Evaluation of Shear Wave Velocity from Bender Elements Using Cross-correlation

Abstract

Abstract The use of bender elements to determine shear wave velocity, and hence shear modulus, is becoming increasingly popular, with commercial systems now available. The most common method of interpretation is to make use of first arrival; however, automation of this procedure has been impossible for the wide range of velocities, sample sizes, stress levels, etc. encountered in soil and weak rock testing. A method for automating the velocity measurement was proposed by Mohsin and Airey in 2003 and relied on the use of cross-correlation. Since that time, many papers have been published on the method of interpretation, and these have often dismissed cross-correlation out of hand or have suggested that it does not work. These criticisms are based on either the fallacy that the input and output waves must look similar in order for the method to work or the selection of the wrong peak in the cross-correlation. We have been using the technique in the laboratory routinely for several years and have found that it can provide results that are at least as good as those obtained via other methods for soils varying from soft clay to silty gravel and well-cemented sand and at stress levels from 10 kPa to 2000 kPa. The greatest advantage of the technique is that it removes much of the subjectivity from the analysis, and in our opinion it is the only viable currently proposed method that is suitable for automation, and thus for standardization. This paper provides details of the procedure, pulse type and frequency, and calibration and discusses issues that affect the development of a standard. Results for a range of soil types, stress levels, and shear wave velocities are presented to demonstrate the reliability of the technique.