INVESTIGATING AMPLITUDE DEPENDENT SOIL PROPERTIES USING A VIBROSEIS TRUCK AND A MICRO-ARRAY OF ACCELEROMETERS

 

LAWRENCE, Z., BODIN, P.,  LANGSTON, C.A., GOMBERG, J., BRACKMAN, T.B., Center for Earthquake Research and Information, University of Memphis, Memphis, TN, 38152; PEARCE, F., and JOHNSON, P.A., Nonlinear Elastic Materials Team, Los Alamos National Laboratory, Los Alamos, NM 87545, zlawrenc@memphis.edu.

 

Amplitude dependent (i.e., nonlinear) properties of soil are an important issue regarding seismic hazard analysis.  I present my ongoing research investigating amplitude dependent soil properties using a large Vibroseis truck with a nearby micro-array of accelerometers in Garner Valley, California and Austin, Texas.  Pearce et al (2004) presented initial results from Garner Valley showing a shift in the resonance of the upper soil layer to lower frequencies, interpreted to represent amplitude dependent modulus reduction within the soil.  The approach used in that study viewed wave propagation as a standing wave (analogous to a resonant column test used in laboratory studies) and used the Vibroseis truck to produce a harmonic source that contained a range of frequencies to ensure the resonant frequency of the near-surface soil layer was excited.  In contrast, my approach focuses on using a harmonic source that is held at a fixed frequency and analyzing basic observables of traveling wave propagation away from the Vibrosies truck, such as phase velocity measurements and time and frequency amplitudes as a function of distance.  Initial results show wave propagation to be quite complex; for instance, the presence of higher-order harmonics, the wavefield exhibiting beating in space, and a high degree of spatial variability of waveforms within a small array area.  Wavefield complexity may be due to complexities in the Vibroseis source radiation, inconsistent instrument response due to coupling issues during large ground motions, and nonlinear wave propagation.  Nonetheless, frequency spectrum, ground motion attenuation, and horizontal phase velocity appear to exhibit some degree of amplitude dependence.