Reply To: Particle periodicity running with MPI.
OpenLB – Open Source Lattice Boltzmann Code › Forums › on OpenLB › General Topics › Particle periodicity running with MPI. › Reply To: Particle periodicity running with MPI.
Dear Jan,
For your last point of suggestion, since I don’t know how to call the applyPeriodicityToPosition
part of the code and whether to add it directly to the main
function or another location. The new periodicity settings you provided should be applied to the new particle system (SuperParticleSystem3D<T, PARTICLE> spSys(cuboidGeometry, loadBalancer, superGeometry)
). However, the functionality of this part of the code cannot implement the reverse action force of particles on the fluid, so legacy code is used. The particle system I am using is (SuperParticleSystem3D<T, PARTICLE> supParticleSystem(superGeometry)
), and regarding the setting of overlap, it may be that the two lattices are too large? However, this can still run in serial mode.
auto materialperiodicBoundary = std::make_shared
< PeriodicBoundary3D<T, PARTICLE>
> (superGeometry, true, true, false);
supParticleSystem.addBoundary(materialperiodicBoundary);
supParticleSystem.setOverlap(2. * converter.getConversionFactorLength());
auto dragModel = std::make_shared<SchillerNaumannDragModel<T, DESCRIPTOR, PARTICLE>>(converter);
NaiveForwardCouplingModel<T, DESCRIPTOR, PARTICLE> forwardCoupling(converter, sLattice, superGeometry, dragModel);
int overlap = 2;
LocalBackCouplingModel<T, DESCRIPTOR, PARTICLE> backCoupling(converter, sLattice, superGeometry, overlap);
int subSteps = 10;
Best regards,
Rookie