Specialists in Geological, Geotechnical, Tunnel and Underground Engineering

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing

Frans van der Merwe

Geotechnical Engineer

Computational Geotechnics, July 2018

 Overview Soil Behaviour HS CM Parameters Empirical C
Overview
» Soil behaviour,

» The Hardening Soil (HS) Constitutive Model (CM),

» Parameters,

» Empirical Correlations,

» Examples.

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 3
Soil Behaviour
 Overview Soil Behaviour HS CM Parameters Empirical C
Soil behaviour to consider
(Lees, 2016)
• Soil Type
• Soft clay and dense sand behave very
different.
• Normally different CM for fine grained and
granular soils.

• Tensile Strength
• little to no σtu so this needs to be included in
the constitutive model

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 5
 Overview Soil Behaviour HS CM Parameters Empirical C
Soil behaviour to consider
• Strength changes during
shear
• Loose soils compress dense soils dilate both
towards the critical state. i.e. shear hardening
in loose soil and shear softening in dense
soils.
• Further shear will result in further reduction to
the residual shear strength and progressive
failure in stiff clays (London Clays)

• Stress-dependency of
stiffness and strength
• Increase in confining stress results in
decreased voids and therefore a denser
consistency and a higher stiffness

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 6
 Overview Soil Behaviour HS CM Parameters Empirical C
Soil behaviour to consider
• Stress-path dependency of stiffness and strength
• On primary loading soil shows highly non-linear stiffness
• On the unload-reload portion of the curve a sudden reduction in stiffness will
be noted as soon as the deviator stress moves from the reloading curve to the
primary loading curve
• Stiffness decreases with deviatoric load (triaxial) but it increases under
compressive load (Oedometer) due to increasing density of the soil.

increases

decreases

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 7
 Overview Soil Behaviour HS CM Parameters Empirical C
Soil behaviour to consider
• Permanent deformations
• Many materials have a significant elastic
range, where elastic reversable
deformation occurs. Soils, does not, and
strain both plastically and elastically
occurs at quite low stress below the
failure stress.

• Bonding and structure

• Natural soils develop fabric and inter-
particle bonding “structure” that gives
additional strength and stiffness. Many
constitutive models are based on
laboratory tests but so many
reconstituted tests are done?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 8
 Overview Soil Behaviour HS CM Parameters Empirical C
Soil behaviour to consider
• Intermediate principal stress
• Many constitutive models are based on triaxial test results, where the
intermediate principal stress σ2 equals the minor principal stress σ3. However
when in FE analysis the constitutive model operates in general stress space

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 9
 Overview Soil Behaviour HS CM Parameters Empirical C
Soil behaviour to consider

• Anisotropy
• Most soils are anisotropic (different when
measured along different axes) to some
extent due to stress history
• Assuming isotropy can over or under
estimate strength and stiffness

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 10
 Overview Soil Behaviour HS CM Parameters Empirical C
Ground behaviour to consider
• Strain-dependent stiffness
• Mair (1993). At small strains soil stiffness
is high and it decays to a lower value as
strain increases

- Creep
- Soft clays, silts and peats creep after
dissipation of excess PWP.

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 11
 Overview Soil Behaviour HS CM Parameters Empirical C
Triaxial Test - SUU

Deviatoric Loading

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 12
 Overview Soil Behaviour HS CM Parameters Empirical C
Triaxial Test - SCU

Isotropic Consolidation
Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 13
 Overview Soil Behaviour HS CM Parameters Empirical C
Triaxial Test - SCD

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 14
 Overview Soil Behaviour HS CM Parameters Empirical C
Which Triaxial should one
use?

• Short Term – saturated clay?

• Long Term – saturated clay?
• Effective Parameters?
• Undrained Stiffness?
• Drained Stiffness?
• Poisson’s Ratio?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 15
 Overview Soil Behaviour HS CM Parameters Empirical C
Oedometer

One-dimensional compression. Stiffness increases due to

compressive load

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 16
 Overview Soil Behaviour HS CM Parameters Empirical C
Effective Stress Path Plots

Does q need to
be q’?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 17
Hardening Soil (HS) Constitutive
Model (CM)
What is a constitutive model

Overview Soil Behaviour HS CM Parameters Empirical C

(CM)?

