Review of shear-wave velocity prediction equations based on piezocone penetration test data: example from Yangtze River floodplain deposits at Nanjing, Jiangsu Province, China

Abstract

Small-strain parameters of soils play an important role in modern engineering geology. A research database of seismic piezocone test (SCPTu) results has been collected to develop empirical correlations for shear-wave velocity ( V s ) prediction of recent floodplain deposits at Nanjing in the lower reaches of the Yangtze River. First, several existing cone penetration test (CPT)– V s or CPTu– V s correlations for general soils have been investigated. Four indices were used in the assessment, including the ratio of the estimated to measured V s values, the root mean square error, the ranking index and the ranking distance. It is observed that all the selected correlations are biased towards overestimation or underestimation of V s values for such floodplain deposits, demonstrating the need for a site-specific V s prediction model. Then, new correlations for Yangtze floodplain deposits were developed through multiple linear regression, the performance of which has been assessed through comparisons of predicted and observed V s values. The recommended correlation is entirely based directly on CPT data ( q c , f s and z ). The verification and calibration of this newly developed correlation for floodplain deposits constitute a useful contribution to the practice of geologists and engineers not only in the Yangtze delta region, but also in other similar depositional environments worldwide.

Relationship Between Shear Wave Velocity and Piezocone Penetration Tests on the Brazilian Continental Margin

“…In order to obtain a good match between predicted and measured shear wave velocity values, many authors (e.g., Onyejekwe et al, 2016;Zhang & Tong, 2017;Tong et al, 2018) have used a more complete analysis with the evaluation of four indexes: Prediction Bias (K), Ranking Index (RI); Ranking Distance (RD) and Root Mean Square Error (RMSE).…”

“…Another parameter to evaluate the fitness of the measured to the estimated data is the Ranking Distance (RD), originally proposed by Cherubini & Orr (2000). On a plot with mean (m) values on the x-axis and standard deviation (s) on the y-axis, RD represents the distance from the optimum condition point, where m K = 1 and s K = 0, to the point calculated using the correlation (Tong et al, 2018). Equation 11 expresses the Ranking Distance.…”

“…On the one hand, it is necessary to understand the nature of the database that generated soil-type specific equations, before applying them to site-specific CPTu data. On the other hand, generic correlations (Robertson, 2009;McGann et al, 2015;McGann et al, 2018;Tong et al, 2018) can be applicable for all soil types as they introduce parameters that can account for differences in behavior, such as I c and B q (McGann et al, 2015).…”

“…In this study, four relatively recent published correlations are analyzed using the local CPTu data: two generic correlations (Robertson, 2009;Tong et al, 2018), and two for specific soil types (Long & Donohue, 2010;Taboada et al, 2013). This selection was based on a preliminary fit analysis for the generic correlations and for the correlations with similar offshore conditions for the specific soil type.…”

“…A substantial number of studies have been conducted since the early 1980s to develop and evaluate empirical correlations between small-strain elastic properties, especially the shear wave velocity or the maximum shear modulus in soils, and the data from cone or piezocone penetration tests (e.g., Rix & Stokoe II, 1991;Hegazy & Mayne, 1995;Mayne & Rix, 1995;Andrus et al, 2007;Long & Donohue, 2010;McGann et al, 2015;Zou et al, 2017;Tong et al, 2018). These studies have shown that cone resistance, sleeve friction, confining stress, depth, and the behavior and geological age of the soil influence the correlations.…”