Interpretation of CPT Results in Fine-Grained Soils

While the mechanics of the CPT advance in granular soils is generally considered to represent drained penetration, the CPT in fine-grained soils is largely interpreted as undrained penetration. This is especially the case for soft saturated clays which generally occur in conditions of relatively high water table. In stiff, highly overconsolidated clays, such as a surfi- cial clay crust, there may be some questions as to the exact drainage conditions surrounding the cone; however, as will be shown, the results generally can be assessed using similar procedures along with other fine-grained soils.

Undrained Shear Strength

Results of the CPT have been used extensively to provide an estimate of the undrained shear strength of clays using both the tip resistance and sleeve friction. Table 4.6 summarizes different approaches for estimating undrained shear strength in clays.

su from qc

Most current correlations for determining s_u from CPT tip resistance q_c have the form where

o_vo = in situ total vertical stress

N_k = empirical factor

The value of N_k depends on the reference value of s_u and can vary from about 10 to 40 but typically averages about 14 in a large number of clays. In soft clays, the reference has often been the filed vane test, whereas in stiff clays, the reference may be other field tests, e.g., plate load or laboratory tests, such as triaxial compression or unconfined compression. N_k is related to the rigidity index, I_r = G/s_u, of the clay. Baligh (1975) showed theoretically that for a 60° cone, the value of N_k only varied from about 14 to 18 over a wide range of I_r values.

An average value of N_k = 17 was suggested by Lunne et al. (1976) to provide a good comparison with corrected field vane results for a number of marine clays in Scandinavia. Stark & Delashaw (1990) obtained a value of N_k of 12 from unconsolidated-undrained (UU) triaxial tests. Jamiolkowski et al. (1982) summarized a large number of test sites correlating q_c to corrected field vane results and showed that N_k decreased with increasing plasticity with a range from about 5 to 25 for soft to medium clays. If soils with low plasticity (P.I. < 10) are eliminated, the mean value of N_k is approximately 15, which corresponds closely with that presented by Baligh, as previously mentioned.

Table 4.6 Summary of methods for estimating s_u from CPT/CPTU

General correlation	References
	Lunne et al. (1976) Jamiolkowski et al. (1982) Rad & Lunne (1988) Stark & Deleshaw (1990) Anagnostopoulos et al. (2003)
	Jamiolkowski et al. (1988) Larsson & Mulabdic (1991)
	Robertson & Campanella (1983b) Campanella et al. (1985) Larsson & Mulabdic (1991)
	Senneset et al. (1982) Campanella et al. (1982) Mayne& Chen (1993, 1994)
	Robertson (2009)

su from qT

Undrained shear strength may also be estimated from the corrected tip resistance, q_t, from the CPTU, which may be used to provide an estimate of

The value of the cone factor N_KT is a function of the rigidity index of the soil, I_r, (I_r = G/ s_u). For typical values of I_r in clays (i.e., 50-500), theoretical values of N_KT only range from about 9.1 to 12.2. Values of N_KT reported in the literature range from about 7 to 30 with reference values of undrained shear strength coming from a variety of laboratory and in situ tests. The value of N_KT is also somewhat dependent on sensitivity and therefore may need to be reduced to N_KT = 10 (approximately) for sensitive clays. Wroth (1988) showed that for a given soil, N_KT is a constant and generally independent of OCR (with a typical value of about 12 over the range of 1 < OCR < 8).

