Influences of integral displacement methods on inverse analysis of accelerograph arrays for cyclic shear stress-strain response

The inverse analysis of accelerograph arrays for cyclic shear stress-strain response is widely used in in-situ monitoring and physical model tests, but the influences of the key factors such as integral methods and distribution functions still lack knowledge. Four representative one-dimensional shear beam distribution functions are selected, and a set of dynamic centrifugal model tests are used to clarify the influences of the integral methods and distribution functions on the inverse analysis of shear stresses and shear strains, and the features of the acquired hysteresis loops and modulus damping ratios are further analyzed. The results show: (1) The integral methods exhibit a visible impact on the inversion of the shear stresses and shear strains, and using the acceleration curves processed by the integral methods to obtain the shear stresses is an important condition for effectively ensuring the smoothness and closure of the hysteresis loops. Compared with the ARI method, the USGS method has a non-negligible influence on the phase and amplitude of the original acceleration curves. (2) The shear stresses and shear strains obtained by the three shear beam distribution functions of linear, cubic spline and weighted residuals are very consistent, and the influences of the distribution functions can be ignored. However, the results obtained by the cosine method distribution function are relatively discrete, which is not suitable for selection. (3) The dependence of the shear modulus on the integral methods is slight, but the damping ratios are evidently affected. The development trend of the damping ratios obtained by the ARI method conforms to the general understanding, while the USGS method is contrary to it. The research methods and conclusions may provide important guidance and method support for effectively and reliably obtaining the cyclic shear stress-strain response of in-situ site and geophysical tests and verifying the constitutive relationship models.