PostprocessorConvergence

This Convergence derives from IterationCountConvergence and compares a Postprocessor value $var element = document.getElementById("moose-equation-d9edad23-0599-4195-9acb-09872b067c9f");katex.render("y", element, {displayMode:false,throwOnError:false,macros:{"\\eqc":"\\,,","\\eqp":"\\,.","\\pd":"\\frac{\\partial #1}{\\partial #2}","\\pr":"\\left(#1\\right)","\\ddt":"\\frac{d #1}{d t}"}});$ to a tolerance $var element = document.getElementById("moose-equation-a51f98a7-dd35-49e8-8f4f-778bafbe061a");katex.render("\\tau", element, {displayMode:false,throwOnError:false,macros:{"\\eqc":"\\,,","\\eqp":"\\,.","\\pd":"\\frac{\\partial #1}{\\partial #2}","\\pr":"\\left(#1\\right)","\\ddt":"\\frac{d #1}{d t}"}});$ :

var element = document.getElementById("moose-equation-eba07ed4-4ba6-4e2e-93fd-a29fcaa10c6b");katex.render("|y| \\leq \\tau \\,.", element, {displayMode:true,throwOnError:false,macros:{"\\eqc":"\\,,","\\eqp":"\\,.","\\pd":"\\frac{\\partial #1}{\\partial #2}","\\pr":"\\left(#1\\right)","\\ddt":"\\frac{d #1}{d t}"}});

For this to work as expected, the execute_on parameter of the post-processor must include values that trigger execution before the desired check. For example, for assessing convergence of the nonlinear solve, the value NONLINEAR_CONVERGENCE should be used. For assessing convergence of a MultiApp fixed point solve, the appropriate execute_on depends on when the MultiApps are executed and on the post-processor type. For example, for Residual, TIMESTEP_BEGIN is appropriate for MultiApps executing on TIMESTEP_BEGIN, and MULTIAPP_FIXED_POINT_CONVERGENCE is appropriate for MultiApps executing on TIMESTEP_END. See SetupInterface for details on different execution points.

Typically the post-processor used should attempt to approximate the error in a system, such as AverageVariableChange.

The parameter "max_diverging_iterations" may be used to specify to diverge after the specified number of consecutive iterations for which the post-processor value is "diverging". By default, "diverging" means the value is getting larger:

var element = document.getElementById("moose-equation-b79250f7-a2c3-45df-8572-b8065c6994f2");katex.render("|y_\\ell| > |y_{\\ell-1}|", element, {displayMode:true,throwOnError:false,macros:{"\\eqc":"\\,,","\\eqp":"\\,.","\\pd":"\\frac{\\partial #1}{\\partial #2}","\\pr":"\\left(#1\\right)","\\ddt":"\\frac{d #1}{d t}"}});

but the parameter "diverging_iteration_rel_reduction" can be used to generalize this condition by specifying some minimum reduction value $var element = document.getElementById("moose-equation-789366a0-5124-49e8-9171-64a5161421c8");katex.render("\\tau_\\text{reduction,min}", element, {displayMode:false,throwOnError:false,macros:{"\\eqc":"\\,,","\\eqp":"\\,.","\\pd":"\\frac{\\partial #1}{\\partial #2}","\\pr":"\\left(#1\\right)","\\ddt":"\\frac{d #1}{d t}"}});$ such that the "diverging" condition is the following:

var element = document.getElementById("moose-equation-4fdb1b8b-93cb-43a1-a563-580cbfd735dd");katex.render("\\frac{|y_{\\ell-1}| - |y_\\ell|}{|y_{\\ell-1}|} < \\tau_\\text{reduction,min}", element, {displayMode:true,throwOnError:false,macros:{"\\eqc":"\\,,","\\eqp":"\\,.","\\pd":"\\frac{\\partial #1}{\\partial #2}","\\pr":"\\left(#1\\right)","\\ddt":"\\frac{d #1}{d t}"}});

This can be used to terminate iteration when convergence is proceeding too slowly. Note that this does not make sense to use in conjunction with a step convergence criteria, which seeks that a quantity, such as some variable, changes very little from iteration to iteration. Instead, this is intended for non-step convergence criteria such as checking whether the current iterate's residual norm is small enough.