Soil constitutive modelling is a mathematical description of the

stress-strain behaviour of soil in response to stresses.

A good soil constitutive model must be able to capture all the

facets of real soil behaviour with as little as possible input to
distinguish between elastic and plastic deformation and should
be easily obtained from simple laboratory tests (Hing, 2009).

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 19
Why is it important to use the

Overview Soil Behaviour HS CM Parameters Empirical C

appropriate constitutive
model?

• Complex soil-structure interaction problems. The analysis

accuracy is driven by constitutive models.
• Soils are usually the softest and weakest elements in a soil-
structure models and therefore governs the deformation and
Probability of Failure (POF)
• Behaviour over a range of stresses needs to be accurately
modeled by the CM
• Concrete/Steel is very stiff and one can therefore use simple
linear elastic constitutive models

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 20
Prof. Pieter Vermeer and Tom

Overview Soil Behaviour HS CM Parameters Empirical C

Schanz developed the HS
Model

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 21
 Overview Soil Behaviour HS CM Parameters Empirical C
Application of constitutive
models
Primary Use:

• Determining the limit state/strength of soil

• Computing failure load
• Assessing forces on construction due to failure

Secondary Use:
• Assessment of deformation of granular materials,
especially when deformations are due to soil failure

Should not be used:

• Used in soft soils with care

• Calculation of settlements in cohesive materials
• Special care needed in undrained conditions / short
term stability of soft soils

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 22
 Overview Soil Behaviour HS CM Parameters Empirical C
Applicability

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 23
 Overview Soil Behaviour HS CM Parameters Empirical C
HS CM

Double hardening model: Both Volumetric and deviatoric hardening.

Deviatoric hardening surface evolves until MC failure line is reached

Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 24
 Overview Soil Behaviour HS CM Parameters Empirical C
Strength Parameters
» Strength Parameters (Failure Criterion) and
Dilatancy Angle,
» Stiffness Parameters,
» Small Strain Stiffness,
» Other Miscellaneous.

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 25
 Overview Soil Behaviour HS CM Parameters Empirical C
Comparison MC and HS
Parameter HS Soil Mohr-
Coulomb
Small strain Eo and γ0,7 none
stiffness
Elastic Eur and vur E and v
Constants E50 and m
Failure φ’ and c’ φ’ and c’
Criterion
Dilatancy Ψ and emax Ψ
Cap Surface Eoed None
OCR
Ko

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 26
Shear Parameters
 Overview Soil Behaviour HS CM Parameters Empirical C
Shear Parameters (c’, φ’)
Before Failure can take place interlocking and frictional
resistance needs to be overcome.
Which value is required in HS Model peak or constant volume?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 28
 Overview Soil Behaviour HS CM Parameters Empirical C
Shear Parameters (cu/su)
Short term shear strength of clayey soils (drain valve closed and
permeability low). What is φu?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 29
 Overview Soil Behaviour HS CM Parameters Empirical C
Dilatancy Angle (ψ = Psi)

» Which soils will have a ψ?

» What test should one do to
establish the dilation angle?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 30
Dilatancy Component of

Overview Soil Behaviour HS CM Parameters Empirical C

effective friction angle

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 31
 Overview Soil Behaviour HS CM Parameters Empirical C
Dilatancy Angle

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 32
Stiffness Parameters
 Overview Soil Behaviour HS CM Parameters Empirical C
Rf – Failure Ratio

In Plaxis Rf = 0.9 is used as the default

qf
Where Rf =
qa
qf = σ1’f – σ3’f

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 34
 Overview Soil Behaviour HS CM Parameters Empirical C
Rf – Failure Ratio
qa is derived as follows

Stress-strain curve may be approximated:

