RIJKSWATERSTAAT COMMUNICATIONS

THE USE OF ASPHALT

IN HYDRAULIC ENGINEERING

BY
TECHNICAL ADVISORY COMMITTEE ON WATERDEFENCES

Thc Hague, 1985

All correspondence and applications should he addressed to

RIJKSW ATERST AAT

DIRECTIE WATERHUISHOUDING EN WATERBEWEGING
HOOFTSKADE 1
P.O. BOX 20907
2500 EX THE HAGUE
NETHERLANDS

The views in this artiele are the author's own.

Recommended catalogue entry:

Use
The use of asphalt in hydraulic engineering / Technical advisory committee on water-
defences ; Rijkswaterstaat. - The Hague : Rijkswaterstaat, 1985. - 308 p. : app., ill. ; 24
a great contribution to this. In January 1984 the guidelines completed and now, one year
Bibliogr. : p. 225.
Preface

Asphalt products have been used in the Netherlands in hydra uIie engineering for a long
time on a large scale, especially after the great disaster in 1953 when a large part of
western Holland was flooded by the sea. After the disaster a great number of dikes had to
be repaircd very quickly and this was possible with the use of asphalt as arevetment
material. Asphalt could be placed much faster thcn thc materials most cornrnonly uscd in
those days. Further more asphalt is; when desired, more watertight than clay which is
also scarce. Except for bitumen no foreign construction materials are necessary.

SkilI, experience and knowledge were only slightly available in those days and had to be
gained in the course of time. After some years it was thought useful that directives or
recommendations were available. A first attempt to achicve this was made by the
Working Group for Sealed Revetmcnts who finished its work in 1961 with thc
'Preliminary Report'. In 1979 the Technical Advisory Committee on Waterdefences
decided to create new dircctivcs for the use of asphalt in hydraulics. The time was
thought right because of the great developments and knowledge obtained since 1961.
The building of the Delta Works including the Eastcrn Scheldt Storm Surge B~rrier was
a great contribution to this. In January 1984 the guidelines cornpletcd and now, one year
later, .an English translation is published.

The guidelines have been composed by a group in which representatives of public

offices, companics and research eentres worked together. It was atternpted to make the
content as complete as possible. Experience and knowledge were gathered, inves-
tigations carried out and design mode Is made. It is inevitable however, that there are
some shortcomings. Therefore it is intended to adapt the guidelines from time to time
and to update them with new information. Comments and reactions of the reader are
therefore very welcome.

The guidelines werc written for Dutch purposes. This means that mostly applications to
dikes, breakwaters and bed protections are csmsidered. Other types as for instanee
reservoir and dam applications are only mentioned without specific details.
Also, reference is made to Dutch standards and directives only. Despite these aspects
these guidelines wiII certainly contribute to a better design, execution, management and
maintcnance for hydraulic asphalt constructions, not only in the Netherlands but also in
other countries, the reason for the translation into the English language.

The translation, which was made possible by the Dutch Ministery of Public Works and

3
Bitumarin B.V., was done by B. Wade. Many thanks to Dr. P. C. Barber (Ceemaid
Serviced Ltd.), C. Davies (MetropolitanBorough of Wirral), Hydraulic Research Ltd.,
J. Harrison (Bitumarinc Ltd.), D. D.. Davidsen (CERC-WES, Vicksburg, MS), Ir.
J. P. J. van der Heide (The Association of Dutch Asphalt Contractors) and Ir. J. A. van
Herpen (Oranjewoud Engineering Consultants) for their contribution and rcmarks. Last
but not least mrs. D. van Eerd and mrs. W. Verhoeven should be mentioned.

I am convineed that these guidelines will help you to a better use of asphalt in hydraulic
engineering.

Leiden, January 12th, 1984

Prof. Ir. P. A. van de Vclde

Formerly Chairman of the
Technical Advisery Committee
on Water Defcnces

4
Contents

page

Preface 3

o Introduction . 15
0.1 Format for guidclines ' . 15
0.2 Organization . 16
0.3 Thc tasks and working methods ofWorking-Group 4A . 18
0.4 Recommendations for the future . 19
0.5 Investigations : . 20

PART A COMPOSITION AND PROPERTIES OF ASPHALT MIXES

1 Mix considerations ~. . 25
1.1 Mix components '. . . . . . . . . . . . . . . . . . . . . . . 25
1.2 The degree of filling of the mix 25
1.3 Voids in asphalt mixtures : ;.............. ' 26
2 The mineral aggregate ,. ...... 27
2.1 General . 27
2.2 The stone fraction 27'
2.3 The sand fraction . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . 28
2.4 FiBer.. . . . . .. . . . . . . . . . .. . .. . . . .. . . . . . . . . . . . . . .. . . . .. . 29

3 Bitumen. 30

4 Additives . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

5 Mix composition 35

6 Mix properties 36
6.1 General '.. 36
6.2 The degree of permeability . . . . . . . . . . . . . . . .. .. . . . . . . . . . . . 36
6.3 Mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.3.1 The stiffness modulus 37
6.3.2 The struin at break. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.3.3 Permanent visco us deformation . . . . . . . . . . . . . . . . . . . . . . . . . . A3
6.3.4 The Poisson-ratio . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . 44
6.3.5 Tcmperature sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.4 Stability .'. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
6.4.1 Stability of the asphalt : ,.. 45
6.4.2 Stability ofthe revetment as a who Ie . .. . . . .. . . . .. . . . . . . .. . 45
6.5 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
6.5.1 Aging................................................ 45
6.5.2 'Stripping' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
6.5.3 Resistance to erosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
6.5.4 Biological resistance 46
6.5.5 Chemical damage , 48
6.5.6 Other forms of damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
6.6 Workability 49
6.7 Environmcntal aspects 51

7 Use of asphalt products in hydraulic structures 52

7.1 Asphalt concrete 52
7.2 Mastic 52
7.3 Grouting mortars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
7.4 Dense stone asphalt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.5 Open stone asphalt 53
7.6 Lean sand asphalt ',................ 53
7.7 Membranes ~. . . . . . . . 53

8 Quality con trol . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

8.1 Mix-design tests .. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . 54
8.2 Production con trol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
8.3 Completion checks 55
8.4 Sampling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
8.4.1 'Construction mate ria Is ' ................. 56
8.4.2 Mixes - " .. 56

PARTB MATERlAL TECHNOLOGY

9 Asphaltic concrete '. . . . . . .. . . . . . . . . . . . . . . . . . 61

9.1 Basic materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
9.I.1 Crushed stone . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
9.1.2 Gravel '. .. 61
9.1.3 Sand 61
9.1.4 Filler 62
9.1.5 Bitumen 62
9.1.6 Other constructional aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
9.2 Mixes , '," . .. . 63
9.2.1 Mix-design 63
9.2.2 Mix-design tests . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.3 Mix properties 66
9.3.1 Mcchanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
9.3.2 Permeability ',' . . . . . . . . . . . . . . . . . . . . . . . 67
9.3.3 Durability ............................................ 67

10 Asphalt mastie ~. . . . . . . . . . . . 68
10.1 Basic materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
10.U Sand 68
10.1.2 Filler 68
10.1.3 Bitumen 68
10.1.4 Other constituents 68
10.2 Mixes. . . . . .. . .. . .. . . 68
10.2.1 Mix-design 68
10.2.2 Mix-design tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
10.3 Mix properties 71
10.3.1 Viscosity 71
10.3.2 Mechanical properties ; .. . . . 74
10.3.3 Permeability . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
10.3.4 Durability . . . . . . . . . . . . . . ... . . . . . .. . . . .. . .. . . . . .. . . . .. . 74

11 Grouting mortars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
IU Basic materials '. . . . . . . . . . . . 76
II.U Sand, filler, bitumen.. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . 76
11.1.2 Gravel, crushed stone ........... 76
11.1.3 Other constructional aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
11.2 Mixes. 76
11.2.1 Mix-design 76
11.2.2 Mix-design tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
11.3 Mix properties .... ~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
11.3.1 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
11.3.2 Viscosity '.' . . . . . . . 79
11.3.3 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
11.3.4 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

12 Dense stone asphalt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

12.1 Basic materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
12.1.1 Crushed stone, gravel 80
12.1.2 Sand, filler and bitumen :................. 80
12.2 Mixes................................................ 80
12.2.1 Mix-design 80
12.2.2 Mix-design tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
12.3 Mix properties 82
12.3.1 Stability and viscosity 82
12.3.2 Stiffness and breaking strength 84
12.3.3 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
13 Open stone asphalt 85
13.1 Basic matcrials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
13.1.1 Crushed stone '. . . . . . . . . . . . . . . . . . . . . 85
13.1.2 Gravel -..... 85
13.1.3 Sand :...... 85
13.104 Filler -. . . . . 85
13.1.5 Bitumen .'............... 85
13.1.6 Other constructional aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
13.2 Mixes. . .. 86
13.2.1 Mix design : , .. -. ... 86
13.2.2 Mix design tests 87
13.3 Mix properties 87
13.3.1 Stability and viscosity 87
13.3.2 Mechanical properties .-. . . . . . . . . . . . . . . . . . 88
13.3.3 Water perrneability 88
13.304 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
14 Lean sand asphalt 91
14.1 Basic matcrials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
14.1.1 Sand 91
14.1.2 Bitumen .-.... 91
14.1.3 Other constructional materials . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
14.2 Mixes................................................ 91
14.2.1 Mix-design 91
14.2.2 Mix-design tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
14.3 Mix properties :.................... 92
14.3.1 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
14.3.2 Mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
14.3.3 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
14.3.4 Durability ... _.. . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 94
15 Bitumen membranes 96
15.1 Basic materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
15.1.1 Filler 96
15.1.2 Bitumen 96
15.1.3 Additives - :.............. 96
15.1.4 Other constructional aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
15.2 Design and design tests .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
15.3 Membrane properties 98
15.3.1 Stability :............. 98
15.3.2 Viscosity 98
15.3.3 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
15.3.4 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
PARTC TECHNICAL ASPECTS OF THE DESIGN

16 ~he functional requirements of an asphalt revetment lOl

17 The dike body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 103

18 Design basis . 106
18.1 Safetyaspects . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. 106
18.2 The decrease of structural strength ; . . . . . . . . . . . . . . . . . . . . . . 107
18.3 Theorctical background of safety aspects . . . . . . . . . . . . . . . . . . 107
19 Loads . .. . .. 109
19.1 Hydraulic boundary conditions .:..... . . . . . . . . . . . . . . . . . .. 109
19.1.1 Water levels 109
19.1.2 Waves ,. . .. IlO
19.1.3 Currents 113
19.2 Settlement and scouring 114
19.2.1 Settlement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
19.2.2 Scouring liS
19.3 Other types of 'Ioads' ". . . . . . . . . . . 117
20 Design methods 119
20.1 Dense asphalt revetment designcd to resist hydraulic uplift
pressures ... . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 119
20.l.l Hydraulic uplift pressures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
20.1.2 Design '.' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
20.1.3 Specific constructienul features ,............. 125
20.2 Design of a plute-type asphalt revetment against wave impacts 127
20.2.1 Wave impact loads 127
20.2.2 .Construction schcmatization 130
20.2.3 Asphalt and subsoil properties . . . . . . . . . . . . . . . . . . . . . . . . .. . 130
20.2.4 Design. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. 133
20.2.5 Practical application ofthe wave impact formula :.. 134
20.3 Design of underwater bed proteetion against hydraulic uplift
pressures caused by currents and waves 138
20.3.1 Uplift pressures caused by currents . . . . . . . . . . . . . . . . . . . . . . . 138
20.3.2 Hydraulic uplift pressures due to wave action 138
20.4 Design of surface- or pattern-grouted stone ' layer against
wave attack . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . 139
20.5 Design of an asphalt revetment against currents 141
20.6 Designing an asphalt revetment against irregular settlcment
and scouting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
20.6.1 Irregular settlement . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . 144
20.6.2 Scouting 146
20.7 Deterrnination ofthe maximum slope 148
PARTD m(ECUTION
21 General aspects 155
21.1 Production :............ 155
21.2 Transport 157
21.3 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 158

22 Asphaltic concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 159

22.1 Production 159
22.1.1 The mixing plant. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. 159
22.1.2 Storage of building matcrials ... . . . . . . . . . . . . . . . . . . . . . . . .. 159
22.1.3 Mixing time and temperature . . . . . . . . . . . . . . . . . . . . . . . . . . .. 159
22.1.4 Hot mix storage 159
22.2 Transport 159
22.2.1 Means of transport 159
22.2.2· Onsite storage , 160
22.3 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 160
22.3.1 Application techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 160
22.3.2 Application temperature ; . . . . . . . . . . . . . . . . . . .. 166
22.3.3 -Seal coat 166

23 Mastic... .. .. . 167
23.1 Production 167
23.1.1 The mixing-plant. ..................................... 167
23.1.2 Storage of building materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
23.1.3 Mixing time and temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
23.1.4 Mix storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
23.2 Transport ;............... 167
23.2.1 Means of transport 167
23.2.2 Onsite storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
23.3 Application . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 168
23.3.1 Application techniques .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
23.3.2 .
Application temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 169
24 Grouting mortars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
24.1 Production 170
24.1.1 The mixing-plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
24.1.2 Storage of building materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
24.1.3 Mixing time and temperaturcs . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
24.1.4 Mix storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
24.2 Transport 170
24.2.1 Means of transport :...................... 170
24.2.2 Onsite storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
24.3 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
24.3.1 Application techniques :.............. 171
24.3.2 Application temperature ................................ 173

25 Dense stone asphalt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

25.1 Production 174
25.1.1 Tbe mixing-plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
25.1.2 Storage of building materials .,. . . .. . . . .. . .. . . . .. 174
25.1.3 Mixing time and temperatures . . . . . . . . .. . . . . . . . . . . . . . . . . . 174
25.2 Transport 174
25.2.1 Means of transport , ,.,................. 174
25.2.2 Onsite storage , , . . . .. . . . . . . . . . . . 174
25.3 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
25.3.1 Applicaion techniques ',' . . . . . . . . . . . . . . . . . . . . . 175
25.3.2 Application temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
26 Open stone asphalt 177
26.1 Production 177
26.1.1 The mixing-plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
26.1.2 Storage of building materials , . . . . . . . . . . . .. .. 177
26.1.3 Mixing time and tcmperatures ,... 177
26.1.4 Mix storage , , ,........... 177
26.2 Transport ,." ,........... 178
26.2.1 Means of transport ,........... 178
26.2.2 Onsitc storage ,, ,.................. 178
26.3 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
26.3.1 Application techniques . . . . . . . . . . . . . .. . . . . . . .. . . . . . . . . . . 179
26.3.2 Application temperature ,. . . 181

27 Lean sand asphalt 182

27.1 Production ... ,....................................... 182
27.1.1 Thc mixing-plant. ..................................... 182
27.1.2 Storage of building materials ... . . . . . . . . . . . . . . . . . . . . . . . . . 182
27.1.3 Mixing time and temperatures .. , .. , .... , . . . . . . . . . . . . . . . . 182
27.1.4 Mix storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
27.2 Transport 183
27.2.1 Means of transport 183
27.2.2 Onsite storagc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
27.3 Application ;....... 183
27.3.1 Application tcchniques 183
27.3.2 Application temperature ,......... 186
28 Membranes ........................................... 187
28.1 Production " , ',' . . 187
28.2 Transport :....................... 187
28.3 Placing , , .. , . . . . . . . . . . . . . . . . . . 188
29 Joints in bituminous revetments ,................. 189
29.1 Introduetion 189
29.2 Joints onto the same material . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
29.3 Joints between different types ofrevetment materials 191
29.4 Revetmentjoints onto structures . . . . . . . . . . . . . . . .. . . .. . . . . 193
29.5 Joints onto toe structures .:............................. 194

PARTE MANAGEMENT AND MAINTENANCE

30 Introduetion . 199
31 Damage prevention . 203
32 Types of damage ; : . 205
32.1 Asphaltic concrete . 205
32.2 Mastic . 206 .
32.3 Grouting mortars . 207
32.4 Dense stone asphalt . 208
32.5 Open stone asphalt . 209
32.6 Lean sand asphalt : 211
32.7 Membranes : . 212
33 Methods of rep air .. ; . , . 213
33.1 Asphaltic concrete . 213
33.2 Mastic . 216
33.3 Grouting mortars ",' . 219
33.3.1 Pattem grouted stone . 219
33.3.2 Fully grouted stone . 219
33.4 Dense stone asphalt . 220
33.5 Open stone asphalt . 220
33.6 Lean sand asphalt . 222
33.6.1 Lean sand asphalt filter layers . 222
33.6.2 Lean sand asphalt reclamation bunds . 222
33.7 Membranes . 223
Referenees ........................................... 225

Appendix I Design of a plate-type asphalt revetment against wave impacts 231

Appendix 11 The lifting of a relatively watertight bed proteetion by wave
action . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . .. . .. . . . . . 255
Appendix III Designing an .impermeable asphalt dike revetment against
hydraulic uplift pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Appendix IV The effects ofsettlement and scouring 277
Appendix V Boussinesq approach for monolithic constructions 292
Appendix VI A soil mechanics model for calculating a slope constructed
from lean sand asphalt 288

Appendix VII Sliding along a membrane 292

Appendix VIII Mastic flow through a pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 298

Appendix IX The slope test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 305

o Introduetion

0.1 Format for guidelines

These guidelines are intended for persons and organizations who wish to inform
themselves about the use of asphalt in hydraulic engineering. Designers, asphalt tech-
nologists, managers, government officials and con tractors will find the information they
need in these guidelines. The general form ofthe guidelines is such that they can also be
used as teaching material.

Originally the subject of the guidelines was the use of asphalt produets as a dike
revetment material. Although this aspect is still covered, in fact the final version has a
much wider scope and includes, namely, all hydraulic engineering applications of
asphalt mixes under Dutch conditions. The title now is, therefore, more suitably 'The use
ofasphalt in hydraulic engineering'. ,
As the title suggests the reader is guided through all aspects invO'lved in the application of
asphalt mixes in hydraulic engineering. lt gives guidelines, design methods, background
information and recommendations. Absolute specifications are not given. Similarly
tender specifications have not been included. For the latter reference should be made to
the work of the Dutch Foundation for Rationalization and Automatization in Road
Construction (RA W), which also publishes standard tender specifications for several
kinds ofhydraulic works, including bituminous applications.

Originally the intention was to limit the guidelines to a number of recommendations

without any detailed explanation. The objectives of the guidelines, however, and the
complexity of the subject and the lack of knowledge about many aspects did not lend
themselves to a neat package. lt was, therefore decided to publish these comprehensive
guidelines af ter recciving commen~s from future users of the book.

The guidelines are divided into several parts, each comprising a separate topic:
Part A 'Composition and properties of asphalt mixes'. This part deals with the general
technology of the asphalt mixes used in hydraulic engineering. The object is to provide
knowledge and insight into the subject and to give the basic information needed for the
other parts. Material parameters, used in Part C, are also presented here.

Part B 'Material technology'. The asphalt mix types most commonly used in hydraulic
engineering are discussed and evaluated specifically in this part. The subjects covered
include the basic materiais, the way in which the mix composition is decided, the ne-
cessary mix design tests and the mix properties.

15
Part C 'Technical aspects of the design'. This part deals with functional requirements,
the starting points and the loads which develop on and in the material, rclated to the
design ofhydraulic engineering asphalt constructions. Design methods are also discussed
here.

Part D 'Exccution', Execution of the work is considered, for each type of asphalt, sub-
divided into production, transport and placing,

Part E 'Management and Maintcnancc'. This part deals with the management and
maintenance of asphalt revetments. Methods of repair are given after a discussion about
the causes, tracing and prevention of the various types of damage which can occur.

In appendices the theoretical backgrounds to several parts of the guidclines and the
development of design models are given. .

The asphalt types considered in the guidelines are:

- asphaltic concrete
mastic
grouting mortars
dense stone asphalt
open stone asphalt
lean sand asphalt
membranes .

Subdivision of the directives between the several aspccts, is uscd to make the subject
more generally easy to grasp and apply. For practical purposes, however, a subdivision
into material technology, design aspects, execution, management and maintenance
°
might be desirabie per mixtype. A separate table, Table (page 21) had, therefore, been
included in which the mixes are given, directly related to the areas ofapplication. If, for
cxample, only one particular mix type is being considered then it is only necessary to
consult those sections ofthe guidelines indicated in the relevant column of'thc tabie.

0.2 Organlzation

In 1979 the Dutch Technical· Advisory Committee on Water Defences (Technische

Adviescommissie voor de Waterkeringen, TA W) decidcd to publish a set of directives on
asphalt dike revetments. This aspect of hydraulic engineering falls within the scope of
TA W Working-Group 4: 'Dike Revctments' and th is group has, thcrcfore, taken the
work under its wing. The detailed work needed was allocated, in 1980, to Sub-group 4A:
'Guidelines for Asphalt Revetments on Dikcs'.
Thc members of this group originate from governmcntal departments and business
organizations involved in the application of asphalt in hydraulic engineering.

16
As of Ist January, 1981, Ir. J. A. van Herpen was appointed as project leader to
coordinate the work involved and to write the guidelines.
At the time ofpublication (ofthe Dutch version) ofthe guidelines the Working-Groups
were composed as follows: .

Working-Group 4 'Dike Revetments'

Prof. Ir. P. A. van de Velde, Former professor at the Delft University of
Chairman Technology, the Netherlands
Ir. E. H. Ebbens, Netherlands Water Defences Research
Secretary Centre
Ir. W. Bandsma State Road Engineering Department
Prof. Drs. W. van Dijk Delft University ofTechnology
Ir. J. A van Herpen Delft University ofTechnology
Dr. Ir. P. A. Kolkman Delft Hydraulics Laboratory
Ir. H. L. Koning Delft Soil Mechanics Laboratory
Ir. R. C. Koole Pavement Consultancy Services
Ir. P. C. Mazure Netherlands Water Defences Research
Centre
Ir. W. Meulenberg The Office of the Dike-reevc, 'Het
Noordhollands Noorderkwartier'
Ir. P. Ruijgrok Delft Soil Mechanics Laboratory
Ir. H. Visser The Walcheren Polder Authority
Ir. G. M. Wolsink Delft University of Tcchnology

On Ist April 1983, Ir. W. J. Heijnen (Delft Soil Mechanics laboratory) and Ing. J. T. de
Vries (Netherlands Water Defences Research Centre) resigncd from the Working- .
Group.

Sub-group 4.11'Directives for Asphalt Revetments 011 Dikes'

Ir. E. H. Ebbcns, Netherlands Water Defcnccs Research
Chairmnn .. Centre
Ir. J. A. van Herpen. Delft University of'Tcchnology
Sccretary/Projcct Leader
Prof. Drs. W. van Dijk Delft University of Tcchnology
H. J. A. J. Gruis State Road Engineering Department
Ir. J. P. J. van der Heide The Association of Dutch Asphalt
Con tractors
Ing. K. A. van den I-loek State Delta-Works Department
Ir. H. L. Koning Delft Soil Mechanics Laberatory
Ir. G. L. M. Mulder Bitumarin B.V.
Dr. R. C. Rcintjes Volkcr Stevin Wegen en Asfalt B.V.

On lst January 1981, Ir. R. F. G. M. Zijlmans (Netherlands Water Defences Research

17
Centre) was replaced by Ir. E. H. Ebbens of the same organization and on lûth Sep-
tember 1981, Ing. K. A. C. Mouw (State Delta Works Department) was replaced by Ing.
K. A. van den Hoek. On 25th May 1982, Ing. J. J. van der Plas (The Association ofDutch
Asphalt Contractors) rcsigned in faveur ofIr. J. P. J. van'der Heide and on 26st Juli 1982,
Ir. W. Bandsma (State Road Engineering Department), resigned in favour ofH. J. A. J.
Gruis.
On the resignation ofIng. J. T. de Vries on lst April 1983 (Netherlands Water Defences
Research Centre) the duties of secretary were taken over by Ir. J. A. van Herpen.

In addition to thc members of Sub-group 4A many other people and organizations

collaborated in the realization of the directives, two in particular, should be mentioned
. by name, Ir. H. Roos (Bitumarin B.V.) and Ing. C. C. Montauban (State Road Engin-
eering Department).

0.3 The tasks and working methods ofWorking-Group 4A

Working-Group 4A was instructed to prepare guidelines for the design, execution,

management and maintenance of asphalt revetments on dikes. To this end the following
programme was executed:
I. Collcction of available information on the subject of the use of asphalt in hydraulic
engineering.
2. Preparation ofthe dircctivcs,
3. Initiatien and supervision of research, with the object of increasing the knowledge
. available on the subject.

In order to realise the directives in a reasonable time a project leader was appointed for a
3 ycar period, beginning lst January 198 I. With thc help ofsome ofthe mcmbcrs ofthe
Working-Group a programme was dr~wn up listing all the aspects ofthe application of
asphalt in hydraulic engineering. Thc first step was to assembie all available information
and then for each type of asphalt a subdivision was made into design methods, material
technology, execution and maintenance.
After the data had been collected an attempt was made to formulate practical rul es for
designing asphalt revetments. This information was then presented in an interim report
to future users - principally dike managers - and members of Working-Group 4, for
comments.
The 'Wave problems on Dikes' and 'Water Movements in Dikes' Working-Groups 1 and
2 of TA Wand also specialists from the Delft Soils Laboratory and the State Road
Engineering Department were consulted on certain aspects.
The guidelines were eventually, drawn up on the basisofthe information obtained and a
draft was presented to the various groups mentioned above for further comment.
Results of research executed by'Working-Group 4A have been used in the guidelines.
This research is discussed further at the end of th is introduction.

18
0.4 Recommendations for the future

During the preparation of the guidelines it became apparent that, for various reasons, it
is possible to improve the present methods of application of asphalt in hydraulic engin-
eering. A number ofrecommendations are therefore given below to assist in future work
of research and development in this field.

Recommendations:

In conneetion withconstruction design a good description is required of the

physicallmechanical propertics of the various asphalt mixes under typical loading
conditions. Any gaps in information should be filled by investigations as soon as
possible.
When investigating the propertics of an asphalt mix it is always preferabie to
determine the mix composition by extraction froma sample.
Dynamic loading tests on the subsoil are recommended to obtain an impression ofthe
behaviour of the subsoil, for a range of moisture contents, on which the asphalt
construction will be placed.
The possibilities ofusing gravel in asphaltic concrete in hydraulic structures should be
investigated further. At present only crushed stone is prescribed in the Dutch Eisen
1978 (6) (sec also Section 9.2.1).
An asphaltic concrete revetment construction consisting of several layers is
recommended if, under the usual methods of compacting cracking in a single layer
cannot be avoided or if the required voids ratio cannot be obtained. It is
recommended that the appropriate method of compaction is determined at the start
of the work.
More information should be obtained about the durability and resistance to currents
of open stone asphalt and lean sand asphalt.
Lean sand asphalt appears to be a very good material to use for bunds. When more is
known about the 'soil mechanical' propertics the application in this field can certainly
be extended.
Better methods fordesigning mastic mixes and grouting mortars are desirabie.
It is recommended that methods and equipment are developed for tracing damage so
that management and maintenance can be done more effectivcly. A rational system
for the management and maintenance ofrevetments and bank proteetion can then be
gradually developed.
Calculation models for construction design are, generally, still in the stage of
development. More practical and usabie models will assist efficient design ing.
The re-cycling of revetment matcrials should be investigated.

19
OS Investigations

In order to implcment the above recommendations eertain investigations will be

neeessary. A nu mb er of investigations have been formulated by Working-Group 4A,
some in cooperation with ether organizations.

These investigations include:

An investigation to determine the best method ofcompacting asphaltic concrete dike

revetments. At th is moment part ofthis investigation has been carried out by the State
Road Engineering Department.
An investigation, relatcd to a ncw mixture design method for mastic, will be
completed shortly. This investigation is being carricd out in cooperation with the
State Road Engineering Department.
Investigations into the resistance of open stone asphalt against wave attack. These
include:

l. A recently complcted investigation in the Delta Flume of the Delft Hydraulics

Laboratory, De Voorst, for Bitumarin Ltd.
2. A recent investigation in the by-pass channel of the navigation dam at Lith, the
Netherlands, in cooperation with the State Delta-Works Department.