Results obtained in soft clays indicated that s_u from laboratory direct simple shear tests, the value of N_KT can be roughly related to the liquid limit. For the range of LL between about 30 and 200, their data suggested that

where

LL = liquid limit (in decimal)

While there was considerable scatter, values of N_KT ranged between about 14 and 20 with an average of about 16. Using results from triaxial compression tests, the test data showed an average value of N_KT of about 11 which was also related to the liquid limit as follows:

A typical value of N_KT = 15 is often used for preliminary evaluation of s_u unless local experience or other local correlations are available. In soft clays, a value of N_KT =15 has been shown to provide good agreement with field vane tests (e.g., Cai et al. 2010; Chung et al. 2010).

su from u

Robertson & Campanella (1983b) suggested that based on cavity expansion theory, the undrained shear strength could be estimated from

(spherical cavity expansion with u measured at cone tip; u, position) or

(cylindrical cavity expansion with u measured behind cone tip; u₂ position) where

Ди = excess pore water pressure = u_T - u_G

In general, then, s_u may be estimated from the empirical expression: where

N_u = an empirical factor

Observed values of the factor N_u vary over a much narrower range than either N_K or N_KT and typically range from about 4 to 10 but are a function of the location of the porous element. The value of N_u also depends to some degree on the sensitivity, the pore pressure parameter at failure, A„ and the rigidity index.

In stiff, highly overconsolidated clays, (e.g., OCR > 8) some problems may be encountered using Equation 4.65 if pore water pressures are measured behind the cone tip (Type 2). Because of high pore pressure gradients established between the cone tip and cone base, the pore water pressures measured at the cone base during the cone sounding may be low in highly overconsolidated clays. This is usually indicated by a significant rise in pore water pressure when the sounding is stopped, and a dissipation test is performed. This means that pore water pressures measured during penetration will be low, leading to estimates of s_u that are too low.

su from qT and u

Senneset et al. (1982) and Campanella et al. (1982) suggested that the undrained shear strength could be estimated from the “effective” corrected cone tip resistance, i.e., (q_t - u₂). The general expression for s_u may be given as

Reported values of N_ke typically range from about 3 to 12 depending on the reference value of s_u.

Mayne &c Chen (1993, 1994) presented an effective stress model for estimating s_u from CPTU using this approach. In this model, the value of N_qu is related to the effective stress friction angle, cp', the critical state failure, M, the volumetric strain ratio, X, and a modified Cam- clay equation factor. For most typical clays, the value of N_qu would vary from about 6.1 to 7.1.

su from Q

Robertson (2009) suggested estimating the normalized undrained shear strength from the normalized cone tip resistance assuming a value of N_KX = 14 as

where

su from fs

Estimates of undrained shear strength using CPT friction sleeve measurements have also been made, as first suggested by Begemann (1965); however, it appears that this practice is currently less common than using the tip resistance. A number of reported correlations between f_s and s„ are summarized in Table 4.7. It has been shown that in some cases,

Table 4.7 Some reported correlations between CPT sleeve friction and undrained shear strength

Soil	Correlation	Cone type	Reference strength	References
		M	N/A	Sanglerat (1972)
CL-ML		M	CIUC	Gorman et al. (1973)
Bangkok clay		M	Vane	Brand et al. (1974)
CL-ML		M	CIUC&UU	Gorman et al. (1975)
CH-CL		E	Vane	Marr& Endley (1982)
CH-CL		E	N/A	Tumay et al. (1982)
ML-CL		M	N/A	Cancelli et al. (1982)
Bangkok clay		M(?)	PMT	Bergado & Khaleque (1986)
CL		E (?)	uu	Zervogiannis & Kalteziotis (1988)
CH-CL stiff		E (?)	Remolded UU	Quiros & Young (1988)
ML		E	UU UC	Takesue et al. (1995)
ML-CL stiff		M E	UU	Anagnostopoulos et al. (2003)

the sleeve friction may actually more closely represent remolded strength. For example, Robertson et al. (1986) presented data for clays in the Vancouver area, which generally showed that the sleeve friction values from an electric cone were very close to remolded strength. This difference may be related to sensitivity.

su from Ϭ'p

It is also possible to estimate the undrained shear strength from the pre-consolidation stress, Cp, discussed in Section 4.8.3. For many clays, there is a unique relationship between undrained shear strength and stress history, which may be given approximately as

Therefore, using the CPT/CPTU to first estimate o', one can use Eq. 4.38 to estimate s_u.