Viscoelastic flows are important for many industrial processes, such as adhesive joining, polymer extrusion and additive manufacturing. Numerical simulations enable virtual evaluation and product realization, which can support the design phase and reduce the amount of costly physical testing. However, such applications are challenging to simulate. Thus, efficient, robust and user-friendly simulation methods are needed.
In this thesis, a Lagrangian–Eulerian simulation framework for viscoelastic flow is presented. The constitutive equation is solved at Lagrangian nodes, convected by the flow, while the momentum and continuity equations are discretized with the finite volume method. The volume of fluid method is used to model free-surface flow, with an injection model for extrusion along arbitrary nozzle paths. The solver combines an automatic and adaptive octree background grid with implicit immersed boundary conditions. In contrast to boundary-conformed mesh techniques, the framework handles arbitrary geometry and moving objects efficiently. Furthermore, novel coupling methods between the Lagrangian and Eulerian solutions as well as unique treatment of the Lagrangian stresses at the fluid-fluid interface are developed. Consequently, the resulting method can simulate the complex flows associated with the intended applications, without the need for advanced stabilization techniques.
The framework is validated for a variety of flows, including relevant benchmarks as well as industrial adhesive joining applications. The latter includes robot-carried adhesive extrusion onto a car fender as well as a hemming application. The results agree with the available experimental data. As such, the research presented in this thesis can contribute to enable virtual process development for joining applications.