A preliminary investigation into the mechanical propertics of open stone asphalt

executed at the Delft University of Technology.
Cooperation in an investigation into the soil mechanical propertics of lean sand
asphalt carried out by the State Delta-Works Department. This investigation will be
completed shortly.
Investigations into the resistance to currents and durability oflean sand asphalt are in
preparation.
General investigations into the mechanical propertics of asphalt mixes are bcing
carried out by the State Road Engineering Department.
Investigations into the possibility of recycling asphalt mixes have been started in
cooperation between Rijkswaterstaat and private industry.

Results, already available from the above investigations, have been included in the
guidelines. Results still to come will be published and will be included within revised
editions of these guidelines.
It should be noted that the above investigations will not cover all the points
recommended for study, and some thought should be given to the remaining aspects for
which information is required. .

20
Table 0 Guideline reference table - The table relates the varia us uses of asphalt in hydraulic
engineering to chapter and paragraph readings for the various mate rial types.

mlx type
grouting lean sand
mortars asphalt
asphal- dense open core
tic mas- ruil) pattem stone- stone mate-
concrete tic grouting grouung avphalt asphatt rial 1,])('["') membranes
Part B basic materialv 9.1 10.1 11.1 12.1 IJ. I 14.1 15.1
Mate-
milt de s Ign 9.2.1 10.2.1 11.2.1 12.2.1 13.2.1 14.2.1 15.2.1
rial
Feehno- mi, de sign tests 9.2.2 10.2.2 11.2.2 12.2.2 13.2.2 14.2.2 15.2.2
lcgy
mie properties 9.3 10.3 11.3 12.3 13.3 14.3 15.3
Pafte devigning a dense
Feehui- avphalt revetment
cal against hydraulic
design uplift pre'iosureo; 20.1 20.1 20.1 20.1 20.1
designing a plate
type a sphalt tevetment app.
against wa ....e impacts 20.2 20.2 20.2 20.1 2u.2 VI 20.2.2
designing under-
w ater bed protee-
tion again ..t upbf't
by waves and
currents 20.3 20.3 20.3 203 20.3
de..igning a 100..('
element re..etment
against w .1,,(' auack 20.4
devigning art a sphalt
revetment against
cutrent 20.5 20.5 20.5 20.5 20.5 20.5 20.5
designing an asphalt
revetrnent agamst
senlement end
scouting 20.6 20.6 206 206 206 206 206
determination of the
maximum slope 20.7 20.7 20.7 20.7 20.7 20.7 20.7
ether 'Ioads' 19.3 19.3 19.3 19.3 19.3 19.3 19.3
Part [) production 22.1 23.1 24.1 25.1 26.1 27.1 28.2
Exe ..
transport 22.2 23.2 24.2 25.2 26.2 27.2 28.3
cunon
placing 22.3 23.3 24.3 25.3 26.3 27.3 29.2
joints. 29.2 29.2 29.2 29.2 29.2 29.2 29.2

c
bet .....een dif-
ferent mate-
't'"' dali 29.3 29.3 29.3 29.3 29.3 29.3 29.3
"Q

"ij onto !otruc-

~ tures. 29.4 29.4 29.4 29,4 29.4 29.4 29.4
:;
toe con-
!!.tructions. 29.5 29.5 29.5 29.5 29.5 29.5 29.5
Part E po~ible damage 32.1 32.2 32.3 32.4 32.5 32.6 32.7
Manage-
Methods. of repair 33.1 33.2 33.3.2 33.3.1 3J.4 33.5 33.6.2 33.6.1 33.7
ment
anJ
Main-
lenance

21
PART A

COMPOSITION AND PROPERTIES OF ASPHALT MIXES

Sumrnary

Part A deals with the general aspects ofthe composition and properties ofasphalt mixes
which are important in thcir application to hydraulic structures. Thc object of Part A, is
to introduce the subject to the reader and enlighten him, making thc manual more easy to
read, and also to present data about the materials which are relevantto Part C which
deals with design techniques. .

Part A comprises:
A discussion of the individual components of asphalt and their influencc on thc
material as a wholc, Attention is paid to those mix properties which are most
important to hydraulic engineering.
A review of the most frequently used asphalt mixture types in hydraulic structures
and how these types are applicd.
Finally, a discussion of quality con trol including mix-design tests, construction con-
trols and completion checks prior to hunding over thc work to the Client. Included
here is a discussion ofthe way in which samples are taken.

24
1 Mix considerations

1.1 Mix components

Asphalt is a mix ofvarious componcnts:

bitumen;
mineral aggregate;
additives, if required.

The mineral aggregate is composed ofcrushed stone, gravel, sand or filler or a combin-
ation.
The choice ofthe most suitable composition for a particular application depends mainly
on the requirements which the material has to meet and the associated mix properties,
see Section 6.1.
The mix properties are specified by the composition, th at is, the relative proportions of
the various components, the properties ofthe components themselves, and the properties
which result from the application andcompaction method.

1.2 The degree of filling of the mix

The mine ral aggregate mix contains voids. Initially, the bitumen coats and binds the
various aggregate components together. Ifmore bitumen is applied than is necessary for
coating and binding then the pores will gradually be filled.
Mixes, in which the bitumen only serves as a binder, are referred to as 'underfilled'
mixes, sec Figure 1.1a. The properties of such a mix are directly related to the properties
of the stone skeleton (4). If the proportion of bitumen is increased the voids in the
mineral become filled and the inf1uence of the bitumen on the properties of the mix is
increased while that of the stone skeleton is reduced. With mixes in which the pores are
almost filled with bitumen, sec Figure l.lb, both the stone skeleton and the bitumen
contribute to the mix properties. This type of mix must be compacted, either
mechanically or under its own weight.
'Overfilled' mixtures are those in which the volume ofbitumen is greater than th at ofthe
voids in the mineral aggregate. In such a mix the properties ofthe bitumen predominatc,
the mine ral providing only a certain amount of stiffening, see Figure l.lc. This type of
mix is impermeable and requires na compaction.

25
a- underfilled mix b- intermediate filling c- overfilled mix
_ bitumen E?ZZI mineral aggregate 0 voids

Figure 1.1 The degree of bitumen filling in the mineral aggregate.

1.3 Voids in asphalt mixtures

The term 'voids' refers to the volume of pores in the compacted asphalt (5). The voids
ratio, VIM (voids in mix), is given by:
d -d
VIM = 100( ma) vol %
dm
density of the mix without voids (kg/rn'')
density of the mix with voids (kg/rn ')

In general, the smaller the voids ratio (HR) the more resistant is the mix to crosion and
the grcater its durability, A mix with a small voids ratio is better 'sealed' against external
inf1uences such as oxygen, light and water. If'water pcnetrates between the bitumen and
the mineral (through the matcrial) there is a 1055 of adhesion which is rcferred to as
'stripping' .
Exposure to the atmosphere and light ages the bitumen. In this respect, the size of in-
dividual pores and dogree ofinterconncction between voids are also important.
The voids ratio and the distribution of voids also determine whether or not thc mix is
sand and watertight. Although water impcrmcability is not always a functional rcquire-
ment it gives a good indiention of the durability of a mix.

To illustrate:
A mix containing sand with 5%, by rnass, of bitumen anda voids ratio of 25% is
sand-tight, A mix, however, of open stone asphalt with 80%, by mass, of stone and
20%, by mass, of mastic with the same voids ratio is not.
An asphaltic concrete with a voids ratio of 3% can be considercd as absolutely
watertight. In th is case the voids are not interconnected. For the sarnc reasens a
mastic with the relatively large voids ratio of 10% is also watcrtight.

26
2 The mineral aggregate

2.1 General

It is important that there is good adhesion between the bitumen and the mineral
aggregate. There are two kinds of adhesion:
l. Physical adliesion:
This type of adhesion is better if the surface of the mineral aggregate is rough.
2. Chcmical adhesion:
Since bitumen is weakly acidic, better adhesion is obtained with mineral aggregates
which are slightly basic. This property is generally to be found with minerals which
contain a limited amount ofsilica oxide.
It is also very important that dry aggregate is used.

A low voids ratio is ensured in an asphalt mix which is not overfilled by adjusting the
coarse fraction in relation to the sand fraction and by using a well-graded mineral
aggregate: the spaces between the larger particles are then filled by the smaller.
Increasing the fillcr fraction, provided that it does not expand the sand/-stone skeleton,
can produce greater internal stability, Tbe internal stability oftbe material in place can
also bc increased by using an angular material instead of round, for example crushed
stone instead of gravel, crushed sand instead of natural sand.
Tbc quantity of bitumen needed to bind the mineral aggregate depends on the specific
aggregate surface. The specific surface is inversely proportional to the second power of
the partiele diameter.
Mineral aggregate can be subdivided according, to grain size. In the Netherlands the
following terminology is used:
stone fraction, larger than 2 mm;
sand fraction, between 2 mm and 63 firn;
filler fraction, smaller than 63 firn.

2.2 The stone fraetion

The stone fraction comprises crushed stone, that is a rough broken material, or a smooth
round material such as gravel.
Crushed stone and gravel must be able to withstand impact and abrasion during mixing
and compaction. This property depends on:
partiele shape; cubes are the best;
- the strength ofthe mate rial.

27
Angular mate rial has a higher interloeking strength.
The maximum particIe size in an asphalt mix is related to Iayer thiekness and the
production and working mcthods. Generally a large grain size gives better stability but
reduces workability. In addition the danger of segregation is greater. For the stone
properties requircd, reference should be made to the Dutch specifications, Eisen 1978
(6).

2.3 The sand fractlon

Natural sand, crushed sand or a mixture ofboth are used for bituminous mixes. The sand
grain size distribution plays an important part in stability, voids ratio and the binder
requirements for a mix -.Sometimes the grain size distribution is specified. If this is not
the case, laboratory investigations are carried out to determine the sand gradation which
most economically satisfies the requircd mix properties. Often th is can be achieved by
mixing different varieties ofsand.
Sand can be characterized by the mass percentages of sieve fractions lying betwcen, for
example, the 2 mm, 500 urn, 180 Itm and 63 Itm sieve sizes (numbers la, 30, 85 and 24c
respective y). The composition can be shown graphically using the sieve fractions in,
wh at is referrcd to as, the 'sand trianglc', see Figure 2.1.

The curved line in the sand triangle

shows the general limits of natural
Dutch sands

I. sharp sand
11. very fine sand
III. medium fine sand
IV. medium coarse sand
V. very coarse sand

percentage- .:.clh;:.:ro~u9::.;;h...:S...:,ev;:.:._':-:80,...:1l:...",_
on slev. 63 IJm

Figure 2.1 The sand triangle (5).

28
Sand is said to be better gradcd ifit lies within the shaded area ofthe triangle. This sand
is referred to as Sand A. According to Eisen 1978, the grain size fractions are:

Percentage by mass
Through
Sicve On
size Sieve Desired Lirnits
2rnrn 500llm 25 10-50
500flrn 18Ollrn· 40 30-60
180 urn 63 urn 35 20-45

2.4 Filler

Filler (5):

I. fills the voids of the stone-sand mix producing a more uniformly gradcd matcrial;
2. forms, together with the bitumen, the binder required;
3. has a stiffening effect on the binder which increase; the viscosity and reduces the risk
of segregation.

In general rather f1exible mixes are required for hydraulic structures. 'Wcak' to 'very
weak' filler is uscd, therefore, which requires little bitumen binding and as a result more
'frcc' bitumen is obtained to provide f1exibility. The preferenee is for filler which is hy-
drophobic and has a basic renetion. so that the adhesion between the bitumen and the
stone is improved.
The quantity of filler and its voids ratio largely determine the quantity of bitumen
required. For these reasons, and for optimum workability, the filler must have strictly
controlled requirements such as:
- a constant absorptivity of bitumen;
- a constant nature and quality.
Limestone fillers should bc used for preferenee.

Methods of characterising filler can include (63):

I. the Rigden dry compaction test, in which a particular weight of filler is compact cd by
a standard method and then its volume recordcd.
2. the Van der Baan test to determine the bitumen number. In this test the volume of
water required to bring the filler to a particular consistency is recorded (95).

For the propertics required reference should be made to Eisen 1978 (6).

29
3 Bitumen

Bitumen is a very viscous, non-volatile material which principally consists of hydro-

carbons or their derivatives (5). The viscosity of bitumen depends on the temperature; it
is what is referred to as a thermoplastic mate rial. Although it does not react to water and
most chemieals it does dissolve in lighter hydrocarbons. lts performance under load is
strongly dependent on its temperature and the duration ofloading.

The bitumen can be defined by using empirical parameters:

In the normal temperature range by the penetration and the softening point ring and
balI. The penetration (pen 25°C) is the intrusion into a bitumen sample measured in
units of 0.1 mm, of a standard probe with a weight of 100 gram at a temperature of
25°C during a period of 5 seconds. The softening point (Tr&b) is the temperature, in
°C, at which a slice of the materiaI, held firmly in a ring under standard test
conditions, undergoes a standard deformation under the weight of a metal ball,
By temperature scnsitivity is understood the extent to which the viscosity depends on
the temperature. This dependency is described by the penetration index (PI).
The higher the penetration index, the lower the temperature sensitivity.

20-S00·A
PI (pen, pen) = I + 50 . A

in which:
log 800 - log pen
A
T,&k -25

pen = penetration at 25°C (units of 0.1 mm)

Tr&b = softening point (OC). The penetration at the softening temperature for
normal bitumen is about 800pen (800 X 0.1 mmo = ,80 mm)

The PI value and the softening point can be determined very simply from two
penetration values using Figure 3.1.

For standard bitumen PI generally lies between +

I and -I. These biturnens are
often referred to in the trade by the limits in which the penetration must lie, for
exarnple, bitumen 801 \00.

30
 I
I -3
-2

Ë 102 / o ,
~
ei
8 2
3
6
c
.2
ë 6

I PI
7

10' Plot the penetronen measured at

8 two or rnere temperc tur es .
6 Draw a strai<;ht line through these
points end reed oft the terepercture
Ier a penetrcucn of 800 pen
( s cf terunq- point ring ond bali )
Draw a parallel line through pcrot A
end reed olf the penetration • index
A on the PI -sccte.

10 20 30 40 50 60 70 80
~ terepercture (oe)

Figure 3.1 Determination of the softcning point (5).

In the low temperature range the Fraass breakingpoint, referred to as 'the brittie
temperature' indicates the consistency.
This number indicates the temperature at which a 0.5 mm thicklayer of bitumen
cracks under a ben ding Ioad.
The higher temperature range is important for the mixing and application of asphalt
mixes. In this conneetion a certain viscosity is essential; the foUowing va lues for
kinematic bitumen viscosity have been determined for different operations (5):

• spraying, about 20-50 mm-/s

• mixing with mineral aggrcgate, about 150-300 mmê/s
• pumping from a bitumen truck, about 1200 mm-/s
• placing and compacting, depending on the type ofmix, about 300-2000 mm-/s.

Remarks: The dynamic viscosity (lJ) is measured in Pa . s (Pascalseconds); the kinematic

viscosity (I') in m2/s.
The relationship between the dynamic and kinematic viscosity is given by v = lJl p, where
pis the density of the bitumen in kg/rn", see Table 3.\.

31
I Pascal = I NIm:!
i Poise = 0.1 Pa . s
The stokes unit is also used for kinematic viscosity.
I Stokes = 10--1 m2/s

Temperatures related to cqual viscosities are referred to as Equi ViscosityTemperatures

(EVT). So, for example, there is an EVT to which a bitumen must be heated to produce
the mixing viscosity of 170 mmê/s, This temperature is 160 to 170°C for bitumen 45/60
and 150 to 160°C for bitumen 80/100, see Figure 3.2.
Whcn bitumen is exposed to the atmosphere it hardens, which lowers thc penetration
and raises the softening point ring and ball, Hardening is among others caused by the
evaporation of the volatile clements and oxydation. The higher the temperature the
faster the process.
Considerabie harderring takes place during asphalt mixing and application. Hardening
which develops during the course of time is referred to as aging.
500.00~

200.000
\\ \ -.\ -, \ ~ -. --- straight run bitumen
100,OO~ ___ cut-back bitumen
50.000
......... tar-bitumen
20.000
__ _ _ road tar
1~.000

..
,:;-
5.000
E 3,000
.5 2.000

1.0~~

\ '1 ',.'\,. \ -, -, 1\ -, 1\

25 5~ 75 10~ 125 150 175 200 225 2~Q 275

terr-perctcre I Oe)

Figure 3.2 The viscosity of different bitumens as a function of temperature (5).

The quality ofa bitumen must be sueh that hardening is not excessive.lnvestigations into
harderring can be carried out in 'penetration after loss on heating' tests, in which a bitu-

32
men sample is heatcd at 163°C for aperiod of5 hours. For thermostability there should
be very little difference between penetrution before and after the test.
After production at the refinery the bitumen must not be heated abovc 200°C since then
the material propertics can change and certain cracking processes can develop.
Bitumen must not be too hard; the working viscosity should be sufficiently low (about
0.2 pa'. s at 140 tot 160°C) and the Fraass breaking point must not be reached under
critical mechanicalloads at low ternpcratures (8).
For requirements and tests related to bitumen, reference should be made to Eisen 1978
(6) or NEN 3902 (21).

Table 3.1 Density of penetration bitumen at various temperatures (5).

Temperature (0C)
25° 100° llO° \20° \30° 140° 150° 160° \70° 180° 190~ 200°
Bitumen
(pen) Density (1000 kg/rn-)
270/330 1,01 0,97 0,96 0,95 0,95 0,94 0,94 0,93 0,93 0,92
160/210 1,02 0,97 0,96 0,96 0,96 0,95 0,94 0,94 0,93 0,93 0,92 0,92
80/100 1,03 0,97 0,98 0,97 0,97 0,96 0,96 0,95 0,95 0,94 0,93 0,93
45/60 1,04 0,98 0,98 0,97 0,97 0,96 0,96 0,95 0,94 0,94 0,93
20/30 1,05 0,99 0,98 0,97 0,96 0,96 0,96 0,95 0,95 0,94

33
4 Additives

Additives are used to improve adhesion by lowering the surface tension between the
bitumen and the mineral aggregate. This limits stripping.
The effect of the additive is noticcable in the short term but, after about two years, is no
longer apparent. Adhesion-improving additives have been used incidentically in
hydraulic structures. It should be borne in mind that the viscosity can be affectcd.
Propertics such as the viscosity can be affected by the addition of polymers, both
chemically and physically. For the time being, polymers are not used in the hydraulic
applications.

Additives can be added during production and/or before application of asphalt mixes.
When adding to the bitumen, segregation must be preventcd.

34
5 Mix composition

The mix composition is mostly designed on a choice, within certain limits, of partiele
grain size distribution and bitumen content, on the basis of investigations into the
mechanical properties, and by application ofthe basic materials allowed and available.
For different applications the composition is, generally, specificd in thc form of rough
proportions by mass. Thc dctailed composition can thcn be determined from laboratory
and in-situ tests.

There are two general principles involved for dctermining mix composition (9):
I. Design for stability, if a mix of a certain strength is required. Thc bitumen content is,
in this case, adjusted to thc grading of the mineral aggregate,
2. Design for a particular viscosity if a f1exible mix is required. In this case a certain
ovcrfilling ofvoids is esscntial.

35
6 Mix properties

6.1 General

As already stated the choice of mix compositión depends on investigations into the mix
properties. The most important propertjes for use in hydraulic engineering are:
dogree of pcrmeability
mechanical properties
stability
durability
workability and compactability

6.2 The degree ofpermeability

A dike revetment - ofwhich asphalt can form a part - must be completcly sandtight.
Sometimes a cover layer is also required to be watertight, The voids ratio ofa mix and the
size and oricntation of the voids detennine the degree of penneability of the mix as a
who Ie. The factors which affect the voids are given in Section 1.3.
If a sand-pcrmeable asphalt mix is choscn for arevetment then the sand scal must be
provided by a filter construction. In the case of watertightness a stricter sealing criteria
wiII be essen tial for asphaltic mixes used for sealing water reservoirs than for dike-
rcvetmcnts.
For exarnplc; a 5 cm thick plate ofbituminous mix, with a voids ratio of3% should not
aIIow anywater through when subjected to a water pressure of3 atm for a few hours (12).
For mixes with such a 100y voids ratio special thought must be givcn to the compaction
operatien because of the possibility of initiating cracks in the material.

6.3 Meeltanical properties

Asphalt mixes and bitumen appear to have similar mechanical properties. An asphalt .
mix, in the same way as bitumen, is a visco-clastic material which under short duration
loading and at low ternpcratures appears to be clastic; under long duration loads and
higher temperatures it is viscous. Thc visco-elastic property ofasphalt is an advantage for
applications such as in dike revetments; it is stiffunder short duration loads such as wave
impacts but yiclds under long duration loads such as those occurring during settlement.

Remarks
The foIIowing sections give nomograms for detennining the stiffness modulus, the initial

36
strain at break and the Poisson ratio of an asphalt mix. ft should be borne in mind that
these nomograms only give general values. If more accurate values are required it is
better to carry out separate investigations on each mix, including, for example, three or
four-point bending tests and creep tests.

6.3.1· Tlzc stiffness modulus

For an elastic material deformation is proportional to the stress applied. The modulus of
elasticity, E, also referred to as Young's Modulus, is often used. This modulus is
independent of temperature and loading duration.

E = alf:
a = stress (N/m2)
e = strain (-)

There is a similar relationship for bitumen, which, however, is strongly dependent on

temperature (T) and loading duration (t). The stiffness modulus, S, is used to charac-
terize the material (5).
a
S(t,1) = li

For increasing values of tand T the values of S reduce.

The nomogram, prepared by Van de Poel, which can be used to determine the stiffness
modulus of bitumen is shown in Figure 6.1.

Since an asphalt mix is visco-elastic the stiffness modulus can also be applied. From
research it appears that there is a conneetion between the S-modulus of bitumen (Shit)
and that of an asphalt mix (Smix)' This relationship is determined, amongst other things
by the volumetrie percentages of the mineral aggregate and the bitumen. The
relationship between Smixand Shit can be shown in what is known as a master curve. An
example ofsuch a curve is given in Figure 6.2.

Over the years various laboratories have carried out research and have developed
nomograms for determining the stiffness modulus of asphalt mixes. The most recent
example of these is give in Figure 6.3.

37
 10' 10· 10' 25,10

.. ..
2.10'
.,
-- -
I I
+7 10'
STIFFNESS MODULUS. 11Im2
+6 -<:::: 'tt
+5 10' +5
t4
10' 10' +4
xt3
10 IQ' +3
êZ +2~,0_'
+1
1:>-' +2
o IQ_, +1

•
t= 0, .• o
~ -1
t;; -I

..
z -2~
"' -3
I,
I,
\
1
TEMPEAATUAE DIFFEAEIleE. oe

100 ec IJ 10 2'J Yl .0 'lel b1 1.:) 110 so 1)0 I'C 'n 1~ '40 -so '60 110 180 '''0 lOO

-10 ABOVE T800.... ".1",,1·',1,1,1,' .1 ".I",,!' "1,,,,1,' " ,I,,, ,1", I" " " ' ·1" ,,,..,,, ,.J, ! I BELOWT800 ...
I,
The penetration index (Pil has been defined bv:
Example
20-PI log pen at T 1 - IOypen at T 2 I,
-- = 50 ---'-------=- Opcrating conditions
10+PI T1-T2 I, Température 100C
The stiffness modulus, defined as the ratio stressl
\ Loading time 0.02 secouds
Itrain, is a function C?ftime of loading (frequency). tem-
I
perature difference with TBOO pen, and PI.
\ Characteristics of the bitumen in the mix
TBOO pen is the temperature at which the penetratien would I
be 800. This is obtained by extrapolating the experimental T 800 pen ( ternperature at which the penetration is 800
log penetratien versus temperature line to the penetratien
value 800.
At low temperatuJes and/or high frequencies the stillnes.
\,
,I
0.1 mm) is 64°C
PI I penetration index) is 0
modulus of all bitumen, a,ymptotes to a limit of approximately Connect .02 seconds on time scale with température
3 x 10? Il/m2. \, difference 64-100C on température scale,
Units; \, Record stiffness on network at PI=O
I tl!m2 = 10 dyn/cm2 = The stiffness of the bitumen determined with this
I, 2
1.02 x 105 kgflcm2 = 1.45 x 10-4 tbl in.2 Nomograph is Sb = 2.0 x lcf3 N/m .
1 N ./m2 = 10 P \,
FAEOUWey, He.
K5LA, AuguIt 1953, 3rd edit ion 1972
IJ t, 4 Z

10'0

TIME OF LOADltlG I

Figure 6.1 Nomogram for determining the sliffness modulus ofbilumcn (Van de Poel) (64).
 Smix- Sbit curve
tor open stone aSl?halt

Smix

106 L-----'-;:- ----'-:;:-- -"-;-- --"--;:;-- -'-::;- ---'-:

102 103 la' 105 106 107

Sbit-

Figure 6.2 Smix - Sbit relation for open stone asphalt (master curve).

The boundary con di tions required for Figure 6.3 are the volumetrie percentages of
bitumen and mineral aggregate. These ean be derived from the mix eomposition in the
following way:

if 11Is + m, + mr+ 11Ib = 100%

in which:
Vb = volumetrie percentage ofbitumen
Vg = volumetrie percentage of rninèral aggregate
111 = mass percentage
d = density (kg/rn")
da = density ofasphalt mix with voids (kg/rn'')
dm = dcnsity of asphalt mix without voids (kg/rn'')
HR = voids ratio of the asphalt mix
The indices s, Z, f and b refer respeetively to stone, sand, filler and bitumen.

39
 STIFFNESS MODULUS
OF THE BITUMINOUS
MIX IN/m2)
80 90
...... b, ::::::". -...
4
70 .... <, <, ,........ <,
MINERAL
AGGREGATE GO.... r-, r-, r-, 1""- "i'>,. r-, 1"- 2
r-, I.........
% VOLUME lVg)
..........
b-. ~ ~---~-----
40 I'.... r-, r-, r-, 1""- ~I

,
r-. i">. , <, ,
20 " 1" " " 95

.' -, <, 1""-"~ -, <, <, '" "r-, 4

I" "'t f"-. r-, r-, 1"-

<, <,
O~" 'J".,
f'. " 2

I" .........
~ .~
r-, r-....... r-, r-, 1""-
<; i"....f'.. r-, r-,
sn FFNESS MODULUS
I R-~"I""
'"
""- 85
OF THE BITUMINOUS
BIN DER (N/m2)
, ...... ,
I Vb I' !:'\ , <,
"-
I <,
I" "
"'- 4
100 40 30 20110 5 -, '"'" ~ r-, <, .......... 1 .......... ......... .........
75-

2
~'\.'\ \ .'\,Ji '1 ~
<,
.........1""- r-,r-, -, 2
-, I\. 1\ 1"- <, <,
.","r-,"" -,
<,

109
"" -, f\.'\\ ""\ \

I I'" -, r-,
G5
-
:=
8
G , I ~ <, <,
::-
"-
'-"~ ~\~I .,1, ......
r-,
4
,,~ <,
Ï'.. 1""- 50
-
-
4

2,""---.---~
~ '" ,,"- ~ 11'.. ~
, -
-
108
8
G
~

CI.~
" ~
~
30

10
=
:
-
-
4 ""-'"
TE::·····.:···1
MiNEÀÀL
\\ ~
~'\ ~

I~
2 80.5 A~G'ÀÈGÀTE
. ~:-:-:"~i '\
."" f'\."' ~ ........

