1" 9 11
Volatile Content After Testing, percent 0.3 0.8
Absorbed Water, percent - 0.9
Retained Stability, percent

NOTES: ' Binder and residual asphalt content based on weight of dry soil.
For test condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water
content. Conditions B and C involve specimen drying before immersion. See text for detailed description of testing
conditions.
3 ~ l a s t i Mixtures:
c Marshall stability taken at flow value of 50.
 TABLE VC

Effect of Curing on Properties of Laterite Soil

IDENTIFICATION O F SOIL DAH-3

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800
Residual Asphalt Content,' percent 6.5
Total Binder Content ,l percent 8.O
Curing Condition2 A B C
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 15.2 15.5 15.4 15.7 15.1 15.1
Volatile Content After Compaction, percent 14.2 14.2 14.6 14.5 14.0 13.9
Wet Density, lbs/cu f t 124.2 124.2 124.2 124.4 124.1 124.1
Dry Density, lbs/cu f t 102.9 102.8 102.6 102.8 102.9 103.1
~ i Voids,
; bercent.
- - lbs
Marshall Stability. 1253 488+ 2875 705 4980
Marshall Flow, 0.0 1" 33 10 35 10 15
Volatile Content After Testing, percent 14.5 15.9 7.3 14.6 0.5 6.6
Absorbed Water, percent 1.O 7.3 7.5
Retained Stability, percent P
38.9 24.5 41.4

NOTES: 'Binder and residual asphalt content based o n weight of dry soil.
r condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A)water
2 ~ o test
content. Conditions B and C involve specimen drying before immersion. See text for detailed description of testing
conditions.
3Plastic Mixtures: Marshall stability taken at flow value of 50.
 TABLE VD

Effect of Curing on Properties of Laterite Soil

IDENTIFICATION OF SOIL NIA-1

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC -800
Residual Asphalt Content ,l percent 6.5
Total Binder Content ,l percent 8 .O
Curing Condition2 A B C
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 9.4 9.7 9.6 9.6 9.8 9.7
Volatile Content After Compaction, percent 8.7 8.7 9 -3 9.2 9 .O 9.1
Wet Density, lbs/cu ft 132.1 132.8 133.4 133.5 133.5 133.6
Dry Density, lbs/cu ft 114.7 115.8 115.2 115.4 115.6 115.6
Air Voids, percent 6.8 6.3 5.3 5.3 5.6 5.3
Marshall Stability, lbs. 395+3 3 18+3 1578 1238 4850 4415
Marshall Flow, 0.01 " 50+3 17 25 11 14
Volatile Content After Testing, percent 9.2 10.1 4.3 5.5 0.3 0.8
Absorbed Water, percent 0.5 1.O 0.8
Retained Stability, percent 80.5 78.5 91 .O

NOTES: Binder and residual asphalt content based on weight of dry soil.
2 ~ o testr condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water
content. Conditions B and C involve specimen drying before immersion. See text for detailed description of testing
conditions.
3 ~ l a s t i Mixtures:
c Marshall stability taken at flow value of 50.
 TABLE VIA

Effect of Curing and Binder Concentration on Properties of Laterite Soil

- -

IDENTIFICATION OF SOIL --- .

I CON-1
Compactive Effort -- Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
T v ~ of
# A
e ~inder- -
RC-800
Residual Asphalt Con tent ,l percent
-
Total Binder Content,' percent
Curing Condition2 A B
Testing Condition2 - - -
compacted Immersed Compacted
- -. --
Immersed
pp
Compacted Immersed Compacted 1 Immersed
Mixing Water Content, percent 12.1 11.5 ~
1 1.7
-- -
-
1 1.7
Volatile Content After compaction,
- . -
percent 10.9 --
10.9 10.9 10.9
m
o Wet Density, lbs/cu ft - i 30 .oP 130.0 129.7 129.7
Dry Density, lbs/cu ft 112.2 ii2.i -
111.9 112.0
Air Voids, percent - --
4.7 - 4.7- -- 4.9 4.9
Marshall Stability, lbs -- -
A --
823t3 65 3 t 3 2425 918
Marshall Flow, 0.0 1" - .- - - -- -- - -- - 40+3- -- -
40t3 14 33
Volatile Content After Testing,- percent 11.4 1 1 .9 5.2 10.0
Absorbed ~ a c r ; ~ e r c e n t
Retained Stability, percent

NOTES: ' Binder and residual asphalt content based on weight of dry soil.
For Condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water content. Condition B involves speci-
men drying before immersion. See text for detailed description of testing conditions.
3 ~ l a s t i cMixtures: Marshall stability taken at flow value of 40.
 TABLE VIB

Effect of Curing and Binder Concentration on Properties of Laterite Soil

IDENTIFICATION OF SOIL CON-2

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800
Residual Asphalt Con tent ,l percent 5 .O 6.5
Total Binder Content ,l percent 6.2 8O
.
Curing Condition2 A B B
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 9.7 9.6 9.6 9.6 9.5 9.4 9.6 9.6
Volatile Content After Compaction, percent 9.4 9.3 9.2 9.2 9.3 9.3 9.1 9.1
Wet Density, lbs/cu f t 131.3 131.2 132.4 132.4 129.5 1 29.4 130.6 130.6
Dry Density, lbs/cu ft 114.8 114.8 115.9 116.0 111.8 111.7 113.0 112.9
Air Voids, percent 4.9 4.9 4.4 4.4 4.9 5O. 4.3 4.3
Marshall Stability, lbs 563 568 201 8 89 3 485+3 443+3 1475 1140
Marshall Flow, 0.0 1" 40 50 19 25 50+3 50+3 18 23
Volatile Content After Testing, percent 9.4 9.7 4.4 6.7 9.3 10.0 4.3 5.4
Absorbed Water, percent - 0 -4 - 2.5 0.4 - 1.2
Retained Stability, percent - 100.9 - 44.3 - 91.3 - 77.3

NOTES: ' Binder and residual asphalt content based on weight of dry soil.
2
For condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water content. Condition B involves speci-
men drying before immersion. See text for detailed description of testing conditions.
3Plastic Mixtures: Marshall stability taken at flow value of 50.
 TABLE VIC