ε
(σ1 – σ3) =
a+bε
ε
a + bε = , a and b might be determined by replotting the stress-
(σ1 – σ3)
strain curve on transformed axes using stresses and strains at 70% and
95% of the failure strain

Where a is the intercept of the

and b the slope resulting from
the line

Therefore:
1
(σ1 – σ3)ult =
b

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 35
 Overview Soil Behaviour HS CM Parameters Empirical C
Rf – Failure Ratio
Example:

(σ1- σ3) f = 180kPa, ε70%f =16%, ε95%f = 28%, (σ1- σ3)70%f = 126kPa and (σ1- σ3)95%f =
171kPa

ε70%f = 16/100 = 0,0012

(σ1− σ3)70%f 126

ε95%f = 28/100 = 0,00163

(σ1− σ3)95%f 171

0,00163 − 0,0012
b= = 0,00358
(28/100)−(16/100)

1
(σ1 – σ3)ult = = 279kPa
0,00358

180 Soft Bangkok Clay

i.e. Rf = = 0,65 (Surarak, 2011)
279
Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 36
 Overview Soil Behaviour HS CM Parameters Empirical C
Application of constitutive
models

• All stiffness moduli are updated according to current stress level.

• Input stiffness are values at reference stress, e.g.pa’= 100kPa.

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 37
 Overview Soil Behaviour HS CM Parameters Empirical C
Stiffness, Ei, E50, Eur

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 38
 Overview Soil Behaviour HS CM Parameters Empirical C
Stiffness, E50
» Stiffness decreases with increasing deviatoric stress

» Ei difficult to establish form the triaxial test so rather calculate E50

» E50 is a secant modulus from the triaxial test at 50% of the

maximum deviatoric stress (qf)

From which type of triaxial test?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 39
 Overview Soil Behaviour HS CM Parameters Empirical C
Stiffness, Eur
» Plaxis Default setting Eurref =3E50ref

Is this true in all cases?

Must be determined from Triaxial test results

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 40
 Overview Soil Behaviour HS CM Parameters Empirical C
Stiffness, Eur
» Normally consolidated clay
(OCR=1) 5 to 8 x E50ref

» Overconsolidated Clay
(OCR>1) 4 x E50ref

» Normally consolidated sand

4 to 5 x E50ref

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 41
 Overview Soil Behaviour HS CM Parameters Empirical C
Oedometric Stiffness, Eoed
» Stiffness increases with increasing
compressive load as soil “densifies”

» Plaxis default Eoedref = E50ref

Not always the case see CUR2003-7

Determine the correlation from oedometer and triaxial tests

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 42
 Normalized/Reference

Overview Soil Behaviour HS CM Parameters Empirical C

Stiffness
σ3+ccotφ m
» Eur = Eurref( )
σref+ccotφ

σ3+ccotφ m
» E50 = E50ref( )
σref+ccotφ

σ +ccotφ
» Eoed = Eoedref( σ 1 +ccotφ )m
ref

Elastizitätsmodul?
Steifemodul?
Verformungsmodul?

Elastic Modulus? Quotient of tension

and elongation,
Stiffness Modulus = From Oedometer,
Deformation Modulus = PLT.

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 43
 Normalized/Reference

Overview Soil Behaviour HS CM Parameters Empirical C

Stiffness Sand m=0.5
v ((1-v)/v)0.5 (1-v)/((1-2v)(1+v) E50ref/Eoed ref
0.20 2.00 1.11 1.8
σx′ m 0.25 1.73 1.20 1.4
» E50 = E50ref( ) 0.30 1.53 1.35 1.1
σref 0.33 1.42 1.48 1.0
0.35 1.36 1.60 0.8
σ′ 0.40 1.22 2.14 0.6
» Eoed = Eoedref( σy )m 0.45 1.11 3.79 0.3
ref 0.49 1.02 17.11 0.1

