| Funding body | DFG, Project 517723402 |
| Project partner | Institute of Mechanics |
| Project duration | 3 years, start 2024 |
| Project management |
Dr.-Ing. Jan Machaček
Antaeus Bettmann, M.Sc. Prof. Dr.-Ing. habil. Ralf Müller , Institute of Mechanics |
| Contact | Dr.-Ing. Jan Machaček |
The simulation of installation processes, landslides, or ground improvement measures remains a major challenge for current numerical methods. In particular, the complex deformation states that arise in connection with significant topological changes cannot yet be captured satisfactorily.
The “Particle Finite Element Method” (PFEM) enables the modelling of such complex topological changes while also accounting for sophisticated non-linear material models. It combines the established Finite Element Method (FEM) with continuous remeshing, thereby reducing excessive mesh distortion. The general algorithm of a PFEM simulation is illustrated in Figure 1.
In addition to simply remeshing existing particles, particles can be added or removed to improve mesh quality.
Example 1: Cone penetration testing
One of the aims of the project is the realistic simulation of installation and penetration processes. The strong topological changes that occur in the process – for example due to the displacement of the soil as a result of a penetrating CPT probe – can be effectively captured using PFEM through continuous remeshing of the computational domain. Another advantage of PFEM lies in its support for adaptive mesh refinement: during remeshing, mesh resolution can be adjusted locally, enabling an optimal balance between computational accuracy and efficiency. This adaptive remeshing process is demonstrated in the following animation, using the example of a simulated pressure sounding.
Example 2: Wave propagation in a linear elastic rod
Another objective of the project is to extend PFEM by incorporating inertial effects, thereby enabling its application to dynamic problems. A key challenge lies in the transfer of dynamic state variables, such as position, velocity, and acceleration, to newly generated particle configurations.
Figure 2 compares simulation results of compression wave propagation in a linear-elastic rod. The results show that the implemented PFEM algorithm exhibits only very minor deviations from the standard FEM solution.