107
8 G.4 I
I
"~ I~,~U~.;~N AIR Va ~
~
,
G VOLUME% "-
5 10 20 30 100

,." ""'N'''' MODUCU"

0' TH'
BINDER 2 x 108 N/m2
""OV,",D f'''N''''
MODULUS
BITUMINOUS BINDER
Vbo VOLUME OF BINDER 13.1% OF THE MIX % VOLUME (Vb)
VOLUME OF MINERAL AGGREGATE 80.5% 1.1 x 1010 N/m2

Figurc 6.3 Nomogram for dctcrmining the stiffness modulus of asphalt mixes (Bonnaure et al)
(65).

40
If the dcnsity of the various eomponents is not known precisely the following general
values can be used for initial guidanee:

bitumen: dm = 1020 kg/rnê

granite: dm = 2850 kg/rn>
basalt: dm = 2950 kg/rnê
limestone: d~ = 2700 kg/rnê
river and dune sand: dm = 2650 kg/rn"
The nomogram, Figure 6.3., has an aeeuracy of the order of 1.5 to 2 and is valid for
Sbit> lOGand Smix > lOS NIm:!.
For lower values of Sbit and Smix other factors sueh as partiele shape and partiele size
distribution ofthe mineral aggrcgate play an obvious role. The stiffness modulus can be
detcrrnincd from eithcr statie or dynamic tests.
For mixes with a voids ratio ofupto 5% the stiffness modulus can be estimated from the
relationship (66):

2,5 Cv)n
Smix = Sbit ( I + n' I _ Cv

in whieh:

11
4, 10
=0.83 log ( ~
1
°)
C = volume of the mineral aggregate
v volume of thc mineral + bitumen

Sbit = stiffness modulus of bitumen (NIm:!)

6.3.2 The strain at break

In order to calculate an asphalt construction it is neeessary to know the strain at break as

well as the stiffness modulus.
Asphalt mixes are sensitive to fatigue and the value of the strain at break reduees the
more oftcn the material is loadcd. The following general fatigue relationship canbe
applied:

in whieh:

N the number ofload repetitions, ofsize a, at whieh the material fails

a magnitude of the applied stress
k and a = eonstants for a particular type of mix with a certain stiffness
modulus;

41
 the factor a, in gcncral, lies between 3 and 7 and stands for normal,
slightly fat mixes, usually at 5. .

The fatigue relationship between the number of load applications and the initial strain at
failure can be determined from dynamic tests or from Figure 6.4.
The Miner's Law is uscd, for conditions involving varying loads. This states that for each
applicd N loading cycles of a certàin load value on the material, which has a loading
cycle number at failure of N, the amount of damage wiII be proportional to 11/ N. Failure
.will occur when the summatien of the damage amounts reaches a value of l. For a
combination of Nj applied loading cycles of amplitude 0j the failure limit is reachcd
when the following condilion is satisfied (67):

A nomogram has been developed (68) in which the initial strain al failure related to the
asphalt mix parameters Smix, Vb and PI, can be estimated in relation to the number of
load applications, see Figure 6.4. In the nomogram loads which can be considered as
producing constant strain are scparared from those producing constant stress, see the two
different Smix scales.

s.Mf(l@\S,lIlod.Jl,J.s
of tbe ",ic

E IJ: sMfl"fSS mod...lus of tne 1I',i.:

3.3 x 1Jt N/pa2
! 1
tit ..l1Ien vlIh.Jnle: 13 -I.
" 13 1S 17 19 per'ldration indn: -0.7 • 1:.5
inltiit duin for iI lIfe ti'lll! of:
'0' (yellS at eh strus tn.t:1.11:l-4
vs. VClU~ETR:~ BITL:MEN(-/.l
1:>1tJdn at [te stn,n tut: ".5 1:
CCNHNT INITIAl STRAIN
1·1 u(oured bitumen

Figurc 6.4 Nomogram for dctcrmining thc failure strain of asphalt (68).

42
The basis for the nomogram are the formulas:

I. constant strain

f:o = (4,102 x PI - 0,205 x PI x Vi, + 1,094 x Vi, - 2,707) x S,;:;O.36 x lVO.2

2. constant stress

f:o = (0,300 x PI - 0,015 x PI x Vi, + 0,080 x Vi, -:- 0,198) x S,;:;O.28 x lVO.2

in which:

fO = initial fatigue struin

PI penetration index of recovered bitumen
Vb = volume of binder
Sm = stiffness modulus of the mix
N = number ofloading cycles which cause the material to failure
If the loading stops after some time the asphalt (the mechanical properties) will reeover
to some extcnt. This aspect is known as healing.

6.3.3 Permanent ViSCOllS deformation

Under longer duration loads asphalt is viscous and permanent deformation can oceur.
The stiffness moduli in this range are low and eannot, in general, be obtained from
Figure 6.3. To determine these parameters it is neeessary to carry out statie and dynamic
tests, such as creep tests, on each particular mix.

Cl
.5 Pm
'U
Cl
.s

_ deformation

Figurc 6.5 Thc Marshall diagram.

43
Much use is made ofwhat is referred to as the Marshall test to determine the resistance to
deformation (6). This test gives only values for comparison with other mixes and not
specific quantities. In these laboratory tests, samples are subjected to specified loads and
the deformation up to failure is rccorded.
The load at failure is referred to as the Marshall-stability, Pm; the deformation which
develops to failure as the Marshall-flow, F~. The relationship between these parameters,
recorded in the test, is given in the Marshall-diagram, see Figure 6.5.

6.3.4 Tize Poisson-Ratio

The Poisson-ratio gives the rclationship between the strain in a sideways direction and
the strain in the direction ofthe load. It can be estimated using values of Sbitumen and the
voids-ratio ofthe mix in Figure 6.6 (11).

1.0

08

0.6

0.4

0.2

Poisson
ratio

0.1

0.08

005

0.04 L- __ -1.. ..L-_--'--_l--L- __ -'

8 10 20

~ verets ratio

Figurc 6.6 Thc Poisson-ratio.

6.3.5 Temperature scnsitivity

In order to deterrnine the properties of asphalt the surrounding temperature must be
known. It is not sufficicnt, to consider only the maximum possible temperature, since,
although the stiffness modulus is the lowest, the strain at failure is the highest. A solution
can be to use the stiffness modulus and the strain at break for a range of temperatures
and to choose the least favourable condition.
The temperatures can vary between some degrees below zero in sharp frosts or, if the
cover is under water -2°C (supercooled seawater), to more than + 50°C in direct
sunlight.

44
6.4 Stability

6.4.1 Stability of the asphalt

If bitumen is laid in a layer on a slope, it will, because of its own weight and viscous
properties, tend to flow down the slope, see Figure 6.7'and Appendix IX. An internal
shear force must be mobilized to resist this flow.
Asphalt is a mixture ofmineral aggregate and bitumen and the internal force which tries
to resist the viscous flow is also developed by the friction between aggregate particles and
is thus, to some extent, dependent on the normal pressure which the particles exert on
each other. The bitumen has a lubricating action as a result of which the intern al
resistance is less than that found in the aggregate alone. The quantity of bitumen in the
mix, therefore, has considerable influence.

Figurc 6.7 The viscous behaviour of bitumen on a slope.

6.4.2 Stability of the revetment as a whole

The revetment should be so heavy and extensive that it cannot move as a who Ie under
the loads acting.

6.5 Durability

The revetment must, with the course of time, continue to fulfil its function. The charac-
teristic mechanical properties should not deteriorate too much within a reasonable time.
The following aspects will affect this.

6.5.1 Aging

Bitumen hardens with exposure to light and atmosphere. Material properties, such as the
stiffness modulus, strain at failure and viscosity also change. This effect is more notice-
able when the temperature is higher and the voids ratio larger. With a dense asphalt mix,
such as asphaltic concrete, aging only takes place in very thin surface layer.

45
Thc hardening of bitumen which occurs during mixing, handling and placing must be
taken into account when designing the revctment. In general a loss of bitumen pene-
tration of 10 to 25% is to be expected, dcpending on thc type of mix-plant.

6.5.2 'Stripping'
Water can strip the bitumen from the surface ofthe mine ral aggregate and causc thc mix
to deteriorate.
Thc danger of stripping is less wh en the voids ratio is low. From tests (12) it has been
found that, for mixes with a voids ratio ofless than 6% there is no danger ofstripping. For
mixes with a voids ratio of less than 10% there is little dangcr.
A method ofmeasuring the deterioration ofa mix is thc 'Immersion Compression Test'
which originated in American road engineering (39). In this test the compressivc strcngth
of blank samples is compared with thosc that have been kcpt under water of a certain
temperature for a given period. Thc ratio found between thc compressive strengths ofthe
blank and immersed samples is called thc retained stability.

6.5.3 Resistance to erosion

Flowing water can erode asphalt mixes and, especially when solid matter is carried along
the extcnt of crosion as aresuit ofimpact forces on the asphalt surface can be substantial.
The stresses in the asphalt caused by such impacts, will increasc with the hardness of the
binder. The resistance of thc material to these stresses, howcver, is not directly propor-
tional to the degrec of hardness. From these considerations thc following nile has been
formulated:
The lower thc minimum temperature at which erosion can be expected the softer thc
bitumen should be (12).

6.5.4 Biological resistance

Organisms can, to a greater or lesser extcnt, affect asphalt.

In thc tidal zone damage is often found which has been caused by algae and seaweed. In
higher zones ofthc slope certain p1ants can damage asphalt with their roots and runners
(13).
Algac damage (wire weed).
As an algae layer dries out it shrinks and exerts a shear force on the rcvetment which
can damage the asphalt surface, sce Photo I.
This can bc prcvented by:
I. A seal coat of bitumen, emulsion, road tar or tar bitumen. Thc latter two have thc
added advantage that they are very rcsistant to oil products but have thc disad-
vantage th at they can, to a certain extcnt, damage the environment.
The surface of thc rcvetment is made smooth by thc seal coat and there is lcss

46
 opportunity for nlgae to attach itself, In or near the tidal zone it is therefore beller
if the seal coat is not blinded with chippings or shell grit.
2. Bladder weed. A covering of bladder weed slows down the drying out process of
the algae.
3. Chemical treaunerus. These methods, unfortunately, are environmentally unac-
ceptable.

Photo I An esphaltic concrete reverment being auacked by algae

Marine borers.
These animals which are to be found in the lower part of the lidal zone, exen during
their growth pressure farces in the cracks and holes where they are established.

47
 Musseis can then grow in the spa ces which develop and sa the damage progresses.
This damage can be prevented by:
I. A seal coat as described above.
2. A thin layer of bitumen which kills the organîsms.

Plant damage.
Plurus cao exert considerable force with their roots, and runners, sec Photo 2.
When there are seeds in the subsoil, plants can grow through the asphalt layer. This
effect is strcngly influenced by the type ofsubsoil; in dredged spoil there will be na or
very few seeds. The environment can, also, be important, partienlar in salt or sweet
water areas. lt is possible to treat the soil with a planticide. but in this case tne environ-
ment can be damaged.
A smooth espbalt surface in which there are na cracks or holes offers very few
auachrneru points for seeds which come onto the dike. A good seal coat cao,
therefore. be an important deterrent.

Photo 2 Plant growth damaging an asphaltic concrete revetment

6.5.5 Chemica! damage

Bituminous matcrials are chemically inert apart from with some carbohydrates. The

48
concentration ofcarbohydrates must be very high before they can cause real damage. A
surface treatment of road tar or tar bitumen can be used for proteetion but this often is
environmentally unacccptable.

6.5.6 Other forms of damage

Other forms of possible damage to the revetment are:

I. traffic, both during construction and operation;
2. vandalisrn, recreational activities, etc.;
3. floating debris;
4. vesscls, anchors;
5. ice.

6.6 Workability

In order to obtain the best possible adhesion between the mineral aggregate and the
bitumen the asphalt must be uniformly mixed. To obtain th is the bitumen must have a
low viscosity. This can be achicvcd by:
a. Heating thc bitumen to 150-200°C (viscosity 150-300 mm-/s); th is is then referred to
as 'hot-mix asphalt'. Hot-mix asphalt is prepared in an mix-plant in which the
bitumen is heatcd to a specified viscosity and then mixed with dried and pre-heated
mineral aggregate.
b. The bitumen can bc diluted with a solvent, for example, a suitable petroleum dis-
tillate or can be emulsificd in water; the solvent or the water has to vanish out of
the mate rial beforc the asphalt mix reaches stability.
Under water or in a humid environment these methods are unsuitable. For this
reason they are rarely used in hydraulic structures, except as an emulsion for seal
coats and tack coats. Also the danger of stripping is greater than with hot-mix asphalt,

After production the mix is transported to the site and placed. The specificatien for mix
composition generally depends largcly on thc workability, since total mechanised
handling and compaction. using heavy equipment, is generally not possible.
The viscosity of the mix required for handling is strongly dependent on the viscosity of
the bitumen and, therefore, requiremcnts for the tempcrature are given. Hot-mix asphalt
must be plueed at the correct temperature on account of the time needed for handling
and to achieve an optimal compaction.
For thickcr asphalt layers the cooling pcriod will he longer and, therefore, more time is
availablc for placing and compaction. In th is respect the weather conditions are
important. When mechanical compaction is involved it should be borne in mind th at the
outer and inner surfaces of the layer will be cool while the inside is still hot. The thicker
the layer the stronger is this effect. The possibility of building up the revetment in several
layers can then be considered.

49
Overfilled mixes need not be compacted but underfilled mixes can be compacted in
order to reduce the voids ratio. This process can be achieved by either mechanical
compaction or compaction under the own wcight of the matcrial.

Mechanical compact ion can be carried out by tarnping. or by using rollers or vibrators.
The degree ofcompaction can be affected by:
the bitumen content;
the content and type of filler;
the mineral aggregate content;
the partiele shape of the mineral aggregate;
the partiele size distribution;
the temperature of the asphalt and the surroundings;
the layer thickness;
the weight and type of compaction equipment.

Compaction is more difficult on steep slopcs; the component of thc weight of the roller
vertical to the slope is smaller and the handling of equipment is more complicated.
In order to neutralize the roller weight parallel to the slo pc, the equipment can be
attached to winches located on the crest. Compaction can be carricdout effectively,
without these special provisions on slopcs more flat than I : 4, sec also Section 22.3.1.
From an acsthetic point ofview a smooth finish to the revetment is desirabie. Sometimes,
however, there is a tendency to prolong the compact ion process in an attempt to obtain a
certain smoothness. When th is happens the asphalt can become too cool and (initial)
cracks can develop. Rolling must be siopped if the asphalt becomes 'live', th at is to say,
ifwaves in the asphalt propagate in front ofthe roller.
Ifthe mix is not very workable (initial) cracks can develop aftcr, only a few passes ofthe
roller, especially at the bottom of the slope whcrc the material may start to flow. Rolling
must then be halted to allow the material to cool down. At the same .time the mix
cornposition should be adjusted.
Generally it seems that the compactability is reduced as the stability of the mix is
incrcascd, .

Compaction due to the weight of the material applies with what are known as, gap-
graded mixes. These are mixes for which there is a 'gap' in the mineral aggregate grading
curve; a particular partiele fraction, usually the fine stone fraction, is missing. The coarse
stone fraction farms the stabie skeleton of which the voids are completely filled with
sand mastic - in the case of a dcnsc asphalt mix - or only coated in the case of an open
asphalt mix.
The object is to obtain a mix which draws on the skeleton ofthe coarser stone to provide
satisfactory stability against flow during handling and which, thanks to the viscous
mortar, needs only a small amount of compaction energy to produce the required voids
ratio. The stability of the mix in the completed stage must also be taken into account.
This may possibly be in contradiction to the foregoing requirements.

50
Because they need not be compacted gap-graded mixes can be used at locations where
compaction would be difficult.

6.7 Environmentai aspects

The following points relate to the possible darnaging effect of asphalt to the environ-
ment:

The binder in the mix is generally bitumen.

Normal straight-run or blown bitumen contains polycyclic aromatics (pca's) which
are dangerous to health. The quantities involved are so small, however, that there is
no dangor to the environment, even when the bitumen is used in water catchment
areas.
Sometimes tar is used. This material contains, however, a much higher proportion of
pca's than bitumen and generally is regarded as dangerous to the environment.
Application in water catchment areas is not advisable.
Bitumen is generally worked at high temperatures, and then vapour is given offwhich
if inhaled excessivcly can lead to nausea, Under normal conditions vapour concen-
trations are below the danger level.
Care must be taken for burns (82, 91).
Mineral aggregates can also, depending on their place of origin, contain dangerous
materials, An example is the use of f1y ash originating in waste incinerators as filler.
Normal sourees in the Netherlands, however, have not caused significant problems.

51
7 Use of asphalt produets in hydraulic structures

The most important mix types applied in hydraulic structures are:

I. Asphaltic concrete
2. Asphalt mastic
3. Grouting mortars
4. Dense stone asphalt
5. Open stone asphalt
6. Lean sand asphalt
7. Asphalt membranes
Mixes are defined by the constituents, the nature ofasphalt mix and the binder material.

7.1 AsphaItie concrete

Asphaltic concrete is probably the best known mix type. It is a mixture ofcrushed stones
or gravel, sand and filler in whieh the pores are practically completc1y filled with bitu-
mcn.The voids ratio is 3 to 6%.
In general thc material must be compacted and is unsuitable for application underwatcr
or in the tidal zone. In view ofthe sm all voids ratio rcquircd, sec Section 9.2.1., asphaltic
concrete can be considered to be impermeable.
Asphaltic concrete is applied as a watertight dike revetment above the mean high
water level, and as a lining for canals, reservoirs etc.

7.2 Mustie
Mastic is a mixture of sand, filler and bitumen. There is more bitumen available than
necessary for filling the voids in the sand filler mixture. The mix, therefore, is naturally
dense and need not be cornpacted. Mastic can be poured at working temperatures and is
used, for asphalt slabs above and under water for lining or as bed and toe protection.
When cold, mastic forms a viscous quasi-statie mass.

7.3 Grouting mortars

Grouting mortars are hot-type mixes of sand, filler and bitumen of which there is
more than required to fill the voids in the mineral; stone and gravel can be added if
necessary. These mortars are used for grouing stone revetments above and below wat-
er-level, and also for slab construction.

52
7.4 Dense stone asphalt

Dense stonè asphalt is a gap-graded mixture of stone, sand, filler and bitumen. The
amount of bitumen slightly overfills the mixture. The material is, therefore, water
impermeable.
It is used as bottom and slope proteetion and also in toe construction.

7.5 Open stone asphalt

Open stone asphalt is a gap-graded mixture of mastic and stone - a stone frequently
used is limestone 20/40 mmo Mixing is carried out in two stages. First mastie is prepared
and secondly it is mixed with limestone. The mastie binder only coats and eonneets the
limestone particles together.
It is an 'underfilled' mix and, beeause of its open strueture, should not be plaeed under
water except in the form of prefabricated mattresses ..

7.6 Lean sand asphalt

Lean sand asphalt is a mixture ofsand, often locally obtained, with 3 to 5% bitumen. It is
a greatly 'underfilled' mix and the function of the bitumen is simply to coat the sand
grains and bind them together. After some time the permeability is very similar to the
sand from whieh it is made.
It is used as a eore material for reclamation bunds, filter layers and as permanent or
tcmporary eover layer above and bclow water-level.

7.7 Membranes

Membranes are th in impermeable watertight layers of bitumen which are prepared

in-situ or prefabrieated. Membranes are used as impermeable linings for eanals, banks
and water courses.

53
8 Quality control

Quality control is essential in order to guarantee the quality of the work during the
execution and on completion. Con trol includes mix design tests.before the start ofwork,
production con trol during the execution and checks on completion of a part or the
complete work.
In road construction there are already many instructions and directives for carrying out
quality con trol, for example, those given in Eisen 1978 (6), V.D.e.W. (18) and A.B.C.W.
(19). There are, however, no such official instructions forthe use ofasphalt products in
hydraulic structures. The construction requirements and the working conditions often
appoint con trol checks required (20).

8.1 Mix-design tests

After the tender has been drawn up the definitive mix composition is decided, based on,
(extensive) laboratory investigations. The type of investigation depends on the mix
properties required and the application in prospect. Stability and voids ratio, for
example, will be more relevant to asphaltic concrete revetments, whereas viscosity wiII
be more relevant to a mastic slab.
During the investigations the suitability ofthe basic matcrials given in the tender is also
examined.
The investigations should lead to a technical and economie optimum mix, taking into
account the basic requirements and the matcrials available. Next the planned produc-
tion and working methods are ehecked to confirm that they will yield the best results.
The produetion method for the material and the execution method often determine the
quality and cost of the work.

The mix-design tests on the basic materials are more or less the same for all asphalt
mixes; they include, for example:
an assessment of general information such as the type of material and identification
of its source. The identification of its souree must include the name of the sup-
plier, location ofthe source, the nature ofthe material and the quantity for which the
identification is valid. Eaeh delivery ofconstruction matcrials conform an identifica-
tion ofits souree should be accompanied by a written authorization
investigations into pollution aspects
sieve analyses: stone/gravel; sand; filler
determination ofthe bitumen number and sensitivity ofthe filler to water
determination ofthe penetration and the softening point ofthe bitumen
determination ofthe density ofthe mineral aggregate.

54
 ...
The design tests for the various mixes differ for each type of asphalt. Refcrence should be
made to Part D, Chapters 9 to 15 inclusive.

8.2 Production control

Production con trol is necessary to guarantee the quality of the final product.
It includes:

1. Investigation of the quality of the construct ion materiaIs, for example:

sicvc analyses of:
• stone/gravel
• sand
• filler
dctermination ofthe bitumen number, the voids ratio, and fraction <63 urn ofthe
filler
determination ofthe penetration and softening point ofthe bitumen
determination ofvoids in sand/filler mixes, with the Engelsmann apparatus.
2. Check on composition ofthe mixture.
3. Check on properties of the mixture.
4. Check on production of the aspha1t mix.
5. Check on handling, working and placing ofthe mix.

Production controls are normally carried out by the contractor. The principal can also
carry out quality control during the execution of the work. In principle, however, from
the very beginning. the principal should assume that the contractor has mastered the
execution process and maintains good quality control. The control by the principal can
include checking the methods used for production control. This can be in the form of
tests on samples (by an independent organization). For these tests:

the resu1ts should be available relatively quickly

sampling and testing methods should be comparable to those used by other parties
involved in the work
supplementary agreements can be made concerning the test procedure.
The instructions, such as those givcn in A.B.C.W. (19), can be used a general guidelines
for asphalt mix production con trol.

8.3 Acccptancc eon trol

The principal carries out acccptance con trol to confirm that the basic requirements have
been satisfied, Ifthe work has not been carried out satisfactorily then penalties should be
imposed.
Acceptance control for road works are dcfined in Eisen 1978 (6) and V.U.C.W. (18). A'
part of this con trol can also be applied to hydraulic structures but obviously cannot be
interpreted in exactly the same way.

55
 '"
Acceptancc control is normally carried out after a part or all of the work is ready. Ifthis
is not possible the checks must be carried out during the work.
The parameters which are most frequently checked are given, per mix type, in Table 8.1.

8.4 Sampling
Good sampling technique is essential for rc1iable production and acceptance contro!.
Samples are required of:

a. basic materiais;
b. asphalt mixes.

8.4.1 Construction materials

Sampling and tests must be carried out sufficiently early so that results are available
before the material is used.

G RAVEL, STONES AND SAND

Sampling of these matcrials should, for the time bcing, be according to the rules given in
the Dutch standard NEN 3542. The sampling frequency is discussed in A.B.C.W. (19).
Samples should be taken when the material arrives at the site, preferably while it is still in
the transporter equipment or during unloading. Whilst there are no final standard rules
on sampling, a specificatien is required for each construction.

FILLER
A.B.C.W. (19) can be used as a guideline on sampling frequency. Samples should be
taken, on arrival, before the material is dumped in the silo. In order to verify the quality,
samples can be taken out of the supply to the mixer.

BITUMEN
The sampling method should be as given in the Dutch standard NEN 3940 (21) and
A.B.C.W. (19).

8.4.2 Mixes

Sampling is needed for:

- production con trol and to check the production con trol

- acceptance contra!.

Depending on the particular aspect being investigated, samples should be taken on one
hand, at the mix-plant, and on the other hand, on site.

56
There are two methods for mixer sampling:

al or in the skip
out of the loaded truck.

The methad of samples can be taken in accordance with A.B.C.W. (19) or any other
systems specified in the particular tender.

The method of sampling, out of the completed werk. is often dependent on the
accessibility ofthe material in the structure and the possibility of obtaining samples with
a saw or by taking cores. Usually cores are taken for road construction works, but this is
nol atways possible Ier hydraulic structures. Taking cores is, if possible, in general the
prefered method of sampling (Photo 3).

Photo 3 Taking a sample out of a esphaltic concrete revetrnent by core boring

Table 8.1. Completion checks for various asphalt mixes.
Completion check
mix
cornposition
c
.9 -0

.,
ë E <J
U
.,c:: '"...
'"
'" ë '"
ti '"
0
c c:: ~:I ~
_"" <:l
0
:-.,
.:.:
u
0
U ~'" .9
0
c- '-6
o .....0-

:s... c::
0
E
'"
c
'";.""
c:;...
'"
'-
0
..c
el)
c
'ëh ~
s.s
>,-

C 0
.0-
e.o5
c ...
.- 0
0
>, 3 -0
'(3 ë.. eo c ",-
"';;c possible sampling
mix ~ :E
0
'Vi ;. ""
-0
'"'" :~ ::l
<TE'" 0'-
",-0 method"
asphaltic x x x x core boring
concrete
asphalt Xl x x core boring
mastic bulk sample
penetratien x x x2 bulk sample
mortar
dense stone x x x core boring
asphalt bulk sample
open in-situ x x x x core boring
stone
prefab. x x x x bulk sample
asphalt
lcan x core boring
sand-asphalt bulk sample
mem- in-situ x x bulk sample
brane
prefab," x x

I Check on layer thickness by core borings or by special thickness probcs, also carried out under water.
2 Desirabie but often impossible to obtain.
3 Testing of the specifications of the manufacturer;
Execution tests, specially joints;
Test on functional requirements:
- impermeability ofthe whole surface, includingjoints
resistance to root growth
- ability to deforrn without leaks devcloping
- durability
- f1exible joints with other structural components.
4 Core borings are preferrcd; these, however, are not always possible. An alternative is bulk
samples per quantity of in-situ matcrial.

58
PARTB

MATERlAL TECHNOLOGY
Summary

Part B discusses aspects of material technology of the various asphalt mixes which arc
most used in hydraulic structures.
These are:

asphaltic concrete;
asphalt mastic;
grouting mortars;
dense stone asphalt;
open stone asphalt;
lcan sand asphalt;
asphalt membranes.

To make this part of the manual more readily applicable it has been subdivided into the
different asphalt types. So that separate sections can be read without cross-reference
sometimes information has been repeated.
Three separate aspects have been discussed for each asphalt type, namely:
The basic matcrials
The matcrials, of which the asphalt type is composed, are discussed together with
rclatcd construction parts.
Mixes
The composition ofthe asphalt and the way in which the mix is designed is described.
The method of mix-design testing is also discussed. These are the investigations which
must be carricd out to determine the most suitable mix composition before the work
can be started,
Mix propertics
The choice of a particular asphalt type is based on the mix properties. Since the
components and the composition vary, the properties ofeach mix type also vary. The
relevant mix propertics are described for each mix type. These properties form the
basis for Part C of the manual which deals with the technical design.

60
9 Asphaltic concrete

9.1 Basic materials

9. I.l Crushed stone

Crushed stone can be frorn local sourees or importcd.

Limestonc gives a good adhesion, but can be expensive if not locally available.
The requircmcnts for crushed stone, used in bitumineus mixes for hydraulic structures,
are given in Eisen 1978 (6).

9.1.2 Gravel

Because of the fa ct that gravel possibly gives less adhesion and the interrial stability of
the mix is lower in the hot condition, for sorne time the use of gravel was abandoned.
Since the lower internal stability never has been proved sufficiently, however, there is,
for the present, no reason not to use gravel.
Reference should be made to Eisen 1978 (6) as far as this is related to hydraulics.

