recuero
commented
3 years ago For the model reported with second-order tet meshes, when the bodies come into contact, there is a failing libmesh check. This makes simulation stop also in opt mode. Backtrace below.
frame #1: 0x00000001088e4dc6 libmesh_dbg.0.dylib`libMesh::libmesh_terminate_handler() at libmesh.C:321:5
frame #2: 0x00007fff6bea3887 libc++abi.dylib`std::__terminate(void (*)()) + 8
frame #3: 0x00007fff6bea61a2 libc++abi.dylib`__cxxabiv1::failed_throw(__cxxabiv1::__cxa_exception*) + 27
frame #4: 0x00007fff6bea6169 libc++abi.dylib`__cxa_throw + 113
frame #5: 0x00000001002441a0 libcontact-dbg.0.dylib`libMesh::DofObject::dof_number(this=0x0000000156daf358, s=1, var=11, comp=0) const at dof_object.h:993:3
frame #6: 0x000000010027c2d8 libcontact-dbg.0.dylib`MechanicalContactConstraint::nodalArea(this=0x000000018931e618, pinfo=0x00000001a81e7150) at MechanicalContactConstraint.C:1724:27
frame #7: 0x000000010027825f libcontact-dbg.0.dylib`MechanicalContactConstraint::getPenalty(this=0x000000018931e618, pinfo=0x00000001a81e7150) at MechanicalContactConstraint.C:1743:16
frame #8: 0x0000000100279174 libcontact-dbg.0.dylib`MechanicalContactConstraint::computeContactForce(this=0x000000018931e618, pinfo=0x00000001a81e7150, update_contact_set=true) at MechanicalContactConstraint.C:556:24
frame #9: 0x0000000100282c14 libcontact-dbg.0.dylib`MechanicalContactConstraint::shouldApply(this=0x000000018931e618) at MechanicalContactConstraint.C:499:9
frame #10: 0x0000000102f8cee7 libmoose-dbg.0.dylib`NonlinearSystemBase::constraintResiduals(this=0x0000000156016e18, residual=0x0000000167c4d1e0, displaced=true) at NonlinearSystemBase.C:1179:24
frame #11: 0x0000000102f86844 libmoose-dbg.0.dylib`NonlinearSystemBase::computeResidualInternal(this=0x0000000156016e18, tags=size=3) at NonlinearSystemBase.C:1586:9
frame #12: 0x0000000102f85022 libmoose-dbg.0.dylib`NonlinearSystemBase::computeResidualTags(this=0x0000000156016e18, tags=size=3) at NonlinearSystemBase.C:732:5
frame #13: 0x00000001027d01ca libmoose-dbg.0.dylib`FEProblemBase::computeResidualTags(this=0x0000000156014818, tags=size=3) at FEProblemBase.C:5524:8
frame #14: 0x00000001027cf776 libmoose-dbg.0.dylib`FEProblemBase::computeResidualInternal(this=0x0000000156014818, soln=0x0000000114c5b360, residual=0x00007ffeefbfe3d0, tags=size=3) at FEProblemBase.C:5362:5
frame #15: 0x00000001027cf59c libmoose-dbg.0.dylib`FEProblemBase::computeResidual(this=0x0000000156014818, soln=0x0000000114c5b360, residual=0x00007ffeefbfe3d0) at FEProblemBase.C:5317:3
frame #16: 0x00000001027f24d6 libmoose-dbg.0.dylib`FEProblemBase::computeResidualSys(this=0x0000000156014818, (null)=0x0000000114c5ad40, soln=0x0000000114c5b360, residual=0x00007ffeefbfe3d0) at FEProblemBase.C:5292:3
frame #17: 0x0000000102fa1807 libmoose-dbg.0.dylib`ComputeResidualFunctor::residual(this=0x0000000156018310, soln=0x0000000114c5b360, residual=0x00007ffeefbfe3d0, sys=0x0000000114c5ad40) at ComputeResidualFunctor.C:23:15