Effect of Curing and Binder Concentration on Properties of Laterite Soil

IDENTIFICATION OF SOIL DAH-3

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800
Residual Asphalt Content ,l percent 5 .O 6.5
Total Binder Content ,l percent 6.2 8 .O
Curing Condition2 A B A B
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 15.6 15.6 15.7 16.2 15.2 15.5 15.4 15.7
Volatile Content After Compaction, percent 14.7 14.6 14.7 14.8 14.2 14.2 14.6 14.5
Wet Density, lbs/cu ft 126.7 126.6 126.6 126.6 124.2 124.2 124.2 124.4
Dry Density, lbs/cu ft 105.8 105.8 105.7 105.7 102.9 102.8 102.6 102.8
Air Voids, percent 3.3 3.5 3.4 3.4 4.3 4.4 4.1 3.9
Marshall Stability, lbs 1213 700+3 3588 718 1253 488+3 2875 705
Marshall Flow, 0.0 1" 30 15 26 33 10 35
Volatile Content After Testing, percent 14.4 15.5 7.6 14.4 14.5 15.9 7.3 14.6
Absorbed Water, percent 0.9 7 .O 1.O 7.3
Retained Stability, percent 57.7 20.0 38.9 24.5

NOTES: ' Binder and residual asphalt content based on weight of dry soil.
or test condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water content. Condition B involves
specimen drying before immersion. See text for detailed description of testing conditions.
3 ~ l a s t i Mixtures:
c Marshall stability taken at flow value of 50.
 TABLE VID

Effect of Curing and Binder Concentration on Properties of Laterite Soil

IDENTIFICATION OF SOIL NIA-1

Compactive Effort Gyratory Compaction: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800
Residual Asphalt Content,' percent 5 .O 6.5
Total Binder Content , l percent 6.2 8 .O
Curing Condition2 A B A B
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 9.4 9.5 9.7 9.7 9.4 9.7 9.6 9.6
Volatile Content After Compaction, percent 9.1 9.2 9.3 9.3 8-7 8.7 9.3 9.2
Wet Density, lbs/cu ft 135.2 135.3 135.0 134.7 132.1 132.8 133.4 133.5
Dry Density, lbs/cu f t 118.2 118.5 118.2 117.9 114.7 115.3 115.2 115.4
Air Voids, percent 5.7 5.7 5 -8 5.9 6.8 6.3 5.3 5.3
Marshall Stability, lbs 545 388 2438 1430 395+3 3 18+3 1578 1233
Marshall Flow, 0.0 1" 36 39 18 26 50+3 50+3 17 25
Volatile Content After Testing, percent 9.2 10.1 4.4 6O. 9.2 10.1 4.3 5.5
Absorbed Water, percent 0.6 1.9 0.5 1.O
Retained Stability, percent 71.2 58.7 80.5 78.5

NOTES: Binder and residual asphalt content based on weight of dry .soil.
For test condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water content. Condition B involves
specimen drying before immersion. See text for detailed description of testing conditions.
3 ~ l a s t i Mixtures:
c Marshall stability taken at flow value of 50.
 TABLE VII

Effect of Mixture Aeration on Properties of Laterite Soil

- - - - --- -

IDENTIFICATION OF SOIL I CON- 1

~ o m ~ a c t i vEffort
e -- -
rato or^ Compactor: 3 0 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800
Residual Asphalt content,' percent
Total ~inder-content? percent
Curing Condition2
Testing Condition2

Wet Density, lbs/cu ft -- -

130.0 130.0 116.2 116.2
Dry Density, lbs/cu ft
-
- --
112.2
-- - - -
112.1 104.1 104.1
Air Voids, percent 4.7 4.7 18.7 18.7
Marshall Stability, lbs - 823t3 653+3 1155 Fell Apart
Marshall Flow, 0.01 " 40+3 " 40t3 21 Fell Apart
Volatile Content After Testing, percent 11.4 11.9 6.8 Fell Apart
Absorbed Water, percent - 0.6 - Fell Apart
I I I

Retained Stability, percent - 79.3 - 0

NOTES: '
Binder and residual asphalt content based on weight of dry soil.
2 ~ otest
r condition A specimens mixed, compacted, and immersed at approximately optimum (AASHO T 180-A) water
content. Condition D involves mixture aeration before compaction. See text for detailed description of testing condi-
tion.
3Plastic Mixtures: Marshall stability taken at flow value of 40.
 TABLE Vlll

Effect of Mixing Water Content, Mixture Aeration and Curing on Properties of Laterite Soil

5
Compactive Effort
DAH-3
Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder RC-800
Residual Asphalt Content , I percent 6.5
Total Binder Content ,l percent 8.O
Curing Condition2 A C F
Testing Condition2 Compacted Immersed Compacted Immersed Compacted Immersed
Mixing Water Content, percent 15.2 15.5 15.1 15.1 18.2 18.3
volatile Content After Compaction, percent 14.2 14.2 14.0 13.9 13.8 14.2
Wet Density, lbs/cu ft 124.2 124.2 124.1 124.1 124-6 124.3
Dry Density, lbs/cu ft 102.9 102.8 102.9 103.1 103.6 103.O
Air Voids, percent 4.3 4.4 4.7 4.8 4.4 4.4
Marshall Stability, lbs 1253 488 t3 4980 2060 4318 25 50
Marshall Flow, 0.01 " 33 50t3 10 15 8 10
Vdlatile Content After Testing, percent 14.5 15 -9 0.5 6.6 0.9 6.5
Absorbed Water, percent 1.O 7.5 6-7
Retained Stability, percent 38.9 41.4 59.1

NOTES: Binder and residual asphalt content based on weight of dry soil.
2For test condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water
content. Condition C involves specimen drying before immersion. Condition F involves mixture aeration before
compaction and specimen drying before immersion. See text for detailed description of testing condition.
3 ~ l a s t i Mixture:
c Marshall stability taken at flow value of 50.
 TABLE IXA

Effect of Mixing and Compaction Water Content on Properties of Laterite Soil

IDENTIFICATION OF SOIL CON- 1

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram
- pressure
Type of Binder