9.1.3 Sand

In principe every kind of sand can be used. A well-graded sand is desirabie since
otherwise high percentages offiller must be added in order to reduce the voids ratio. This
can lead to mixes which are difficult to handle and are more expensive.
A sand, that lies in the optimum area ofthe sand triangle, Figure 2.1, is well-graded and
satisfies the rcquirements for Sand Type A, given in Eisen 1978 (6).
The following points relate to sands Iying outside the optimum area:
There is no experience with sand that lies to the right of the line for natural sand, sec
Figure 2. I. As far as it is known such natural sands do not occur in the Netherlands.
Sands with à grading stich asthat in Area 11in Figure 2.1 contain the finest fractions
and are very uniformly graded. Because ofthis the mineral mix will have a high voids
ratio and a relatively large amount of additional filler and bitumen would be required
to produce a dense asphalt mix. As aresuit the workability would be adversely
affectcd.
Sands lying in the lower left hand corner of the sand triangle produce good results
provided that the filler content, as indicated by a sand-filler compaction test (En-
gelsmarin method), is increased. Then workable mixes are obtained with a low voids
ratio.

61
- Sands laying in Area V ofthe sand triangle, Figure 2.1, can also pro duce mixes with
good workability. The use ofsuch sands, however, is rather expensive.

Stability during the construction phase can be improved if some of the natural sand is
replaced by crushed sand; this is a more angular matcrial. On the other hand the
workabilitycan be influenced adversely by this.

9.1.4 Filler

'Very weak' fillers are used in hydraulic structures, being fillers with a very low voids
ratio which require relativcly very little bitumen. Ideally limestone fillers should be used
because they give better adhesion in a humid environment.
From experience it appears that mix workability is increased by using 'weak' rather than
'very weak' filler, while there is no noticeable affect on the required voids ratio.
For further information, reference should be made to Eisen 1978 (6).

9.1.5 Bitumen

Concerning the required flexibility of the revetment it has been found that bitumen
80/100 pen is the most suitable in the Nctherlands. For bitumen requirements, sec Eisen
1978 (6).

9.1.6 Other constructional aspects

Asphaltic concrete is, preferably, covered by a seal coat. This can consist of a sealing
layer and/or a surface dressing. Thc function ofthe seal coat is to:

I. cover surface cracks and open texture;

2. limit temperature affects due to the sun;
3. improve the aesthetic appearance.
The surface dressing comprises a sprayed layer of bitumen cmulsion, Type Unstable,
about 1.0 kg/rn", which, above the tidal zone is blindcd with chippings, for cxample, 10
to 12 kg/rn'' ofsize fraction 4/8 mm, or shell grit (16).
The seal coat is preferably preceded by a proteetion layer which must be applied as
quickly as possible after the asphaltic concrete has been placed. This layer comprises
sprayed bitumen emulsion, Type Unstable, about 1.0 kg/rn-. For aesthetic and practical
reasons it is rccommcndcd that the emulsion is spraycd from thc top of the slope
downwards.

62
9.2 Mixes

9.2.1 Mix design

The as ph alt ie concrete uscd in hydraulic structures is a hot-preparcd mix of crushed

stone or gravel, sand, 'wcak' or 'very weak' filler and bitumen 80/100 pen. The object is
to design a mix which has the lowest possible voids ratio and still is sufficiently stable,
The voids ratio of the mix depcnds, irrespective of compaction, on the voids ratio of the
mineral aggregate and the bitumen content.
It appears that the minimum voids ratio in the sand-stone mixture, without filler, is
between SOand 60% by mass of normal crushed stone or grave!.
For a sand-filler mix there is also a minimum voids ratio for a certain sand-filler
proportion (Figure 9.1).

In principle the filler content can be chosen so that the voids ratio in the sand-filler mix is
minima!. In practicc, howevcr, for economie reasons, a lower filler content of7 to 8% m,
of the complete asphalt mix, is often chosen. Around the minimum voids ratio value
changes in the filler content have only a lirnited effects on the voids ratio (Figure 9.1). It
is also practical to choose a filler content somcwhat to the left ofthe minimum since filier
contents to the right of the minimum value shown on the curve result in an undesirable
increasc to the volume of the voids.

34

33

~
Ë
~ 32

-c
c
"~ 31
~
<C

2 30
~
"
.~

29 f1ller centent
selec.ted

28

10 ~o 50
.---. moss percentage- of fitter in tht> ~nd - Mier milt

Figurc 9.1' Rclationship between voids ratio in a sand-filler mix and thc filler content in thc mix.

63
The composition of the mineral aggregate content is determined in the way described
above.
The bitumen content selected depends on:

I. The form of the curves which relate voids ratio ofthe mix to the bitumen content, sec
Figure 9.2.
2. The required voids ratio compared with the voids ratio found in Marshall samples,
compacted by 50 to 100 strokes on one side, sec Figure 9.2.

Whether or not a mixture with the chosen bitumen content is also suitable to be placed
and compactcd, also on slopcs, can only be found out in practise. Only limitedinforma-
tion can be obtained about th is aspect by doing slope tests in the laboratory.
With asphaltic concrete used in hydraulic structures, containing for example, Sand A
and/or fine sand it appears that increasing the bitumen content, and thus lowering the
viscosity, lowers the voids ratio to a certain limiting value. A mix with a bitumen content
of 7.5% m in 100% m mineral is difficult to handle because it is insufficiently stabie
in the hot condition.

13

12

11

~ 10

5 strokes

10 strokes

5
~ ~ 20 strokes

_ ~ 50 strckes
100 stroi<es

6.0 6.5 7.0 7.5 80

percentage cf bitumen.
by rr-c s.s , in 100% minerel

Figure 9.2 One-sided compaction-Marshall samples (16).

Investigations by the Commissie Verdichting Asfaltdijktaluds (Dutch commission for

thc compaction of asphalt dikc revctments) (16) indicatc the following suitable mix
composition (see also Eisen 1978):

64
 Asphaltic concrete for hydraulic structures sh9~ld be a hot-prepared mix of crushed
stone 6/16 mm, sand, very weak filler and bitumen 80/100 pen for layers less than or
equal to 150 mm thick.For thicker layer crushed stone 6/22 mm should be used.
In practice, in the last few years, a bitumen of6.5% m has often been used to improve
the workability, especially the stability in the hot condition.
When asphaltic concrete is in place the voids ratio should not be greater than 6% m on
slopes steeper than I : 4 and not greater than 5% m on slop es of 1: 4 or less.

Table 9.1 Asphaltic concrete mix composition us cd in hydraulic structures.

% by mass
thru sieve on sicvc required min. max. toleranee
C 22.4 2mm 50 48 55
2mm 63 firn 42 37 47 :!:5
631lm - 8 7 10 :!:1
bitumen content
in 100% mineral 7 6.5 7.5 :!:0.5

The voids ratios above give, in general, mixes which are durable and have good wor-
kability.
In practice often for aesthetic reasons, mixes are compacted too long and too heavily so
that initial cracks develop (59, 89). This can normally be prevented by using a good
compaction method, sec Sectien 22.3.1. Ifthe required voids ratio is so small or thc layer
thickness is so large that cracking cannotbe avoided even with a good compaction
method then the revetment must be laid in several Iayers, In general. thc thinner the
layer the beller the compaction. Naturally good cohesion must then be provided bet-
ween the layers. This can lead to problems during exccution if dry sands blows onto the
surface being preparcd. Dry sand and other'pollutants must, thcrcfore, first be removed.
At locations where dry sand is not a problem the method using severallayers produces
good results and is to be recommended when compaction problems are anticipated. The
layers can be built up of the same material or from different matcrials which are
matchcd. For th is aspect reference should be made to the various cover layer construc-
tions used for dams and roadworks.
For conditions in the Netherlands the first solution is often adequate.

9.2.2 Mix-design tests

In order to find suitable mix compositions with a voids ratio as low as possible and which
are sufficiently stabie when cold it is necessary to carry out laboratory tests.
When these two criteria cannot be satisficd 'in combination, generally the voids ratio
requirement is given first priority. In practice, as far as it is known, stability problems
under normal cireumstances have never occurred .

65
In Eisen 1978 only the investigation of voids ratio is required; after compaction ac-
cording to a Marshall test, the voids ratio should be not more than 4%. In order to
achieve better compaclion results and to have more time available for compaction the
mix should be such that it can be compacted at the highest possible tcmperatures.
Mix-design tests can include a numbcr of'sections, such as:
investigation of the voids ratio of the sand-filler mix with, for cxamplc, the Engels-
mann apparatus;
determination of the voids ratio of thc asphaltic concrete mix;
comparative investigations by Marshall tests.

Mix propertics can be investigatcd by varying the mass percentages of thc various
components within the given lirnits. Normally, however, only the effect of the bitumen
content is investigatcd.

From the rcsults of these invcstigations the mix can bc designed. Wh en the mix has been
dccidcd thc inf1uence of the tolerances should be invcstigated, for examplc by carrying
out two tests on mixes with the designed mineral composition; one with 0.5% by mass
more bitumen and one with 0.5% by mass less bitumen. If it appears that problems may
develop then the mix composition should be modified.
A written statement must be given ofthe composition ofthe asphalt to be dclivered to the
site. Thc tolcrances, for use with this composition, as given in Table 9.1, should also be
declared. If it appears that the composition of the asphalt has to be changed during the
execution, th en these changes should be reported in writing and the tolerances should be
restatcd,
Reference should be made to Section 8.1 for the investigations to be carried out on the
basic materiais.

9.3 Mix properties

9.3.1 Mechanica! properties

During mixing and handling thc bitumen hardens and, as a result, the penetration
decreases and the softening point increases (6). In order to deterrnine exactly the extent
ofthis effect a sample of the bitumen must be recovered.
The stiffness modulus of the bitumen can be determined from the Van de Poel nomo-
gram, see Figure 6.1. The penetratien-index and the softening point of the recovered
bitumen and the relatcd loading period and temperature determine the value of the
stiffncss modulus. The Poisson ratio ofthe asphalt mix can be estimated using Figure 6.6.
The stiffness modulus of the mix can be estimated with Figure 6.3. for values of
Sbit> lOG N/m2 and Smix > 108N/m2• Preferably, however, the S-modulus should be
determined by experiment.

66
Figure 6.4. gives general values for the initial strain at failure. This value can also be
determined for cach stiffness of modulus by .experiments.
Since the Marshall test only gives an indiention of the stability and flow properties of
mixes it is only usabic for comparison purposes, for example, in conneetion with con-
struction and con trol.

9.3.2 Permeability

Permeability is largely dependent of the voids of ratio.

It appears that, for voids ratio below 3%, no water cao be forced through a test plate 5 cm
thick, under a pressure of3 atm. For mixes with voids ratios ofabout 8% the permeability
coefficient is about 10.8 mIs (12).
Usually the voids ratio of asphaltic concrete, used for slope rcvetmcnts is between 5 and
6% and the material considered to be water impermeable. If this pcrmeability does not
appear to give a satisfactory se al th en water permeability tests should be undertaken.

9.3.3 Durability

AGING AND HARDENING

During mixing and handling, the bitumen hardcns, a phenomenon which can be
measured using a thermostability test, see Chapter 3.
Changes in the penctration can vary from 10 to 25%, depending on the type of mix plant
and size of the batch. The conditions set out in Eisen 1978 (6) should be adhered too. In
view of the relativcly low voids ratio and tbc high density of asphaltic concrete, aging of
the bitumen, during the lifetime of the revetment, only takes place in a very thin surface
layer; the phenomena is, therefore, oflittle importance. It can be reduced even further by
a seal coat.

LOSS OF ADHESION DUE Ta WATER; 'STRIPPING'

With a dense mix such as asphaltic concrete there is little danger of stripping especially if
the revetment is given a seal coat.

CHEMICAL RESISTANCE
Bituminous materials are chemically resistant except to a limited number of carbohy-
drates. Beforc these can cause any damage to a dense asphaltic concrete p1ix the
concentrations in the surface water must be very high. A seal coat of road tar or
tar-bitumen can limit the effects. The environmental aspects, discussed in Section 6.7,
are then also relevant here.

IlIOLOGICAL RESISTANCE
Sec Sectien 6.5.4.

67
10 Asphalt mastic

10.1 Basie materials

10.1.1 Sand

Provided the mix propertics are satisfaetory, in principle, all sand types ean be used.
For the sand requirements refcrenee should be made to Eisen 1978 (6).

10.1.2 Pilier

Filler should be weak or very weak. Sinee the filler strongly determines the viseosity of
the mastie it should have a uniform quality and a constant absorption capacity. Sec also
Eisen 1978 (6).

10.1.3 Bitumen

The type and quantity of bitumen greatly influenee the viseosity of the mastic. A
bitumen which is no harder than pen 80/100 is reeommended. See also Eisen 1978 (6).

10.1.4 Other constituents

Polymers are sometimes added to restriet flow in eold eonditions. It is advisable, in sueh
instanees, to use a different mix composition.

10.2 Mixes
10.2.1 Mix design

Mastic is a mix of60 to 70% by mass sand, 15 to 25% by mass filler and about 20% by mass
bitumen. lt is an overfillcd mix whieh, in itself, has the properties of a highly viseous
fluid, Mastie mix design is bascd, therefore, not on stability but on viseosity.
Mastic requiremcnts, see (23), includc: .

I. The mix must be pourableat high temperatures. This depends on the viseosity at
placing temperatures; in practice the viscosity is between 10 and 200 Pa . s.
The mastie must have sueh a viseosity that, for cxample, it can be poured through a
pi pc.
If placing tempcratures are too high vapour-filled bubbles dcvclop in thc material
when it is applicd under water. These subsequently heal by mastie flow.

68
 Calculations show that, dcpending on the pipe and nozzle diameters, thc rnastic
viscosity should not be higher than 200 Pa. s for it to flow through a vertical pipe
under its own weight and without the excess pressure becoming too large (23), see also
Appendix VIII. On the other hand, however, the viscosity must not be so low that
water can intrudc into the nozzle. Naturally both upper and lower limits to viscosity
depend on the water depth.
Wh en working with a closed hopper bucket or a crane the viscosity should be 30 to
100 Pa . s under water and 50 to 200 Pa . s above water. The viscosity should not be so
low that, on slopes of I : 10 above water and I : 7 under water, excessive hot flow
develop down the slope. On the other hand the flow shuuld be sufficient to form a
continuous layer.
A viscosity between 80 and 150 Pa . s at the placing temperature is sufficient to ensure
this. (sec Table 10.1).
2. After cooling down to normal ternperatures the mastic viseosity should be sufficiently
high to ensure that it only flows within particular limits, For this temperature, slope
and layer thickness have to be taken in account. From calculations it appears th at, for
a 10 cm thick plate on à slope of I : 10 to flow a few decimetres in 7 years, the viscosity
would have to be at least 109 Pa. s, sec also Appendix IX. (See Table 10.1).

Table 10.1 Rcquircd asphalt mastic viseositics at various tcmpcrature stage:

Requircd viseosity
(Pa. s)
Asphalt mastie min. max.
In thc cquipmcnt: \70-100 (0C)
pipe 30 150
buekct, eranc: under water 30 150
above water 50 200
During cxceution: \70-100 (0C)
hot flow of a platc on
a slopc of I : 10 80 1000
Aftcr cxceution: lOeC)
eold flow of a platc 10°_1010

Some examples of mastic mixes used in the Netherlands for slab construction are:
Grevclingen Dam
• sand 63-65% by mass
• filler 15-17% by mass
• bitumen 280/320 20%
Eastern Scheldt Storm Surge Barricr
• sand 62 % by mass sand 64 % by mass
• filler 20.5% by mass filler 17.5% by mass
• bitumen 160/210 17.5% by muss bitumen 80/200 18.5% by mass
Viscosity at 130°C, 30 to Pa . s.

69
10.2.2 Mix-design tests

Mix-design tests to determine the mix composition are mostly based on the Kerkhoven
Method.
First of all the optimum sand/filler proportion is found using the Engelsmann apparatus
and based on this value a sornewhat smal1er filler content is selectcd, sec Section 9.2. I.
The tests with the Engelsmann apparatus are carried out to determine the smallest voids
ratio in the sand/filler mix in order to ensure that the related bitumen content is the
smallest possible. The bitumen content is determined from the Vsf-faetor, whieh, in its
turn, is found from Figure 10.4., sec Section' 10.3. I.
The viscosity at high tcmperatures, 120 to 170°C, is found using the Kerkhoven viseosity
meter (Figure 10.1.). With this meter, the time taken for one litre of mastic to flow out is
measured, in seeonds.

I
I 153

I 195

I 3 00

I .>: -
25

~
sa
I
i
I reecsur es in mm

Figure 10.1 Kerkhoven viscosity apparatus.

Since the flow is laminar the viscosity can be calculated using:

lJ = 43.5' 10-5 p g r ïfor the type ofviseosimeter shown in Figure 10.1)
in whieh:

11 viseosity (Pa. s)
p bulk density (kg/rn'')
g aeeeleration due to gravity (mi SZ)
( outflow time (s)

The following can be investigated:

70
I. the influence of variations in mix composition within given lirnits, on the mastie
properties.
2. the influence ofvariations in the grading ofthe sand on the mastie properties.

In praetice the scnsitivity to temperatures (Figure 10.2), of the mix is important in

conneetion with the execution. This property is determined by the mix composition and
the properties of the components.

temperature (Oe)

Figure 10.2 Viscosity of u mustie mix rclatcd to tcmpcraturc.

Remarks

The Kerkhoven method ofmix design has a number oflimitations:

the tests give only the viscosities at the test temperatures and not at the normal
tcmperatures
the measuring accuraey is not very great.

The designer and the user should have a more usabie metlied to supplement the
Kerkhoven method. Research into th is is al most completed.

10.3 Mix properties

10.3.1 Viscosity

Mcasurements of mustie viscosity have been made by Kerkhoven (22). With this the
measured relative viscosity, 11mhlbit (11m = viscosity of the mix. 1Jbit = viscosity of the
bitumen) is related to thc bulk volume of the sand/filler mix (Vsf),

71
The Vsf can be determined using:

100 F+M
Vsf = 100 _ Hm . F + M + B' 100%

in which:

Hm voids ratio ofthe cornpacted sand/filler mix

F volume offiller
,\1 volume ofsand
B = volume of bitumen

The Vsf represents the solid phase of the mix, that is, the bulk volume of the sand/fillcr
mix in which the voids are filled with fixed bitumen, see Figure 10.3.

-,---
.. tr oo " liquid phase

bitumen
1--- - - 1-- ----- 1-
B
fixad
F
• - - - - --
.
M
B
sand • solid phase
F+M filler

Figurc 10.3 The different phases of u mustie mix.

Mcasuremcnts have been made, using the Kerkhoven method, on a large number of
mastic samples made up from four different sand types, including dune and river sand,
mixed in different proportions with filler and bitumen 50/60 and tested at 140°C and
40°C (22).
Kerkhoven used the Marshall compaction hammer to compact the sand/filler mix, a
method which produced a lower voids ratio than the Engelsmann apparatus. The latter
method, however, gives better agreement with values obtained in practiee (4), sec Figure
10.4.
It appears that the viseosity ofa mastic, as a first approximation, ean be deterrnined using
the Vsf formula. A simple dryeompaction test, based on Engelsmann, ean then be used
to detcrmine the voids ratio ofthe mineral aggregate. The relative volumetrie proportion
of filler to bitumen only influences the viseosity at higher temperatures and only then
when the proportion is 0.20 to 0.25 or less. These low proportions are, in fact, scldom

72
found. With a low rclative volume offiller to bitumen it is possible to design a mix with
a relatively low viscosity at working temperatures and a high viscosity at normal tem-
peratures.

F/B _ Volumeoffiller
- Volume of bitumen

10l,
I1
8 11
--- viscosity at working temperatures(FIS >0.25)1 I
6
- . - viscosity at working temperatures[FIS<0.2~l "
l,
- - - viscosity at lower temperatures 1 1
~1
\V sand-liller. Marshall compaction 11//
I I
2 CD sand - lilier • Engelsmann compaction 1 1
/ /
/ 1
FIS = volume ol filler / /
103 volume of bitumen / /
/ /
8 / /
6 /
I
/
:ö / /
-
F"

E
F"
l,

I
I
/5. /

I
I

2 /
/ I
/ I
/
102
CD/
I 3.

8 cD
6

t:

101
50 60 70 80 90 100 110
Vsf

Figure 10.4 The effect ofVsfon the viscosity of mastic mixes.

The viscosities at working and normal temperatures are very closely related. An easily
workable mix will, in general, have greater flow at low temperatures. The flow charac-
teristics of mixes at normal temperatures, therefore, can only be compared when the
mixes have the same viscosities at working temperatures.

73
10.3.2 Meeltanical properties

Thc stiffness modulus and the strain at break of a mastic mix can be detcrmincd, very
generally. using the methods discussed in Sections 6.3.1 and 6.3.2.

10.3.3 Permeability

Mastic is an overfilled mix and is, thus, dense. Air bubbles can form in the mix as aresuIt
ofthc preparation process which generally do not affect thc impermcability. Thcrc may
weil be a voids ratio of 7 to 12% but the voids are not in contact with each other.
Air bubbles can, howevcr, have an effect on the viscosity. An increase in the air content
can cause a reduction of the rclative viscosity.

10.3.4 Durability

Because thc mastic is overfillcd, the mix is, in itsclf, scaled and therefore very durablc .

. .
74
11 Grouting mortars

Thrcc ways in which an asphalt mortar can bc uscd to grout stonc are:

I. Surface grouting. This is also rcfcrrcd to by thc Gcrman term 'Vcrklammerung'. In

this mcthod a ccrtain quantity of grouting mortar is applicd uniformly to thc wholc
surface, sec Figurc 11.1a. Thc construction is not complctcly sealed sincc thc mortar
only pcnctrates thc surface laycr.
2. Pnttcm/partial grouting. In this mcthod thc whole laycr ofstonc is pcnctrated on a
prcdctcrmincd pattcrn. As a rcsult, in fact, lumps of stonc are forrncd, sec Figurc
11.1b.
3. Fully grouting. In th is mcthod thc voids in thc stonc are complctcly fillcd with
grouting mortar forming a continuous homogeneous rcvctmcnt, sec Figurc 11.lc.

a : sur tcce grouting

b pcttorn grouting

I I

~~-

Ic: fully grouting . l

Figure 11.1 Different methods of grouting,

75
11.1 Basic materials

11.1.1 Sand,jiller, bitumen

Sec Chapter 10, Mastic.

11.1.2 Gravel, stone

Sec Eisen 1978 (6).

11.1.3 Ot/Ier constructional aspects

I. The grouted stone

Crushed stone is normally specified by the mass of the 'smallest', and 'largest' stone sizc,
for examp1e quarry-run 10/60 kg.
The stone must, in the first pl ace, satisfy the general conditions for dumped stone (49):

it must be sufficiently hard;

it must be frostlweather resistant;
it must have 1ess than 3% by mass below the minimum stated size or, when it is to be
dumped under water, 1ess than 1%;
there must be no flat pieces; the re1ationship between the breadth and the length, and
the heigth and 1ength must be 1arger than 113;
it must be preferab1y uniform1y graded;
there must be very litt1e breakage during transport and handling.

An important aspect to be considered is the effect of different propertics such as stone

size distribution and the percentage of flat stones, and the viscosity of the aspha1t mortar
on the grouting depth and spreading.

2. Filters
A sand and watertight construction cannot a1ways be guaranteed even with fuUy grouted
stone and, therefore, a1so in re1ation to the execution of the work a filter is
recommended. When a filter cloth is used tbc temperature of the grouting mortar, which
wiJl touch the cloth, shou1d not be too high. For a po1ypropy1ene cloth the maximum
temperature is about 140°C.

11.2 Mixes

11.2.1 Mix-design
A grouting mortar is an asphalt mastic or an homogeneous mix ofmastic with gravel or
stone. The use of gravel or stone is dependent on the dimensions of the stone to be
penetrated. Gravel or stone serve to:

76
1. reduce the quantity ofmastic needed;
2. prevent thc mortar sagging onto the slopc;
3. limit, sometimes, the grouting depth.

Mix-design is often a question of experience. Ouring execution the mix viscosity should
be sufficiently low, in the range 30 to 80 Pa . s, for the mortar to be handled and the stone
grouted without it sagging onto the slo pc.
After execution the viscosity should be so high that flow at ambient temperatures is
limitcd. Calculations indicate that, to reduce the flow ofmastic in rubble 80/100 kg, on a
slope of I : 5 in a 7 year period to a minimum, a mortar viscosity of 109 Pa". sis needed. In
general the viseosity required wiII lie between 106 and 109 Pa. s, depending on the
tempcrature, slope angle and dimensions ofthe stone (23).
The flow properties depend on the eomposition ofthe grouting mortar in relation to the
dimensions of the stone and the slope-angle, If the flow of the hot mastie in the voids
between the stones is regarded as Iaminar flow through a pipe, the following relationship
ean be used to find the grouting depth (78):

in whieh:

I grouting depth (m)

d specifie stone size, d20 (m)
1)0 = initial viscosity of the mastie (Pa" s)
C = constant, determined by experiment (Ns/rn'')

The design ofthe grouting mortar is dircctly rclated to the wcight and dimensions ofthe
stone to be grouted. The type of applieation is also important, forexample, when
grouting is to be carried out under water sinee the mix cools down quickly and the time
available for grouting is shorter. Experienee gained wh en construeting the breakwaters
at IJmuiden indieates that, in order to obtain a grouting depth oftwo layers ofstone, the
relationship between the maximum partiele size in the mix, dS5' and the minimum
partiele size of the stone d15, should be (27):
- above water-level, d151 dS5 = 5 to 10;
- below water-level, d15/ d85 = 10 to 20.

The design of the mastic is earried out in the way described under mix design tests, sec
Seetion 10.3.1. The optimum sand/filler proportion is determined from tests with
the Engelsmann apparatus, the grouting depth required being achieved by varying the
bitumen, gravellrubble, sand and filler proportions, the mixing temperature and the bi-
tumen hardness.
In order to aehieve a satisfaetory grouting depth a minimum of 50 to 55% mastic is

77
required in ihc mortar, depending on the partic1e size distribution ofthe gravel or stone
and the other cornponents (60).
The maximum partic1e size ofthe stone or gravel ofthe grouting mortar is determined by
the mixing equiprncnt to be used. For normal asphalt mix-plants the maximum particic
size is 50 to 60 mmo For coarser matcrials special equipment is needed.
For good 'surf ace' penetration at least 113 ofthe voids ofthe revetment should be filled
(the upper layer being completely fillcd) (37). For 'pattern' penetration 50 to 80% of the
revetment surface should be covered. With fully grouting all the voids are filled.

A grouting mortar has the tendency to scgregate, sec Figure 11.2. Wh en used for slab
construction it is essential to distinguish between the properties of the mastie and the
properties of the segregated grouting mortar.
The performance of a segregated grouting mortar is determined by the mineral partic1e
skeleton and the degree of self-compaction. If th is compaction is limited thc mastic
properties predominate. With good compaction, however, it is thc partiele skeleton
which predominates.

I I
rncs nc

mastIe. gravel + stone

,I segregoted groutln9 mortor )

~
Figure 11.2 Segregated grouting mortar.

As an exarnple a viscosity of 40 to 80 Pa. s is required for grouting stone 5/40 kg. A

grouting mortar which is often used is given in Table 11.1.

Table 11.1 Grouting mortar often used for stone 5/40 kg.
by mass %
gravcl4/16 30
sand A 61
very weak filler 9
bitumen 80/l00 on 100% mineral

11.2.2 Mix design tests

Mix design tests, in general, inc1ude:

1. An investigation of construction materiais, sec Seetion 8.1.
2. Laboratory investigations of the mix, the mastie, see Seetion 10.2.2, which inc1udes,
the effect of variations in mix composition, within given limits, on the properties of
the asphalt mortar.

