Code

Boltzmann integration#242

Code

Merged

trevilo merged 80 commits intomainpecos/tps:mainfrom boltzmann-integrationpecos/tps:boltzmann-integrationCopy head branch name to clipboardJun 6, 2024

+2,443-57

Commits

Commits on Oct 20, 2023

• basic code to launch bte in tps code

  milindasf

  committed
  d053c88

Commits on Oct 21, 2023

• initial dev. of spatally adapted v-grids

  milindasf

  committed
  e3eb112

Commits on Oct 22, 2023

• multiple grids batched v-space solver

  milindasf

  committed
  6ed4cc0

• multiple v-space grids added, and boltzmann to tps push code added.

  milindasf

  committed
  7a2d75b

Commits on Oct 25, 2023

• tps bte batched solver integration with Parla.

  milindasf

  committed
  01e51bf

• k-means clustering updated

  milindasf

  committed
  31ec57f

Commits on Oct 27, 2023

• Add spatial coordinate getter in tps2boltzamann

  uvilla

  committed
  00f02dc

• Now the fetch method saves a Paraview file

  uvilla

  committed
  7c7f37b

• Fix spatial coordinates fe order

  uvilla

  committed
  32877b6

Commits on Oct 31, 2023

• vtk output added.

  milindasf

  committed
  0066290

• Te clusters are sorted based on emax

  milindasf

  committed
  a11bfa5

Commits on Nov 1, 2023

• rel error computation updated

  milindasf

  committed
  837566f

• minor fix

  milindasf

  committed
  19e08fc

Commits on Jan 5, 2024

• Add the dof index to the reaction interface/rename the the python bte solver

  uvilla

  committed
  6da87b1

• Add GridFunctionReaction

  uvilla

  committed
  c337e61

• Merge branch 'main' into boltzmann-integration

  uvilla

  authored
  c7f6091

• Preparing the data structures to support the Boltzmann integration

  uvilla

  committed
  3ea95f1

• make enforcestyle

  uvilla

  committed
  53002e6

• make style

  uvilla

  committed
  248aa45

Commits on Jan 6, 2024

• Complete interface

  uvilla

  committed
  e456262

• make enforcestyle

  uvilla

  committed
  baddac5

Commits on Jan 7, 2024

• Small edits

  uvilla

  committed
  804da18

• Let GridFunctionReaction return 0 if the gridfunction is not set.\n Use Arrhenius rates in the Mock Boltzmann solver

  uvilla

  committed
  1bb3216

Commits on Jan 8, 2024

• Few extra tweaks, including computing the number of reaction involving BTE

  uvilla

  committed
  89c3b54

• make enforcestyle

  uvilla

  committed
  062a1ed

• Bugfixes

  uvilla

  committed
  c94674c

• Fix compilation typos

  uvilla

  committed
  e65671a

• make enforcestyle

  uvilla

  committed
  5ac5617

• make style

  uvilla

  committed
  bb2e7c8

• Avoid compilation error on Tioga

  uvilla

  committed
  b154312

• Avoid compilation error on Tioga

  uvilla

  committed
  1771a79

• Restore MFEM_HOST_DEVICE in the constructor/deconstructor of Reaction/Chemestry

  uvilla

  committed
  e7d3755

• Fix incorrect allocation and setting of external rates data on device

  uvilla

  committed
  65e892d

Commits on Jan 9, 2024

• Fix compilation on device. No dynamic_cast

  uvilla

  committed
  332e4e7

• Fix variable name typo

  uvilla

  committed
  88856bc

• Bugfix for gpu-cpu build

  uvilla

  committed
  d5733d5

Commits on Jan 24, 2024

• code clean up for 2w-coupling, grid setup moved to a new function.

  milindasf

  committed
  7d5e170

Commits on Jan 29, 2024

• basic tps+ BTE coupling, main loop moved to Parla + MPI

  milindasf

  committed
  11e0733

• timing results for BTE and TPS steps added.

  milindasf

  committed
  5bdcaf7

• Export timestep and currentTime

  uvilla

  committed
  d7fbc5a

• Make style

  uvilla

  committed
  ce6c857

Commits on Feb 6, 2024

• basic tps + 0d bte 2-way coupling with parla

  milindasf

  committed
  b6f2460

Commits on Feb 14, 2024

• hierarcical clustering added for the bte solve, active grid select based on ionization threshold

  milindasf

  committed
  9761207

Commits on Feb 16, 2024

• now we evolve the tps code until steady state once the BTE rate coefficients computed.

