Investigation and interpretation of cone penetration rate effects

The Cone Penetration Test (CPT) is the most popular in-situ test for onshore and offshore site investigation. The standard penetration rate employed of 20±5 mm/s is well embedded in practice since cone penetration is expected to be fully drained in sand and fully undrained in clay at this velocity. Interpretation of geotechnical parameters from the CPT methods for sand and clay has a relatively high reliability. However this reliability is poor in intermediate soils, such as silts, clayey and sandy silts and residual soils, as penetration takes place under partially drained conditions in these materials. Very little guidance is available to practitioners for derivation of geotechnical parameters during partially drained penetration. This thesis addresses the need for such guidance by examining the rate dependency of penetrometer resistance and consequently also provides information to assist estimation of the penetration resistance and rate dependency of penetrometers with different diameters, such as piles.
The Thesis examines the rate dependency of penetrometer resistance using (i) a comprehensive series of penetrometer testing in the field and laboratory and (ii) Finite Element (FE) analyses that simulate penetrometer penetration at various velocities in a range of soils modelled using a non-linear elasto-plastic constitutive model. Interpretation of the experimental penetrometer tests is supported by a series of laboratory element tests on samples of the in-situ soils. The observations made from the experimental and numerical components of this research are combined with the existing database of measurements obtained in previous studies to formulate conclusions.
An extensive series of piezocone penetration tests was performed at three sites in Western Australia, to extend the current sparse database of variable rate piezocone tests in the field. Undrained, partially drained and fully drained conditions were observed by performing tests at rates down to 100,000 times slower than the standard rate; variable rate T-bar and ball penetration tests were conducted for comparative purposes. The effect of fines content on CPT end resistance was also examined in a series of variable rate penetration tests on kaolin-sand mixture performed in both the centrifuge and in pressurised chambers at 1g. Laboratory tests on intact and reconstituted samples assisted interpretation of the field and laboratory scale penetrometer tests.
A series of parametric numerical analyses was carried out employing the spherical cavity expansion method using the FE code PLAXIS. A coupled consolidation calculation with a non-linear elasto-plastic soil model was used to simulate variable rate cone penetration. These analyses allowed evaluation of the specific effects on penetration resistance of soil stiffness, friction angle, in-situ lateral effective stress, permeability, cone diameter and stress level. Use of a normalisation velocity term incorporating drainage length, horizontal permeability and stiffness are found to be effective to define drainage transitions.
The numerical and experimental studies show the velocity range over which conditions can be expected to be partially drained for a standard CPT; this range is defined in terms of a normalised velocity. These studies also show that ratio between drained and undrained penetration resistance is strongly dependent on the soil stiffness and friction angle and hence varies widely with soil type; typical responses for a range of common soil types are identified. Practical recommendations when conducting CPTs in intermediate soils are provided to allow improved derivation of geotechnical parameters.