The CPT in unsaturated soils Associate Professor Adrian Russell (UNSW) Mr David Reid (Golder Associates) Prof Nasser Khalili (UNSW) Dr Mohammad Pournaghiazar (UNSW) Dr Hongwei Yang (Uni of Hong Kong)
Outline Background and motivation Part 1: Experiments Calibration chamber testing Unsaturated Sydney sand Unsaturated silty sand (decomposed granite) Part 2: Cavity expansion theory Understand drainage conditions Correct chamber to field equivalent values Part 3: Application & examples Part 4: Implications for soil classification
Background What are unsaturated soils? Solid particles with air and water occupying the voids Widely spread and need to be dealt with in many engineering problems Foundations, pavements, slopes, dams, compacted soils Behaviour influenced by suction Strength is increased due to suction (capillary forces). A collapsible structure may exist. s u a u w
Background cont. What does suction do to soil? Suction increases the particle to particle contact forces thus increases the effective stress σ = σ + χs The effective stress may then be used in the usual way for many unsaturated soils. e.g. c and φ are suction independent constants at large strains τ = c + σ tanφ Drying causes suction and strength to increase Wetting causes suction and strength to reduce
Motivation Why study the CPT in unsaturated soils? Measuring suction requires specialised equipment and can be slow Osmotic technique Pressure plate apparatus
Motivation cont. Why study the CPT in unsaturated soils? Tensiometers or vibrating wire piezometers
Motivation cont. Why study the CPT in unsaturated soils? Very little research has been performed on in situ testing of unsaturated soils The CPT may enable in situ determination of suction How are CPT results affected by suction?
The first (published) CPT observations For unsaturated Ottawa sand, when S r is less than 10%, q c can be twice the saturated value 3 qc /(qc )Sr=1.0 srlim /(srlim )Sr =1.0 2.5 2 1.5 Hryciw and Dowding (1987). Geotechnical Testing Journal. 1 0.5 0 0.2 0.4 0.6 0.8 1 Degree of saturation, S r
depth, z (m) More CPT observations Seasonally dependant moisture content in Perth sand caused q c to vary by a factor of 2 0 1 2 3 q c (kpa) 0 5000 10000 15000 20000 solid symbols - end of wet season open symbols - end of dry season Lehane et al (2004). Géotechnique. 4 5 6 7
Suction controlled calibration chamber Prepare uniform specimens with known properties Induce known suction 1 CPT : 1 month Pournaghiazar et al (2011). Canadian Geotechnical Journal.
The assembled chamber
Part 1: Experiments
Percentage Passing Moisture content, w The CPT in unsaturated Sydney sand 100 90 80 70 60 50 40 30 20 10 0 0.01 0.1 1 10 Grain Size (mm) p cs 3 Dr 3.7 ln 0. 9 pa e max = 0.92 e min = 0.60 cs 36 25 s s e 0.45 1 0.55 s s e 0.3 0.25 0.2 0.15 0.1 0.05 0 1 1 10 100 1000 10000 100000 for for for 1 s s e s s e 25 1 25 s s e Suction, s (kpa) Filter paper Pressure plate e 0.68 Russell and Khalili (2006a). International Journal for Numerical and Analytical Methods in Geomechanics.
Specimen formation 1. Dry specimens formed by dry pluviation (sand raining) 2. Density controlled by drop height and flow rate 3. Dry specimens then flooded 4. Flooded specimens then allowed to drain freely 5. Suction then imposed by axis translation
Depth (m) CPT results Repeatability and uniformity 0 Cone resistance, q c (MPa) 0 5 10 15 20 0.1 Pournaghiazar et al (2013). Géotechnique 0.2 0.3 0.4 D r = 0.33 D r = 0.61 0.5 0.6 0.7 Original Repeated
Depth (m) CPT results Saturated versus dry 0 Cone resistance, q c (MPa) 0 5 10 15 20 saturated dry 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Confining stress=50 kpa Confining stress=100 kpa
CPT results Suction influence (loose sand D r = 0.33) s = 25 kpa s = 50 kpa s = 100 kpa
CPT results Suction influence (medium dense sand D r = 0.61) s = 30 kpa s = 50 kpa s = 100 kpa
Percent Passing (%) The CPT in unsaturated Lyell silty sand (decomposed granite) 100 90 80 70 60 50 40 30 20 10 0 0.0001 0.001 0.01 0.1 1 10 Particle Size (mm) p 3 D 4.9 ln 1. 0 cs r pa e max = 0.69 e min = 0.26 cs 36 s s e 1 0.55 for for s s e 1 1 s s e Yang and Russell (2015). International Journal for Numerical and Analytical Methods in Geomechanics.