Curing Condition2 A E
Testing Condition2 Compacted Immersed Compacted Immersed
Mixing Water Content, percent 11.6 12.0 9.9 10.0
Volatile Content After Compaction, percent 11.6 11.4 9.4 9.4
Wet Density, lbs/cu ft 128.4 128.4 128.3 128.6
D_rv Density, lbs/cu ft 110.2 110.3 112.1 112.4
Air Voids, percent 5 -2 5.3 7.4 7.2
Marshall Stability, lbs 593 400t3 1190 263+3
Marshall Flow, 0.0 1" 32 40t3 20 40t3
Volatile Content After Testing, percent 11.6 12.4 9.5 13.3
Absorbed Water, percent
-.
-
0.7 - 3.3
Retained Stability, percent - 74.2 - I 22.1

NOTES: 'Binder and residual asphalt content based on weight of dry soil.
or test condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water con-
tent. For test condition E specimens mixed, compacted and immersed at approximately two percent below optimum water
content. See text for detailed description of testing conditions.
3 ~ l a s t i Mixtures:
c Marshall stability taken at flow value of 40.
 TABLE IXB

Effect of Mixing and Compaction Water Content on Properties of Laterite Soil

IDENTIFICATION OF SOIL CON-2

Compactive Effort Gyratory Compactor: 30 gyrations, one degree angle, 100 psi ram pressure
Type of Binder MC-800
Residual Asphalt Content,l percent 5 .O
--
.

Total Binder Content ,l percent -

6.2
Curing Condition2 A E
Testing Condition2 Compacted Immersed Compacted Immersed
Mixing Water Content, percent 9.7 9.6. 7.4 7.7
Volatile Content After Compaction, percent 9.6 9.6 7.9 7.8
Wet Density, lbs/cu ft 130.7 130.7 132-4 132.4
Dry Density, lbs/cu ft 114.0 114.1 117.3 117.4
Air Voids, percent 5.2 5.2 -- - 5.7 5.9
Marshall Stability, lbs 423 408 635 558
Marshall Flow, 0.0 1" 33 40 23 32
Volatile Content After Testing, percent 9.3 9.9 7.5 8.2
Absorbed Water, percent - 0.4 - 0.5
Retained Stability, percent - 96.5 - 87.9

NOTES: Binder and residual asphalt content based on weight of dry soil.
For test condition A specimens mixed, compacted and immersed at approximately optimum (AASHO T 180-A) water
content. For test condition E specimens mixed, compacted and immersed at approximately two percent below optimum
water content. See text for detailed description of testing conditions.
 APPENDIX FOR CHAPTER 12

DETERMINATION OF LOAD-DEFLECTION
CHARA CTERISTICS OF PA VEMENT SYSTEM
B Y MEANS OF THE DEFLECTION BEAM

PROCEDURE:
toe to drop slowly to the test surface. Check the probe beam
Creep Deflection and Measurement of and dial gauge stem clearances and make adjustments of the
the Deflection Basin. rear support, if necessary.
1. Select and mark location to be tested in both inner and 5. Turn Buzzer switch to the "on" position. Set the face
outer wheel paths. of the dial micrometer t o zero.
2. With the loading vehicle (Fig. 133) aligned paralled to 6. Move the loading vehicle straight ahead (noting if there
the longtitudinal axis of the road way, center its rear wheels is an initial backward movement of the dial) a distance of one
on the test points. foot and record rebound. Repeat for distances of 1.5, 2.0,3.0,
3. From the rear of the loading vehicle insert the probe 4.0, 5 .0,6.0 and 20 feet. See Figs. 135 and 136.
beam (Fig. 134) between the dual tire and place the beam on 7. The maximum rebound deflection is calculated as the
the test surface so that the beam toe is approximately centered difference from the initial reading (zero) and the final reading
between the tires and on line with the rear axle. at 20 feet multiplied by 2.
4. Hold the rear end of the probe beam down against the 8. The slope of the deflection curve is obtained by
beam guide, then release the beam lock and allow the beam plotting the rebound deflections at the various distances of
load to toe.

. 'c.9,
Fig. 133. Loading Device. Two Axle Truck with rear dual-tired
wheels each loaded to 9,000 lbs.
Charge appliquie: Camion a deux essieu avec neus arriers
jumelis, chacun chargi a 9000 livres.

Fig. 134. Deflection Beam. Similar to Benkelman Beam used

in the States. This Beam is a design of the Road Research
Laboratory, United Kingdom.
Poutre de diflection: Semblable a la poutre de Benkelman
utilisie aux ETATS-UNIS. Cette a i t 6 congue par le Road
Fig. 135. Position of Beam for measuring maximum rebound Fig. 136. Rebound deflections measured at various distanc
deflect ion. from the point of loading.
Position de la poutre pour la mesure de la deflection Deflection maximum mesurie B differentes distances du poi
maximum. de chargement .
 PA VEMENT CONDITION SUR VEY

TABLE 54. TRAFFIC DATA FOR ROAD SECTIONS TESTED.

