ABSTRACT
This paper describes an approximate method to assess the static and seismic stability of a landfill that is located in an area underlain by cavitose bedrock. The static stability is assessed by plane strain finite element modeling of a section comprising a layer of compacted waste and a layer of competent residual soil overburden underlain by cavitose bedrock. The most critical combination of competent overburden thickness and cavity radius is considered. The nonlinear variable moduli model with Mohr-Coulomb failure criterion is used to represent the behavior of the residual soil. Variable-moduli model parameters are determined based on unconsolidated undrained (UU) triaxial shear tests and are verified by back-prediction of the test results using axisymmetric finite element modeling of triaxial soil samples. In evaluating the static stability, a strength-to-stress ratio is defined as the ratio of the failure shear stress to the calculated shear stress. This ratio is used to locate areas in the model that are close to failure and to assess the overall stability of the soil/cavity system. Seismic stability is assessed by using an approach that combines the results of the static finite element analysis and a seismic stress profile obtained through an equivalent linear seismic response analysis (SHAKE-91). The seismic shear stress profile is selected as that corresponding to the scaled seismic event that results in the most severe shear stress/strain response. Several actual and synthetic seismic events are considered in the selection of the controlling event. In assessing stability, the seismic shear stresses obtained from SHAKE-91 and the static shear stresses obtained from the finite element analysis are superposed and a new stress state is defined. New strength-to-stress ratios are then calculated and the seismic stability of the soil/cavity system is evaluated.
The applied methods provide a rough assessment of the stability of a very complex system without resorting to advanced finite element procedures which require considerable computer resources and without using sophisticated constitutive models which require extensive laboratory testing.