Specimen formation 1. Moist cured soil was statically compacted in 90 mm thick layers 2. Density was controlled by static pressure 3. Two preparation types: i. Suction increased by axis ii. translation Tested at as-compacted moisture content (piezometers measured suction)
CPT results Suction influence (loose, low confining pressure) Target for s = 100 kpa After t = 790 hours Target for s = 300 kpa After t = 1625 hours s = 60 kpa, e = 0.65 Yang and Russell (2015).
CPT results Suction influence (loose, moderate confining pressure) s = 120 kpa, e = 0.64
CPT results Suction influence (loose, high confining pressure) s = 240 kpa, e = 0.59
CPT results Density influence (loose & med. dense, mod. confining pressure) s = 120 kpa, e = 0.64 s = 120 kpa, e = 0.52
CPT results Horizontal stress influence (loose, moderate confining pressure) s v = s h = 60 kpa, e = 0.65 s v = 60 kpa, s h = 100 kpa, e = 0.62
Part 2: Insights from cavity expansion theory s Russell and Khalili (2006b). Computational Mechanics. c Expanding cavity r R s r Pournaghiazar et al (2012a). International Journal for Numerical and Analytical Methods in Geomechanics Elastic plastic region Elastic region Yang and Russell (2015). International Journal for Numerical and Analytical Methods in Geomechanics
Spherical cavity expansion and the CPT The pressure required to expand a cavity in a soil approaches a limiting value This cavity pressure is related to cone penetration resistance q c We study what effect suction has on cavity pressure to infer what effect suction has on q c
Drainage condition s 0 or ( s) or w 0 0 Easiest to deal with in practice Most realistic as penetration occurs quickly
Cavity expansion in Sydney sand Two major discoveries 1. The pressure required to expand a cavity s L may significantly increase with suction 2. Constant suction and constant moisture content results are indistinguishable srlim /p 0 s c /p 0 120 100 80 60 40 20 0 p 0 = 25kPa p 0 = 100kPa p 0 = 400kPa 0 50 100 150 200 250 s 0 (kpa)
Cavity expansion in Lyell silty sand Three major discoveries Partial saturation permits soil volume change around cavity Constant suction, constant s and constant moisture content results are similar. Constant s is a good approximation Saturated and unsaturated responses are very different due to suction hardening
Initial mean effective stress, p' 0 (kpa) Lyell silty sand - general trends 0 50 100 150 0 1000 2000 3000 4000 5000 Constant χs condition e 0 =0.64 Cavity wall pressure, σ L (kpa) For initial state at MW and initial suction =20 kpa For initial state at MD and initial suction= 20 kpa For initial state at MW and initial suction= 40 kpa For initial state at MD and initial suction= 40 kpa For initial state at MW and initial suction= 80 kpa For initial state at MD and initial suction= 80 kpa 200 250 e 0 =0.51 e 0 =0.38 300
Part 3: Application & examples
Link between q c and s L Bishop et al. (1945) noted a direct proportionality between penetration resistance and the cavity pressure q c,chamber q c,field s L s L, Pournaghiazar et al (2012b). Géotechnique Letters
Convert CPT data to field equivalent values The chamber to cone radius ratio was 29 1.3 1.2 1.1 R/B=0.4 Zero displacement boundary R/B=0.3 σ' c /σ' c, 1 0.9 0.8 0.7 0.6 0.5 R/B=0.6 R/B=0.3 R/B=0.4 R/B=0.5 Constant stress boundary 0 5 10 15 20 25 30 35 B/c p' 0 = 500 kpa p' 0 = 300 kpa p' 0 = 200 kpa p' 0 = 100 kpa p' 0 = 50 kpa
