Three-Dimensional Finite Element Simulation of Buried Pipelines Subjected to Reverse Fault Motions

In common practices, the simplified beam-spring model is applied for modeling the pipe behavior against fault displacement. On the other hand, due to the ease of modeling, most simulations have been focused on strike-slip faults and very rare studies have paid attention to the simulation of pipes crossing reverse faults. In the present study, the behavior of the buried pipes subjected to reverse fault motions have been investigated by using three-dimensional continuum finite element modelings. The ABAQUS software has been utilized in the simulations. By this software, the analyses have been performed by using the explicit method. To provide better adaptation between simulation and the behavioral properties of pipe and soil, shell elements and solid elements have been used for the modeling of pipe and soil, respectively. The material non-linearities associated with pipe-material and soil is modeled by considering elasto-plastic behavioral model for soil and pipe. In addition, interface elements have been considered between the soil and the pipe elements. As for the first stage of numerical modeling, the numerical simulation procedure was validated by simulating a large-scale physical model of a pipe crossing a reverse fault. Comparison of the results (in terms of axial compression strains of the pipe) obtained from the simulations with those of the physical model indicates a good match. In the next stage, the behavior of a pipe with a reverse fault motion is investigated from two different approaches. To this aim, the current approach, i.e. three-dimensional continuum modeling was compared with conventional beam-spring model, and the results of the simulations are compared. The results show that the beam-spring model gives logical answers only for small amount of fault displacements while for large fault motions, the model cannot consider correctly the justified behavior of the pipe. The reason is because of the governing local buckling of the pipe at large fault displacement, which cannot be well considered in the beam-spring model. In other words, the beam-spring model can only take the global buckling into consideration; however, this approach is not suitable to study the pipe behavior for large fault displacement, and thus, the problem should be studied by considering the continuum body of the soil as well as the pipe body. In this study, the effect of the diameter to pipe thickness ratio was investigated by using the 3D simulations. The results show that as the diameter to thickness ratio is varied, the failure mechanism of the pipe is changed too. As the diameter/thickness ratio increases, a local buckling is generated at small level of fault displacement, and hence, the resistance of the pipe against the local buckling decreases. In addition, the pipe deformation pattern is different. For the thicker pipe, the pipe deforms in a longer distance around the fault; however, the thinner pipe is crushed at the location of the differential fault displacement. As the other parameter that is effective on the pipe deformation pattern is the soil dilatancy. The numerical modeling indicates that as the soil dilatancy increases, the axial strains of the pipe augments too. The increase in dilatancy from zero to 30 degrees causes a double increase in the pipe strain level. The effect of fault dip angle on the pipe deformation is also investigated numerically. To do this, two faults with different dip angles of 40 and 70 degrees were considered in the modelings. It was found that as the dip angle of the fault is smaller, the level of the axial compression strains increases too. The rate of increase
in the axial strain to the fault displacement is higher too. The deformation pattern of the pipe is investigated, which released that the pipe is much more deformed and damaged for smaller fault dip angle (40 degree). As a conclusion, it can be briefly deduced that: 1) in order to study the deformation of pipe crossing reverse faults, 3D numerical modeling approach are more justified than the simplified beam-spring approach; 2) To reduce the pipe damage, the fill around the pipe should be filled with fines-grained soils, which have low values of dilatancy; 3) As the dip angle of the fault increases, the pipe should be selected as to be thicker in order to prevent local buckling of the pipe.