78
11.3 Mix properties

11.3.1 Stability

A grouting mortar should be so stabie that:

I. Any subsequent grouting, when cold, is limited to a minimum.
2. The mix is not sagged onto the slope.

By adding crushed stone or gravel a good quality mix can be produced which has only a
limited cold flow. This can be investigated by large scale slope tests. These tests should be
of sufficiently long duration and the test site weIl chosen.
Above water-level slopes ofupto I : 1.7 can be grouted, below water-level the maximum
slope is I : 3. By grouting carried out in layers these slopes can be increased to I : land
I : 1.5.

11.3.2 Viscosity

The viscosity of'grouting mortar is determined in the same way as the viscosity of mastic,
see Section 10.3.1. The working viscosity will, in generallie between 10 and 100 Pa . s.

1ol.3.3 Penneability

Grouting mortar, being overfilled, is, in itself impermeable. Pcrmeability of the revet-
ment is dependant on the degree of penetration. One can only speak of 'impermcablc',
however, in relation to fully grouting and even th is cannot be guaranteed in all cases.
When designing for hydrostatic uplift complete water impermeability should be
assumed.

When a permeable revetment is required, then preferably not more than 70% of the
crushed stone surface should be grouted. Good mix composition and execution is
essential to ensure that the grouting mortar does not remain in the top layer or sags and
thus forms a dense layer.

11.3.4 Durability

The durability of the grouting mortar is determined by that of the stone and that of the
mortar itself. For the stone the requirement is that it is, in itself, durable (49). For the
dura~ility ofthe mortar reference should be made to Section 10.3.4.

79
12 Dense stone asphalt

12.1 Basic materials

12.1.I Crushed stone, gravel

Since dense stone asphalt is a slightly overfilled mix it is, on the who Ie, not important
what kind of stone is used. Good adhesion is obtained with limestone.
Limestone 20/40 mm is often used. Use ofthis size is based on the fact that:

it is the maximum size which can be used in conventional asphalt mixers (maximum
about 60 mm);
it is a type which is generally available;
it is suitable for the layer thicknesses typically used;
The stone or gravel requirements are based on (49):

the oversize;
the undersize;
the proportion of filler;
the quantity of flat pieces;
the frost and weather resistance;
hardness;
purity;
a uniform grading;

For further details, sec Eisen 1978 (6).

The stability ofvery thick layers, both during the hot and cold phase is increased by using
heavier stoncs, A dense stone asphalt with stones of a maximum 10/60 kg was for
example applied for the construction of the breakwaters at l.lmuidcn. A special mix-
plant is then required.

12.1.2 Sand.filler and bitumen

See Section 10.1.

12.2 Mixes

12.2.1 su« design

Dense stone asphalt is a slightly overfilled mix with sornc gap-grading, aresuIt ofwhich

80
is that compact ion takes places undcr its own weight. Gap-grading is important for
underwater applications where the cooling phaseand, thus the time available for placing
the mix is relativcly short.
The ability to form a dense stabie revetment by self-compaction depends on:
the viscosity of the asphalt mortar (mastie or grouting mortar);
the proportion of stones to asphalt mortar,

The stones form a skeleton which gives a certain stability to the mix in the hot ph ase,
while the mortarhas a binding and load-sprcading funetion. The viscosity ofthe mortar
is important in both the hot and cold phases,

A rclativcly low viscosity is essenrial at werking temperatures to ensure that the stone
asphalt has good workability.
After cooling down the viscosity must be rclatively high in order to prevent any
excessive cold flow.

Generally. deponding on stone size and grading, the mix comprises 50 to 70% by mass
stone and 50 to 30% by rnass mortar. With lower stone contents a stiffer mortar should be
used. For mortar design, sec Section 10.2 and 11.2.
The layer thickness is also a detcrmining factor in the choice of stone size.
Examples of a few dense stone asphalt mixes used are given in Table 12. I.

Table 12.1 Percentages by mass of dcnse stone asphalt and mortar mixes used for the Eastern
Scheldt storm surge barricr and IJmuiden breakwaters.

IJmuiden
Eastcrn ScheIdt below water abovc water
Asphalt mix (rnass %)
- limestonc 10/60 kg - 46.0 -
- limestone \110 kg - 20.0 74.5
- limestone 20/40 mm 60.0 - -
- mortar 40.0 34.0 25.5
rnortar (mass %)
- wcll-gradcd mine ral - 66.0 67.0
- local sand 64.0 - -
- vcry wcak liIler 17.0 15.0 15.5
- asphalt bitumen 801\00 19.0 19.0 17.5

12.2.2 Afix design tests

Mix design tests can inc1ude:

Invcstigations of basic materials, see Section 8.1.
Investigations ofmix composition, produetion and execution.

The laboratory design of the mix, generally is as follows:

81
 The stone size is selected, in principle, based on the layer thickness required. A
number of mixes are then made with various stone contents and different mortar
compositions and, thus, different viscosities. If the stone size is too large for labo-
ratory invcstigation then tests can bc made at the mix-plant.

In order to obtain an impression ofthe hot-flow characteristics slopc tests are carried out
at the required slope,
For asphalt morlar investigations reference should be made to Sections 10.2.1 en 11.2.1.

12.3 Mix properties

12.3.1 Stability mul viscosity

Dense stone asphalt is an overfilled mix of asphalt mortar (mastic with or without
crushed stone or gravel) and crushed stone which gets its stability from a limited amount
of scgrcgation,
The stability and workability are, in fact, two contrary propertics which are dependent
on:
stone content;
maximum stone size;
stone size distribution;
the viscosity of'thc asphalt mortar at working ternpcratures.

By incrcasing the stone content and the stone size and using a low viscous asphalt mortar,
the stones settie more quickly under their own weight on the slope, there by producing,
more quickly a larger interlock in the stone skeleton. In this way it is possible to design a
mix which is easier to work with and in which the stones settle quickly producing a
greatcr stability in the hot condition than with a mix of smaller stones and a more viscous
mastic.
Use of larger stones givcs a rough surface which has more holding points for waves and
currents.
At normal temperatures the material is viscous, except under very short pcriod loads,
Application on stccp slopes can, bccause ofthe limited stability, be difficult.
The Iines in Figure 12.1 show viscosity related to percentage of stone for various mixes
under the same conditions.
Table 12.2 gives the composition of the va rio us mixes referred to in Figure 12.1.
Tests have been carried out by the Delta Department of Rijkswaterstaat with stone
asphalt.Iayer thickness 0.5 m, on a slo pc of I : 4, at normal tcmperatures. In these tests, a
limestone/mastic mixture, rclativc proportions 65/35, even after a few weeks, showed no
noticeablc deformation.
A mix with limcstone/mastic proportions of 60/40, however, did show some defor-
mation. The compositions ofthe mixes, which we re tested, arc given in Table 12.3.

82
 asphaltic concrete.
stone 5 -1Smm

light dense stone asphalt.

stone 20-1.0mm

heavy dense stone csphctt •

stone 50-200mm

_ ·Ie stone

Figure 12.1 Stab~lity (workability) and composition ofsome asphalt mixes (14).

Tablc 12.2 Composition ofthe mixes used in Figure 12.1.

percentages by rnass
heavy dense light dense Asphaltic
Mix composition stone asphalt stone asphalt concrete
Stone 5/15 mm 48,0
Graded sand 44,5
Very weak filler 7,5
Asphalt bitumen
80/ 100 pen on 100% mineral 7,5
Stone 50-200 mm 75,0
Stone 20-40 mm 55,0
Asphalt mustic 25,0 45,0
- graded sand 67,0 69,0
- very wcak filler 15,5 15,5
- asphalt bitumen 80/100 pen 17,5 15,5
Working temperature 1l0°C 140°C 140°C
Slope I: 2 I: 3 I: 3

12.3.2 Stiffness mul breaking strengtli

With very thick layers and large stones, there Is
generally no need to consider tearing
under wave attack, Stiffness and strength, in this cas~, are only secondary factors (22).
Stîffness and strength are, however, of importance for light stone asphalts in thin layers
and, in thisease, it is essential to evaluate these properties. There has been, however,
little investigation in this field and aetual values are, generally, not available. The
properties of the mastie, to a large extent, dete~mine the stiffness and strength of the
stone asphalt.
See also part 6.3.

83
 Table 12.3 Composition ofstone asphalt mixes, percentage byweight (24).

weight
mix composition percentage
Mix I limestone 20/40 mrn 65
mastic 35
mastic: Sand A 62
Filler 19
Bitumen 80/100 19
Mix2 lirnestone 20/40 rnrn 60
mastic 40
mustie: Sand A 62
Filler 18
Bitumen 160/210 20

12.3.3 Durability

The mix is dcnse and, therefore, has a good durability. In this respect the density ofthe
layer surface is very important.

84
13 Open stone asphalt

Open stone asphalt is used in the tidal zone and above high water and, also, in the form
of pre-fabricated mattresses under water. The product is patented under the name
'Fixtone',

13.1 Basic materials

13.1.1 Crushed stone

Because ofthe need for good adhesion, only limestone is used in open stone asphalt. The
gradation most frequently used is 20/40 mmo See also under dense stone asphalt, Section
12.1.

13.1.2 Gravel

In view of the lower adhesion obtained with gravel it is not used in open stone asphalt,

13.1.3 Sand
In principle every kind of sand can be used provided it does not contuin any foreign
matter.

13.1.4 Filler
Preferably the filler must be weak, Limestone filler is preferred. Sec also Eisen 1978 (60).

13.1.5 Bitumen

Normally bitumen 80/100 pen is used.

13.1.6 Ot/Ier constructional aspects

a. Filter material
Open stone asphalt is so perrneable that it cannot be eonsidered as sand-tight and a
filter must, thcrefore, be used. This can be in the form of:
à granular filter;
a filtercloth
a filter of lean sand asphalt.

85
 The filtering propertics of lean sand asphalt and filtercloth best kept in hand. A
filtercloth can, if it acts as reinforcement, have a negative effect, sec Section 20.6.
b. Reinforcing is generally used with open stone asphait mattresses to prevent breaking
during handling. The reinforcing comprises metal gauze or similar material. Steel
cables or synthetie ropes are plaeed underthe mattress for earrying. These are
removed wh en the mattress is in place. Lifting eables whieh are ineorporated into the
mattress ean also be used.
Mattresses must be manufactured very earefully insueh a way that there can be no
direet contact between the reinforeing and the filtercloth. Rust, from steel cables has a
negative effect on the filtercloth ageing process. The minimum. thickness of the
mattress, when using limestone 20/40 mm, is 12 cm; the reinforcing is then 6 cm
above the fiitercloth.

13.2 Mixes

13.2.1 Afix design

Open stone asphalt is an underfiIIed mix oflimestonc and mastic. Manufacture is in two
phases: first the mastic is prepared and then it is mixed with the preheated stone fraetion.
By using this method of mixing the mastie can be spread very homogeneously over the
surface of the crushcd stone and all the contact surfaces between the stones can be
fixed with a minimum quantity of the mastic. The production of open stone asphait
requires aceurate con trol of the mix composition; espeeially the mastic is very important
for the overal quality. In addition, it is important that the stone does not contain any un-
dersized pieces.
The stone/mastic ratio in the mix is largcly based on the viscosity ofthe asphalt mastie. If
the viscosity is low, mixing is casier but, thcn, only a limited quantity ofmastic is retained
by the stones. The excess ofthe mastie sags and seals the lowest part ofthe layer. With a
stiffmastic mixing is more diffieult but more mastie is retained and the tendcncy to sag is
reduced.
The lower limit ofthe asphalt mastie/stone ratio is determined by the minimum quantity
ofmastic which is needed to completcly coat the stones. This quantity depcnds, to a large
extent, on the mastie viseosity and the mineral grading. The upper limit of the asphalt
mastie/stene ratio is determined by thc extent 10 whieh the mastie tends to sag (20).
In general the composition, is:
- mastie 20% by mass
- limestone 20/40 80% by mass.
For mastic design, sec also Sectien 10.2.
In praetice, the mastie/stone ratio is determined by a sag test in whieh mastie is added to
the mix and the quantity notcd at which the mix is still 'open'. A particular quantity of

86
thc mix is plueed in a tall tube and, after cooling, tbc amount ofasphalt mastic above and
below is recorded using extraction methods and visual observations. In practice the
mastic percentage lies between 17 and 21% by mass.
Typical open stone asphalt mixes are given in Tablc 13.1.

Table 13.1 Composition of open stone asphalt mixes used for the Eastem Scheldt Project.

composition percentage by rnass

Mix. component desired min max.
Limestone 20/40 trim 81 79 83
Asphalt mastic 19 17 21
Asphalt mastie
- for mattrcss
Sand 67 63.5 70.5
Filler 14 12 16
Bitumen 19 17.5 20.5
Asphalt mastic
- for fi\terconstruction
Sand 64.0 60.5 67.5
Filler 17.0 15.0 19.0
Bitumen 19.0 17.5 20.4

13.2.2 Mix design tests

Mix design tests can include:

1. Investigations of basic matcrials, see Section 8.1.

2. Investigations of the mix
a. For mastic, see Section 10.2.2.
b. For open stone asphalt the object of the investigations is to limit the tcndency to
scgrcgate at handling temperatures.

13.3 Mix properties

13.3.1 Stability and viscosity

The stability of the mix is determined by:

1. the stone skeleton;

2. the ~dhesion between the mustie and the stone.

The stone skeleton of the mix has less stability than a loose pile of stones but the mastic
add a viseous property which, depending on the deformation speed, gives extra

87
stability so that the total stability is larger than for loose stoncs. Wh en deformation occurs
quickly the material reacts elastically.
An impression of the stability ean be obtained by moulding a quantity of the mix using
wooden formwork at the slo pc required and noting how it performs at working tem-
pcratures and in cold conditions.
It appears that:
I. Undcr hot conditions (I100-120°C) mixes with limestone are stabie on steeper slopes
(I : I) than mixes with gravel (I : 1.5). This is due to thc greatcr interlocking of the
limestonc fragments.
2. For long-pcried norm al conditions (30°-40°), howevcr, gravel is more stabie. This is
bccause the contact surfaces between gravel particlcs, being round, are largcr than for
thc angular limcstonc and the adhcsion, thcrcforc, is greatcr. With good compaction,
however, the stability oflimestone mixes can bc improved. In practice compaction is
not carried out.

The viscous behaviour of the mix is strongly determined by the viscosity of the mastie.
The upper limit of thc mastie percentage is dctcrmined by the degrce of saggin aftcr
placing or by the minimum voids ratio required, sec Section 13.2.1. For mast ie a viscosity
of40 Pa. s at 140°C is dcsirablc. Thc limits to viscosity, depcriding amongst other things
on thc measuring mcthod, lie between 30 and 80 Pa . s.

13.3.2 Mechanica! properlies

Mix propertics such as breaking strain, stiffness modulus etc., va lues whieh are neeessary
for ealculating an open stone asphalt rcvetment can be cstirnated, generally, with the
help of Figures 13.1 and 13.2.
These figures are based on invcstigations into the propertics of open stone asphalt (72).

13.3.3 lValer penncability

Thc voids ratio of open stone asphalt is, generally, between 20 to 25%. The material is,
thus, very permcabie. The water permeability whieh is ofthe same order ofmagnitude as
th at of loosely dumpcd stone of the same nature can be measuredusing a permeability
test carried out with an apparatus designed for large samples.
The permeability dcpcnds on:
the stone grading;
thc mastie/stene ratio.

Invcstigations indicate that, under normal conditions, thc flow in open stone asphalt is
between laminar and turbulent. If open stone asphalt is used with a filtercloth, the
pcrmeability ofthe strueture will be largely deterrnined by thc pcrmcability of the cloth.

88
 '08
",-

s
-.z Ë
In

,~7

J I I

'02
, ,
103
! I I " I

Sb.! IN/m2)
! t

Figure 13.1 Smix-Sbilrelationship for open asphalt, determined from the results of ereep and
dynamic tests.

..
o
c

~--
~
,~ 10-3

:~

- Smlx(N/m2)

Figure 13.2 The relationship between the stiffness modulus for n = 10, the initial strain and the
number of loading cycles required to break.

13.3.4 Durability

Open stone asphalt, as the name suggests, is an open mix and the durability thcreforc, is
of great importancc.

89
The durability is also deterrriined by th at of the stone for which there are standard
durability requirements.
For the mix itselfthe following must be taken into account:

a. Aging
This manifests itself in
reduced adhesion between the stone and the mastie ('stripping');
- aging ofthe mastic layer itself;
- sensitivity ofthe total mix.
b. Erosion. This can be investigated with flow tests.
c. Biological attack, This should be investigated with plant penetration tests.
d. Chemical attack, by, for example, polluted water which:
reduces adhesion;
- attacks the mortar.

Some petroleum produets weaken bituminous materiais. This process will occur more
quickly with open mixes. Marshall tests on cylinders of open stone asphalt, which have
been placed for a certain time in a bath containing hot water with 1.5%m oil, indicated
only a limited change in the mechanical properties (30).
The laboratory of the Bundesanstalt für Wasserbau at Karlsruhe has carried out various
tests related to temperature resistance durability (62):

I. A sample, placed on a slope of I : 3, was subjected to the equivalent of 3000 hours

sunshine at a surface temperature of60°C. No flow phenomena were noted.
2. A sample, placed on a slope of I : 2, was subjected to ultraviolet rays, equivalent of
4000 hours ofsunshine and, again, no flow phenomena were noted.

Freeze/thaw tests on open stone asphalt cylinders did not show any damage to the
material.

90
14 Lean sand asphalt

14.1 Basic matcrials

14.1.1 Smul

See Eisen 1978 (6).

From expericnce it appears that, in order to obtain the same quality oflean sand asphalt
with different types of sand, different bitumen percentages have to be used. Factors
which influence th is are, amongst other things, the specifie surface and the shape of the
sand particles.
When lean sand asphalt is to be used as a filter material it is, in conneetion with the
permeability, best to use the sarnc sand as that upon which the filter is to be laid for
permeability performance.

14.1.2 Bi/lil/Wil

A bitumen which is too hard can lead 10 higher working temperatures. For application
under water, however, this is undesirable. At temperatures over 100°C the water will
boil. This can neutralize the adhesion required between the bitumen and the mineral, see
Scction 14.3.4.

14.1.3 Other constructional materials

Gravel and filler can be used for certain applications.

Gravel can be used if:

a greater stability is required;_

a faster mixer production is desired;
a transition must be made onto coarser material, such as quarry stone.

The use of filler, and the associated change in bitumen content, gives, in gcneral, a
strenger material with grealer resistance to erosion which can be used as a permanent
rcvetmcnt. The water permeability is greatly reduced and, since compaction is then
necessary it cannot be applied under water.

14.2 Mixes

14.2.1 Mix design

Lcan sand asphalt is generally a mix of sand and bitumen, the most frequently used

91
bitumen content being 2 to 6% by mass. The bitumen functions as a binder and tends to
concentrate on the contact surfaces between the particles. In order to obtain a certain
film thickness, the total surface area ofthe mineral particles has to be taken into account
as well as the bitumen content (34).
So, for example, in a hot mix lean sand asphalt, with a mean partiele diameter of 0.1 mm
and a bitumen content of 4% m, the particles have a film thickness of 1 to 2 urn.
The pattern of the voids of the original sand is changed by the addition of the bitumen;
the extent, however, is unknown. Adding of filler above a certain level results in
enlargmcnt of the voids. Only by heavy compaction the voids between the sand grains
are fillcd in such a way with a stiffer mortar, that they are reduced noticably.
Mix design, aimed particularlyon durability and water permeability, is based mainly on
experience. Factors which can influe?ce the properties ofiean sand asphalt are (32): .
1. The bitumen content. This is the critical factor, since:
a. The strenth increases as the bitumen content is decreased;
b. The resistance to wear increases with increasing bitumen content;
c. The permeability decreases with increasing bitumen content;
d. When altering the bitumen content the deformation characteristics change.

A bitumen content of 3% m is definitely the lower limit. It must not be any lower,
since then the durability is strongly rcduccd. A bitumen content of 5% m is the
maximum since a higher percentage produces a lcan sand asphalt which is too 'fat'
which, depending on the application, reduces strength too much. A bitumen content
of 3% m is adequate for teinporary work; for permanent work and filter layers the
content should be between 4 and 5% m.
2. Degree of compaction. A high degree of compaction increases the wear resistance and
the deformation and strength properties but lowers the permeability.

14.2.2 Mix design tests

In addition to investigations into basic materials mix design tests can comprise a
comparison with known lean sand asphalt. These investigations can incl~de:
1. Permeability;
2. Erosion.tests (by rolling or by water-jet);
3. Retained stability;
4. Marshall tests;
5. Triaxial tests.

14.3 Mix properties

14.3.1 Stability

The stability, in relation to forces aplied slowly is less than for natural sand. With forces

92
which are applied more quickly the stability is greater, see also Section 20.7.

14.3.2 Mechanical properties

The stiffness modulus of the mix is dependent on the duration of loading and the
temperature and is related to die mineral content and the stiffness modulus of the
bitumen.
Because of the high voids ratio of Ican sand asphalt the bitumen hardening process is
faster than for most dense mixes. Creep tests, carried out on samples taken from al: I
scale test which comprised 96% m fine sand and 4% m bitumen at temperatures of 15°
and 20°C, indicated the values shown in Figures 14.1 and 14.2 (32).

14.3.3 Permeability

Because ofthe limited quantity of bitumen in the mix the sand grains are only covered by
a thin film of bitumen a few microns thick. The bitumen tends to concentrate on the
contact surfaces between the grains, This is the cause ofthe fact that, depending on the
compaction, grain composition and shape, the iean sand asphalt has a high permeability.
With the low degree of compaction which is found in practice the permeability tends to
be similar to the sand of which it is made. With dry lean sand asphalt an 'initial
resistance' has to be overcome before the ultimate permeability can be reached.

\oadtng 0.01 MNIm2

o lab ec-e at 20°C

o lob core et G
lS C

Figure 14.1 StifTncss moduli oflcan sand asphalt,

93
 10
loodlng 0.01 MN/m2

o tab ccre at 200t

C lab care ot 15°C

c
o
u

o
~
"

103
loadlng durenen (s)

Figurc 14.2 Dcformation as a function ofthc loading duration,

14.3.4 Durability

The durability ofthe lean sand asphalt depends on the permanent binding effect ofthe
bitumen.
For applications under water working temperatures above 100°C have an undesirable
effect on durability, sec Sectien 14.2.2.
The following phcnomcna and properties are relevant to durability ofiean sand asphalt.

a. Aging is relativcly fast because of the open character of the mix.

b. The adhesion betweeri the sand and the bitumen is reduced when water is forced into
the mix ('stripping'). "fhis effect can be measured with a retained stabilty test (39) or
some other water sensitivity test. If the stability is measured over a period of time it
appears that it eventually reduces to a value ofabout 30%, sce Figure 14.3, depending
on the mix composition and the physical conditions (32, 33). A higher bitumen
content and a thicker film on the grains tends 10 improve this effect.
c. Good chemical resistance particularly the reaetion to polluted water.

94
 100

~ 7S
~
~
50
~
ë
~
2S

0 12 14
__ time (we-eks)

Figure 14.3 The effect of stripping on lean sand asphalt.

95
15 Bitumen membranes

Bitumen membranes are used for making canals, watercourses and banks watertight.
There are two methods of execution:

in situ;
pre-fabricated.

15.1 Basic materials

15.l.l Fil/er

Some times filler is used to increase the stability of membranes. The filler has to be in
conneetion with the other matcrials us ed.

15.1.2 Bitumen

The selection of thc bitumen type depends on the application - for example, at low
working temperatures or for a certain desired deformation. For application of blown
bitumen, sec (35) and Section 15.2.1.

15.1.3 Additives

plastic rubber to inerease stability;

matcrials to prevent plant penetration.

15.1.4 Other constructional aspects

Prefabrieatcd membranes include:

I. Reinforcing (sometimcs also used in membranes made in situ). Generally the
reinforeing is a synthetie cloth whieh has the following properties .
strength in warp and weft direetions;
strain at break;
strength in rclation to temperature effeets;
workability (can be handled easily).
2. Thin sheets, also refcrred 10 as inlays
to resist root pcnetration;
to proleet sealing strips.

96
3. A wenring surface finish, amongst other things to increase the friction.
This can be achieved by scattering the surface with: .
- sand;
- fine stone particles.
4. lt is advisable to use blown bitumen for welding thc membranes together.

15.2 Design and design tests

At present membranc design is done empirically, using 'trial and error' methods. Mix
design involves tests on the membrane in the conditioris in whieh it will be used.
There are two types of membranes:

I. Membranes prepared in situ.

The membranc consists oftwo or three layers of bitumen, R85/50 or R85/40, with a
total thickness of 8 to 12 mm, sprayed onto the subsoil (35, 69). The quantity of
bitumen depends on the subsoil texture and ranges from 6 kg/rn" for fine subsoils to 8
kg/rn" for rough, The maxi~um slope angle ofthe subsoil shouldbe 1 : 2.5.
To proteet the membrane it should be covered with at least 30 cm soil or, if available,
gravel. Ifvegetation is anticipated théprotcction layer thickness should be increased
to 50 cm so that weeds can be removed without darnaging the membrane. The subsoil
should be well-compacted and smooth without stones or other obstacles and should
be pretreated with a planticide.
The disadvantagcs ofmembranes, prepared in situ, are:
a. they are pro ne to plant penetration;
b. they are prone to settlement of the subsoil;
c. adhesion between the layers of bitumen cannot be guaranteed;
d. the quality is only moderately reliablc;
e. the subsoil must be preparcd very carcfully,
2. Prefabricatcd membranes.
Because of the disadvantagcs of in situ membranes prefabricated membranes have
been developed. Prefabricated membranes often comprise a bearer or reinforcing of
synthetic cloth between layers of bitumen or bitumerr/fillcr-mix. Because of the
reinforcing the mcmbranc is lcss likcly to settlcment damage and, becausc of the
anti-plantpcnetration sheetit is not ~ecessary to treat thc subsoil with a planticide.
Thc subsoil must be smooth and free ofsharp objects. In order to reduce the dangor of
the proteetion layer sliding off the membranc, it is sprinkled with sand or fine rubble
(29).
The proteetion layer can be of earth, crushed stone, concrete asphalt etc. Depending
on the type of proteetion an 'non-woven' can be used to reduce the effects of the
temperature and mechanicalloadings on the membrane.
An example of a prefabricated membranc, patented under the named 'Hypofors', is
shown in Figure 15.1 (36).

97
 5
6_-----1IIIIIi~ I. paper
2. bitumen
3. reinforcing
3 4. bitumen
2 _ 5. polyester film
6. bitumen
7. polyethylcne strip
8. sprinkled sand

Figure 15.1 Hypofors (prefabricated mcmbranc-patented).

15.3 Membrane properties

15.3.1 Stability

It is necessary to know the stability of themembrane in relation to the sliding forces to

which it is subjeered. Invcstigations can be made with the help of a rhcometcr,

15.3.2 Viscosity
The viscosity ofthe bitumen is important in relation to:
I. the stability of the membrane and the proteetion layer;
2. the workability.
For the viscosity of bitumen reference should be made to Chapter 3.

15.3.3 Permeability
The membrane and the seams must be and must remain watertight under the maximum
specified loads. This property can he investigated using a vacuum meter.
The memhrane must also rernain impcrmcable under deformation.

15.3.4 Durability

The memhrane must he durahle. In this respect the following propertics are relevant:

Ageing of the hituminous layer.

Biological rcsistancc; tests may he necdcd on plant penetration.
Chernical resistance; resistance against agressivc percolatien.
Memhranes are gene rally protected with a cover layer.