'TRAFFIC CATEGORIES
TP S E C AGE
 FA CTORIAL A N A L YSIS

No experimental measurements are exactly reproduceable. nb

As the sensitivity of measuring instruments increases, the sb = nanc (Y'J - 9' variance due to b (6)
magnitude of the variations between measurements will J
,decrease, but some discrepancy will always remain. ne
Variations in experimental data arise from two sources. sc = nanb C (yfl$ - Q2 variance due to c
(7)
k
Those due t o identifiable factors, known as "assigned causes"
The variance due t o error, Sz, is obtained by difference.
and .those due t o non-identifiable factors, known as "chance
The quantities Yil, Yjl' and Y ~ " are the mean values of
causes." As experiments become increasingly elaborate, a
effects a, b and c measured at their ith, jth and kth level
greater percentage of variations can be attributed t o "assigned
respectively.
causes" with less incidence of variations from "chance"
Since the standard error of each effect is required, the
sources. Factorial analysis is a method used t o distinguish
number of degrees of freedom are determined.
which of the hypothesized factors have an effect on the
For each factor:
measured variable.
DFM = Nm-1, where Nm equals the number of (8)
MATHEMATICAL ANALYSIS levels for effect M.
While for the error,
A variable y has been measured as a function of three 3
distinct variables a, b and c. Various values of a, b and c have D F E R R = (N - 1) - C D F M , where N represents (9)
M the total number of
been used and the value of y has been recorded for all
combinations. data points.
The hypothesis t o be tested: is y a function of a, b and c, Subsequent division of each variance by its number of
expressed as: degrees of freedom yields the standard error squared for each
Y = f(a,b,c) . . . . . . (1) effect, otherwise known as its mean square.
A simple model is constructed such that any value of y is F ratios are generated by dividing the mean square of each
equal t o the mean value of y for all measurements, plus effect by the mean square of the error.
deviations due to the main effects a, b and c and some
experimental error. TESTS OF SIGNIFICANCE
Outcomes of the F test via factorial analysis are identical t o
those found in discriminant analysis and their interpreta-
where tion is as follows:
Yijk = any experimental measurement of y (1) When the computed F value is less than the statistical
Y = overall y mean F values at the one and five percent significance
levels, the factor has no effect on the variable y.
Yai = deviation of y due to factor a at level
i (1 < i S n a ) (2) If the computed F value is between the one and five
= deviation of y due t o factor b at level percent statistical F values, the factor may have an
Ybj effect on the variable y but its significance is
j (l<j<nb)
inconclusive.
= deviation of y due t o factor c at level
Ycj (3) A computed F value greater than the statistical F
k ( 1 <k<na)
values at both the one and five percent levels
2 ijk = experimental error indicates that the factor influences the value of Y.
It is assumed that the total variance is equal to the sum of PROGRANIMING
the individual variances due t o factors a, b and c and the
Systematic reduction of factorial analysis data was pro-
experimental error.
cessed by means of a computer program. Written in FOR-
STotal = Sa + S b + S c + S z . . . (3) TRAN I1 programming language, it operates on an IBM-1620
with the mathematical definitions I1 computer with 60K storage, card reader and typewriter
output. No additional peripheral equipment is required. A
listing of the program is given at the end of this section.
The program is designed t o handle three factors at either
na two, three or four levels, depending upon user choice. It is
sa = nbnc & (Y 'j - 9' variance due t o a (5) restricted in the sense that all three factors must be handled
1
simultaneously with at least two levels each.
 An abbreviated flow-chart of the program is as follows: LEVEL 1 LEVEL 2 LEVEL 3 LEVEL 4
CBR 60-70 70-80 80-90
LL 10-15 15-20 20-25 25-30
1 OMC 6-9 9-12
READ NO. OF LEVELS OF
The data is recorded in a factorial table given on the input
EACH EFFECT
sheet, at the end of this section. Effect one refers t o CBR

47
I
READ DATA -
values, effect two t o liquid limit and effect three t o optimum
moisture content.

Input:
Included w i t h the input cards are the following:
CALCULATE MEANS FOR EACH
(1) Description of Effects Cards: - Factors one, two and
EFFECT AT EACH LEVEL ALSO three are alphabetically identified by entering the
OVERALL MEAN initials CBR, LL and OMC in columns one, five and
nine respectively.
1 (2) No. of Levels Card - Three levels of CBR, four values
of liquid limit and two values of OMC are indicated
CALCULATE VARl ANCES AS
by entering the numbers three, four and two in
GIVEN IN EQUATIONS 4-7
columns two, four and six.
I (3) Data Block - once the data has been recorded in the
factorial table on the input sheet, it is typed directly.
COMPUTE DEGREES OF FREEDOM,
MEANS SQUARES AND F RATIOS Output:
An abbreviated output is produced in which, for each
effect, the number of degrees of freedom (DF), the variance
OUTPUT RESULTS (SS), mean square (MS) and F ratio (F) associated with it
appears. No F value for error (ERR) is given.
To use the results, the F ratio of each factor is compared
with tabulated F values found in any standard set of statistical
tables, for one and five percent significance levels. Two
different degrees of freedom are needed when using the F test.
APPLICATION One is provided by the factor itself, the other by the error
(ERR).
Consider as an example, the test t o see which factors, either
CBR, liquid limit (LL) or optimum moisture content (OMC)
BIBLIOGRAPHY
of the sub-base material have an effect on the deflection of a
flexible highway pavement. (1) Johnson, P. O., 1949. Statistical methods in research.
Different values of deflection are measured at three levels Wiley, 377 p.
of CBR, four levels of liquid limit and two levels of OMC. In (2) Quenouille, M. H., 1953. The design and analysis of
all, twenty-four measurements are made, at the following experiments. Charles Griffin & Company, Ltd., Lon-
parameter combinations: don. 356 p.

"0808
C
C FACTOR AlVALYSlS
C
DIMENSION BOMC%4nl BLL%4nl BCBR%4nl TABLE%64,4n
DIMENSION NAME%4n
C
C l NPUT
C
READ 503, NAME
READ 500,NOMC,N LL,NCBR
NCARDS#NCBR*NLL
DO 600 J#1 ,NCARDS
READ 501, %TABLE%J,Kn,K#1,NOMCn
TOT W
TOTDW
CBRDEVW
DEVLLW
DEVOMCW
DO 400 J#1,4
BOMC%Jn#O
BLL%JnW
BCBR%JnW

OMC MEANS

L L MEANS

LSTA RT#1
LSTOPWCB R
DO 2 J#1 ,N L L
DO 3 K#LSTART, LSTOP
DO 3 L#1 ,NOMC
BLL%Jn#TABLE%K,Ln/FLOAT%NCBR,NOMCn&BLL%Jn
LSTART#LSTART&NCB R
LSTOP#LSTOP&NCB R

CBR MEANS

LSTA RT#1
LSTOP#NCBR *%NLL-1 n&1
DO 4 J#1 ,NCB R
DO 5 K#LSTART,LSTOP,NCBR
DO 5 L#1 ,NOMC
BCBR%Jn#TABLE%K,Ln/FLOAT%NOMC,NLLn&BCBR%Jn
LSTA RT#LSTA RT&1
LSTOP#LSTOP& 1

GRAND MEAN

TOTAL DEVIATIONS

OMC DEVIATIONS
CBR DEVIATIONS

ERROR DEVIATIONS
DEVERR#TOTD-%DEVOMC&CBRDEV&DEVLLn
DFLLWLL-1
DFCBRWCBR-1
DFOMC#NOMC-1
DFERRWLL"NCBR"N0MC-1 -%NLL&NCBR&NOMC-3n
SMOMC#DEVOMC/DFOMC
SMCBR##CBRDEV/DFCBR
SMLLWEVLLIDFLL
SMERR#DEVERR/DFERR
FOMC#SMOMC/SMERR
FCBR#SMCBR/SMER R
FLL#SMLL/SMERR
NDFLLMILL-1
NDFCBR#NCBR-1
NDFOMCWOMC-1
NDFERR#DFERR