Convert CPT data to field equivalent values An important difference in stress fields around cones must be accounted for σ' va = overburden stress constant Average vertical stress behind the cone : s s va vb q c R 2 D Non uniform stress in front of cone σ' v = σ' vb is constant σ' vb = applied vertical stress at base of specimen (a) Field (b) Chamber
Convert CPT data to field equivalent values It was found that chamber measured q c values were always within 5% of field equivalent values for R D =29 1.2 1 Correlations of Mayne and Kulhawy (1991) Size effect, q c /q c, (σ' c /σ' c, ) 0.8 0.6 0.4 0.2 Correlations of Jamiolkowski et al. (2003) Spherical cavity- uncorrected vertical stress Spherical cavity - corrected vertical stress σ' h /σ' vb = 0.40 Bellotti (1984) Cylindrical cavity 0 0 10 20 30 40 50 60 70 R D
Ratio between q c and s L Sydney sand: q c = 5.7s L Lyell silty sand: q c = 8.3s L
Mean effective stress (kpa) Synthesis of trends Sydney sand Cone resistance, q c (kpa) Saturated and unsaturated results coincide and show power law between q c and p' 0 50 0 10000 20000 30000 40000 Saturated = 0.33 D r Unsaturated = 0.33 D r Unsaturated = 0.61 D r D r Saturated = 0.61 0.85 p exp2. qc 45 78D r 100 Equation applicable for saturated and unsaturated conditions 150 200 250 D r = 0.33 D r = 0.61 Pournaghiazar et al (2013). Géotechnique
Application Sydney sand Two q c values obtained for different p' values but equal D r (although knowing the actual D r is not necessary) are linked according to: q q c2 c1 p 2 p 1 0.85
Application Sydney sand For two tests conducted in unsaturated conditions at the same D r : q q c, s2 c, s1 p p 2 1 For two tests, one in a saturated soil and the other in an unsaturated soil at the same D r : s s 2 1 0.85 qc, s p s qc,0 p uw 0.85
Example 1 Determination of D r and in Sydney sand Suppose q c = 4000 kpa was obtained for a certain depth from a CPT conducted in unsaturated Sydney sand, where p = 25 kpa. It is further supposed that χs = 25 kpa These values correspond to p' = p + χs = 50 kpa D r = 0.42 and ' = 39 are determined However, if suction influences are ignored and p' = p = 25 kpa is incorrectly assumed, values of D r = 0.63 and ' = 43 are determined Failure to account for suction results in overestimation of density and strength
depth, z (m) depth, z (m) Example 2 Determination of s in Perth sand 0 1 q c (kpa) 0 5000 10000 15000 20000 solid symbols - end of wet season open symbols - end of dry season 0 1 s (kpa) 0 20 40 60 80 2 2 3 4 3 4 qc, s p s qc,0 p uw 0.85 5 5 6 6 7 7
depth, z (m) depth, z (m) Mean effective stress (kpa) Example 3 - Siesmic cone in Perth sand 0 1 2 G vh0 (kpa) 0 100000 200000 300000 solid symbols - end of wet season open symbols - end of dry season 0 20 40 60 G vh0 (kpa) 0 100000 200000 300000 solid symbols - end of wet season open symbols - end of dry season G p a = 1400 p + χs p a 0.6 3 4 5 6 80 100 120 140 160 0 1 2 3 4 5 s (kpa) 0 20 40 60 80 7 180 6 7
Synthesis of trends Lyell silty sand More complex because: Presence of suction hardening Presence of hydraulic hysteresis CPT while saturated - undrained (constant volume, u w varies) CPT while unsaturated - drained (changing volume, s constant) These are fundamental differences We can not readily extend what we know for saturated silty sands