1−v Silt m =0.75

» σy’ = σx’ ( v ) v ((1-v)/v)
0.75
(1-v)/((1-2v)(1+v)
ref
E50 /Eoed
ref

0.20 2.83 1.11 2.5

0.25 2.28 1.20 1.9
0.30 1.89 1.35 1.4
E(1−v)
» Eoed = 0.33 1.70 1.48 1.1
(1+v)(1−2v) 0.35 1.59 1.60 1.0
0.40 1.36 2.14 0.6
0.45 1.16 3.79 0.3
E50ref 1−v (1+v)(1−2v) 0.49 1.03 17.11 0.1
» = ( v )m
Eoedref (1−v) NC Clay m=1
v ((1-v)/v)1.1 (1-v)/((1-2v)(1+v) E50ref/Eoed ref
0.20 4.00 1.11 3.6
0.25 3.00 1.20 2.5
0.30 2.33 1.35 1.7
0.33 2.03 1.48 1.4
0.35 1.86 1.60 1.2
0.40 1.50 2.14 0.7
0.45 1.22 3.79 0.3
0.49 1.04 17.11 0.1

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 44
 Normalized/Reference

Overview Soil Behaviour HS CM Parameters Empirical C

Stiffness Sand m=0.5
v ((1-v)/v)0.5 (1-v)/((1-2v)(1+v) E50ref/Eoed ref
0.20 2.00 1.11 1.8
σx′ m 0.25 1.73 1.20 1.4
» E50 = E50ref( ) 0.30 1.53 1.35 1.1
σref 0.33 1.42 1.48 1.0
0.35 1.36 1.60 0.8
σ′ 0.40 1.22 2.14 0.6
» Eoed = Eoedref( σy )m 0.45 1.11 3.79 0.3
ref 0.49 1.02 17.11 0.1

1−v Silt m =0.75

» σy’ = σx’ ( v ) v ((1-v)/v)
0.75
(1-v)/((1-2v)(1+v)
ref
E50 /Eoed
ref

0.20 2.83 1.11 2.5

0.25 2.28 1.20 1.9
0.30 1.89 1.35 1.4
E(1−v)
» Eoed = 0.33 1.70 1.48 1.1
(1+v)(1−2v) 0.35 1.59 1.60 1.0
0.40 1.36 2.14 0.6
0.45 1.16 3.79 0.3
E50ref 1−v (1+v)(1−2v) 0.49 1.03 17.11 0.1

Not Favoured
» = ( v )m
Eoedref (1−v) NC Clay m=1
v ((1-v)/v)1.1 (1-v)/((1-2v)(1+v) E50ref/Eoed ref
0.20 4.00 1.11 3.6
0.25 3.00 1.20 2.5
0.30 2.33 1.35 1.7
0.33 2.03 1.48 1.4
0.35 1.86 1.60 1.2
0.40 1.50 2.14 0.7
0.45 1.22 3.79 0.3
0.49 1.04 17.11 0.1

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 45
 Overview Soil Behaviour HS CM Parameters Empirical C
Power m
The power m indicates the dependency of stiffness on the confining stress

m from zero to unity

m=1 for a normally consolidated clay,

m=0,5 for a medium dense sand
m=0 perfectly elastic behaviour (Surarak, 2011)

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 46
 Overview Soil Behaviour HS CM Parameters Empirical C
Power m
The power m indicates the dependency of stiffness on the confining stress

1.80
y = 1.1189x + 0.9521
1.60
1.40
1.20
log (Ei / pa )

1.00
Soft Bangkok Clay
0.80
(Surarak, 2011)
0.60
0.40
0.20
0.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70
log (σ3 / pa )

What is the power m from the graph above?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 47
 Overview Soil Behaviour HS CM Parameters Empirical C
Shear Hardening

2 q 2q
Yield function (Cone): fs = q − − γp
Ei 1 − Eur
aq
γp = -(2ε1p –εvp) ≈ -2ε1p

1 q q γp is regarded as the plastic shear strain related

ε1p = − to the mobilised shear resisitance
Ei 1 − q Eur
q a

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 48
 Overview Soil Behaviour HS CM Parameters Empirical C
Shear Hardening

