luchete80 / WeldForm

Weakly-Compressible Smoothed Particle Hydrodynamics Parallel Solver for Elasto Plastic and thermal coupled Solid Mechanics
GNU General Public License v3.0
9 stars 5 forks source link

Solver is stopping only with "Leaving" msg #110

Open luchete80 opened 3 years ago

luchete80 commented 3 years ago

In elastoplastic Compresion test has been observed following thing

Total CPU time: 1035.54, Neigbour search time: 0.299, Pr Accel Calc time: 27.586Las Acel Calc Time778.778 Max plastic strain: 0.0116599in particle24169 Steps count in this interval: 114Total Step count8982 Total Neighbour search time in this interval: 0Average Neighbour search time in this interval: 0 Output No. 82 at 0.00400035 has been generated Current Time Step = 4.3984e-07 Total CPU time: 1038.99, Neigbour search time: 0.299, Pr Accel Calc time: 27.617Las Acel Calc Time778.903 Max plastic strain: 0.00103344in particle37026 Steps count in this interval: 114Total Step count9096 Total Neighbour search time in this interval: 0Average Neighbour search time in this interval: 0Leaving

luchete80 commented 3 years ago

There has been used standard Verlet integration (has been added recently) with NO XSPH

image

luchete80 commented 3 years ago

image

luchete80 commented 3 years ago

Has been changing

    dom.XSPH    = 0.5; //Very important

but still the same

luchete80 commented 3 years ago

Here is the first neighbour search

Total CPU time: 134.883, Neigbour search time: 0.125, Pr Accel Calc time: 3.444Las Acel Calc Time100.715 Max plastic strain: 0.0263977in particle4603 Steps count in this interval: 18Total Step count1020 Total Neighbour search time in this interval: 0Average Neighbour search time in this interval: 0 Output No. 59 at 0.00285072 has been generated Current Time Step = 2.749e-06 Total CPU time: 135.536, Neigbour search time: 0.125, Pr Accel Calc time: 3.444Las Acel Calc Time100.73 Max plastic strain: 0.0263977in particle4603 Steps count in this interval: 18Total Step count1038 Total Neighbour search time in this interval: 0Average Neighbour search time in this interval: 0Leaving

luchete80 commented 3 years ago

Has ben found same behavior in Cantilever beam at Both Ends WITH Artifficial Viscosity WeldForm\New_Tests\w_ERROR

luchete80 commented 2 years ago

This is due to some problem in list and cell generation in Nb search.

Compression test for 40k part with manual time step: Output No. 2 at 0 has been generated Current Time Step = 2.5e-06 Total CPU time: 0.278528 , Nb: 0.063, StAcc: 0.015, PrAcc: 0Ls Acc: 0.047Contact: 0Msh: 0, BC: 0, mv: 0 Max plastic strain: 0in particle0 Steps count in this interval: 1Total Step count1 Total Nb search time in this interval: 0.063Average Nb search time in this interval: 0.063 Avg Neighbour Count27 id free surf0 0, 0, 0, 0, id free surf0 0, 0, 0, 0, id free surf0 0, 0, 0, 0, Leaving

luchete80 commented 2 years ago
  for (size_t a=0; a<Particles.Size(); a++)
  {
      i= (int) (floor((Particles[a]->x(0) - BLPF(0)) / CellSize(0)));
      j= (int) (floor((Particles[a]->x(1) - BLPF(1)) / CellSize(1)));
      k= (int) (floor((Particles[a]->x(2) - BLPF(2)) / CellSize(2)));

        if (i<0)
        {
                if ((BLPF(0) - Particles[a]->x(0))<=hmax) i=0;
                        else std::cout<<"Leaving"<<std::endl;
        }
        if (j<0)
        {
                if ((BLPF(1) - Particles[a]->x(1))<=hmax) j=0;
                        else std::cout<<"Leaving"<<std::endl;
        }
        if (k<0)
        {
                if ((BLPF(2) - Particles[a]->x(2))<=hmax) k=0;
                        else std::cout<<"Leaving"<<std::endl;
        }