Abstract
A simplified numerical procedure to model a deep-penetration process, press-replace method (PRM), was used to simulate a cone penetration test (CPT) in undrained conditions with finite-element software. A detailed description is given on how PRM can be implemented in CPT. The performance of PRM was validated by comparing results obtained with solutions from three other numerical approaches: large-deformation finite element (LDFE), steady-state finite element (SSFE), and material point method (MPM). For a wide range of soil rigidities, initial soil stress anisotropy and roughness conditions, normalized cone factors from PRM using a Tresca model, agreed within 10% of LDFE, and the average value was greater than LDFE by 4%. Total stress, principal stress directions, and magnitudes from PRM were also found to be similar to LDFE results. The PRM approach was then applied to a modified Cam-clay soil model. Excess pore-pressure distributions calculated around the cone agreed well with solutions by SSFE. For a smooth cone with various initial stress states, PRM calculated cone tip resistances that were approximately 7% more than MPM. (C) 2018 American Society of Civil Engineers.