Example:
E50 = 10MPa, Eur = 30MPa, Rf = 0.9, qa = 400kPa, q = 100kPa

2x10
Ei = = 18.2MPa
2 −0.9

1 100 100
ε1p = − = 0.0039
18200 1 − 100 30000
400

2 100 2x100
fs = − + 2ε1p =0.0159
18200 1 − 100 30000
400
Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 49
 Overview Soil Behaviour HS CM Parameters Empirical C
Shear Hardening

Example:
E50 = 10MPa, Eur = 30MPa, Rf = 0.9, qa = 400kPa, q = 200kPa

2x10
Ei = = 18.2MPa
2 −0.9

1 200 200
ε1p = − = 0.0153
18200 1 − 200 30000
400

2 200 2x200
fs = − + 2ε1p = 0.0612
18200 1 − 200 30000
400
Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 50
 Overview Soil Behaviour HS CM Parameters Empirical C
Volumetric (Density
Hardening)
The shear hardening yield surfaces does not
explain the plastic volume strain that is
measured in the compression test, mostly
observed in softer type soil and therefore a
second yield surface in the form of a cap
needs to be introduce

q
!2
fc = 2 + (p’)2 – pp2
M

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 51
 Overview Soil Behaviour HS CM Parameters Empirical C
Four Stiffness Zone

Both Volumetric and deviatoric hardening.

Deviatoric hardening surface evolves until MC failure line is reached

Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 52
Empirical Correlations
 HS CM Parameters Empirical Correlations Examples
Density

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 54
 HS CM Parameters Empirical Correlations Examples
Density
2500

y = -0.0812x2 + 13.95x + 1424.8

2000 R² = 0.9741

Dry Density (kg/m3)

1500

1000

500

0
0 10 20 30 40 50 60
SPT-N

Is this correct?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 55
 HS CM Parameters Empirical Correlations Examples
UCS

Is this correct?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 56
 HS CM Parameters Empirical Correlations Examples
Unsaturated SPT - Sand

Mohamed et al, 2012

For clays the effect is

significantly larger

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 57
 HS CM Parameters Empirical Correlations Examples
SPT’s in unsaturated soils
The reason why SPT records change seasonally is the reduction of
matric suction in soil. Matric suction is an apparent increase in
cohesion in unsaturated soils. The matric suction decreases as the
soil approaches saturation. There have been a few investigations
lately in an attempt to propose some relationships between field test
results and moisture content to account for that. For the time being,
care should be taken in relying on field tests above water table.

This is most of our soils…..

Be careful….

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 58
 HS CM Parameters Empirical Correlations Examples
Consistency ito CPT
» DIN 1054: Verification of the Safety of Earthworks and Foundations
D
U = 60 (Uniformity Coefficient)
D10

Compaction Relative Density Cone

resistance of
CPT (MN/m2)
U≤3 U>3
Very Loose D<0,15 D <0,2 qc < 5,0
Loose 0,15 ≤ D <0,3 0,2 ≤ D <0,45 5,0 ≤ qc < 7,5
Medium-dense 0,3 ≤ D <0,5 0,45 ≤ D <0,65 7,5 ≤ qc < 15
Dense 0,5 ≤ D <0,75 0,65 ≤ D <0,9 15,0 ≤ qc < 25
Very Dense 0,75 ≤ D 0,9 ≤ D qc > 25,0

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 59
 HS CM Parameters Empirical Correlations Examples
Consistency ito SPT

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 60
 HS CM Parameters Empirical Correlations Examples
Consisitency ito SPT
» DIN 1054: Verification of the Safety of Earthworks and Foundations
D
U = 60 (Uniformity Coefficient)
D10

