When using tri-mesh, have segmentation error

[seongwonHam]

Hello,
I have some trouble and it make me confuse about tri-mesh.

When I can get the result well using rectangular element mesh file, but it didn't work just changing the mesh as triangle element.
The error code like below was appeared. I try some times mesh changing to solve this problem, but it did not work yet.

[Mesh]
  file = tri_1011.e
[]

[GlobalParams]
  displacements = 'disp_x disp_y'
[]

[Variables]
  [./temp]
    initial_condition = 300
  [../]
[]

[AuxVariables]
  [./eigenstrain_yy]
    order = CONSTANT
    family = MONOMIAL
  [../]
  [./eigenstrain_xx]
    order = CONSTANT
    family = MONOMIAL
  [../]
  [./total_strain_yy]
    order = CONSTANT
    family = MONOMIAL
  [../]
  [./total_strain_xx]
    order = CONSTANT
    family = MONOMIAL
  [../]
[]

[Contact]
  [./leftright]
    secondary = 'buffer_outer'
    primary = 'IPyC_inner'
    model = glued
    penalty = 1e9
  [../]
[]

[ThermalContact]
  [./block_contact]
    type = GapHeatTransfer
    primary = 'buffer_outer'
    secondary = 'IPyC_inner'
    variable = temp
    quadrature = true
    gap_conductance = 1e9
    emissivity_primary = 0
    emissivity_secondary = 0
  [../]
[]

[Modules]
  [./TensorMechanics]
    [./Master]
      [./all]
        strain = SMALL
        incremental = true
        add_variables = true
        eigenstrain_names = eigenstrain
        generate_output = 'stress_xx stress_yy'
      [../]
    [../]
  [../]
[]

[Kernels]
  [./heatCond]
    type = HeatConduction
    variable = temp
  [../]
  [./HeatTime]
    type = HeatConductionTimeDerivative
    variable = temp
  [../]
[]

[AuxKernels]
  [./eigenstrain_yy]
    type = RankTwoAux
    rank_two_tensor = eigenstrain
    variable = eigenstrain_yy
    index_i = 1
    index_j = 1
    execute_on = 'initial timestep_end'
  [../]
  [./eigenstrain_xx]
    type = RankTwoAux
    rank_two_tensor = eigenstrain
    variable = eigenstrain_xx
    index_i = 0
    index_j = 0
    execute_on = 'initial timestep_end'
  [../]
  [./total_strain_yy]
    type = RankTwoAux
    rank_two_tensor = total_strain
    variable = total_strain_yy
    index_i = 1
    index_j = 1
    execute_on = 'initial timestep_end'
  [../]
  [./total_strain_xx]
    type = RankTwoAux
    rank_two_tensor = total_strain
    variable = total_strain_xx
    index_i = 0
    index_j = 0
    execute_on = 'initial timestep_end'
  [../]
[]

[BCs]
  [./IPyC_top_disp]
    type = DirichletBC
    variable = disp_y
    boundary = IPyC_outer
    value = 0.0
  [../]
  [./x_min_disp]
    type = DirichletBC
    variable = disp_x
    boundary = x_min
    value = 0.0
  [../]
  [./x_max_disp]
    type = DirichletBC
    variable = disp_x
    boundary = x_max
    value = 0.0
  [../]
  [./x_bot_temp]
    type = DirichletBC
    variable = temp
    boundary = buffer_inner
    value = 3000.0
  [../]
  [./Pressure]
    [./outer]
      boundary = buffer_inner
      function = 1e10*t
    [../]
  [../]
[]

[Materials]
  [./Isotropic_block1]
    type = ComputeIsotropicElasticityTensor
    youngs_modulus = 12e9
    poissons_ratio = 0.33
    block = 1
  [../]
  [./Axisymmetric_block2]
    type = ComputeElasticityTensor
    C_ijkl = '1109e9 139e9 0 39e9 5e9'
    fill_method = axisymmetric_rz
    block = 2
  [../]
  [./small_stress]
    type = ComputeFiniteStrainElasticStress
  [../]
  [./thermal_prop_buffer]
    type = GenericConstantMaterial
    prop_names = 'density thermal_conductivity specific_heat'
    prop_values = '3200 0.5 707'
    block = 1
  [../]
  [./thermal_expansion_strain_buffer]
    type = ComputeThermalExpansionEigenstrain
    stress_free_temperature = 300
    thermal_expansion_coeff = 3.5e-6 #ref Development...
    temperature = temp
    eigenstrain_name = eigenstrain
    block = 1
  [../]
  [./thermal_prop_IPyC]
    type = GenericConstantMaterial
    prop_names = 'density thermal_conductivity specific_heat'
    prop_values = '2200 4.0 710'
    block = 2
  [../]
  [./thermal_expansion_strain_IPyC]
    type = ComputeThermalExpansionEigenstrain
    stress_free_temperature = 300
    thermal_expansion_coeff = 5.5e-6 #ref Development...
    temperature = temp
    eigenstrain_name = eigenstrain
    block = 2
  [../]
[]

[Preconditioning]
  [SMP]
    type = SMP
    full = true
  []
[]

[Executioner]
  type = Transient
  solve_type = 'PJFNK'

  l_max_its = 100
  nl_max_its = 100
  nl_rel_tol = 1e-5
  nl_abs_tol = 1e-5
  l_tol = 1e-4

  start_time = -0.01
  n_startup_steps = 1
  end_time = 1
  dt = 0.01
  dtmin = 1e-5

  automatic_scaling = true
  scaling_group_variables = ' disp_x disp_y; temp'

  petsc_options_iname = '-pc_type -pc_factor_mat_solver_package -ksp_gmres_restart'
  petsc_options_value = 'lu superlu_dist 101'
[]

[Outputs]
  exodus = true
[]

[Postprocessors]
  [./eigenstrain_xx]
    type = ElementAverageValue
    variable = eigenstrain_xx
    execute_on = 'initial timestep_end'
  [../]
  [./eigenstrain_yy]
    type = ElementAverageValue
    variable = eigenstrain_yy
    execute_on = 'initial timestep_end'
  [../]
  [./total_strain_xx]
    type = ElementAverageValue
    variable = total_strain_xx
    execute_on = 'initial timestep_end'
  [../]
  [./total_strain_yy]
    type = ElementAverageValue
    variable = total_strain_yy
    execute_on = 'initial timestep_end'
  [../]
[]

tri_1011.txt
<< this file is exodus file, but I can not upload the .e file so convert as .txt file

it seem like this:

please somebody tell me why signal 11 was occurred only with tri-mesh.

Thanks.

[GiudGiud]

Hello

SIG11 is a seg fault so something is going horribly wrong.
I run in debug mode with your mesh and I hit this assert:

Assertion `side < elem->n_sides()' failed.
side = 4
elem->n_sides() = 3

The mesh cannot be read by libmesh, I think you may have a degenerate QUAD in this tri mesh.
How did you generate this?

Guillaume

[HasanCaprz]

Hello Guillaume,

We created the mesh file by using CUBIT Coreform. And performed Trimesh action while meshing. The thing is that we do not have any problem with rectangular mesh but encounter with this error in trimesh. Actually we even tried trimesh action with a simple geometry and it still resulted with Segmentation Fault error. We are just trying to find out the way of using trimesh in our case.

Hasan

[GiudGiud]

I looked at the mesh more and could not find anything obviously wrong.

Given that this is a very simple mesh, you can generate it directly in moose with our triangulator or with gmsh.
Otherwise you can post on libmesh Discussions or you can contact Cubit support about this.

[HasanCaprz]

Thanks for your suggestions, we will definitely do that.