OUTPUT

PRINT 502
PRINT 100,NAME%1n,NDFOMC,DEVOMC,SMOMCIFOMC
PRINT 1OO,NAME%2n,NDFLL,DEVLL,SM LL,FLL
PRINT 100,NAME%3n,NDFCBR,CBRDEV,SMCBR,FCBR
PRINT 1OO,NAME%4ntNDFERR,DEVERRtSMERR
FORMAT%GX,A4,1 15,8XtE14.7,2X,E14.7,2XtE14.7n
FORMAT%312n
FORMAT%~F~.OU
FORMAT%///,4X,8HVARIABLE,1 1Xt2HDF,14X,2HSS,14X,2HMS,14X,1 HF,/n
FOR MAT%4A4n
STOP
END

FUNCTION FLOAT%NOI\IE,NTWOn
F LOATWONE"NTW0
RETURN
END
 F A C T O R I A L ANALYSIS

DE SCRlPTlON OF EFFECTS CARD (EFFECT I , EFFECT 2,EFFECT 3 )

No OF LEVELS ( EFFECT I, EFFEOT 2 , EFFECT 3 )

b--.-- ------+
-t
EFFECT 1 --

1- LEVEL I -f-- LEVEL 2 LEVEL 3 LEVEL 4 ----

7 M -
L l I I 1 I I I I l I I 1111l I I I Ill1 I
L
E k L 2
W-
I I I I I I I I I 1 1111l 1 1 I 11111 1 \ 1 11
L
2
Y
L
w -
,
L ~ 3 [ I I I I I I I I ] ( I I I I ) I I I 1 1 1 1 1 - 1 1 1 1 1 1 L1
L4 I I I I I 1 1 1 I I I I I l l I l l 1 I I
+ 1 1 1 1 1 1 1 1 1 1 1 1 [ 1 , 1 1 1 1 1 I 1 1 1 1 1 I I I 1 I
L to-
o
L 2 [ [ ( 1 1 1 1 [ [ [ I I [ ( I 111 1 1 I I I I I I l l l l l l ;
E u-
L kW --
~ 3 1 1 I I I ) I I)( 1 1 1 1 I I I 1 I I I I I 1 1 1 1 1 1 :
1 1 1 1 1 1 1 1 1 1 : I I I 1 I'
1 1 1 1 1 1 1 1 1 1 1 1 1 ] I l l I I
E
V
E
L L 2
W-
I I I I I I I I I I I I I I I I I I [ I I I I 1 I I I ) I I [ 1 ] ] ]
L
4 ~-
W ' I I I I I ~ ~ ~ ~ ~ I I I I ~ ~ ~ ~
3 ~
I '4 I 1 I I I I I I I I ) I ( I I I I I I I ( I ] I I ( I I I I I ( I ( ]
 TYPICAL COMPACTION CUR VES

140

130

120

110

100

90

-
9
80

- 70
n
m 60
U
50
AASHO STANMRO
40

30

20

10

-
u
D
a - 140

'D-
-
LL

--
-
z130

..
F
t
c
- $120
W

-
n.
+
K
0 110
0-

"- 100
2 4 15 8 10 I.? 14 16 W 20 22 24 26 28 30 32 34 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
MOISTURE CONTENT % MOISTURE CONTENT X

T P IV-2 SAMPLE C
150 150

140 140

130 130

120 120

110 110

100 100

90 90

-Z 80 - 80

70 70
u
60 60
AASliO STANDARD AASHO STANDARD
50 50

40 40

30 30

20 20

-u
0
10
- 10

u
V 0 z 0

N-
-
LL
140
N
,140
LL
U
-
a
130 N-
U
-
a
130
* *
0
V)
0 'z
N- N- V)

a-
g I20 a.
g I20

9- * m- *0"
0" I10 I10

C- F-

100 100
2 4 6 B 10 12 14 16 I8 20 22 24 26 28 30 32 34 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
MOISTURE CONTENT O/o MOISTURE CONTENT %
I50

140

I30

I20

I10

100

90

80

70

60
AAPlO STANDARD
50

40

30 FELD DENSITY

20

I0

140

130

120

II0

00
2 4 6 8 10 12 14 16 Is 20 22 24 26 28 30 32 34 4 6 8 10 I2 14
' 16 18 20 22 24 26 28 30 32 34 36
MOISTURE CONTENT Vo MOISTURE CONTENT Vo

T P IV-6 SAMPLE 0 SAMPLE A

150

140

130

120

110

100

90

-8 80

70

60
A A P l O STANDARD A A W O STANDARD
50

40

30

20

-
0
0
10

0 0

140
N
u-

N-
-
O
P

130
>
0
N- U)

0.
X I20
?- >
g 110 --
i - 1 7
,
- -

-
rr/
a \
* I

00 -7 1 , I I - 1
4 6 8 I0 I2 14 16 18 20 22 24 26 20 30 32 3J 36 2 4 6 8 10 12 14 16 Is 20 22 24 26 28 30 32 34
MOISTURE CONTENT Vo MOISTURE CONTENT Vo
 110 FERRSOLS
w
AASHO MODlflED
100

90 A AS
* A
- 80 GHnNA S T D

70
AS YOU)ED

60
AAWO STANDARD
50

40 0 AS
SDAK

30

2 4 6 8 .O 12 14 16 M 20 22 24 26 28 30 32 34
MOISTURE CONTENT % MOISTURE CONTENT Yo

I50

140

130

120

II0

100

90

-a-' 80

70

60
AAWD STANDARD
50

40

30

20

-
0
10

0 0

- 140

z- IA

-
-
0

I30
*
0
=!
N- ln
m.
# I20

9- >
g 110

80 I
I

1 !
1 l! i
. ! 1
too
4 6 8 13 I2 14 16 18 23 22 24 26 28 30 32 34 2 4 cj H 10 I;? 14 16 K3 20 22 24 26 28 30 32 34
MOISTURE CONTENT 4b MOISTURE CONTENT Yo
 140