frame #18: 0x00000001096591ce libmesh_dbg.0.dylib`::libmesh_petsc_snes_residual(snes=0x0000000218813620, x=0x0000000218809020, r=0x00000001b000be20, ctx=0x0000000114c5b9f0) at petsc_nonlinear_solver.C:163:35
frame #19: 0x000000010e6fcfae libpetsc.3.14.dylib`SNESComputeFunction + 334
frame #20: 0x000000010e711bd4 libpetsc.3.14.dylib`SNESSolve_NEWTONLS + 404
frame #21: 0x000000010e70253d libpetsc.3.14.dylib`SNESSolve + 1837
frame #22: 0x000000010965e051 libmesh_dbg.0.dylib`libMesh::PetscNonlinearSolver<double>::solve(this=0x0000000114c5b9f0, pre_in=0x0000000156ba1380, x_in=0x0000000114c5b240, r_in=0x0000000114c5b810, (null)=0.001, (null)=25) at petsc_nonlinear_solver.C:951:10
frame #23: 0x00000001097034c8 libmesh_dbg.0.dylib`libMesh::NonlinearImplicitSystem::solve(this=0x0000000114c5ad40) at nonlinear_implicit_system.C:178:27
frame #24: 0x0000000103291d6d libmoose-dbg.0.dylib`TimeIntegrator::solve(this=0x0000000167ce6398) at TimeIntegrator.C:66:16
frame #25: 0x0000000102fa28f8 libmoose-dbg.0.dylib`NonlinearSystem::solve(this=0x0000000156016e18) at NonlinearSystem.C:200:23
frame #26: 0x00000001027fa7c0 libmoose-dbg.0.dylib`FEProblemBase::solve(this=0x0000000156014818) at FEProblemBase.C:5042:10
frame #27: 0x00000001022da134 libmoose-dbg.0.dylib`FEProblemSolve::solve(this=0x0000000188010980) at FEProblemSolve.C:197:12
frame #28: 0x00000001022d5496 libmoose-dbg.0.dylib`PicardSolve::solveStep(this=0x0000000188010a68, begin_norm_old=1.7976931348623157E+308, begin_norm=0x0000000167c60840, end_norm_old=1.7976931348623157E+308, end_norm=0x0000000167c9c160, relax=false, relaxed_dofs=size=0) at PicardSolve.C:445:22
frame #29: 0x00000001022d38c7 libmoose-dbg.0.dylib`PicardSolve::solve(this=0x0000000188010a68) at PicardSolve.C:236:28
frame #30: 0x0000000102d89b9c libmoose-dbg.0.dylib`TimeStepper::step(this=0x0000000167d3c8a8) at TimeStepper.C:163:43
frame #31: 0x00000001022d7ea0 libmoose-dbg.0.dylib`Transient::takeStep(this=0x0000000188010818, input_dt=-1) at Transient.C:431:20
frame #32: 0x00000001022e08fb libmoose-dbg.0.dylib`Transient::execute(this=0x0000000188010818) at Transient.C:314:5
frame #33: 0x00000001035d8cd8 libmoose-dbg.0.dylib`MooseApp::executeExecutioner(this=0x0000000157008e18) at MooseApp.C:1023:19
frame #34: 0x00000001035d40bf libmoose-dbg.0.dylib`MooseApp::run(this=0x0000000157008e18) at MooseApp.C:1150:5
frame #35: 0x0000000100002f96 contact-dbg`main(argc=3, argv=0x00007ffeefbffa18) at main.C:36:8
Retrieving the dof number appears to be ultimately causing the issue:
dof_id_type dof = node->dof_number(_aux_system.number(), _nodal_area_var->number(), 0);
sapitts
Bug Description
A three-dimensional problem with second-order tets results in segmentation fault when the bodies come into contact, as reported in #16773.
Steps to Reproduce
The problem built in #16773 reproduces this problem.
Impact
Reduced robustness of three-dimensional
Tet10(possiblyTet4) mechanical contact.