TY - JOUR
T1 - Stress wave velocity in soils
T2 - apparent grain-size effect and optimum input frequencies
AU - Dutta, T. T.
AU - Otsubo, M.
AU - Kuwano, R.
AU - O'Sullivan, C.
PY - 2019/12
Y1 - 2019/12
N2 - There is a lack of consensus in the literature on the influence of the median particle size on stress wave velocity in cohesionless soils. For assemblies of spherical particles with Hertzian contacts, the stress wave velocities should not depend on particle size. However, a link between particle size and stress wave velocity has been reported in laboratory experiments. In this study, to identify the reasons for the discrepancies, wave velocity measurements were performed using planar piezoelectric transducers on four different sizes of alkaline glass beads and natural silica sands. The experimental results indicate that shear and compression wave velocities are independent of the median particle size. In accordance with the dispersion theory, both the experiments and discrete-element simulations demonstrate that the maximum frequency that can propagate through a granular assembly (i.e. the low-pass frequency) reduces with increasing median particle size. The relationship between the low-pass frequency and the input signal frequency determines the quality of the received signal and hence the accuracy of the interpreted stress wave velocity data. To accurately estimate shear wave velocities, the selected input frequencies should match those frequencies which exhibit the largest gain factors and the input frequencies should not exceed more than half of the low-pass frequency. To determine the compression wave velocity, it is suggested that the start-to-start method be adopted and an input frequency which is slightly lower than the low-pass frequency be chosen.
AB - There is a lack of consensus in the literature on the influence of the median particle size on stress wave velocity in cohesionless soils. For assemblies of spherical particles with Hertzian contacts, the stress wave velocities should not depend on particle size. However, a link between particle size and stress wave velocity has been reported in laboratory experiments. In this study, to identify the reasons for the discrepancies, wave velocity measurements were performed using planar piezoelectric transducers on four different sizes of alkaline glass beads and natural silica sands. The experimental results indicate that shear and compression wave velocities are independent of the median particle size. In accordance with the dispersion theory, both the experiments and discrete-element simulations demonstrate that the maximum frequency that can propagate through a granular assembly (i.e. the low-pass frequency) reduces with increasing median particle size. The relationship between the low-pass frequency and the input signal frequency determines the quality of the received signal and hence the accuracy of the interpreted stress wave velocity data. To accurately estimate shear wave velocities, the selected input frequencies should match those frequencies which exhibit the largest gain factors and the input frequencies should not exceed more than half of the low-pass frequency. To determine the compression wave velocity, it is suggested that the start-to-start method be adopted and an input frequency which is slightly lower than the low-pass frequency be chosen.
KW - Discrete-element modelling
KW - Geophysics
KW - Stiffness
UR - http://www.scopus.com/inward/record.url?scp=85083169806&partnerID=8YFLogxK
U2 - 10.1680/jgele.18.00219
DO - 10.1680/jgele.18.00219
M3 - Article
AN - SCOPUS:85083169806
SN - 2045-2543
VL - 9
SP - 340
EP - 347
JO - Geotechnique Letters
JF - Geotechnique Letters
IS - 4
ER -