98
PARTC

TECHNICAL ASPECTS OF THE DESIGN

Summary

Part edeals with the technical aspects of the design of asphalt revetments used in
hydraulic engineering.
Such revetments must satisfy a number ofrequirements in order to function properly. A
review of these functional requirements is given; the designer should take no te ofwhich
of these requirements to apply to a particular situation.
The functional requirements are partly deduced from the loads which can be exerted on
an asphalt revetment. The loads can be distinguished into loads related to the hydraulic
conditions such as water-levels, waves and currents, loads arising from the settlement of
a dikebody and the subsoil and erosion ofthe foreshore and other loads. The influence of
these loads on an asphalt structure are discussed.
Dike design is based on a numberof starting points. In the Netherlands for some
situations these are established; in other situations, however, not. A summary of the
present 'state of the art' is given.
A very important factor in the design and construction of an asphalt revetment is the
nature ofthe dike body and the subsoiI. A paragraph is devoted to the aspects which play
a role in this subject. An attempt is made to indicate in which way the situation can be
influenced favourable.
An asphalt revetment is normally dimensioned on the basis of the loads which result
from the hydraulic conditions, settlements and scouring and, sornctimes, on other loads.
The dimensioning methods for the most relevant conditions are given. The following
situations are discussed:

design of a relatively impermeable asphalt revetment against hydraulic uplift;

design of a plate-type asphalt revetment against wave loads;
design of an under water bed proteetion against uplift pressures caused by currents
and waves;
design of a surface- or pattem-grouted stone revetment against wave action;
design of an asphalt revetment against currents;
design of an asphalt revetment against irregular settlements and scouring;
the determination of the maximum slope angle.

100
16 The functional requirements of an asphalt revetment

The function ofa dikc is to proteet the area behind from floading. In order to proteet the
dike body, which is generally built up from soil, against erosion it is often provided with
arevetment. Since the revetment is, at the same time, a par! of the dike it can also fulfil
other functions as, for cxample, watertightness. In general, for design purposes, the
revetment may not be used to increase the safety of the dike body, in itself, against
slipping.
An asphalt revetment must, in order to function satisfactorily, fulfil various require-
ments; requirements which stem"partially from the loads which can be cxerted. These
requirements are:

I. The revetment must be so that material from the dikc body cannot pass through it.
Sometimes the requirement is that the revetment should be watertight.
2. The rcvetment must be able to withstand:
waves generated by wind and ships;
- currents and material catried with it such as sand, stones and driftwood;
- uplift water pressures (only applicable to a relatively watertight revetment);
3. The revetment must be able to adjust to settlements and scouring, within limits,
and must remain in contact with the subsoil.
4. The revetment must be stabie.
It should not slip down from the slope, also, during construction. It should, also, be
stabie, as a whoIc, so that it cannot be carried away by the loads which act upon it.
This implies a certain dimension and weight.
5. The revetment must be weather and water resistant to erosion, corrosion, light, wind,
temperature, and ice. The environment should not be able to damage the revetment
or vice versa.
6. Thc revetment must be durable, that is, it must continue to function throughout its
design life.
7. The revetment should, preferably, be aesthetically acceptable. This condition, how-
ever, if it is contrary to other requirements, should never be overriding.
8. In addition the following should be taken into account and guardcd against:
biological damagc by plants, animals and sea organisms;
chemical damage by pollutcd or salt water;
possible land traffic, during construction and when completcd;
vandalism;
rccrcational activities;
vesscls and anchors.

101
The above requirements must, in principle, be satisfied. This is possible by composing
the revetment of one or more materiais. Often, because the various requirements
demand different revetment properties, a compromise must be sought.
The combination ofthe revetment and the dike body determine whether or not the dike
functions satisfactorily, also in the coarse of time. Also it should be possible to construct
the revetment under tolocal conditions. The dike body itsc1f should be such th at a
reasonable cheap, well-functioning revetment can be achieved.

102
17 The dike body

An important factor to take into consideration when dimensioning and constructing a

dike revetment is the nature ofthe dike.body, that is, the sandbed."
The following aspects are important:

a. The bearing capacity of the dike body determines among others the performance of a
revetment under wave attack and other forces, and, therefore, plays a large role in the
dimensioning. If the bearing capacity is large then often the thickness of the rcvct-
ment can be reduced.
The properties of the soi! such as the modulus of clasticity, the modulus of subgrade
reaction and Poisson's ratio are important, sec Appendix V. They themselves are in-
fluenced by the amount of compacting of the dike body. These propertics can be
determined from, for example, plate bearing tests.
b. A high degree of compaction can, amongst other things, a/vert the softening of a
saturated or almost saturated soil by impact loads, for cxample wave attack, which
can cause loss ofbearing capacity,
A relative proctor density of95-100%, down to a depth ofabout 2 m, can, in general,
in sand reduce the possibility of softening to an acceptable minimum.
Bad, permcabIe, wet soil is prone to weakcning; the presence of mud, in this context,
is undesirable.
c. The perrneability ofthe sandbed is important in conneetion with groundwater flow in
the dike body and, consequential uplift pressures under a relatively watertight
revetment. lt is also important in conneetion with likely softening ofthe sand body.
d. The placing above water of an open asphalt mix on a saturated sandbed can, through
the influx of water, result in the early development of stripping. Vibration compac-
tion can soften a loosely packcd saturated sandbody. Under impermeabie mixes, as
asphaltic concrete, uplift pressures can develop whiie the asphalt is still soft. This
situation can, for cxample, occur by the deliverywater from hydraulic filling work.
e. The dike body should have sufficient bearing capacity to support construction
activities, If the sandbed has little resistance to dcformation it is difficult to compact
and, construction equipment can cause track impressions ('rutting').
After the sandbed has been compacted and made smooth it should not be driven over
or care should be taken to ensure that it is not disturbed.
The dike body is often formed of sand which is reasonably easy to compact. A

* Although this chapter is directly related to the dike body the contents are also applicable
wherever an asphalt revctrncnt is constructcd on an earth structurc.

103
 good cornpacteö subsoil gives fewer problems while compacting the revetment, initial
cracking is limited and the voids ratio required easier to achieve.
In some areas in the Netherlands it has been common practice to only smoothen the
sandbed with a bulldozer. This, however, only produces a lirnited improvement on
the original density, On several dike projects. tests have shown that, even at consid-
erabie depth, the compaction was low. It is. in fact, beller to build up the sandbody in
thin layers. with bulldozers. The dike face can th en he profiled accurateIy, also with
bulldozers. Ir this does not produce the compaction required, then a vibration roller
should be used. For th is process it is recommended that the sandbed is first sprayed
with water (Photo 4).

Photo 4 Compacting the sandbed using a vibratcry roller

The best dike body construction method, however, is to dump an excess of mate rial.
and then. after ruis has been compacted, to produce the profile required by grading.
The application of a day underlayer is not recommended because of its weak
consistency nnd the possibility of Trost heave'. Ifan asphalt revetment is to be built on
an existing clay revetrnent this should be of a good quality, otherwise it should be
removed. lnvestigations ofthe quality are desirabie.
There are generally na problems with underlayers of inert mine waste or lean sand
asphalt. There can be some problems when laying an asphalt revetment on mine
waste in eenneetion with obtaining an equet revetment thickness.

104
 Subsoil can be improved by:
mechanical compaction;
physicallmechanical treatment; an improvement of the partiele skeleton in
combination with mechanical compaction. This method has not yet, generally
been used in hydraulic structures, but perhaps it will be in the future.
f. After constructing the revetment the dike body will tend to settle. If.it has not been
weIl compacted or if there are clay or peat layers in the subsoil, the settlement can be
large and irregular. If the bed, is at the same time highly permeable then it is possible
that thc grain stress only reeovers slowly and that the bearing capacity of the bed
temporarily appears to be insufficient. This effect must certainly be taken into
account with dayey subsoils; good drainage in this case is essential (17). With very
permeable material the situation does not develop.

Thc bearing capacity of a ground mass can be measured using C.B.R. tests, soundings, o'r
plate-bearing tests. It is reeommended that laboratory investigations are carried out to
determine values of permeability, proetor density, friction factors, etc. Sufficient meas-
urements should be made to obtain rcpresentative values.
For the present the Rijkswaterstaat makes the following recommendations for
eompaeting subsoil: The minimum compaction should be 95% ofthe maximum praetor
density. The ave rage compaction should be 98%.

105
18. Design basis

18.1 Safetyaspects
The philosophy of dike revetment design sterns from several starting points. Although
these starting points vary from situation to situation, in the Netherlands, generally. the
following can be posed:
In the report of the Delta Commission (83) for sea dikes, the most extreme conditions
are established by the design level. If the water-level reaehes this design level then,
according to the Delta Commission, there should still be a considerable factor of
safety against breaching. This implies that a revetment, as far that it is essential for the
functioning for the dike, must be able to resist a design water-level with a large
probability. During a storm in which thc design level is reached not all parts of the
dike are loaded to a maximum. The maximum and most critica I wave impact loads
developa little below still water-level. The maximum uplift pressure under the
revetment will occur some hours after the maximum water-level has been reached.
Dune toe protection.
No general basic rules have been formulated for the use of asphalt in dune toe
protection. The effect of dune toe proteetion on erosion during the design conditions
is still to a large extent unknown.
When designing against dune erosion the basis should be the same as that used for sca
dikes. The most important function of dune toe proteetion was often to limit dune
coast erosion.
River dikes.
The following is stated in the report of the River Dikes Commission (84): 'The
commission recommends that improvements to river dikes should be based on a
leading water-level rclatcd to a Rhine discharge of 16,500 m3/s at Lobith. This dis-
charge occurs with a frequency of about 111250'. In the directives for river dikes
(Leidraad Rivierdijken), published by the Dutch Technical Advisory Committee on
Water Defences (Technische Adviescommissie voor de Waterkeringen) in 1984, it is
stresscd, as in the report by the Delta Cornmission, that there should be a large factor
of safety against possible brcaching in design water-level conditions.
Canal dikes and ernbankments, etc.
Dikes and embankments along canals, lakes, reclaimed arcas and polder reservoirs
form a type ofprotection works ofwhich the object is to proteet against a permanently
high water-level. These dikcs are often of limited dimensions. With the exccption of
bow waves from passing ships, there is generally very little wave aetion on these
structures. Exceptions to this are the dikes around large Iakcs, such as the IJsselmeer,

106
 where there can be considerabie ware attack and high water-levels can develop in
relation to wind set-up.

18.2 The decrease of structural strength

The fact that the strength of a structure can decrease with time must be taken into
account in the design.
There are two different ways of estimating this problem:

1. The structure can be 'over-designed' so th at it continues to function satisfactorily

throughout its life.
2. Structural strength can be preserved by regular con trol and maintenance procedures.

The choice between the two depends on the experience of the designer: does he accept
that a certain amount of risk is delegated to the con trol and maintenance by other partics,
and also by economie factors: lighter structures cost less to build but more to maintain.
At the same time, the safety of the structure during the construction period should be
borne in mind and the possibility ofstructural Iailure in this period should be anticipatcd
in the design and method ofconstruction. A solution can often be found by choosing a
favourable time ofthe ycar, construction sequence and method ofexecution.

18.3 Theoretical background of safetyaspects

A dike rcvctmcntis dimensioned with .the help of theoretical modeIs in which certain
basic variables are used to determine design dimensions, such as the revetment thickness.
The basic variables describe, on one hand, the different loads on the structure: waves,
water-levels, structural weight, ctc.; and, on the ether hand, the structural properties:
material properties, geometry, etc.
If the basic varia bles are specific deterministic quantities there are no problems related
to the calculation, apart from the choice of a good theoretical model.
The theory statcs that thc structure in that case will not fail ifthe - deterministic -loads
are less than the - determenistic - strength.
The difficulty lies in the fact that thc basic variables are stochastic quantities, that is,
variables. Loads and stresses are, therefore also stochastic properties. The problem can
be defined as follows: the structure must be designed in such a way that the probability
of failure is acceptably smal!. For this type of design there are three different methods
available (2):
1. The (semi) - deterministic method
In this method the basic load-related parameters are assumed to be adequately high
and thus 'safe' values. Often aver,?-geor safe values are selectcd for the strcngth
properties. The choice of 'safe' loading is mainly based on experience.

107
2. The quasi-probabilistic method
In this method 'safe', what are referred to as 'characteristic' values are assumed for all
basic variables. The characteristic load is, for exarnple, defined as the load which has
5% probability of being cxceeded during the lifetime of the structure. The charac-
teristic strength propertics are then those values which have a 5% probability ofbeing
less than the lower limit, sec also Figure 18.1.
In the building and construction world the margin between characteristic loading
and strength can be given by using partial safety coefficients. Partial safety
coefficients, as defined in 1.5.0. 2394, are then transposed by multiplication into the
total safety coefficient:
m

YIOlal = n
, i=1
Ypartil i

stre09th probabihty
den'ilty function

- loading J strength

I charocteris.hc
strength prope-rty

Figurc 18.1 Examplc of a choice of characteristic va lues

3. The probalistic method

This method, which is still being devcloped, is the most advanced. All basic variables
are specified asprobability density functions. The probability ofstructural failure is
then determined by integrating the probabilistic density functions of the loading and
strength propertics. The probability density of structural failure is then used as a
criterion for judging design. More information'on this subject is given in (75, 76, 77).

In practice it appears that, at present, the method most used is somewhere between
deterministic and quasi-probalistic. Variations in the loads are taken into account but
not the structural strength properties.
For loading variables values are mostly selected which have an extremcly low probability
of being exceeded. Avcrage or characteristic values are often assumed for strength
properties.

108
19 Loads

A number of different kinds of loads can be exerted on an asphalt revetment. These can
be roughly subdivided into loads caused by hydraulic conditions, loads resulting from
settlement and scouring and loads due to other factors. In general, revetment design is
based on the first two categories and also on other loading types which develop regularly
or can be expected to occur with reasonable probability. Sometimes it is not practical to
design for certain loads, because, although they are very large it is cheaper to repair the
revetment or to take preventative measures.
The most important conditions and the related loads are treated gene rally in thc
following sections. The most common types of loads and methods of design are dis-
cussed in subsequent chapters.

19.1 Hydraulie boundary conditions

A particular water-level can be anticipated for a dike either permanent or temporary.

The type of protection, the absence or presence of a foreshore and its height, and the
water-level and its variation are all important to this aspect. Waves and currents can be
expected to occur in the water in front of the dike.
In order to dimension arevetment for specified hydraulic conditions it is necessary to
know the actual loads which will occur. Ideally, therefore, the waves, currents and
water-levels, which will occur on the dike itself, should be known so that the loads can be
calculated dircctly. Those types of data are however mostly available for dcep water,
some di stance from the shore, and have to be converted to data at the coast itself. Waves,
for example, can be transferred by refraction calculations.

19.1.1 Water-levels

The water-levels in front ofa dike may vary. For dikes on the co ast, in inlets and estuaries
this variation will mainly be caused by tides, winds or both in combination. In the upper
reaches ofrivers the water-level is mainly determined by the discharge; this can also have
effect in lower reaches. In canals, waterways and polder reservoirs water-level variations
are due, principally, to inflow and outflow through locks, to rainfall, and to waves caused
by ships; in lakes and reservoirs water-level variations are caused by the in- and outflow
of water and by wind (I). Wind gcnerated waves also give short-term water-level
variations.
The duration ofa change in water-level can vary greatly. Variations due to wind and ship
waves last only a few seconds, tidal variations on the other hand have periods of several

109
hours, long-term wind effect can last a few hours or days and river discharges can affect
the water-level for days to weeks on end.
A phreatic line is set up in the dike body in conneetion with the water-level in front ofthe
dike. Variations in the water-level produce changes in the phreatic level. The extent of
these changes depends, amongst other things, on the duration and extent of water-level
changes, thc geometry and permeability of the subsoil and the dike body, and the
relative permeabilty of the cover layer. When the water-level outside the dike is lower
than the phreatic line and water cannot flow freely out ofthe dike because the revetment
is relativcly impermeable then uplift pressures can develop under 'the revetment. This
will tend to lift the revetment. In this situation the revetment should be designed as dis-
cussed in Section 20.1.
The soil under the phreatic line is saturated. Weak soil under the cover layer can, under
a sudden lead, bc softened and lose its bearing capacity. Although this phenomenon
should be taken into account in the dike design, in general it is not practical for it to form
a basis for revetment dimensioning.
The maximum allowable slope angle is also dependent on the water-level in front ofand
in the dike, see Sectien 20.7.
F~nally the water-level also affects the wave elimate and the currents at the dike.

19.1.2 JVaves

Waves, at least those dealt with in this section, are generated by wind. The factors
affecting waves include the wind speed, the fetch, the wind duration and the water dcpth.
A wave can be expressed in terms of:

1. The wave hcight (lJ);

2. The wave period (7);
3. Thc speed ofpropagation (c),

The wave length is: L = c- T.

In the area where the wind disturbs the sca a very irregular wave field can devclop which
comprises a whole range of waves moving in several directions with different heights
and periods. When the influcnce of the wind deminishes the waves propagate out of the
wave field in different directions. The short period waves damp out before thc long
pc riod, leaving a more regular swell.
One method of defining a wave field is that wich is referred to as a spectrum, by which
the wave period is rclatcd to a parameter deduced from wave hcight. A wave spectrum is
normally composed from a number of measurements made at severallocations over a
long or short period. From these measurements a series of stochastic quantities are
obtained over a limited time period which can be treated statistically to long-term
predictions. More information on this topic can bc obtaincd from the considcrable
literature availablc and also from Appendix I.
A parameter which is frcquently uscd to dcscribe a wave in deep water is thc signif-

IlO
icant wave height (Jlsig)' This is the avcrage of the highest third of the observed waves.
This value agrees weIl with the wave height which, based on a Raylcigh distribu-
tion ofwaves in a wave field, is exceeded by 13.5% ofthe waves.

Roughly three are as can be distinguished:

1. Deep water
The wave movement is not influenced by the waterdepth; the speed ofpropagation is
related only to the wave period. The movement ofwater particles in the wave at the
bottom, is not notiecabie.
2. Shallow water
The waves are greatly influenced by the bottom; the wave cclerity is determined by
the water depth alone. The horizontal velocity of the water particles in the wave is
independent of the vertical position of the partiele and is very strong at the bottom.
3. Intermediate water
The wave movement is influenced by both the wave period and the water depth.
Generally these categories are expressed as:

- Deep water dl Lo > 0,25

- Shallow water dl Lo < 0,05
Lo is the wave length in dcep water and d the waterdepth.

As the waterdepth (d) deercases a wave ofheight H travelling from deep to shallow water
will increase in height to a certain limiting value given by H Id = y, y is the breaker in-
dex, after which the wave starts to break, sec Figure 19.1.
A wave approaching a dike slope will, depending on the water depth in front ofthe dike,
the wave length, and the wave height have already brok en or will break on the dike face.
In the case ofthe latter the breaking process wiIl be influenced also by the dike slopc.
There are a number of breaker types:

1. Spilling breakers:
The wave front, in th is case, is steep but not vertical. The wave begins to collapse at
the crest and the breaking mass of water runs into the trough of the preceding wave
and up the wave slopc.
2. Plunging breakers:
A 'tongue' extends out from the crest and falls freely into the preceding wave trough
or onto the dike face.
3. Surging breakers:
These develop with low waves and steep dike slopcs. The wave builds up like a
plunging breaker but before the crest tongue has formed the toe of the breaker surges
up the slope and the breaker develops no further.

It is possible to get an impression ofwhich type ofbreaker will devclop with the helpof
the so-called ~ parameter.

111
in which:

IX = slope of dike face (dgr)

Ho= deep water wave height (m)
g acceleration due to gravity (m/s-)
T = wave period (s)
~<0,5: spilling breakers
0,5 < ~<3,3:, plunging breakers
~>3,3: surging breakers

d H = wave height
d = wave depth
L = wave length
Hb = breaking wave height
db = water depth ot breaking I

Figure 19.1 The breaking of a wave as it approaches a dike slope.

A dike revetment must be able to resist the forces exerted by waves. In the design
distinction is made between plate-type structures and structures built up of loose elc-
ments.
Plute-type revetments are, for example, asphaltic concrete revetments, mastic slabs, fully
grouted rubble layers, open and dense stone .asphalt revetments and layers of
lean sandasphalt.
The largest loads which waves exert on a plute-type revetment are impacts from plunging
breakers. The dimensioning method is given in Section 20.2. Structures built up from
loose clements are for example pattern grouted riprap revetments. Riprap slopes, with
surface grouting form, an intermediate type of structure between fully grouted riprap
and pattem grouted riprap.
Water flowing on and off the revetment can cause shear which has an abrasive effect.
This aspect is discussed in Section 20.5.
At the same time pressure differences can be built up by wave movement across a
relatively waterimpermeable revetment. This results in upward prcssures which tend to

112
lift the revetmcnt and reduces the resistance to slipping. The pressure differences can
also produce a hydraulic gradicnt in the groundwater which can cause the move ment of
sand along and/or through the revetment, The result can be a cavity under the re-
vetment.
There are no spccific dimensioning methods to deal with this phenomenon. A saturated
subsoil can be softened by a sudden loading, for exarnple, a wave impact, and lose its
benring capacity.
The water movements in a wave, whether or not it has brokcn, can erode the bed at the
toe ofa dike. In order to prevent this a bed proteetion is often constructed. Ifthe bottom
proteetion is extended to whcre the wave movements are too weak to transport the sand
particlcs, thcrc will bc no problems provided that therc are no other streng currents
present. To guard against erosion the bed proteetion should bc sufficiently long so that
any scouring stops at some distance from the dike toe. This aspect is discussed further in
Sectien 20.6.2.
A dcnsc underwater bottorn proteetion can bc lifted up by differences in prcssure,
above and below, caused by translatory or standing waves. This phenornenon is
discussed further in Sectien 20.3.2.
It is usual to design arcvetment for extreme conditions referred to as the design
boundary conditions. \Vhat should not bc overlookcd, howevcr, is th at wave farces,
although they are not always extremc1y large, can occur frequently and that a lot of
minor damages can eventually lead to failure. Since asphalt is sensitive to fatigue
- the strcngthreduces in relation to the loading frequency - this propcrty should not be
overlooked in the design.

Waves, acting on arcvetment ofloose elemcnts, such as riprap, will exert certain forces.
Thc flow pattern causcd by wave action on thc rcvctrncnt is complicated and, as aresuit,
it is not possible to give a good theoretical assissment of the problem. Formulas for
determining the weight of revctmcnt elcmcnts under wave attack are wholly or partly
empirica!. An example is theHudson formula, sec Section 20.4.

19.1.3 Currents

Currents can be distinguised into:

stationary currents;
quasi-stationary currents, for example, ihosc caused by the movement of a ship; .
non-stationary currents, for example, the run-up and run-down on a dike face causcd
by waveaction.

Flowing water can have an eroding effect on an asphalt revctment apart from th at
caused by material which it is transporting. Pieces of wood, stone, etc., being carried by
the water, can hit the revctmcnt and cause damage, especially in the breaker zone, sec
also Section 7.5.3.

113
Flowing water can get hold on the side of thin asphalt plates to flap. At the same time the
pressure differences which develop on a platetype, watertight bed proteetion can cause it
to lift. Methods of design against this arc given in Section 20.3.

As.is the case with wave actionit is not possible to adequately describe the loads on a
loose clement revetment due to currents, thcoretically. Thc weight of the revetment
clements must be determined using formulas developed empirically.
N on-stcady flows, because of thc fluctuations can induce fatigue loads onto an asphalt
revetment. The effect of currcnts on the different types of asphalt is trcated in Section
20.5.
Finally, currents can crodc thc bottom in front ofa dike and the toe can be undcrmincd.
This can be prevented by using a bottom protcction, sec Section 20.6.2.

19.2 SettIement and scouring

19.2.1 Settlement
Settlcment is caused by the deformation ofthe dike body and the subsoil. The revetment
must, within limits, be able to adjust to irregular settlcment. Since it is a viscous material
the ability to adjust is determined by its composition. Overfilled mixes, such as mastic,
are very flexible, whereas mixes in which the voids are not completely filled with bi-
tumen, such as asphaltic concrete, are less. .
If the speed of irregular settlement is great then the asphalt platc cannot adjust
immediatcly and a cavity will develop under the cover layer. Thcn, ifthe bending stress
caused in the material exceeds the maximum allowable, the structure will fail. This
manifcsts itsclf in crack development around the edgcs of the settlement. Large scttle-
ments can cause significant changes in the thickness and material propertics of the
original revetment. Simultaneously other forces, for exarnple, a wave impact on the
revetmcnt before it has adjustcd to the settlement of the subsoil can cause extra tension
and possible structural failure. It is very important th at such settlernents remain as small
as possible. A methad of designing a plute-type rcvetmcnt to cape with settlcmcnts is
given in Section 20.6.1.
Settlements of the dike body itself, can be lirnited by using appropriate construction
methods: construction of the sandbcd in layers, prcferably with a certain amount of
e?,cess material to be removed by grading, will ensure th at it is well-compacted, sec also
Chapter 18.
Subsoil settlements, to some extent, must bc accepted. If'unacceptably large settlcment is
expected, th en it may be possible to improvc the siibsoil or to excavate and replace it.
It is important th at any subsoil discontinuities, for example, old watercourses, which
could lead to uneven sett\ements, are taken into account (Photo 5, 6, 7). Also joint
construction, should not be overlookcd in relation to this. Because of fa ulty construction,
material can vanish from the dikc body and cavitics can dcvclop undcr the revctmcnt.

114
In genera! an asphalt revetment is as most materials unable to adapt to cavities which
develop quickly. ldenlly the dike design should be such mat cavities cannot appeaL

19.2.2 Scouring

Scouring mostry develops when material in front of the dike is eroded by. for example.
waves or currents. Erosion can propagatc under the revetment. It can be 50 exrensive and
rast that revetments, for exa mple asphaltic concrete and grouted riprap. cannot adjust.
The dike should be öesigned in such a way that scouring cannot develop, with, for

Puoto 5 Deforrnation of all esphaltic concrete revenuent caused by irregular seulement of the dike
body

115
 Photo 6
The eau se of the senlement

Photo 7 Detail oflocal damage due to the settiement

116
example, a suitable toe and bed proteetion. Asphalt mastie, being a fairly viscous
material whieh ean adapt to relatively large scouring, is often used as bed protection.
Design methods for th is iype of application are given in Section 20.6.2., see also
Photo 8.

19.3 . Other types of'loads'

Table 19.1 reviews other important 'loading types' which can affect the various revet-
ment matcrials. The numbers in the table refer to thc, remarks given below. These
factors must be taken into account. In general. however, there are no definite rul es
which can be applicd.

Tablc 19.1 Review of other load types on an asphalt revetment.

pene- dense open lean

asphaltic tration stone stone sand- mem-
concrete ·mastic mortars asphalt asphalt asphalt branes
biologicallchcmical
attack (I) x x x (1.1) x x x x (1.2)
collisions x x x x x (2.1) x
anchors (3) x x (3.1) x (3.1) x x x
floating debrisl
icc (4) x x x (8) x x (10)
recreational activi-
ties/vandalisrn (5) x x (8) x (10) x
traffic: (6)
during execution x x x x x x x
after completing x
long-term leads (7) x x x x x x
Remarks
I For biological/chemical attack sec Section 6.5.
1.1 Affect of the grouting mortar between the stones.
1.2 For membranes made in-situ, which are sensitive to plant penetration from bel ow, the bed
should be pre-treared with a planticide. When tubers are present in the soil, however, plan-
ticidc has been found, gene rally, to be ineffective.
Plant-resistant foils are often built into prefabricated membranes.
2 Impact loads from colliding ships can not only darnagc the revetrnent, but also soften the
subsoil 50 th at it loses its boaring capacity and the revetrnentmay slide off. Dimensioning is
maybe possible by the models givcn in 20.3 and Appendix 1.1.1.
2.1 To illustrate:
Ship tests have been carricd out in the Rhine-Main-Donau Canal on an 18 cm thick layer of
open stone asphalt overlying a 12 cm layer oflean sand asphalt, In the tests a Europa c1ass ship
(76.5 x 12 m; total weight = 1,720 ton) was driven at a degress to dike face of I : 2.5 at a speed
of 1.2 mIs into a dike face; as a result the ship carne about 90 cm out ofthe water. After a few
minutes it pulled itself free. Investigations showcd that the ship had made a shallow groove,
I to 2 cm dcep, 4 cm wide and 12 to 15 m long, in the dike face. No cracks in the surface
were found. Vessc1s travelling at slower speeds and smaller approach anglcs didn't cause
any noticable darnage (29).

