This paper presented research study findings in which the applications of bearing capacity and cavity expansion models in interpreting the cone penetration test results of dilatant cemented sand are addressed. Two bearing capacity and two cavity expansion theories are used to predict the tip resistances of dilatant sandy specimens tested in calibrated conditions. Tip resistance interpretations are compared with experimental results in order to address the prediction capabilities of the models. Tests conducted on cemented and uncemented Monterey No. 0/30 sand in a calibration chamber and on a field cemented deposit have provided the experimental results for this comparison.

Cone penetration in very weakly cemented sands was modeled in a calibration chamber study [1]. Monterey no. 0/30 sand and ordinary Portland cement were used as the main constituents in the specimen preparation. Cone penetration tests were conducted on these specimens using a miniature cone penetrometer under zero lateral strain condition traditionally known as Boundary Condition 3 or BC3. All sandy specimens were tested under normally consolidated conditions. Three types of cement content (0, 1, and 2% by dry weight), three ranges of relative density (45 - 55%, 65 - 75%, 85 - 90%), and three levels of confining pressure (100, 200, and 300 kPa) were investigated. A total of thirty seven calibration chamber tests were conducted. Drained triaxial tests were also conducted on identical cemented specimens. These tests provided shear strength parameters, cohesion intercept and friction angle properties. The tip resistance, sleeve friction and friction ratio values from chamber tests were correlated with stresses, attraction intercept and relative state parameter or relative density. The attraction intercept, which provides an estimate of the tensile strength of the soil specimen, is taken as the ratio of the cohesion intercept and tangent of friction angle. The relative state parameter is defined as the ratio of the state parameter to the difference in maximum and minimum void ratios. Based on chamber and triaxial tests empirical correlations were developed.

Two frequently used bearing capacity theories include Durgunoglu and Mitchell [2] (D & M) theory and Janbu and Senneset [3] (J&S) theory; and cavity expansion models include Carter et al. [4] solution, and Yu and Houlsby [5] solution are considered in this evaluation. These models were developed based on distinct types of soil behaviour such as rigid-plastic, elastic-perfectly plastic, and elasto-plastic models.

Four cone penetration tests were conducted using a research testing vehicle on a bluff of loose to medium dense loess deposits located near the US Army Corps of Engineers Waterways Experiment Station in Vicksburg, Mississippi, USA. Undisturbed block specimens of 0.03 to 0.04 m³ in volume were obtained from two elevations. Soil specimens were subjected to both consolidated isotropic drained (CID) triaxial tests and unconfined compression strength tests. Cone penetration tests conducted at the site with three different penetration rates provided similar tip resistances. This confirmed that the rate of penetration has no effect on the cone tip resistance, which can be regarded as that the drained conditions prevailed during cone penetration testing. Since soil specimens showed drained behaviour during cone penetration, it is assumed that both test results can be interpreted with similar models and correlations that use drained effective stress characteristics.

Comparisons between theoretical and experimental results showed that D & M bearing capacity theory predictions are close to measured test results only at low confining pressures and at low relative densities. However, the predictions are different from measured results at higher confining pressures and relative densities. The rigid plastic assumption in this model is not a realistic case for representing the elasto-plastic soil behaviour. The J & S model for tip resistance calculation requires strength properties and a plastification angle parameter, which represents the volume change or compressibility behaviour in soils. The J & S model does not specify any direct or indirect procedure for determining the plastification angle. The authors provided a formalism in which the plastification angle is correlated either with dilation angle or with relative state parameter. Results obtained by using the plastification and dilation angle formalism provided predictions, which are quite close to the measured tip resistances.

The two cavity expansion models proposed by Carter et al. [4] and Yu and Houlsby [5] incorporated elastic-perfectly plastic soil models in their theories. Both models differ with one another with respect to the type of strain analysis used in the solution. Yu and Houlsby adopted a large strain analysis where as Carter et al. used a small strain analysis in the zone of elastic behaviour. However, both solutions provided similar limiting pressure values, which are quite small when compared to measured tip resistance. The ratios vary from three to six with lower values being obtained at higher relative densities.

A field study conducted on a partially saturated loess bluff at Vicksburg, Mississippi, also reconfirmed the above findings on the bearing capacity and the cavity expansion theories. Among the bearing capacity theories, the J & S theory has provided the best predictions at higher confining pressures. The cavity expansion solutions, on the other hand, have predicted limiting pressures, which are 2 to 20 times lower than the measured tip resistance magnitudes. This means that the cavity expansion theories cannot be used to directly interpret the tip resistances of cone penetration testing without applying correction factors.

The major limitation of the cavity expansion models lies in the representation of the dilational behaviour during shearing. Strain softening behaviour is not considered in the original model formulations, which result in lower predictions. Thus, modified or new cavity expansion theories by considering realistic soil behaviour with strain softening are needed for better interpretations of cone penetration test results.

[1]

A.J. Puppala, Y.B. Acar and M.T. Tumay, "Cone Penetration in Very Weakly Cemented Sand", Journal of Geotechnical Engineering, ASCE, 121(8), 1995.

[2]

H.T. Durgunoglu and J.K. Mitchell, "Static Penetration Resistance of Soils", Space Sciences Laboratory, University of California, Berkeley, April, 1973.

[3]

N. Janbu and K. Senneset, "Effective Stress Interpretation of In-situ Static Penetration Tests", Proceedings of the European Symposium on Penetration Testing, ESOPT I, Stockholm, Sweden, 2.2, 181-193, 1974.

[4]

J. P. Carter, J.R. Booker, A. Crooks, L. Rothenburg, "Cavity Expansion in Cohesive Frictional Soil", Geotechnique, London, England, 36, 349-358, 1986.

[5]

H.S. Yu and G.T. Houlsby, "Finite Cavity Expansion in Dilatant Soils: Loading Analysis", Geotechnique 41(2), 173-183, 1991.

Return to the contents page
Return to the book description