130

120

110

I00

90

- 80

- 70
LT
m 60
0
50
AASHO STANMRD
40

30

20

10

-
0
P 120
m

m
-
-
-
C
-; I10

t
-
UJ
-.
w 2100
W

-
n- t
I
X
90
0

80
'4 6 I3 10 I2 I4 16 18 20 22 24 26 28 30 32 34
MOISTURE CONTENT % MOISTURE CONTENT 46

TP 30 SAMPLE 8
150

140

130

120

110

100

90
a-"
80

(r 70

60

50

40

30

20

-u 10

S O

N
-
LL
O
140

-
-a 130
t
0
N- UJ

m-
g 120

'- i l l 0
C

100 , , , , , , , 8 8 . 8 , , 100
~ ;
,'
/ ! I 1
I

2 4 6 8 10 12 14 16 LB 20 22 24 26 28 30 32 34 :' 4 6 f3 10 12 14 16 I8 20 22 24 26 28 30 32 34
MOISTURE CONTENT 4b MOISTURE CONTENT O/.
 150

140

130

120

110

100

90
a-"
- 80

a 70

60
AAYlO STANDARD
AASHO STANMRD 50

40

30

20

-
0
10

'D 0

,
I40
LL
*-
N

-
O
0

130
t
2-
z
g 120
i-

*- g 1
0
C-

oo
2 4 6 8 I0 12 14 16 It) 20 22 24 26 28 30 32 34
4 6 R 10 12 I4 16 18 23 22 24 26 28 30 32 34 MOISTURE CONTENT 'Yo

MOISTURE CONTENT 'Ye

150

140

130

120

II0

100

90

- 80

70

60
AASHO STANDARD
50

40

30

20

-
0
0
10

s o

m -
C)

N
-
LL
140

m -
G-
-
O
P
130
t
C - o
N- V)

CO-
a.
;120

E- *g I10
O
C-

oo
4 6 8 15 12 I4 16 18 20 22 24 26 28 30 32 34 2 4 6 8 10 12 14 16 It) 20 22 24 26 28 30 32 34
MOISTURE CONTENT Ye MOISTURE CONTENT Yo
 150 140

140 130

130 120

120 110

110 KX)

I00 90

Z
90 -Z 80

80 - 70
u
70
m 60
0
60 50
AAIHO STANMRD
50 40

40 30

30 20

20 10

-
u
10 0

D 0
-
-7

N
-
LL
140 P
m - 140

-
O
a
-
a.
-
LL
-
0
t
!=
130

-
-
El30

N- v,
t
t_
0-
g I20 a. 2120
W

-
n- t
(L

0110

100
2 4 6 8 10 12 14 16 LB 20 22 24 26 28 30 32 34 4 6 8 10 I2 14 16 18 20 22 24 26 28 30 32 34
MOISTURE CONTENT 4b MOISTURE CONTENT Ye

TP 42 SAMPLE 0
140

130

120

110

I00

90

-E 80

- 70

a 60

U
50
A A W STANDARD
40

30

20

I0

-
u
P 120
m -
9-
-
LL

-
-2
t
I10

t I I
,- 500-
W
I
-
n- t
[I

0 90-
0

2 4 6 8 10 12 14 16 IB 20 22 24 26 28 30 32 34
8 0 ~
4 6 8 10 12 14 16 18 20 22 24
I ,
26
,
28
,
30
.
32
,
34
, _

MOISTURE CONTENT % MOISTURE CONTENT Y.

 APPENDIX FOR CHAPTER 16

EXPANSION PRESSURE OF SOILS

(References: AASHO Designation T-190

and Utah State Department of Highways,
Manual of Instruction, Part-8, Materials)
SCOPE bearing end shall rest on the adjustment plug.
This method of test covers the determination of the (2) Place a 112-inch diameter steel ball in the center
expansion pressure of treated and untreated soils for use as hole of the spring steel loading bar of the proving
base, subbase, or roadbed soil. ring assembly.
(3) Place the proving ring in the expansion pressure
APPARATUS: frame, and center the steel ball on top of the
Expansion pressure device with accessories. proving ring under the deflection gauge contact.
(1)
(2) Deflection gauge- A gauge with divisions of (4) Adjust the turntable until the proving ring gauge
0.0001 inch, and an Allen wrench. reads zero. Then set the adjustment plug so that
(3) Proving ring- A proving ring of known capacity the deflection gauge reads zero.
per 0.0001 inch deflection. (5) Increase the pressure on the deflection gauge, in
(4) Mold- A cylindrical-shaped steel mold, 5 inches in steps of 0.001 inch, from 0 to 0.01 inch. The
height with an inside diameter of 4 inches. deflection gauge should check the proving ring
(5) Compaction equipment - Mechanical equipment gauge within a tolerance of -+ 0.002 inch at every
suitable t o compact a soil specimen by means of a point.
2-inch diameter rammer, having a flat, circular face (6) If the gauges do not check within the above
weighmg 5.5 lb. The rammer shall have a free fall tolerance, loosen the top frame bar on the ex-
of 12 inches. pansion pressure device, and adjust the position of
Filter paper- Medium grade filter paper, 4 inches the shims between the frame and the spring steel
(6)
in diameter. bar until the desired check is obtained.
Spacer disk- A metal disk, preferably aluminum, (7) Remove the proving ring, and set the adjustment
(7)
2 112 inches in height with a diameter of 3 15116 plug by means of an Allen wrench until the
inches, suitable t o control the height of the molded deflection gauge is on minus 0.001 inch. (The
soil specimen. deflection gauge will read 0.009).