117
3 An anchorwhieh is dragging or has hooked up on the edge ofthe revetment can cause damage.
Falling anchors can punch a hole in the revetment. If the asphalt structure is designed as a
watertight layer then this type of damage cannot be allowed and anchoring here must be
forbidden. Alternativc1y the revetment should be design cd to rcsist these loads.
3.1 Sunken pipclines, syphons and underwater mains are occasionally protected against anchors
by using a grouted 'rip-rap or mastie cover.
It has been found that a 1350 kg Danforth anchor will not break through a 900 mm thiek layer
ofrip-rap (10/60 kg). The danger ofbreaking through is much less with a fully grouted layer of
stone. An anchor tends to slide on a mastie apron and, therefore, do no harm. Investigations
indicate that the thickness of the proteetion construction, rc1ated to a layer of rip-rap 10/60 kg
with a thiekness of am depending on anchor size, should be (26):
- fully grouted a thiekness ofO.5a m
- mastie slab a thickness of 1I3a m.
4 Floating debris and ice can hit the revetment and damage it. Damage can be avoidcd, as much
as possible, by:
- using arcvetment with a suitable bitumen hardness, a smooth surface, and adequate
thiekness;
- the timc1y removal of debris from the dike face slo pc.
5 Normally a dike is accessible for recreational activities and the design should, therefore,
inc1ude amongst others:
- special parking places ifthe revetment is inadequate against the Ionding ofcars
- a hard binder, a minimum number of'gripping' points and a good quality surface finish to
prevent vandalism.
6 The design should take into account traffic on the dike during both the construction phase and
on completion. Construction access roads should be provided. In addition the size ofthe dike
and its components, the method of construction and the mixes should be adequate. Specific
dimensions can be derived by using conventional road construction rnethods (94).
7 Since asphalt is visco-clastic permanent features, such as pipelines etc. can cause deformations
which eventually may lead to damage. Special provisions should be made.
8 The breaking of stones out of the revetrnent can be limited by using large stones and good
grouting.
9 Since lean sand-asphalt filter layers and cores are ultimatc1y covercd, they are only prone to
damage dunng the construction ph asc. Damage to filter layers duringconstruclion, may not be
allowed.
10 Individual stones can be jerked relativc1y easilyout of the revetment.

118
20 Design methods

In the following sections design methods are given for the most common applications of
asphalt rcvctrnents, including:
A dense asphalt revetment against hydraulic uplift pressures resulting from quasi-
static conditions.
A plute-type asphalt rcvetment against wave impacts.
A dense underwater bed proteetion against uplift pressures caused by currents and
waves.
A surface- or pattern-grouted crushed stone layer against wave action.
An asphalt revetment against currents.
An asphalt revetment adapting to irregular settlement and scouring.
Determination of the maximum dike face slope is also discussed.
Design methods are not given for all types of loading situations. The methods de-
scribed can probably also be applied to loadings other than those discussed.

20.1 Dense asphalt revetment designed to resist hydraulic uplift pressures

Hydraulic uplift prcssures dcvclop under a sealed, absolutcly or relatively water-un-
permcabie dike revctmcnt as a result of diffcrences in water-level inside and outside the
dike body. This pressures can force the revetment off the dike face. To prevent this
and to ensure that the revetment continues to function it is necessary 10 take the
ph en omen on into account in the design.
Air pressure can develop under an airtight dike crest revetment as a result of rising
groundwater levels. This pressure can cause the revetment to crack. The effect can be
avoidcd by ensuring that there is good ventilation.
Air pressures act over the whole width of the crest whereas water pressures tends to act
locally.

20.1.1 Hydraulic uplift presslIres

Hydraulic uplift pressures can be caused by:
Quasi-statie conditions:
• The groundwater level in the dike lags behind the ebb and flood of the,tide. As
soon as the groundwater level is higher than the water-level outside the dike body
there is an hydraulic uplift pressure under the revetment.
• The largest uplift pressures can be expected after a storm surge when the water-
level outside the dike body falls rapidly and the groundwater in the dike falls more
slowly.

119
 • Uplift pressures can develop during construction and for some considerable time
afterwards if in the vicinity of the revetment, sand is moved by means of hy-
draulic transport.

Dynamic conditions:
Uplift pressure develops in the dike body when the water-level outside is lowered
locally, over a very short time period, by a passing ship.
• Uplift pressures develop wh en wind waves produce changes in water-level on the
dike face.

Dynamic conditions which can cause uplift water pressures are not discussed further
here. Often the time interval in which they act is sa small that insufficient water can flow
to develop pressures of any consequence. Further information is given in Section 20.3.

A large number of factors influcnce the development of hydraulic uplift pressures:

the height, duration, and form oftime-dependent boundary conditions such as storm
surges and tides in front of the dike, and the potential on the rearside of the dike: the
polder level, the drainage ditch level, etc.;
the permeability and the differences in permeability ofthe soil in and under the dike
body;
the dike geometry: dimensions; dike face slopc; berms; toe level;
the water storage capacity of the subsoil;
the level of the foreshore in front of the dike;
the level ofany impermeable layers, for example, clay, in the subsoil;
the length of any sheetpiling in the toe;
the presence and type of any drainage systems in the toe or elsewhere in the dike
body.
A drainage system is somctimes installed in the dike to 'contrcl groundwater move-
ments. By using a toe drainage it is possible to completcly or partly prevent the
build-up of uplift pressure. It is important that such drainage systems continue to
function throughout the lifetime of the structure.

Bccausc many ofthe above factors are variabie it is not possible to give general rules for
determining the amount of uplift pressure which can develop. It is always advisable,
certainly for large projects, for sites where the subsoil is not homogeneaus, when a
permeabie layer is present under the revetment, or if the revetment design is very
different fram that schematised in Figure 20.1 to determine the uplift pressures using
electrical analogues (52, 53, 85) or a finite clement method of calculation (79).
The Van der Veer method can be used to obtain a very prcliminary estimation ofuplift
pressures. This methad has, however, drawbacks, the most important being:

- The subsoil must be homogeneous, a situation which rarely occurs in practicc.

120
 The groundwater level has to be estimated.
The revetment must be schematized as shown in Figure 20.1.

The Van der Veer method often gives va lues which appear to be too low in comparison
with electrical analogue results.

THE ELECTRICAL ANALOGUE

A good method for determining the hydra uIie uplift pressures under an asphalt revet-
ment is the electrical analogue. This method, which has been developed extensively,
simulatcs the groundwater flow by electric current. Often the designer does not have an
analogue available and 'advice must then be sought of the appropriate bodies. The
best known analogues in the Netherlands are the ELNAG model and the electrical con-
ductivity paper modeIs, (Teledeltos) (52, 53, 85) ofthe Delta Department ofthe Rijkswa-
terstaat. With these analogues time-dependent boundary conditions such as tides and
storm surges can be used as inputs, sec Appendix III.

THE VAN DER VEER FORMULA

This formula is suitable for a preliminary estimate of uplift pressures. It is based on
two-dimensional groundwater flow in a homogeneous subsoil and the presence of the
maximim uplift pressure at the location ofthe outside water-level. The latter occurrence
is valid if more than 20% of the revetment lies under water, measured from the bottom
edge of the revetment to the level of the phreatic Ene. The height of the phreatic line
must be estimated.

0)=0

Figure 20.1 Schematization of the revetment for the Van der Veer formula.

T!J.e maximum uplift pressure is determined for stationary flow, - constant horizontal
supply of groundwàter - and non-stationary flow (51). The maximum uplift pressure, P,
develops at the waterline:

when _v_ <0.8 to 0.85

a+v
and is given by p = c· 9v where 9v is the difference between the water-level in front of
the dike and the groundwater level in the dike body in the stationary case.

121
The coefficient c is given by:

stationary flow:

c= v 1- --(v
a+v
)"/0
non-stationary flow

with

c= n1 arccos [ ( v
2 a+v
),,/0 - I] with () = arctg (11)
tt
+ 2"

Where the dike face slope is I : 11.

In the stationary case, in which the phreatic line remains at a constant level, the uplift
pressure is notably bigger than the value obtained with the formula for non-stationary
flow.
The affects of a sheet pile wall or a toe proteetion are shown in Figure 20.2.
The coefficient c then becomes:

stationary flow:

c- _ V ( a+q+v
1- v ):</0

non-stationary flow:

I
c= - arccos
tt [2 ( a+q+v
v ):</0 _ I]

.~
. '. v .

a a

Figurc 20.2 Schcmatization of' a toe proteetion and a sheet pilc wall

122
When making a preliminary estimate for tides and storm surges the parameter v, can be
taken as 50% of the difference between the highest and the averuge outside water-levels.
For long-term water-level differences, such as in reservoirs and in high water conditions
in rivers \' is taken as 100% of the difference.

20.1.2 Design

DESIGN CRITERIA
Wh at is the effect of hydraulic uplift pressures on a dense asphalt revetment?
a. If the weight component of the revetment down the dike face is greatcr than the
frictional resistance then the rcvetmcnt will tend to slide locally. The revetment will
hang 'on higher parts and rest on lower parts where the frictional resistance is still
sufficiently large. As aresuIt deformation - stress and struin - will develop (53).
Bccause the asphalt is viscous the deformation will be permanent. With a series of
loading cycles the deformation can become so large that the material will yield. In
addition signs offatigue will appear.
The area of greatest uplift pressure will move together with the outside water level
down the dike slope. The section ofrevetment which originally supported tbe section
above will, at a certain condition, also slide. With successive high waters the revet-
ment will, thus, tend to creep like a caterpillar down the slope.
b. If'the hydraulic uplift pressure is larger than the weight component normal to the dike
face then upward forces will develop which can lift the revetment. In the resulting
cavity underneath the revetment, sand movements can take pi ace which prevent the
rcvetrnent from returning to its original profile. Because these sand movements are
downwards there is a tendency for bulges to develop down the slope and subsidences
on the upper parts. In view of the characteristics of asphalt the buiging can on the
long-term or with regular loading, be considerable and continuing.
The following design criteria can be set out:

I. Sliding criteria
The revctment should be designed so that it does not slide under frequently occurring
loading situations such as spring tides,
2. Uplift criteria
In loading situations which occur rarely, such as storm surges, the component of the
weight of the revetment, normal to the dike face should be greater than the uplift
pressurc caused by the water.
3. Equilibrium criteria
The revetmcnt must be in equilibrium as a whoie.

Remarks

In order to limit the uplift pressures an open revetment can be used in the area on the
dike face where the grcatest pressures occur, mostly in the tidal zone.

123
The uplift water pressures can be schematized as shown in Figure 20.3. The uplift
pressure must be determined by the designer. Determination of the maximum uplift
pressure (<Jwo) is treated in Section 20.1.1. The variation of the pressure can also be
determined using an electric analogue or a finite clement calculation. The term <Jwo' used
in the following formulae represents the maximum hydraulic uplift pressure. The Van
der Veer mcthod and the electrical analogue give the maximum potential difference, p,
in metres of water, at the surface of the revetment. The uplift water pressure, <Jwo• is then
obtained by adding h cos exto the value of pand multiplying the whole by Pwg (80), sec
also Appendix lIl.

\
\

Which indicates th at wh en p = 0, the uplift pressure <Jwo = Pwgh cos ex

cW : Pw· 9 Ip + hcosa 1

Figure 20.3 Schematization of the water pressure under a sealed revetment

The dimensions ofthe revetment can be obtained using the following formulas, sec also
Appendix lIl.

I. Sliding criterion

f·<1v.o
11> (}".g'(fcos a-sin a)

2. Uplift criterion

If thc revetment is not supported below, for example, by a toe construction or by

another revetment then a check must be made to ensure that the tensile strength in thc
asphalt is not excecdcd, sec Appendix lIl. If there is a possibility that it would be
cxceeded then the layer thickness must be increased and/or the di kc face slopc

124
reduced. In this case the maximum value of the hydraulic uplift prcssurewhich would
be reached in the most extreme conditions can be used.

Symbols used:
h revetment thickness (m)
(Jwo maximum uplift pressure (Nz m'')
For the sliding criterion (Jwo is determined for frequcntly occurring con-
ditions; for thc uplift criterion, conditions which occur vcry rarcly are uscd
a slopc of dikc face (dgr)
Pa asphalt bulk dcnsity (kg/rn'')
Pw density ofwater (kg/rn")
g acccleration due to gravity
f coefficient of friction: f = 'tg 9 if 9>0, else f = tg 0
9. angle of intcrnal friction of the subsoil (dgr)
(} angle of friction between the revetment and the subsoil (dgr).

3. Equilibrium criterion
See Section 20.7.

20.1.3 Spccific constructional features

TOE PROTECTION
Mustie, in the form of a horizontal or al most horizontal slab, can be used in front of the
toe of a dike, see figure 20.4.
Ifa mastie slab is sealed directly onto the toe ofa dense revetment very high uplift water
pressures can develop under the whole construction and a very thick revetment would be
required. Ta avoid this an open 'berm' can be used between the slab and the toe or good
drainage must be provided.

slab

open construct Ion

Figurc 20.4 Hydraulic uplift pressures under a slab at the front of a dikc.

125
The uplift criterion applies to the slab.
In the first approximation:

,,>---- (Jw

(la -s- cosa

In which:
aw = maximum uplift water pressure under the slab (N/mZ)
h slab thickness (m)
Ga bulk density of mastic (kg/m-')
g acccleration due to gravity (m/s-)
Cl: = angle of inclination of the slab (dgr)
If it is likely that scouring occurs and the slab has to adjust to this (Figure 20.5), thc
following checks must be made (4):

a. The uplift criterion in the new situation should not be excceded.

N.B.: the uplift pressures will be different in the new situation.
b. The slab should not slide.

slab

.' ."
open eenstrucnon

Figurc 20.5 Mustic slab modifications after scouring.

GROUTEDRtP-RAP(crushed stone)
In thc past it was often recommended that, in that part of the dike face where thc largest
hydraulic uplift pressures would occur and where the sliding criterion would be ex-
ceeded (in the Netherlands, in the tidal zone), not to use revetments like asphaltic con-
crete. Material with a more definitc skeleton structure such as (grouted) tip-rap was
recommended. \Vith such a material the normal strcsses are transferred beller to the toe,
An asphalt revetment with a definite skeleton structure will behave less viscous than a
mix in which all the particles are coated with bitumen. (Another solution, ifit is possible,
is a watcr-pcrmeable revctment.)
A fully groutcd stone layer has relatively large intcrnal stability. Therefore, in the past,
thc design was never based on the sliding criterion; under extreme conditions the de-

126
sign was bascd, simplyon uplift. It is essential that the revetment is well supported
by, for example, a toe construction.
A fully grouted stone layer is, in principle, impermeable, although due to lack of
adhesion between the stone and the grouting mortar, there will be a certain amount of
permeability. No account ofthis should be taken inthe design.

LEAN SAND ASPHALT

The permeability of lean sand asphalt should be similar to or larger than that of the
underlying sand bed in order to prevent the development ofhydraulic uplift pressures.

ASPHALTIC MEMBRANES
Asphaltic membranes must be watertight and must remain so under the water pressures
which develop. The value ofthe limiting pressure can be found from a permeability test;
sec, for example (35) or derived from the manufacturer's specifications. It should not be
possible for uplift water pressures to lift a membrane, that is, the proteetion layer should
be sufficiently heavy. The dimensions of the proteetion layer can be found using uplift
criteria:

aw
h~----
Q .g. cosa

In which:

h = thickness of the proteetion layer (m)

Uw = uplift water pressure (Nz m'')
Q = bulk density of the proteetion layer (kg/rn-')
g = acceleration due to gravity (rn/s'')
0: = slope angle (dgr)
If the membranc is applied on a slope arid covered with an asphalt mix, the sliding
criterion must also be applied. In addition, tensile forces in the membrane should not
be too excessive, sec also Appendix VII.

20.2 Design of a plate-type asphalt revetment agalnst wave impacts

The largest forces that waves can cxert on a plute-type revetment are impacts caused by
plunging breakers.
Plate-type revetments include asphaltic concrete plates, mastic siabs, fully grouted stone
layers, open and dense stone asphalt and layers of lean sand asphalt.

20.2.1 Wave impact loads

A wave impact occurs when a mass ofwater from a plunging breaker strikes the slope at

127
great speed. A wave impact is, in fact, regarded as a pressure which acts over a certain
width, To obtain the appropriate dimensions the impact is schematized as a line load.

P = p'b

where:

P the size ofthe wave impact (NIm!)

p the maximum pressure (NIm:!)
b the width over which the pressure must act in order to represent the
complete wave load (m)
Wave forces are dependent on a large number of factors such as, for example, wave
height and steepness, and slope angle. Preferably they should bedetermincd for each
partienlar situation by obscrvations and investigations. If this is not possible use can bc
made of the data given below. These data have been devcloped from the results of an
invcstigation carricd out by the Delft Hydraulics Laberatory (93) in consulration with
Working-Group 1 of the Dutch Technical Advisory Cornmittee on Water Defences. In
th is investigation into pressure magnitude, pressure width and duration and the loca-
tion ofwave impacts on the slope, only slopes of 1: 3 and 1: 4 were studied, Using va lues
given in, for cxarnple, the 'Voorlopig Rapport 1961' (13) the results can bc cxtrapo-
latcd to other slopes. The new wave impact values can differ from those which were
used extensively in the past from the 'Voorlopig Rapport'.

The various parameters are given by the following relationships:

- Maximum pressure

P=Gw·g·q·/f

in which:

(}w = density of water (kg/rnê)

g acccleration due to gravity (ru/s-)
IJ wave height (m)
q a factor related to the slope (see table bclow)

slopc q
I: 2 2.3
I: 3 2.7
I: 4 2.3
I: 6 2

The schematized width over which the maximum pressure is considered to act:

b = 0.4· IJ
128
 The duration of the pressure effect, t sec, is, depending on the particular model:
slope ::; I : 3; t = 0.06 . HII2
slope 2':1 : 4; t = 0.18 'HII2
The length of the wave impact is dependent on the angle of wave approach to the
. slope and the speed of propagation of the wave. The larger these factors are, the
shorter is the impact length.
-' A breaking wave hits the slope at a distance MI below still water level (SWL), see
Figure 20.6. The 'area' in which /j./z lies is shown in Figure 20.7.

Figurc 20.6 The breaking wave.

'rne point wher e the breoking WO\l@

hits the stcce lies in the shoded areo.

1:4 1:5
stcpe 1: n

Figure 20.7 The point, uh below SWL, where the wave impact hits the slopc, related to the wave
height Hand plotted against the slope 1 : n.

In the above the parameter H is the height of a single wave. In practice the load on the
revetment wil! be due to an irregular wave train containing a large number ofwaves of
different height and frequencies of occurrence. The significant wave height, Hs, which
characterises a particular wave field can be used for determining the wave impact. The

129
number of times that this wave occurs is selected so that the samc total 'fatigue load' is
reachcd as th at caused by the wave field as a wholc, sec Appendix 1.2.The calculation of
wave impact should take into account the fact that only a small number of waves in the
field will cause actually an impact on the revetment.

20.2.2 Construction schematization

It is important that the schematization of the construction approximates as much as

possible to reality. The most simple schematization is that of a plate lying on an elastic
subsoil with a delaycd response. This is developed in Appendix 1.1.1.
The method treated in this Chapter is suitable for wave impacts and alsofor other types
of loads which can act on the revetment. These inc1ude, for examplc, loads caused by
colliding vessels, maintcnance and recreations traffic, long-term loads and loads during
the construction ph ase.
The loads can be:

a. Statie; that is loads which are always present. The way in which a construction reacts
to statie loads depends, amongst other things, on the size of the load, the stiffness of
the revetrncnt and the subsoil, and the thickness of the revetment.
b. Dynamie; that is time-dependent loads. In addition to the factors mentioned above
under statie loads the speed, frequeney and type of loading, the density of the
revetment material and thc damping and dircctly rclated mass ofthe subsoil are also
important. Wh en the rcvetmcnt is frequently loaded the asphalt propertics are alter-
cd: the strain at break reduces. This phenomenon is known as fatigue.

In th is manual the statie solution is applied to the sclccted schernatization, sec Appendix
U.I. The number of loading eycles is taken into account in the calculation of the
breaking strength of the matcrial. The duration of loading is incorporated into the
stiffness modulus of the asphalt mix.
The loading is schematized as a line load. Since, in praetiee, the wave impact is a
distributed load, corrections are applicd.
This schematization is not so suitable for eomplieated constructions (varying layer
thiekness, rnulti-layer systcms, joints etc.). In these situations a more extensive calcu-
lation involving, for example, finite clement methods can offer the best solution.

20.2.3 Asphalt and subsoil properties

When designing it is essential to know the asphalt and subsoil properties. Preferably
data should be obtained by earrying out specific tests for cach design. Ifsuch tests are not
possible, use can be made of the general values given in Tables 20.1 and 20.3. The
matcrials used in thc construction should then be carefully checked against these values.

130
 Tablc 20.1 Moduli of stiffncss and related initial stresses and strains at break for different asphalt types. (With the exeeption of open stone
asphalt, the values are deterrnined using Figures 6.1., 6.3., and 6.4. with whieh values for the lower loading cycles have been
extra pola ted),
Initial strain at failure Initial stress at break
number ofloading cycles number ofloading cyc1es
mix stiffness
type modulus N/m2 I 100 1000 10000 100000 1 100 1000 10000 100000
Asphaltic
concrete _. 7: 109 . 1.2· 10-3 5.2' 10-4 3.4 . 10-4 2.5' 10-4 1.6· 10-4 8.4· 106 3.6' 106 2.4· 106 1.8· lOG i.t lOG
Asphaltic
mastic 1.109 8.6· 10-3 3.4.10.3: '2.2' 10-3 1.4· 10-3 8.6' 10-1 8.6' lOG 3.4· lOG 2.2' lOG 1.4 . lOG 8.6' HP
Densc stonc /

asphalt 4.5' 109 2.3' 10-3 9.2' 10-4 5.8' 10-4 3.7' 10-4 2.3 . 10-4 1.107 4.1' 106 2.6· 106 1.6· 106 I· 106
/

Openstone /

asphalt 7, IO~ 3.4· 10-3 1.3. 10-3 7.9· 10-4 4.8' 10-4 3.0' 10-4 2.4· 106 9.1 . lOS 5.5 . lOS 3.4· 105 2.1 . lo-~
Lean /

sandasphalt 1.109 1.1 . 10-3 4.2' 10-4 2.6' 10-4 1.7· 10-1 '1'.1. 10-1 1· lOG 4.2' 105 2.6 . Io-~ 1.7· 105 1.105

-
Vol
Table 20.2 Mix eompositions of the asphalt types given in Table 20.1.

Composition (by mass %)

(very)
erushed weak bitumen Voids ratio
Mix stone sand filler 80/100 (%v)

Asphaltic
concrete 46.9 39.5 7.5 6.1 5
mastie - 64 17 19 0
Dense stone
asphalt 60 25.6 6.8 7.6 5
Open stone
asphalt 82.9 9.9 4.4 3.1 31
Lean
sand asphalt - 96 - 4 30

The asphalt properties - modulus ofstiffness and stress at break - are temperature and
loading duration dependent. Since the heavy storms, against which the revetment is
mostly designed, occur mostly in the winter season a temperature criterion ofSoC can be
accepted (13). The loading duration is wave height-dependent and can be deduced from
Section 20.2.1. Variations in this parameter, within practicallimits, have little effect on
asphalt properties. .
Some properties of the most used asphalt types are given in Table 20.1.; the mix
compositions of these types are given in Table 20.2. Bitumen 80/100 is used by which an
unfavorable loss in penetration is. taken into account. The temperature is fixed at SOC
and the loading period 3 seconds.
The stresses at break for loading cycles ofiess than 10,000 are found by linear extra po-
lation on a log-scale. Although th is mcthod is not completely correct it gives presumably
safe values.
Table 20.3 General values for the modulus of subgrade reaetion of differentsoil types.

soil type modulus ofsubgrade

reaetion c (N/m3)
sand - medium eompacted (relative Praetor density 87-95) 1 X 107_1 X 108
- well eompacted (relative Praetor density 95-100) I X 108-3 X 108
,
sand + clay 3 X 107-8 X 107
sand + silt 2 X 107-5 X 107
clay - low eompressibility 3 X 107-5 X 107
- high eompressibility < 4x 107
Peat < s x 107
Gravel > 7x 107
Lean sand asphalt > s x 108

132
For strongly deviating mix compositions, temperatures and loading conditions the mix
eompositions can be determined using the method given in Part A, Sections 6.3.\. and
6.3.2. The Poissons ratio of the mix follows from Section 6.3.4.
The parameter defining the subsoil is referred to as the modulus of subgrade reaction.
General values of th is parameter are given in Table 20.3. (10, 50).

20.2.4 Design

Design criteria are selected sueh that the stresses and strains developing in an asphalt
plate ofa certain thick~.ess due to bending moments do not exceed the allowable values.
Plate thiekness can be determined using the calculation model developed in Appendix
l.I. for a plate of constant thickness on a delaycd reacting elastic subsoil.
The formula reads:

h = 0,75. V2716 . (I - I v)
2 • ( P)4 . (~)
(Tb C

in which:

h = thiekness of revetment (m)

ab asphalt stress at failure (N/m2)
P wave impact (Nz'm-)
S stiffness modulus ofthe asphalt (N/m2)
I' Poisson ratio for asphalt
C Modulus ofsubgrade reaction (N/m3)
0.75 = reduction factor, sec Appendix l.I.\.

Usually the revetment isxlesigned for a design parameter ·such as the significant
wave height, lis, which characterises the wave elimate in a severe storm or at an
extremcly high design water level.
If the revetment is also subjected to loads ofa normal daily wave elimate then this should
be taken into account in the layer thickness calculated with the design Hs. A method for
doing this which gives safe values for Dutch conditions is given in Appendix 1.2.3.
Distinction should be made between:
\. Thàt part of the revetment on which only the design condition, Hs' 'acts. No
correction is necessary here.
2. That part on which the design conditions act together with the normal wave climate.
3. That part on which only the normal wave elimate acts (and not the design wave).

That part ofthe revetment which, for example, lies above the (spring) tide zone is classed
in Category \. The revetment in the (spring) tide zone is acted on by the normal wave
elimate but, in general. not by the design wave, Hç; which occurs at higher water levels.

133
This part is c1assed in Category 3. Dikes with a deep foreshore can be c1assed in
Category 2.

20.2.5 Practical application of the wave impact formula

General values of layer thickness are given in Figures 20.8. to 20.12. for some standard
mixes, for various significant design wave heights and subsoil parameters. These valucs
are based on assumptions for application in Dutch conditions, see Appendix 1.2.2.
The assumptions are:

the design storm has a duration of36 hours (3 tidal cyc1es)

the number of wave impacts used for dimensioning is 10% of the total number of
waves in the design storm
the waves in the storm are generated in a wave field which has a Rayleigh distribution
the relationship between the significant wave height H, (m) which characterises the
storm and the avcrage wave period in the storm, T(sec) is r = 3.5 X HsO.5
the wave impact is determined from Section 20.2.1. in which Pw = 1000 kg/rn" and
g = 9.81 m/s-
the asphalt propertics are given in Table 20.1. The number of loading cyc1es, lis,
related to the significant wave height, Hs' necded to determine the strength at break,
is given in the table below. -
A value of 0.35 is taken for Poisson's ratio.

Hs(m) ns
2 9900
3 8000
4 6950
5 6200
6 5670
7 5250
8 4900
9 4630
10 4400

In order to obtain an optimum design or ifthe design conditions differ greatly from these
assumptions, then the calculation of layer thickness can deviate from the method used
for Fig1;1res20.8. to 20.12. .
Adaptations can be made by:
- ..
changing the mix composition;
treating/compacting the subsoil;
changing the dike geometry.

It is then essential to carry out extra checks on site and in the laboratory. This wiII also
indicate the programme for the execution requirements.