Synthesis of trends Lyell silty sand Unsaturated results show power law between q c and p (as also seen for saturated poorly graded sands, i.e. Sydney sand) q 0.65 p exp2. 2 c 163 D Equation not applicable for CPT in saturated soil r Mean effective stress (kpa) 0 50 100 150 200 250 300 350 Cone resistance, q c (kpa) 0 10000 20000 30000 0.07 0.09 0.09 0.12 0.19 0.12 0.140.37 0.42 0.26 0.42 D r =0.25 D r =0.1 D r =0.4 D r =0.6
Application Lyell silty sand Two q c values obtained for different p' values but equal D r (although knowing the actual D r is not necessary) are linked according to: q q c2 c1 p 2 p 1 0.65 c, s 2 c, s1 Only valid when both q c values obtained in an unsaturated soil qc, s q c,0 p p q q s uw 0.65 p p 2 1 s s 2 1 0.65
Q tn Q tn 1000 100 10 6 1 5 4 1 0.1 1 10 0.85 q F r (%) c s v pa fs p Fr (%) a s v qc s v 1. Sensitive, fine grained 2. Organic soils peats 3. Clays clay to silty clay 4. Silt mixtures clayey silt to silty clay 5. Sand mixtures silty sand to sandy silt Soil classification 8 Contractive 3 9 Dilative 2 CPT data for Sydney sand with the effects of suction correctly incorporated in the circular symbols (solid symbols represent saturated or dry tests and hollow symbols represent unsaturated tests). Also shown is data when the effects of suction are ignored in the unsaturated tests using the cross symbols (i.e. where σ' v is taken to be equal to σ v ). 100 6. Sands clean sand to silty sand 7. Gravelly sand to sand 8. Very stiff sand to clayey sand* 9. Very stiff, fine grained* (*) Heavily overconsolidated or cemented
Soil classification cont. Q tn 1000 100 10 7 6 1 5 4 8 9 Dilative Contractive 3 2 CPT data for quartz marine sand with up to 8% fines, with the effects of suction correctly incorporated in the saturated tests below the water table (circular solid symbols) and ignored in the unsaturated tests above the water table (cross symbols, for which σ' v is taken to be equal to σ v ). Q tn 1 0.1 0.80 1 10 qc s v pa F r (%) fs p Fr (%) a s v qc s v 1. Sensitive, fine grained 2. Organic soils peats 3. Clays clay to silty clay 4. Silt mixtures clayey silt to silty clay 5. Sand mixtures silty sand to sandy silt 100 6. Sands clean sand to silty sand 7. Gravelly sand to sand 8. Very stiff sand to clayey sand* 9. Very stiff, fine grained* (*) Heavily overconsolidated or cemented
Q tn Q tn 1000 100 10 7 6 1 5 4 1 0.1 0.60 1 10 qc s v pa F r (%) fs p Fr (%) a s v qc s v 1. Sensitive, fine grained 2. Organic soils peats 3. Clays clay to silty clay 4. Silt mixtures clayey silt to silty clay 5. Sand mixtures silty sand to sandy silt Soil classification cont. 8 Contractive 3 9 Dilative 2 CPT data for gold tailings, at same location in a dam but at different times, with the effects of suction correctly incorporated in the saturated tests when tailings is below the water table (circular solid symbols) and ignored in the unsaturated tests when tailings is above the water table (cross symbols, for which σ' v is taken to be σ v ). 100 6. Sands clean sand to silty sand 7. Gravelly sand to sand 8. Very stiff sand to clayey sand* 9. Very stiff, fine grained* (*) Heavily overconsolidated or cemented
Take home messages Be careful q c,unsat = C q c,sat In unsaturated clean sands C ranges from 1 to 3 in the upper 5m where most likely unsaturated In unsaturated clean sands put s in the effective stress then use established charts (e.g. Pournaghiazar et al. 2013) In unsaturated soils with fines DO NOT use charts developed for saturated soils