Custom Phase Field Actions

To simplify the formation of input files that use standard Allen-Cahn and Cahn-Hilliard equations, custom actions have been created that automatically create conserved and nonconserved phase field field variables and all the corresponding kernels. Additional kernels can still be added using standard MOOSE syntax.

The actions are in the phase field block, under the modules block. Nonconserved variables are created using NonconservedAction. For an example, see

[Modules]
  [./PhaseField]
    [./Nonconserved]
      [./eta]
        free_energy = F
        kappa = 2.0
        mobility = 1.0
        variable_mobility = false
      [../]
    [../]
  [../]
[]

#
# Test the parsed function free enery Allen-Cahn Bulk kernel
#

[Mesh]
  type = GeneratedMesh
  dim = 2
  nx = 40
  ny = 40
  xmax = 40
  ymax = 40
  elem_type = QUAD
[]

[Modules]
  [./PhaseField]
    [./Nonconserved]
      [./eta]
        free_energy = F
        kappa = 2.0
        mobility = 1.0
        variable_mobility = false
      [../]
    [../]
  [../]
[]

[ICs]
  [./InitialCondition]
    type = SmoothCircleIC
    variable = eta
    x1 = 20.0
    y1 = 20.0
    radius = 6.0
    invalue = 0.9
    outvalue = 0.1
    int_width = 3.0
  [../]
[]

[Materials]
  [./free_energy]
    type = DerivativeParsedMaterial
    f_name = F
    args = 'eta'
    function = '2 * eta^2 * (1-eta)^2 - 0.2*eta'
    derivative_order = 2
  [../]
[]

[Executioner]
  type = Transient
  scheme = 'bdf2'

  solve_type = 'NEWTON'

  l_max_its = 15
  l_tol = 1.0e-4

  nl_max_its = 10
  nl_rel_tol = 1.0e-11

  num_steps = 10
  dt = 1.0
[]

[Outputs]
  exodus = true
[]

Conserved variables are created using ConservedAction. For an example, See

[Modules]
  [./PhaseField]
    [./Conserved]
      [./cv]
        solve_type = REVERSE_SPLIT
        free_energy = F
        kappa = 2.0
        mobility = 1.0
      [../]
    [../]
  [../]
[]

#
# Test the conserved action with split solve and 1 variable
#
[Mesh]
  type = GeneratedMesh
  dim = 2
  nx = 50
  ny = 50
  xmax = 50
  ymax = 50
  elem_type = QUAD
[]

[Modules]
  [./PhaseField]
    [./Conserved]
      [./cv]
        solve_type = REVERSE_SPLIT
        free_energy = F
        kappa = 2.0
        mobility = 1.0
      [../]
    [../]
  [../]
[]

[ICs]
  [./InitialCondition]
    type = CrossIC
    x1 = 5.0
    y1 = 5.0
    x2 = 45.0
    y2 = 45.0
    variable = cv
  [../]
[]

[Materials]
  [./free_energy]
    type = DerivativeParsedMaterial
    f_name = F
    args = 'cv'
    function = '(1-cv)^2 * (1+cv)^2'
  [../]
[]

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

[Executioner]
  type = Transient
  scheme = 'bdf2'

  solve_type = 'NEWTON'

  l_max_its = 30
  l_tol = 1.0e-5
  nl_max_its = 10
  nl_rel_tol = 1.0e-12

  start_time = 0.0
  num_steps = 5
  dt = 0.7
[]

[Outputs]
  exodus = true
[]

and

[Modules]
  [./PhaseField]
    [./Conserved]
      [./cv]
        solve_type = direct
        free_energy = F
        kappa = 2.0
        mobility = 1.0
      [../]
    [../]
  [../]
[]

#
# Test consreved action for direct solve
#
[Mesh]
  type = GeneratedMesh
  dim = 2
  nx = 16
  ny = 16
  xmax = 50
  ymax = 50
  elem_type = QUAD
[]

[Modules]
  [./PhaseField]
    [./Conserved]
      [./cv]
        solve_type = direct
        free_energy = F
        kappa = 2.0
        mobility = 1.0
      [../]
    [../]
  [../]
[]

[ICs]
  [./InitialCondition]
    type = CrossIC
    x1 = 5.0
    y1 = 5.0
    x2 = 45.0
    y2 = 45.0
    variable = cv
  [../]
[]

[Materials]
  [./free_energy]
    type = DerivativeParsedMaterial
    f_name = F
    args = 'cv'
    function = '(1-cv)^2 * (1+cv)^2'
  [../]
[]

[Executioner]
  type = Transient
  scheme = 'bdf2'

  solve_type = 'NEWTON'

  l_max_its = 30
  l_tol = 1.0e-4
  nl_max_its = 10
  nl_rel_tol = 1.0e-10

  start_time = 0.0
  num_steps = 5
  dt = 0.7
[]

[Outputs]
  [./out]
    type = Exodus
    refinements = 2
  [../]
[]