Complete V5 acceptance milestone
This commit is contained in:
@@ -1,23 +1,54 @@
|
||||
# Dynamic Grain Packing of a Flexible Container Smoke Case
|
||||
# Dynamic Grain Packing of a Flexible Container
|
||||
|
||||
This folder contains the original dynamic grain-packing presentation plus a
|
||||
compact FemLite V4.8 smoke case:
|
||||
This folder contains the original dynamic grain-packing presentation, the V4.8
|
||||
structural smoke case and a V5.8 compact particle-container coupling benchmark.
|
||||
|
||||
Executable inputs:
|
||||
|
||||
- `flexible_container_ground_contact_smoke.inp`
|
||||
- `dynamic_grain_container_benchmark.inp`
|
||||
- `flexible_container_grain_benchmark.inp`
|
||||
- `check_grain_container_results.py`
|
||||
- `run_case.bat`
|
||||
|
||||
The reference model couples a granular-flow domain to a thin flexible container:
|
||||
grains are released from a cylinder, contact forces are accumulated on the
|
||||
container walls, the membrane receives those forces as face loads, and the
|
||||
container can also contact a rigid ground surface. FemLite V4.8 does not include
|
||||
a granular-flow solver or two-way grain/membrane coupling. The included input
|
||||
keeps the structural part executable: an S4 flexible container panel receives a
|
||||
ramped pressure load and starts in contact with an analytical rigid ground.
|
||||
The reference model releases grains into a thin flexible container. The V4.8
|
||||
smoke input keeps the old structural subset: a shell panel receives a ramped
|
||||
pressure load and contacts an analytical ground plane.
|
||||
|
||||
Run:
|
||||
The V5.8 benchmark removes that pressure surrogate. `dynamic_grain_container_benchmark.inp`
|
||||
defines four discrete particles, contact parameters and time controls, while
|
||||
`flexible_container_grain_benchmark.inp` defines only the flexible S4 membrane.
|
||||
The benchmark driver in `tools/benchmarks/dynamic_grain_container_benchmark.f90`
|
||||
uses the production particle-facet coupling path to:
|
||||
|
||||
- compute particle-facet contact forces from real sphere overlap,
|
||||
- map the opposite force to FEM shell nodes through `fsph`,
|
||||
- advance the flexible container with the mapped nodal loads,
|
||||
- preserve the particle reaction force for particle integration.
|
||||
|
||||
Generated benchmark outputs:
|
||||
|
||||
- `grain_container_history.csv`
|
||||
- `grain_container_summary.csv`
|
||||
|
||||
The checker verifies particle count, active grain-container contacts, mapped
|
||||
node force, container displacement, mass conservation, bounded overlap and
|
||||
nonnegative kinetic/dissipated energy. It also rejects a benchmark structural
|
||||
input that still contains `*DLOAD`, so the V5.8 case fails if it falls back to a
|
||||
pressure-only surrogate.
|
||||
|
||||
Run from this directory:
|
||||
|
||||
```bat
|
||||
run_case.bat
|
||||
```
|
||||
|
||||
Results are written to `results\auto_run`.
|
||||
CTest benchmark:
|
||||
|
||||
```bat
|
||||
ctest --test-dir build-win-gcc-16 -C Release -R example_dynamic_grain_container_benchmark --output-on-failure
|
||||
```
|
||||
|
||||
The compact benchmark is a coupling acceptance case, not a full industrial DEM
|
||||
packing simulation. It intentionally stays small enough for default CTest while
|
||||
checking the real particle-to-FEM force path.
|
||||
|
||||
Reference in New Issue
Block a user