Inverse analysis of cyclic constitutive models for unsaturated soil under consideration of oscillating functions

¹ Department of Civil Engineering, Bauhaus-Universität Weimar, Germany
² Geomechanics and Geotechnics, Kiel University, Germany
³ Institute of Structural Mechanics–Stochastics and Optimization, Bauhaus-Universität Weimar, Germany

^a e-mail: muyiwa.alalade@uni-weimar.de
^b e-mail: kafle@gpi.uni-kiel.de
^c e-mail: fw@gpi.uni-kiel.de
^d e-mail: tom.lahmer@uni-weimar.de

Abstract

In order to assess the probability of foundation failure resulting from cyclic action on structures and to minimize the prediction error, various existing constitutive models considering cyclic loaded dry soils were extended to unsaturated soil conditions by the authors, thus requiring further calibration during application on existing slightly variable soil condition as well as the soil heterogeneities. The efficiency and effectiveness of these models is majorly influenced by the cyclic constitutive parameters and the soil suction. Little or no details exist in literature about the model based identification and the calibration of the constitutive parameters under cyclic loaded soils. This could be attributed to the difficulties and complexities of the inverse modeling of such complex phenomena. A wide variety of optimization strategies for the solution of the sum of least-squares problems as usually done in the field of model calibration exists, however the inverse analysis of the unsaturated soil response under oscillatory load functions has not been solved up to now. This paper gives insight into the model calibration challenges and also puts forward advanced optimization methods for the inverse modeling of cyclic loaded foundation response on unsaturated soils.