  milindasf

  committed
  7e449e4

Commits on Feb 27, 2024

• 6-species collision model added for the tps batched sover,

  Show description for d81feaa
  milindasf

  committed
  d81feaa

Commits on Mar 1, 2024

• adding e-e collisions for the torch interface with lumped cs data

  milindasf

  committed
  170af58

Commits on Mar 2, 2024

• minor changes

  milindasf

  committed
  d518258

Commits on Mar 4, 2024

• negative rate coefficients are set to zero from BTE side

  milindasf

  committed
  4d6b6ef

Commits on Mar 12, 2024

• io crash from multi gpu case is fixed

  milindasf

  committed
  09b3e2f

Commits on Mar 13, 2024

• additional parameters + profile counters for tps-bte profiling

  milindasf

  committed
  3ef9e46

Commits on Mar 16, 2024

• ibrun -np 2 debuging code, it seems to me the crash happens when launching CPU parla tasks in the async fetch

  milindasf

  committed
  8f012b8

• tps bte minimum parla crash example

  milindasf

  committed
  6d0a2cb

Commits on Mar 18, 2024

• tps-bte with Parla multi-gpu version [working]

  milindasf

  committed
  c7aedb0

• single stream no parla + mpi bte solver added

  milindasf

  committed
  f4624eb

Commits on Mar 19, 2024

• datetime stamp and parameters dump added to the timing output file

  milindasf

  committed
  3e180b8

Commits on Mar 20, 2024

• minor changes

  milindasf

  committed
  acfcab5

Commits on Mar 21, 2024

• nprocs added to the profile file

  milindasf

  committed
  38b69be

• bte fetch when cluster samples < total tps points we need to do sampling with replacement

  milindasf

  committed
  ba72be0

• Add logic to read distance fcn from restart if requested

  trevilo

  committed
  94d3b9a

• nvtx tags added for ncu profiling

  milindasf

  committed
  9611699

Commits on Mar 22, 2024

• Bugfix in test/vpath.sh so that tps-bte_0d3v.py is now also dynamically linked

  uvilla

  committed
  1a3f5b4

Commits on Mar 23, 2024

• Exchange face neighbor data before we compute primitives

  Show description for 239a904
  trevilo

  committed
  239a904

• Bug fix: move mesh->ExchangeFaceNbrNodes() and mesh->ExchangeFaceNbrData() out of conditional

  Show description for f3b5a0e
  trevilo

  committed
  f3b5a0e

• Fix U->HostWrite that should be U->HostRead in M2ulPhyS::updatePrimitives

  trevilo

  committed
  d8df437

• Fix uninitialized species in outlet BC ctor

  Show description for b3e801c
  trevilo

  committed
  b3e801c

Commits on Mar 24, 2024

• thesholding E to avoid E=0 case for the steady-state solver, because when E=0 the steady-state solution would be the delta function

  milindasf

  committed
  d772a02

Commits on Mar 25, 2024

• Restore tps.py and add tps-ntvx-profile.py

  

  Umberto Emanuele Villa

  committed
  85d2f49

• Update gpu_constructor.cpp

  Show description for f7b63ee
  uvilla

  authored
  f7b63ee

• enable mpirun

  

  Umberto Emanuele Villa

  committed
  aafa6da

• Merge branch 'boltzmann-integration' of github.com:pecos/tps into boltzmann-integration

  

  Umberto Emanuele Villa

  committed
  68ce47f

• Add a command line parameter for GPU-aware MPI

  

  Umberto Emanuele Villa

  committed
  e45aaf3

Commits on Mar 29, 2024

• Ensure SpongeZone::sigma is properly initialized

  Show description for c3c0e6c
  trevilo

  committed
  c3c0e6c

Commits on Apr 9, 2024

• Ex, Ey component output for csv files.

  milindasf

  committed
  64e0d1e

• Merge branch 'boltzmann-integration' of github.com:pecos/tps into boltzmann-integration

  milindasf

  committed
  7698e8c

Commits on Apr 12, 2024

• Merge branch 'main' into boltzmann-integration

  uvilla

  authored
  8ab1969

• Merge branch 'main' into boltzmann-integration

  uvilla

  authored
  27323e9

• Update outletBC.cpp

  uvilla

  authored
  55b8bfe

• Update tps.cpp

  Show description for dc6e8dc
  uvilla

  authored
  dc6e8dc

• Update tps.cpp

  uvilla

  authored
  1030494

• Update tps.cpp

  Show description for 7593860
  uvilla

  authored
  7593860