Centrifuge and Numerical Modelling of Pile and Penetrometer in Sand

Meen-Wah Gui, Cambridge University, Geotechnical Engineering Group


The use of miniature CPT probes in centrifuge models is reviewed. Methods of normalisation are suggested for the cone data. Dimensional analysis and modelling of models is used to demonstrate the importance of distinguishing shallow and deep penetration mechanisms for cones which are large relative to stratum depth. Soil particle size is shown to be insignificant if less than one twentieth the cone diameter. The effects of stress level, particle crushing, OCR are also mentioned. The phenomenon of the development of the stress history at the shallower depth and the concept of effective diameter of cone have also been introduced.

26 centrifuge tests, 12 triaxial compression tests, and 12 shear box tests were carried out. Centrifuge CPTs were performed using a 10 mm diameter cone penetrometer. To study the effects of diameter ratio, three different sizes of container (210, 420 and 850 mm diameter) were used. Side boundary effect is found to be significant if container/cone diameter ratio is less than 40. The possibility of simulating an infinite body of sand via a membrane is also studied. Tip resistance beneath a slope has been shown to decrease with the increase of gradient. Ambient stress is found to be significant, as expected. The development of stress history was effectively completed after the first 5 cone diameter of pull-out.

A method has been derived for treating the triaxial data for sand which is useful for analysing piles and penetrometers or other situations in which significant stress levels vary from very high to very low. The data is divided into the high, medium and low stress zones (H-M-L). These are separately modelled as zones of plastic hardening, in this case using the FLAC program. The consistency of the modelling is demonstrated by simulating triaxial tests. Application of this H-M-L model in the analysis of a pile or a cone penetration problem has also been demonstrated. Sensitivity analyses were performed by varying the rate of the post-peak softening of the material in the low stress zone. Anisotropy and the localisation of shearing in the rupture zone have been observed. Comparison with centrifuge results has been found to be encouraging.

Finally, the geometry effects in applying CPT data and the reduction of soil rigidity in pile design have also been reviewed. The tip resistance of pile can be best presented if normalised against v(s’v).

Keywords: CPT, pile, centrifuge test, sand, slope test, pull-out test, triaxial test, total stress cell, dimensional analysis, numerical analysis, rupture, development, rigidity.