(8) Sample extruder- A jack, lever, frame, or other

SAMPLE PREPARATION
device that can be used t o extrude compacted soil
specimens from the mold. (1) Sieve from 15 to 25 kg. (depending on the type of
Balance- A balance with a capacity of 5000 gm. material under test) of the representative pul-
(9)
and sensitive t o 1 gm. verized soil over the No. 4 sieve. Discard the
material retained on the No. 4 sieve.
(1 0) Mixing equipment- Miscellaneous tools such as a
mixing pan, spoon, trowel, spatula, etc. (2) Select a representative test sample, weighing ap-
proximately 12000 gm.
(1 1) Straightedge- A 12-inch long, steel straightedge
with one beveled edge. (3) Determine the optimum moisture content of the
sample material in accordance with the standard
(1 2) Drying oven- A thermostatically-controlled oven
method of test for the Moisture-Density Relations
capable of maintaining a temperature of 230
of Soils.
degrees F. *9 degrees F.
(4) Determine the moisture content of the sample
(13) Sieve- A No. 4 sieve of woven wire-cloth
material.
construction with square openings, conforming to
the requirements of AASHO Designation M-92.
PREPARATION OF SPECIMENS
CALIBRATION PROCEDURE OF Each specimen shall be prepared as follows:
THE EXPANSION PRESSURE DEVICE (1) Select a representative sample, weighing from 1200
(1) Place the deflection gauge assembly in position on t o 1400 gm.,from the soil passing the No. 4 sieve.
top of the expansion pressure frame. The single (2) Thoroughly mix this sample with sufficient water
 to obtain the optimum moisture content. brass plate and rod facing up.
(3) Place the spacer disk and a filter paper in the mold, (5) Seat the perforated brass plate firmly on the
and form a specimen by compacting the prepared specimen with pressure applied by the fingers.
soil sample in the mold in three equal layers to give (6) Adjust the turntable until the deflection gauge
a total compacted depth of approximately 2 112 reads zero.
inches. Compact each layer by 25 uniformly (7) Read and record the height of the specimen for the
distributed blows from the rammer, dropping free densitv determination.
from a height of 12 inches above the elevation of
the soil. During compaction, the mold shall rest on (8) Pour approximately 200 ml. of water over the
specimen in the mold, and allow the device to
a uniform, rigid foundation such as is provided by
stand undisturbed for 24 hours or more.
a cube of concrete weighing not less than 200 lb.
(9) Read and record the deflection gauge reading at
(4) During compaction, obtain a sample, and de-
the end of the soaking period. If the water has
termine its "as compacted" moisture content.
drained through the specimen into the pan, pour
(5) Following compaction, carefully trim the com- this water back over the specimen, and allow it to
pacted soil even with the top of the mold by means begin percolating through the specimen.
of the straightedge.
(10) Record whether the water drained freely through
(6) Remove the spacer disk, and weigh the specimen the specimen.
and mold. Subtract the weight - of the mold to
obtain the wet density of the specimen. For each sosil sample to be tested, a total of three expansion
pressure determinations shall be made, as outlined above. The
TEST PROCEDURE average value obtained shall be the expansion pressure of the
soil.
(1) Clean all dust and foreign material from the spring
steel bar and adjustment plug on the expansion
CALCULATION
pressure device.
The expansion pressure shall be calculated as follows:
(2) Place the deflection gauge in position on the top
bar of the expansion pressure device. P = kd
(3) Place the perforated brass plate with rod on the where P = the expansion pressure in psi
bottom of the compacted specimen. k = the constant of the spring steel bar calibrated
(4) Place the filter paper on the turntable of the from the proving ring (psi. per 0.001 inch).
expansion pressure device, and place the mold with d = the deflection in 0.0001 inch shown by the
the compacted specimen on the turntable with the deflection gauge.

PROCEDURES FOR SAMPLE COMPACTION

AND TESTING WITH THE SWELL INDEX DEVICE

EQUIPMENT (Plates 7,8,9,10)

Swell Index Device Proving Ring and Top Bar where it will not be jarred
Compaction Hammer and Cylinder during compaction.
Knife and Straight Edge Place Compaction Ring, with letters CO at the top,
Water (squirt bottle) on the Base and so that the No. 1 on the Base is
Two Dry Porous Stones aligned with No. 2 on the Compaction Ring. Place
Brush (optional) Spacer Ring on Compaction Ring with letters 00 at
118 inch Dia. Pin the bottom (radial grooves are on top) and aligned
Teaspoon with the letters CO on the Compaction Ring. Note:
No. 10 Sieve (optional) In Swell Index Devices where the holes for the bolts
Wrenches attaching the Compaction and Spacer Rings to the
Base are perfectly symmetrical, this careful
alignment is not necessary. Insert the 3 Bolts and
PREPARATION FOR COMPACTION tighten firmly.
1. Disassemble Swell Index Device, with exception of 3. The sample should not contain any stones or dried
Rods which can remain screwed into the Base. Place clay lumps that will not pass through a No. 10 sieve
 (-07 inch diameter). of the sample. When the sample is almost level, do
the final leveling by drawing a metal straight edge
COMPACTION over the sample. The final surface must be level and
1. The compaction procedures t o follow are shown firm ("holes" in the surface of the sample should be
below, along with the compactive efforts and their filled again with soil with light tamping).
relation to Standard and Modified AASHO 7. Rotate Compaction Ring (to break bond between
compaction. Base and Soil) and remove. Clean soil from Base and
from holes in Compaction and Spacer Rings.
No. of No. of Blows Compact ive Energy*
Layers Per Layer (ft.lb./cu.ft.) ASSEMBLY AND START OF TEST
3 7 55,000 (Modified AASHO) Place Spacer Ring on Base with No. 2 on the Ring
3 4 3 1,000 (% Modified AASHO) (radial grooves are on top) aligned with No. 1 on the
1 5 13,000 (Standard AASHO) Base. Place a dry Porous Stone (the stone must be
dry throughout) in the Spacer Ring (top of Porous
*Compactive Energy (ft .lb ./cu.ft .)
Stone should be level with top of Spacer Ring). Place
- No. Layers x No. Blows/Layer x 5.5 lb. x 1 ft . Compaction Ring, with recessed soil on top, on
.002 15 ft Spacer Ring and Porous Stone with letter S on
Compaction Ring lined up with letter S on Spacer
= 2.55 x No. Layers x No. Blows/Layer Ring (see note under PREPARATION FOR COM-
2. Place the apparatus on a sturdy support or on .the PACTION, 2 with regard to alignment). Insert 3
floor for compaction. Bolts and tighten firmly).
3. For the compaction procedure requiring one layer, Place 0-Ring and screw Lucite Container firmly
press the soil with the Hammer into the Rings until down on to P-Ring to ensure water tightness.
the top of the soil is about 118 inch below the top of Place a dry Porous Stone on top of sample inside
the Spacer Ring before applying the blows. During Compaction Ring. Place the Cover, with the radial
compaction space the blows evenly over the surface grooves on the bottom, on the Porous Stone.
by shifting the location of the hammer after each Place Top Bar with Proving Ring on .the steel Rod.
blow. Make sure that the Sleeve for the Compaction (Be sure that Adjustable Rod does not strike the
Hammer rests inside the Spacer Ring so that the Cover, as jarring of the Proving Ring Dial may be
Hammer does not strike the Spacer Ring. Make sure harmful), add Washers and Nuts and tighten firmly.
that the top of the compacted sample after com- Push up on Proving Ring Dial t o see that it appears
paction is 118 t o 114 inch above the Compaction to work properly. The Dial should move about one
Ring. division (.0001 inch) per 5.6 pounds/force.
4. For the Compaction procedure requiring three Set Proving Ring Dial to zero by moving the face of
layers, add 2 to 3 heaping teaspoons per layer and the Dial, then clamp face to dial. Turn Adjustable
press the soil with the Hammer to smooth and Rod down into groove on top of the Cover until a
firm-up the surface before applying the blows. (This Dial reading corresponding to 200 lb/sq.ft. (about
reduces the amount of soil "jumping" out of the one division) is attained. Tighten Adjustable Nut (9)
mold during compaction). Each layer of soil after on Adjustable Rod so that Adjustable Rod has no
compaction should have a thickness of about 114 play. Check to see that pressure on top of sample is
inch so that the final compacted thickness before still 200 lb./sq.ft. as indicated by the Dial reading.
trimming is 118 to 1/4 inch above the Compaction
Record the time and Proving Ring reading on the
Ring. (If the soil is below the level of the Com-
data sheet. Add water to the sample by squeezing
paction Ring, remove soil and recompact). The top
water from the squirt bottle into one of the three
of the first and second layers should be scarified
vertical (0.1 4 inch diameter) holes located at the top
(draw knife across top several times to loosen the
of the Compaction Ring until the water level in the
top 1/16 inch of soil) to ensure a good bond
Lucite Container has reached the Cover. (This
between successive layers.
procedure is used to reduce the amount of air
5. Disassemble. Remove the 3 Bolts. Rotate Spacer entrapped in the Porous Stones and thus to ensure
Ring (to break bond between Ring and soil) and that the sample has access t o water over its entire
remove. top and bottom surfaces.)
6. Trim the top of the sample. Start by trimming the The Swell Index is the pressure at the end of two
edges of the sample first, gradually toward the center hours.
 G U L F OF G U I N E A
FIG 13
SOlLS MAP OF AFRICA
CARTE DES SOLS D'AFRIQUE
e , - dbpris ~'Hoore,1964)
( o f ter ~ ' ~ o o r 1964
LEGEND - LEGENDE
SOLS MINERAUX BRUTS E T SOLS PEU ~VOLUE'S
RAW MINERAL SOlLS AND WEAKLY DEVELOPED SOlLS