Compaction Relative Density (N1)60

U≤3 U>3
Very Loose D<0,15 D <0,2 (N1)60< 3,0
Loose 0,15 ≤ D <0,3 0,2 ≤ D <0,45 3,0 ≤ (N1)60 < 8,0
Medium-dense 0,3 ≤ D <0,5 0,45 ≤ D <0,65 8,0 ≤ (N1)60 < 25,0
Dense 0,5 ≤ D <0,75 0,65 ≤ D <0,9 25,0 ≤ (N1)60< 42
Very Dense 0,75 ≤ D 0,9 ≤ D (N1)60 > 42

What is (N1)60?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 61
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness

» OHDE (1930)/Janbu(1963)

σ we
» Eoed = veσatm ( ) where we=m and veσatm = Eoedref and σatm = 100kPa
σatm
(reference pressure)
» CUR Report 2003-7 (Bepaling geotechnische parameters) recommends the
following: Sand E50ref = Eoedref =Eurref = 1:1:4 or 5, Overconsolidated Clay 1:1:4,
Normally Consolidated clay 2:1:8 and Normally consolidated clayey silt 2:1:5

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 62
Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness
» Duncan Chang (1980)

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 63
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness
» Which one is HS and which one MC?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 64
Recommendations of the

HS CM Parameters Empirical Correlations Examples

“Committee for Waterfront
Structures Harbours and
Waterways”

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 65
Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 66
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness
» Recommendations of the Committee for Waterfront Structures
(2015) according to DIN18196 soil groups

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 67
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness
» Recommendations of the Committee for
Waterfront Structures (2015)

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 68
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness
» Recommendations of the Committee for
Waterfront Structures (2015)

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 69
Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness

What energy?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 70
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness
σ3+ccotφ m
» Eur = Eurref( )
σref+ccotφ

σ3+ccotφ m
» E50 = E50ref( )
σref+ccotφ

σ +ccotφ
» Eoed = E50ref( σ 1 +ccotφ )m
ref

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 71
 Normalized/Reference

HS CM Parameters Empirical Correlations Examples

Stiffness?
» Using Stroud’s (1989)
curves at qnet/qult = 0,5?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 72
 HS CM Parameters Empirical Correlations Examples
From DIN18196 to USCS
DIN 18196 USCS
GE GW/GP/SW/SP
GW/GI GW/SW
GU/GT GM/GC/SM/SC
GU*/GT* GM/GC
SE SP
SW/SI SW
SU/ST SM/SC
SU*/ST* SM/SC
UL CL/ML
UM/UA ML/CL/MH

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 73
 HS CM Parameters Empirical Correlations Examples
From DIN18196 to USCS

DIN 18196 USCS

TL CL
TM CL
TA MH/CH
OU/OT OH/OL

qc (N1)60
<7,5 <8
7,5 -15 8 -25
>15 >25

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 74
vur – unload reload Poisson’s

HS CM Parameters Empirical Correlations Examples

ratio

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 75
 HS CM Parameters Empirical Correlations Examples
Failure versus Deformation

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 76
Applied Footing stress:

HS CM Parameters Empirical Correlations Examples

Drained Loading

q = σ1’ – σ3’
p’ = (σ1’ + 2σ3’)/3 40 + foundation stress (50kPa)

q = 40 -20 =20kPa 20
p’ = (σ1’ + 2σ3’)/3 = 26.6kPa

Change due to building load

q = 90 - 20 = 70kPa
p = ((90) + 2(20))/3 =43kPa

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 77
 HS CM Parameters Empirical Correlations Examples
Dewatering

q = σ1’ – σ3’
p’ = (σ1’ + 2σ3’)/3 40

q = 40 -20 =20kPa 20
p’ = (σ1’ + 2σ3’)/3 = 26.6kPa

Change due to dewatering (negative PWP, assume u = -10kPa)

q = (40 +10) – (20+10) = 20kPa
p = (40+10) + 2(20+10))/3 =36.6kPa, i.e. isotropic hardening

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 78
Excavation (Behind wall

HS CM Parameters Empirical Correlations Examples

Active)

q = σ1’ – σ3’
p’ = (σ1’ + 2σ3’)/3 40
q = 40 -20 =20kPa 20
p’ = (σ1’ + 2σ3’)/3 =
26.6kPa