134
 t 0.8
0.8 0.8

1: 31 I slopo 1: 21 1 : 1.1 I slope 1: 61

.0.6
IJl
IJl

'"c: Hs(m!
'";S
<J 0.1. Hs(m) 0.1.
Hslm)
6
6
ë 5 6
~ 0.2 0.2 5 5
I. I.
c; I.
> 3 3 3
e
0
107 108
modulus of
109
subgrade
2
0
107
reaction
108
IN/m3 )
109
2
:l======
107 108
=2
109

Figurc 20.8 Necessary layer thickness for a revetment of esphaltic concrete, plotred against the
modulus of subgrade reaction and for various significant wave hcight and slopcs (sec
the text above for the basis ofthe graph).

0.8

I slope 1: 31 I slope 1 : 21 1 : 1.1 I slope 1: 61

] 0.6 0.6
VI
IJl

'"c:
Hslml
'"
<J
;S
0.1. Hslml 0.1.
6
ë 6
5
E 0.2 0.2 5
I.
c; I,
> 3 3
e 2 2
0
108 109 107 108 109
modulus of subgrade reaction I N/m3)

Figure 20.9 Neccssary laycr thickness for a revetment of mastic, plotred against the modulus of
subgrade reaction and for various significant wave heights and slopes (sec the tcxt
above for the basis ofthe graph),

135
 t 0.8 0.8 0.8

I slope 1: 31 I slope 1 : 2 11 : 41 1slope 1: 61

0.6 0.6

Hslm) 0.4 0.4

Hslml
6 Hslm)
ë 6
~ 0.2 5 0.2 0.2 6
5 5
ë;; 4 4
> 3 4
l: 3
2 2 ~
0 0 0
107 108 109 107 108 109 107 108 109
modulus of subgrade roccüon (Nlm3 )

Figure 20.10 Necessary layer thickness for a revetment of dense stone asphalt plotted against the
modulus of subgrade reaction and for various significant wave heights and slopcs
(sec the text above for the basis of the graph).

2.0 2.0 2.0

1 slope 1 : 31 I slope 1: 2 11 : 41 I slope 1 : 61

1.6

ijl
Cl Hslm)
c Hs(m)
:Q 0.8 6 0.8 0.8 Hs\m)
:5 6
6
ë 5 5
0.4 0.4 5

tOt======~
4
4
3
----2 -----2 3
o o
107 108 109 107 108
modulus of subgrade reaction (N/m3)

Figure 20.11 Nccessary laycr thickness for arcvetment of open stone asphalt plotted against the
modulus of subgrade reaction and for va rious significant wave heights and slopes
(sec the text above for the basis ofthe graph),

136
 1.0 1..0 1..0

I slope 1: 31 I slope 1 : 2 , 1 : 1.1 I slope 1 : 61

3.0

]
IJl
~ 2.0
c
.Y.
U
6 Hslml
;S 6
5 6
ë 1.0 1.0 5 1.0
'" I. 5
E I.
<;; 4
3 3
e> 2 2
3
2
0 0 0
107 108 109 107 108 109 107 108 109
modulus of subgrade renetion I N/m3)

Figure 20.12 Necessary layer thickness for a revetment of lean sand asphalt plotred against the
modulus of subgrade reaction and for various significant wave heights and slopes
(see the text above for the basis ofthe graph).

The following remarks should be noted for certain specific materiais:

Fully grouted stone

A fully grouted stone layer has, in general since it is composed of severallayers of
stone, considerabie thickness. Because of this it is in most cases not necessary to de-
sign on the basis of wave impacts.
The design of a fully grouted stone layer against wave impact in which the revet-
ment is considered as a plate, is discussed in Appendix 1.3.
Design is based on the normal wave impact formuia using the material propertics of
the grouting mortar (mastic). In order to obtain the design layer thickness the value
calculatcd must be multiplied by a factor lying between 1.4 and 1.75. These factors are
valid if the shear stresses which develop between the mastic and the crushed stone can
be transferred. lt is not immediately obvious to wh at extent th is condition is satisficd,
Open stone asphalt
Because of the open character of open stone asphalt the wave pressure can quickly
propagate through the revetment the result being that the laad on the plate is less than
that indicated in Section 20.2.1.

Lean sand asphalt applied as a care material

The performance under wave impact of lean sand asphalt applied as a core material
can be calculated using the Boussinesq approximation, sec Appendix V.

137
20.3 Design of underwater bed proteetion against hydraulic uplift pressures caused by
currents and waves

20.3.1 Uplift pressures caused by currents

A mastic slab is often used to proteet the bed against erosion caused by currents.
In extreme situations fluctuations in the flow of water and groundwater can create
pressure differences across a bed protection. If the pressure abovc a revetment plus its
weight is less than the pressure underneath, the revetm~nt will tend to be lifted (81).
This should be prevented. If a slab lifts a cavity develops underneath into which water
flows, the cxtcnt depending on the duration of the pressure difference. Because of the
viscous propertics of the asphalt mix it will deform. Depending on the duration and the
quantity ofinflowing water, the deformation can be so large that the slab breaks.
A simple solution to this problem cannot be given since the pressures which dcvelop vary
from situation to situation. For simple cases the water pressurè can probably be es-
timated using simple formulas. For more complicated cases an extensive calculation
program me or an clectrical analogue can offer the solution.
More important than the lifting ofthe revetment by exccss water pressure is the stability
of the edges of a bed proteetion ormattress, sec Section 20.3.2 and 20.5.

20.3.2 Hydrauite uplift presslIres due to wave action

An impermeable bed proteetion slab, lying on sand, can be lifted by differences in

prcssure above and bclow the slab caused by wave action (41, 79). This must be
prevented.
Wave action causes groundwater movements under the bed proteetion which change the
groundwater pressure. These changes, howcver, are not necessarily the same as those
which occur abovc the slab and an upward pressure can result.
Two situations can be identified:

a. The wavelength is longer than the bed protection.

The maximum pressure under the revetment, in this situation, can be estimated using
Barends (79). Obviously the weight of the revetment must be greater than the uplift
pressure. The following rclationship is derivcd in Appendix 11:

(Jw H
h~-·-
9' (Ja 2 (1< L)

in which:
h = thickness ofbed proteetion (m)
(Jw = density ofwater (kg/rn'')

138
 (Ja = bulk density ofbed proteetion material (kg/rn'')
H = wave height (m)
I = length ofthe bed proteetion in the wave direction (m)
L = wavelength (m)
b. The wavelength is much shorter than the length of the bed proteetion.
In order to prevent the slab from being raised its weight must be greater than the maxi-
mum uplift pressure underneath. An uplift pressure is caused by the pressure re-
sulting from the wave action abovethe slab being,locally, less than the groundwater
pressure below. The time-dependent pore water movements, which strongly
determine the groundwater pressures, cause considerable damping to this effect
(41,79), An approximation for this system is given in Appendix 11.With the formula
derived, it is possible to estimate the thickness ofrevetment required. In the example,
worked out in the appendix, it can be seen that the damping due to groundwater
movements is very large.
Under normalconditions the phenomena of lifting by wave action is not of major
. importance; More important is the possibility of scouring at the edges of the bed
proteetion as a result of erosion. If the sand at the edges is unstable i t is recommended
that the watertight bed proteetion is overlapped at the edges with an open sand-tight
revetment.

20.4 Design of a surface- or pattern-grouted stone layer agalnst wave attaek

A much-used formula for calculating a construction built up out ofdiscrete elements, for
example, quarry stone, to resist wave attack is the empirical Hudson formula (44):

(Ja' g.H)
w= KD·À ) -cotg e

in which:
W = weight of an element (N)
(Ja= density ofan element (kg/m'')
g = acceleration due to gravitytm/s'')
H = wave height (m)
.1 = relative density of an element
.1 = {Ja-Ow
{JN

OW density of water (kg/rn'')

KD damage coefficient, which takes into account the shape, degree of
interlock, roughness .and location of the revetment on the slope.
n =slope angle,

139
Remarks

I. TheHudson formula is valid for slopes with a < 33.7°.

2. The formula has been developed for breakwaters, subjected to non -breaking waves.
The effect ofbreaking waves can be introduced by lowering the value of Ko (86).
3. The formula is only valid for revetrnents on the front face ofthe breakwater.
4. The wave conditions are characterized by a single parameter, H. Investigations (87)
indicate that the damage, that is the number ofrevetment elements rcmoved, is fairly
independent ofthe storm duration except when the design wave height is excccded by
over 30%.
5. The Hudson formula is designcd for regular waves. Investigations (87) show that
damage caused by a wave spectrum characterized by a significant wave height Hsig' is
larger than th at caused by regular waves of the sarnc height. This difference becomes
more pronounced as the spectrum width is increased.
6. The wave period is not taken into account.

The value of the damagecoefficient Ko can be increased by grouting a crushed stone

construction with asphalt mortar.

The following applications can be distinguished:

Surface-grouted stone
If the size or weight of crushed stone is barely inadequate to satisfy the wave
conditions thcn the safety of the revetment can be increased by fixing the stone with
an asphalt grout.
Ifabout 30~ ofthe voids in the stone is covercd, the Ko value can be multiplied by I
to 1.5.
Note: In principle thc revetment does not consist of individual clements and th is
method should not, in fact, be applicd. In practice, however, it works weil and gives
satisfactory results.
Pattern-grouting
If about 60% of the total surface is filled the Ko factor can be multiplied by 5 to 7.
From model investigations it appears that a relativcly smaller increase in stability is
obtained by grouting more than 50% of the surface ofthe crushed stone. Grouting 50
to 70% of the total voids in the stone laycr appears to give the best results (27).
The revaluating ofthe Ko factor is very dependent on the execution and care must be
taken to ensure that the grout does not remain in the surface of the stones or sags
completcly through the cover layer.
Prototype tests with high waves, Hsig up to 8 m, demonstrate that, because of the
enormous forces developed by breaking waves in the clefts between the stones, the
grouted lumps themselves can be split and shell-shaped pieces can brake off. This
type of construction, therefore, should not be used in arcas of heavy wave attack; it
has proved to be successful, however, in waves of 3 to 5 m (4).

140
 For reasons of safety it is recommended that three layers ofbroken stone are used, only
the top two being grouted. A few loose stone will probably be washed away by the waves
but th is will not be dangerous. If a lump of grouted stones is washed away the third layer
will still proteet the core since it is held fast by the overlying grouted lumps (27).
Ideally the KD value should be found from model tests.
Grouting produces a smoother revetment surface and, as a result, there is more wave
uprush, The crest, thcrcfore, needs to be higher. An indication of this is given in Table
20.4.

Tablc 20.4 ,
Wavc uprush on ccrtain slope surfaces comparcd with a smooth surfacc.

run-up on thc surface

surface type run-up on a smooth surface
smooth: impcrrncablc \,0
rip-rap 0.5-0.6
pattcrn groutcd 0.6-0.7
fully groutcd 0.6-0.8

20.5 Design of un asphalt rcvctment against currents

The possible forces which can bc caused by flowing water are given in Section 19.1.3. The
performance of various asphalt types under such forces is discussed in the present
section,
Under normal conditions an asphalt revetrnent is very resistant to flowing water.
Considerabie damage however can develop, for examplc, if the water carries hard
objects such as stones. In addition currents can lift the edges of a plate or mattress and
cause to turn them over. This can be prevented by, for cxarnple, increasing the weight of
the edges or by burying them.

ASPHAL TIC CONCRETE

Asphaltic concrete is only attacked slightly by currents. If the water, howcver, carries
hard particles, impacts can occur which damage the revetment material.
In Los Angeles it was found that, in the prevailing temperatures with a minimum of
about lOoe, an asphaltic concrete revetment can resist the erosive effect of debris in the
water ifthe binder is softer than penetratien 50. For Dutch conditions, with a minimum
temperature around the freezing point, bitumen 80/100 is more suitable (12).

MASTIC
Mastic is an overfilled mix with a relatively low stiffness. The stresses caused by impacts
from debris etc. carried by the flow are generally small and can be withstood. It is
important, however, that the edges of the plate are not made to flap by the current. This
can be prevented by:

141
I. burying the cdges of the plate so that the eurrent ean gain no purehase on it.
2. dumping erushed stone on the cdges. These pieees will then penetrate into the mastie
layer until thcy are in equilibrium. This proeess is dependent on the differenee in
density between the stone and the mastie, the shape ofthe stones and the viscosity of
the mastie. (The penetration into a layer of'stoncs takes longer than of loose stones.)
Investigations indieate that a few eentimeters ofmastie slab originally 20 em thick still
remained under the stone, dumped under water, plaeed five years ago (88). -

A mastie slab ean be built up from a number of separate layers laid like rooftiles over
eaeh other. Ifthcre is not good adhesion between these layers, beeause ofthc presenee of
sand pollutions or inadequate heat transfer, flowing water ean get hold ofthe unattaehed
scctions, lift them and break them off (81).

GROUTING ~IORTARS
The following formula is aften used for designing loose erushed stone against sta-
tionary or quasi-stationary flow (25):

in which:

mcdian diameter ofthe revctment material (m)

eurrent veloeity parallel to the axis of the ehannel (mis)
a slope anglc
Q~-Qw
Pw
density ofthe revetment material (kg/rn'')
density ofthe water (kg/rn")
angle of internal friction of the rcvetment
aeeeleration due to gravity
a stability parameter
The stability parameter is dependent on many factors. With a uniform, eontinuous flow,
for eonditions whieh occur in the Duteh waterways it is, on average, 1.4. For other situ-
ations reference should be made to literature (25).
With grouted crushed stone, the stability parameter b ean be lowered. For prcfercnce,
the value of b should be established for each particular situation by model tests.
The effect of currents on fully grouted stone is negligible.

DENSE STONE ASI'HALT

Dense stone asphalt is an overfilled mix and, thus, very resistant to currents. The design
should be sueh that the cdges do not flap.

142
OPEN STONE ASPHALT
Information about the rcsistance of open stone asphalt to currents is, as yet, not
complete. Investigations have produced the following results:

With the stationary and quasi-stationary flow very limited crosion was observed after
34 hours with current veloeities of6 mis, (6). The damage which did occur consisted
mainly of the loosening of limestone particles.
The material has also been tested with currents generated by ships' screws. These tests
were made on a bank revetment of 12 cm lean sand asphalt under a cover of 18 cm
open stone asphalt in the Rhcin-Main-Donau Canal, The currents we re produced by .
the screws ofa cargo ship with a draught of2 mand an 800 HP motor at full strength
for 5 minutes. No damage ocurred.
Investigations have been carricd out into the rcsistance of open stone asphalt to wave
attack, which showed that under normal tide conditions and also over a long period
the material did not erode.

The cdges of an open stone asphalt mattress can flap when the critical stcady flow
velocity (2.5 to 3 mis for a IS cm thick mattress) is exceeded. This can be prevented by
making the edges henvier. This phenomena of flapping can occur with all types of
siabs and mattresses.

LEAN SAND ASPHALT

Indicative tests have shown that loosely dumped lean sand asphalt is resistant to currents
upto 3 mis (32). This resistance to currents can be increascd by ruising the bitumen
content and by a certain amount ofcompaction. When lean sand asphalt is used as a co re
material there is generally only need to take flow precoutions during the construction
phase. The loss of a certain amount of core material during construction can be acceptcd
provided that, on completion, the core is sufficiently large.
When lean sand asphalt is used as a filter Iayer it must only be exposed to direct currents
. during construction. This phase should be as short as possible.
If lcan sand asphalt is used as a permanent revetment there must be no loss of matcrial.
Therefore the waves and currents to which it is exposed should not be to large. Tests
indicate that the maximum current velocity is about 3 mis. More investigations into th is
aspect are desirabie.
Erosion can possibly be restricted by compacting or adapting the mix or providing it with
a seal coat. These measures can affect the water permeability. Crushed stone in the
weight range upto 60-300 kg can be laid directlyon lean sand asphalt as a proteetion
layer. By experience it appears that larger stones cause turbulence which can erode the
lean sand asphalt. Ta prevent th is an intermediate layer of, for example, gravel can be
used. .

143
20.6 Designing an asphalt revetment against irregular settlement and scouring

20.6.1 Irregular set/lemen I

Deponding on the speed of settling and the propertics of the asphalt, an asphalt revet-
ment will not immediately adapt to irregular settlcment. After some time, however,
provided that there is no break, the revetment willlie again on the subsoil. The bending
ofthe asphalt plate due to irregular settlcment and the time taken for it to adjust can be
determined using simp le formulas derived by mechanics (50).
Ifthe surface ofthe settlement is assumed to be circular, the time taken for the revetment
to adjust, I, can be determined using the following formula. By repeating the calculation
a number of times with related values of the stiffness modulus, I can be approximated:

13 16·v·h2
--= 2 4 (sec Appendix IV.I)
S(/,T) 3·Ga·g·(I-u )'11

in which:
Ua = density ofthe asphalt (kg/rn")
g acceleration due to gravity (rn/s-)
1I Poisson's ratio for asphalt
h thickness ofthe revetment (m)
11 horizontal speed ofsettling (mis)
v vertical speed of settling (mis)
time (s)
time at which settling begins: I = 0
S stiffness modulus of asphalt, time and temperature dependent. The
related time, can for example, be fixed at 0.5 I
The speed of settling (v and 11) is assumed to be constant, sec also Figure 20.13.

I a , a

144
The largest bending stress will develop just before the plate touches the bed, see
Appendix IV. I. The bending stress is then:
3 . (la • g. 112 • tf
Ub = 4."

t1 = adjustment time

If the expansion of the settling has stopped before the revetment can adjust:

a = Sec Figure 20.13

If the allowable stress at break, Ob rnax- is smaller than the stress which develops, Ob, then
the revetment will fail before the material can adapt to the subsoil surfacc.
The bend in the plate can also be so big that, after a long time, the deforrnation capacity
is exceeded and the plate breaks.
The presence of a cavity under the revetment can in conjunction with other loads, for
example, a wave impact, cause extra stress. This can lead to failure.
In addition large settlements can lead to a reduction in the layer thickness (viscous flow)
which reduces the strength of the revetment.
If the asphalt revetment is to follow irrcgular settlement without cracking this will
depend on the speed ofsettling and the asphalt propertics.

Asphalt mastic is a very suitable material for adjusting to irregular settlement because
it is reasonably viscous. In the above mentioned formulas the stiffness modulus Scan
be rcplaccd by 37]/t in whieh 7] is the value ofthe viscosity ofthe mastic, sec Appendix
IV.2.
Grouted crushed stone must be able ot adjust to irregular settiement without losing its
cohesion. It is more able to do th is when the voids are completely filled.
It is important that open stone mattresses remain in contact. with the subsoiI. If the
mattress is anchorcd, stresses can devclop in certain sections which in combination
withwave induced forces can lead to failure.
Because ofits limited ability to adjust and in view ofits function great careis needed
when using lean sand asphalt as a filter layer or revetrnent. Failure will not develop,
initially, because oflarge deformation but because oftoo great a deformation speed.
Because of its large bulk wh en lean sand asphalt is used as a core material, its ability
to adapt to settlement is of much less significanee than that of a sand asphalt filter
layer. Thc care should bc designed to act monolithically undcr diffcrcntial scttlement,
see also Appendix V. Thc tension level and the deformation speed should, howcver,
~~~~. .
145
 An asphaltic membranc which adjusts to settlements will stretch, sec Figure 20.14. If
the extension is 111 and the original length I then the strain is MI I (that is .
111 = V11112 + I1x2 - I1x).
This strain must not exceed the maximum allowable value.
Reinforcing in the membrane enables the strain whieh otherwise would concentrate
in one pi ace and could lead to excessive extension.

I
------------,
----------1
I
I

6.x

Figurc 20.14 Extcnsion of a mernbrane.

20.6.2 Scouring

Scouring usually takes place so quickly that the asphalt rcvetrncnt cannot adjust, The
dike should, thcrefore, bc design cd in such a way th at the revctrncnt cannot be
undermined, for cxample, by providing a good toe and bed protection.
Bccausc ofits good viscosity, asphalt mastic can, to a large extent adjust to underrnining,
sec Photo 8, and is, therefore, often applicd as bed protection.
Ifscouring occurs at the end of a mastic apron it will bend, sec Figure 20.15. The time for
the end of the plate to reach the bottom of the hole can be determined from the
following:

t= (sec Appendix V.2.)

in which:

11 = thickness ofthe revetment(m)

11 = viscosity of the mastic (Pa' s)
Ga = density of the mastie (kg/rn")
g acceleration due to gravity (m/s'')
V = Poisson's ratio for asphalt
l' horizontal scouring speed (mis)
z depth ofscour (m)
time when bending bcgins: t = 0 (s)

146
Photo 8 The effects ofscouring on a mustic slab

The speed ofscouring is assumed to be constant.

lil)

-
.-- , ./ ,
I
scourmg speed ,..,.

-- Vlm;I//

-- ~_-=-_-_-""'~~ - -- - ----1
Pigure 20.\5 Scouting of a mastic slab (schematic).

The largest bending stresses will develop ut a distance /(1) from the end of the slab just
befare the end of the slab setnes onto the bed of the scour hole. Ir these stresses exceed
the limiting value the revcunent will break.

(see Appendix V.2.)

in which:

Ub = bending stress (Nim:?)

1 adjustruent time (s)

147
As soon as the end of thc slab touches the bed, in principle, the scouring process stops.
Thc mastic slab, howevcr, does not lic everywhere in contact with the bed and in between
the slab and the bed there ma)' be cavities.
The length over which the slab is not supported is increased since the bed underneath
adjusts to its own natural flatter slopc. The slab will then bend further to complete its
adjustment. Ncw strcsses develop in the slab which in combination with the reduced slab
thickness can lead to failure. In practice a crack can dcvclop at A, sec Figure 20.16.

Figure 20.16 Bending of a mastie slab after the end has settled onto the seour hole.

Failure can also occur in a runstic slab lying on a sand bed which over1ies a cIay layer. In
this situation wave pressures cannot propagate through the clay layer and will build up to
eau se large uplift pressures under the mastic slab.
Sand under the slab can be lost through cracks and as aresuIt undermining can proceed.
llecause of the stiffness propertics of the material cracks are more likely to develop in
winter than in summer. In warmer periods it is possible for the cracks to flow together
and re-scal.
The reduction oflayer thickness due to viscous flow should not be overlookcd.
The length of a mastic slab in front of a dike must be sa large that the toe cannot be
undermined and the slope stability is endangered.

20.7 Dctermlnntlon of the maximum slope

In order to prevent the revetment sliding off the dike body (Photo 9), the slope angle
must be less than the angle of internal friction. For a relatively impermeable revetment,
the slo pc angle in places where water is likcly to occur behind the revetment, should be
na larger than:

tg a <: tg r/J. [(lW]

I - On (sec Appendix 1\1)

148
in which:

a = slopc angle (dgr)

rp angle ofinternal friction of the subsoil (dgr)
êw = densuy of water (kg /m")
Qn = density of wet soil (kg/rnê)

For a slope of cohesiouless material where groundwater can flow out Freely, in order to
prevent slope instability the angle shauld be (84):

- under water: above water:

tg a
tg t,b> ;0--=-'''----,
[I - Qw • (1 + ig'a)]
Q"

in which:

i = potenrial gradient at the surface

The allowable slope is also determined by the iuternal stabilityofthe mix itself, in both
the construction and the completed phnse. The method of construction can ntso be a
deterrnining factor in the choice of slope.

Photc 9 Failure of an asphultic concrete revetmen t due te unstability of the sub-soil

149
ASPIIALTIC CONCRETE
Depending on the mix composition asphaltic concrete is stablc on a slope of about I : 1.7
even when hot (3, 22). After completion the mix is stabIe at the sarne angle. The
construct ion method, particularly the applicability of the construction equipment, is
generally the limiting factor.

lIlASTIC
Mastie is, preferably, used on a horizontal subsoil. For a lean mastie the .maximurn
allowablc slope is about \00 and for a fat mix 5°, depending on the duration and
temperature regime.

GROUTING MORTARS, SURFACE-GROUTING

The method of applying the mastie must be sueh th at in the whole surface the stone re-
mains in plaee.

PATTERN-GROUTING
The grouted areas of stone must bestabIe under the prevailing eonditions. The areas
must satisfaetorily loek up the interlying stones. It is, thus, important th at the grouted
areas remain as lumps. The mix composition and the method ofexeeution must be sueh
th at the grout does not stay in the upper layers or sags completely through the layer.

FULLY GROUTING
Beeause ofits internal skeleton this type of revetment has high stability, sec also Chapter
II.

DEN SE STONE ASPIIALT

Dense stone asphalt is preferably used on a flat subsoil. Sec also Chapter 12.

OPEN STONE ASPIIALT

Open stone asphalt, when hot, is stabIe on slopcs upto about I : 1.5, sec also Seetion
13.3.1.
The stability of open stone asphalt mattresses should be such that they do not slide, as a
whole, down the slope. The maximum allowable slope is about I : 2.5.
It is possible to anchor mattresses to the slopc. In th is case it is essential that they remain
in contact with the subsoil. .

LEAN SAND ASPIIALT

Lean sand asphalt has because ofits viscousity just as other asphalt mixes the tendeney to
creep down a slo pc
I. In constructions where deformations_play an important rolc, such as foundations and
abutments, ereep has to be limited.

150
2. For constructions in which deformation is not important the only limitation is safety.
The resistance to failure ofgranular materials is characterized by the angle ofinternal
friction, 9, and the cohesion, c. ,
The failure mechanism can be studled using, for example, a slip circle analysis or the
Prandtl wegde method.

For present purposes the equilibrium method is used to calculate the stability of the
material in which it is assumed that deformations remain sufficiently sm all if the
relationship between the shear stresses which develop and the maximum allowable shear
stress is equal or less than 0.75. For bituminous products the stability problem is more
complicated because of the viscosity. A time-dependent component must be taken into
account.
Investigations have been carricd out on lean sand asphalt made from sand from the
Eastern Scheldt and 3 to 4% bitumen 80/100. The object was (32):
I. To establish the minimum angle ofinternal friction, 9 and thus the stress situation for
which the deformation speeds are smal!. Under such shear stresses 9 was in the range
20° to 24°.
2. To establish the maximum angle of internal friction, 9 that is, the situation when
deviator-stresses produce deformation speeds which do not become smaller but
remain constant or increase. Under such shear stresses, 9 was in the range 30° to 34 0.
3. To establish relationships between the stress conditions and deformations and
deformation speed between the maximurn and minimum.

There are several ways in which the failure criteria can be established:

I. Bishop's slip circle analysis

In this method the total shear strength which develops at the critical slip surface is
determined. The Bishop method is the most traditional and readily applicable
method, see Appendix VI.
2. A finite element method '
The applicability of what is referred to as the MARC calculation program for
simulating creep in asphalt is studied at the Laboratorium voor Grondmechanica
(Soil Mechanics Laboratory) at Delft.
3. The plate method
This is a modification ofthe Bishop method, the underlying principle ofwhich is that
the sliding of adjacent plates of asphalt can be calculated with the existing method
and that from this the relationship between deformation speed and shear stress can be
found simply. In the method stiff plates can slide over each other by means of
imaginary viscous interlayers. The summation of the separate deformations indicates
the extent of the total deformation.
Large deformations, developing in a short time indicate unallowable loads and thus
failure (32).

151
ASPIIALTIC MEMBRANES
Asphaltic membranes are general1y covered. In situ prepared membranes are often
covered with earth; prefabricated membranes with earth, rubble, concrete, asphalt, etc.
The fol1owing failure mechanisms can develop:
I. The membrane causes the proteetion layer to slide off.
To prevent th is the friction component along the membrane should be larger than
the weight component of the proteetion down the slope, that is:

f>tg Ct (see Appendix VII)

inwhich:

f the friction component between the membrane and the subsoil or proteetion
layer
Ct = the slope angle of the membrane
2. The shear stress transfer over the membrane can cause such large deformations that it
can result in changing geometry or damage of the protection.
The deformations which develop 'as a result of shear forces acting on the membrane
should not exceed certain limits, see Appendix VII.

152