SUR CUIRASSES FERRUGINEUSES

ON FERRUGINOUS CRUSTS

SUR ROCHES VOLCANIQUE

.... . CN VOLCANIC ROCKS

MATERIAUX DETRlTlQUES DES DESERTS E T SOLS SUBDESERTIQUES

DESERT DETRITUS AND SUB-DESERT SO1 L S

SOLS JEUNES SUR MATERIAUX MEUBLES D'APPORT ACTUEL ou TRES RECENT

E T SOLS HYDROMORPHES
J U V E N I L E SOlLS ON RECENT DEPOSITS AND HYDROMORPHIC SOILS

SUR E F F L A T A
ON VOLCANIC A S H

n SGLS ARGILEUX NOlRS

BLACK CLAY SOlLS

SUR ROCHES RICHES E N MINERAUX FERROMAGNESIENS OU SUR ROCHES VOLCANIWES

ON ROCKS RICH IN FERROMAGNESIAN MINERALS OR ON VOLCANIC ROCKS

SOLS BRUNS DES REGIONS ARIDES, SUBARIDES E T TROPICALES, NON DIFFERENCIES

UNDIFFERENTIATED BROWN SOlLS OF ARID, SEMI-ARID AND TROPICAL REGIONS

SUR E F F L A T A
ON VOLCANIC ASH

SOLS FERRUGINEUX TROPICAUX

FERRUGINOUS TROPICAL SOlLS

SUR CLllRASSES FERRUGINEUSES

ON FERRUGINOUS CRUSTS

SUR MATERIAUX ORlGlNELS SABLEUX

ON SANDY PARENT MATERIAL

I5 FERRISOLS
FERRISOLS

SUR CUIRASSES F E R R U G I ~ E U S E S
ON FERRUGINOUS CRUSTS

SOLS F E R R A L L I T I O U E S
FERRALLITIC SOlLS

SUR CUIRASSES FERRUGINEUSES

ON FERRUGINOUS CRUSTS

SUR SEDIMENTS MEUBLES SABLEUX

ON LOOSE SANDY SEDIMENTS

FIG 13
SOlLS MAP OF AFRICA
CARTE DES SOLS D'AFRIQUE
(of ter D'Hoore, 1964 - dbpres DHoore, 1964)

BEST AVAILABLE COPY

FERNANDO POD.."...-..'~ ...-...

sno THOM'E
 FIG 3
AFRIQUE - AFRICA
CARTE G'EOLOGIQUE DE L'AFRIQLJE
GEOLOGICAL MAP OF A F R I C A
STRATIGRAPHICAL AND LITHOLOGICAL SKETCH M A P

Complexe de base (Prgcambrien), surtout granit, gneiss et schistes

Basement complex (Precambrian), mainly granite, gneiss and schists

Formations Pale'ozoique, surtout grds et schistes

Paleozoic formations, mainly sandstones and schists

Formations M6sozoi'que surtout gre's, calcaires et argilites

Mesozoic formations, mainly sandstones, limestones and mudstones.

Formations tertiaires, surtout gres, calcaires et d6p6ts sableux

Tertiary formations, mainly sandstones, limestones and sand deposits

Formations quaternaires, surtout d i p a t s sableux, alluvions et colluvions

Quaternary formations, mainly sand deposits, alluvia and colluvia

lmportantes venues de roches basiques (antgquaternaires)

Extensive pre-Quaternary basic r o c k formations

Laves rhcentes
Recent lavas

After ~ ' ~ o o r( 1
e 964)
~'aprbs DIHoore ( 1964 )