MR CFD
Oops! You are not logged in.

For watching this lesson you should sign in first, if you don't have an account, you can create one in seconds.

Toggle Lesson List

Dynamic Mesh: Advanced CFD Training Package — Ep 10

Twin Screw Extruder: DEM

Lesson
10
Run Time
31m 26s
Published
Sep 21, 2026
Course Progress
0%
Mark as Complete
Add to Watchlist
About This Lesson

Twin Screw Extruder CFD Simulation, Using DEM and Dynamic Mesh, by ANSYS Fluent

Description

This project simulates a twin screw extruder using ANSYS Fluent, modeling water flowing from the inlet at a specific flow rate while carrying suspended particles, with the particle flow rate set 9 times higher than that of the continuous phase. The simulation accounts for both the mutual interaction between the continuous and discrete phases, and the interaction between discrete particles themselves — implemented through a 4-way DEM (Discrete Element Method) module. The twin screw rotates at a constant rotational velocity of 300 rpm throughout the simulation.

The geometry was designed in SpaceClaim, then transferred to ANSYS Meshing to generate an unstructured grid. Notably, the element count doesn't remain fixed — it changes continuously with each time step as the mesh adapts to the rotating screw geometry.

Methodology

The dynamic mesh module was activated with smoothing and remeshing sub-models to generate and modify the mesh at each time step as the twin screw rotates at its constant angular velocity. Turbulence was resolved using the standard k-epsilon model. The DPM model represented the discrete phase — CaCO₃ particles — with the 4-way DEM coupling capturing all relevant interactions: continuous-to-discrete, discrete-to-continuous, and particle-to-particle collisions.

Conclusion

The computational cost of this simulation proved extremely high, driven by the combined demands of DEM and dynamic mesh. The initial mesh was built from 2.2 mm tetrahedral cells but changed continuously throughout the calculation as the geometry evolved. Time step size proved critical to convergence — an inappropriately sized time step could drive the solution to diverge due to negative cell volume detection as the mesh deformed.

Following the simulation, particle tracking was extracted across a 60-degree rotation of the twin screw, with particle residence time contours presented from multiple viewing angles — together illustrating how particles move, mix, and reside within the extruder as the screws rotate, offering insight directly relevant to extruder design and process optimization for particle-laden flows.