Change due to excavation

q = 40 – 0 = 40kPa
p = 40/3 =13,3kPa
Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 79
 HS CM Parameters Empirical Correlations Examples
Infiltration

q = σ1’ – σ3’
p’ = (σ1’ + 2σ3’)/3 40
q = 40 – 20 = 20kPa
20
p’ = (σ1’ + 2σ3’)/3 = 26.66kPa

Change due to infiltration (Positive PWP, assume u= 10kPa)

q = (40 +10) – (20+10) = 20kPa
p = (40-10)+2(20-10))/3 =16.6kPa

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 80
 HS CM Parameters Empirical Correlations Examples
Basal Heave

q = σ1’ – σ3’
p’ = (σ1’ + 2σ3’)/3 100 to 0kPa
q = 100 - 50 =50 50
p’ = (σ1’ + 2σ3’)/3 = 66,6

Change due to excavation

q = 0 – 50 = -50
p = 2(50)/3 =33,33

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 81
 HS CM Parameters Empirical Correlations Examples
HS Small
» If the small strain stiffness overlay after Hardin & Drnevich predicts
a stiffness lower than Gurref (Eurref respectively) the model switches
to hardening plasticity of the standard HS model

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 82
 HS CM Parameters Empirical Correlations Examples
HS Small

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 83
 HS CM Parameters Empirical Correlations Examples
HS Small Example
» Assume E50ref = Eoedref = 40000 kN/m2 for a dense sand and

Eurref = 4Eoedref and vur =0.2

Eur 160000
Gur = = = 114 285kN/m2
(1+2vur) 1.4

Go/Gur = 3

Goref = 342 857 kN/m2

γ0.7 = 0.0001

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 84
Examples - Empirical
 HS CM Parameters Empirical Correlations Examples
Plaxis input general

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 86
 HS CM Parameters Empirical Correlations Examples
General – Drainage Type

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 87
 HS CM Parameters Empirical Correlations Examples
General – Density
γunsat (kN/m3) = Dry density
γsat (kN/m3) = Saturated density

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 88
 General - Dilatancy Angle Cut

HS CM Parameters Empirical Correlations Examples

Off (ψ)
» einit and emax of importance
» If emax is reached dilatancy (At the
critical state)
When e<emax :

sinφm −sinφcv
sinψm =
1 −sinφmsinφcv
When e ≥ emax :
ψm = 0°
Void ratio and volumentric strain
linked as follows:

1+e
-(εv – εvinit) = ln( )
1+einit
Hardening Soil (HS) Model –
Using SPT-N values and Laboratory Testing 89
 General - Dilatancy Angle Cut

HS CM Parameters Empirical Correlations Examples

Off (ψ)
» But how does one get these values? emax and einit

Gs
ρd = ρ
1+e w
0.9
emax
0.8

0.7

0.6

0.5
emin
e

0.4
y = -0.17ln(x) + 1.0606
0.3 R² = 0.9862

0.2

0.1

0
0 10 20 30 40 50 60
SPT-N

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 90
 HS CM Parameters Empirical Correlations Examples
Plaxis input Parameters

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 91
 HS CM Parameters Empirical Correlations Examples
Parameters - Eoedref
Example:

Say we have an SPT-N value of 15 for an SW USCS

Change <7,5, 7,5-1,5, and >15 to (N1)60 as <8, 8,0 – 25,0 and >25

700

y = 23.529x + 11.765
600
R² = 1
500
Eoedref = veσatm
400
ve

300 = (23,529*15
200 +11,765)*100
100 =36 470 (kPa)
0
0 5 10 15 20 25 30
(N1)60

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 92
 HS CM Parameters Empirical Correlations Examples
Parameters - Eoedref
Example:

From Probing Results: Assume SE

v = 146 + 217log15 = 401, i.e. Eoedref = 401x100 =40100kPa

But is this N30 value corrected for overburden pressures and energy?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 93
 HS CM Parameters Empirical Correlations Examples
Parameters – E50ref
Example:

SW as per previous slide Eoedref = E50ref = veσatm = 36 470(kPa) for Sand

(OCR=1) according to CUR2003-7

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 94
 HS CM Parameters Empirical Correlations Examples
Parameters – Eurref
Example

SW as per previous slide Eurref = 4 to 5 timed E50ref = 4,5veσatm =

164 115 (kPa) for Sand (OCR=1) according to CUR2003-7

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 95
 HS CM Parameters Empirical Correlations Examples
Parameters – Power m
Example:

SW as per previous slide m = 0,0003(15)2 -0,0202(15) +0,8398 = 0,604

0.8

0.7

0.6

0.5 y = 0.0003x2 - 0.0202x + 0.8398

R² = 1
m(we)

0.4

0.3

0.2

0.1

0
0 5 10 15 20 25 30
(N1)60

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 96
 HS CM Parameters Empirical Correlations Examples
Parameters – c’ ref
Example

SW as per previous slide c’ref at any density or (N1)60 = 0kPa

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 97
 HS CM Parameters Empirical Correlations Examples
Parameters – φ’ and K0NC
Example

SW as per previous slide φ’ = 0,2941(15) + 27,647 =32° (Lower Bound) to 36°

(Upper Bound)
K0NC = 1-sinφ’ =0.47

45
y = 0.4412x + 29.387
40 R² = 0.9643

35

30 y = 0.2941x + 27.647
R² = 1
25
φ'

20

15

10
5

0
0 5 10 15 20 25 30
(N1)60

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 98
 HS CM Parameters Empirical Correlations Examples
Parameters – ψ
Example

SW as per previous slide ψp = φp – φc = 32 – 30 =2°

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 99
 HS CM Parameters Empirical Correlations Examples
Parameters from Triaxial tests

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 100
Example for Triaxial Tests and
Oedometer
Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

Oedometer

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 102
 Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

1.80
y = 1.1189x + 0.9521
1.60
E50 = 80/0.08 = 1000kPa 1.40
(CID S1) m=1.12 1.20
log (Ei / pa )

1.00
(slope of graph)
0.80
0.60
Derived from methods as 0.40

described by Duncan Chang 0.20

0.00
1980 at σ’3 = 138 (CID S1) 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70
and 414kPa (CID S5) log (σ3 / pa )

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 103
Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

E50 = 80/0.08 = 1000kPa (CID S1) m=1.12

Assuming c’ = 0kPa φ’ = 23.6°
0
σ′ 3+ccotφ m
E50 = E50 (
ref ) ∴ E50ref = 701kPa and Rf = 0.65
pref+ccotφ

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 104
Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

E50ref = 701kPa, how does this correspond to typical values?

“Klei-silt” 1 -7MPa? Probably very soft, so in all likelihood about

right?

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 105
Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

140−80
Eoed = = 600kPa
0,1 0
ccosφ+ σ′sinφ m
Eoed = Eoedref ( )
ccosφ+prefsinφ
Eoedref= 600kPa

150−70
Eurrefoed = = 2 666kPa, i.e. 4,4times stiffer on the unload
0,03
reload cycle

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 106
Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

Is the 4,4times stiffer on the unload reload cycle as expected?

Eurref oed= 5 to 8 times Eoedref

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 107
Triaxial State (Bangkok Clay,

HS CM Parameters Empirical Correlations Examples

Surarak, 2012)

Always ask for the raw data to avoid estimating actual values
from graphs….

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 108
 HS CM Parameters Empirical Correlations Examples
What about CU Triaxial Tests
Be careful vu = 0,5 as no dissipation of PWP’s and therefore not
effective

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 109
 HS CM Parameters Empirical Correlations Examples
What about CU Triaxial Tests
Be careful vu = 0,5 as no dissipation of PWP’s and therefore not
effective

Choosing an f-factor is difficult!!

Hardening Soil (HS) Model –

Using SPT-N values and Laboratory Testing 110
Questions?