Estimation of static loading curve from dynamic penetration test through soil-cone interaction model: application in sands

Abstract

Dynamic cone penetration tests provide a unique parameter associated to soil failure for which interpretation remains often empirical. Assessing information on soil deformability from cone penetration tests has been a challenge for engineers. A newer penetration technique, dynamic cone loading test (DCLT), integrates instrumentation and wave analysis enabling to assess soil stress-displacement curve during penetration (DCLT curve). Currently, a simple Smith-based model is applied to interpret DCLT curve which does not capture nonlinear soil response and takes unusual parameters as input. This paper presents an alternative methodology for interpreting DCLT curve, based on a nonlinear soil-cone interaction model. Two approaches (Unloading Point Method and Maximum Stress Point Method) are applied to determine ultimate soil resistance injected in the proposed interaction model. Apart from providing better description of soil reaction, this interpretation methodology of soil penetration allows to estimate static loading curve from DCLT data and to derive a soil modulus. To evaluate results obtained from the methodology, a series of dynamic and static loading tests were performed in samples of Hostun and Fontainebleau sands reconstructed at different densities. Results showed that static curve derived from DCLT were comparable with those obtained experimentally. Maximum point method seems more well-adapted for dense sands whereas unloading point method provides better agreement to static curves for medium dense sands. In addition, soil modulus determined from the interpretation of DCLT is within expected interval of modulus reported for the same sands.