Input Parameters

postprocessorPost-processor to use for convergence criteria
C++ Type:PostprocessorName
Unit:(no unit assumed)
Controllable:No
Description:Post-processor to use for convergence criteria
toleranceAbsolute tolerance to use for convergence criteria
C++ Type:double
Unit:(no unit assumed)
Controllable:No
Description:Absolute tolerance to use for convergence criteria

Required Parameters

converge_at_max_iterationsFalseConverge at 'max_iterations' instead of diverging
Default:False
C++ Type:bool
Controllable:No
Description:Converge at 'max_iterations' instead of diverging
diverging_iteration_rel_reduction0The relative reduction with respect to the previous iteration, for which the iteration counts as a diverging iteration with respect to 'max_diverging_iterations': an iteration is diverging if (|pp_old| - |pp_new|)/|pp_old| < rel_reduction. The default value of 0 corresponds to checking that the iteration is actually converging, whereas values < 1 can be used to additionally check that the iteration is converging at an acceptable rate.
Default:0
C++ Type:double
Unit:(no unit assumed)
Controllable:No
Description:The relative reduction with respect to the previous iteration, for which the iteration counts as a diverging iteration with respect to 'max_diverging_iterations': an iteration is diverging if (|pp_old| - |pp_new|)/|pp_old| < rel_reduction. The default value of 0 corresponds to checking that the iteration is actually converging, whereas values < 1 can be used to additionally check that the iteration is converging at an acceptable rate.
max_diverging_iterations4294967295Number of consecutive iterations of the post-processor value either increasing or not reducing fast enough, at which to consider the solve diverged
Default:4294967295
C++ Type:unsigned int
Controllable:No
Description:Number of consecutive iterations of the post-processor value either increasing or not reducing fast enough, at which to consider the solve diverged
max_iterations50Maximum number of iterations
Default:50
C++ Type:unsigned int
Controllable:No
Description:Maximum number of iterations
min_iterations0Minimum number of iterations
Default:0
C++ Type:unsigned int
Controllable:No
Description:Minimum number of iterations
verboseFalseEnable printing of additional information, including convergence and divergence reasons.
Default:False
C++ Type:bool
Controllable:No
Description:Enable printing of additional information, including convergence and divergence reasons.

Optional Parameters

control_tagsAdds user-defined labels for accessing object parameters via control logic.
C++ Type:std::vector<std::string>
Controllable:No
Description:Adds user-defined labels for accessing object parameters via control logic.
enableTrueSet the enabled status of the MooseObject.
Default:True
C++ Type:bool
Controllable:No
Description:Set the enabled status of the MooseObject.
implicitTrueDetermines whether this object is calculated using an implicit or explicit form
Default:True
C++ Type:bool
Controllable:No
Description:Determines whether this object is calculated using an implicit or explicit form

Advanced Parameters

Input Files

(test/tests/convergence/postprocessor_convergence/diverging_iterations.i)
(test/tests/convergence/default_steady_state_convergence/pp.i)
(test/tests/convergence/postprocessor_convergence/multiapp_fp.i)
(test/tests/convergence/postprocessor_convergence/nonlinear.i)

(test/tests/convergence/postprocessor_convergence/diverging_iterations.i)

[Mesh]
  type = GeneratedMesh
  dim = 1
  nx = 1
[]

[Functions]
  [test_fn]
    type = PiecewiseLinear
    x = '0    1   2  3  4     5'
    y = '100 10 9.5 11 10 1e-10'
  []
[]

[Postprocessors]
  [test_pp]
    type = FunctionValuePostprocessor
    function = test_fn
  []
[]

[Convergence]
  [test_conv]
    type = PostprocessorConvergence
    postprocessor = test_pp
    tolerance = 1e-7
    max_diverging_iterations = 2
    diverging_iteration_rel_reduction = 0.1
    verbose = true
  []
[]

[Problem]
  solve = false
[]

[Executioner]
  type = Transient
  num_steps = 10
  dt = 1
  steady_state_detection = true
  steady_state_convergence = test_conv
[]

(test/tests/convergence/default_steady_state_convergence/pp.i)

!include base.i

[Convergence]
  [steady_conv]
    type = PostprocessorConvergence
    postprocessor = var_diff_norm
    tolerance = ${ss_tol}
  []
[]

[Postprocessors]
  [var_diff_norm]
    type = RelativeSolutionDifferenceNorm
    execute_on = 'INITIAL TIMESTEP_END'
  []
[]

(test/tests/convergence/postprocessor_convergence/multiapp_fp.i)

!include ../parent.i

[Convergence]
  [fp_conv]
    type = PostprocessorConvergence
    postprocessor = nl_res
    tolerance = 1e-2
    max_iterations = 15
  []
[]

[Postprocessors]
  [nl_res]
    type = Residual
    residual_type = COMPUTE
    execute_on = 'MULTIAPP_FIXED_POINT_CONVERGENCE'
  []
[]

(test/tests/convergence/postprocessor_convergence/nonlinear.i)

[Mesh]
  type = GeneratedMesh
  dim = 2
  nx = 10
  ny = 10
[]

[Variables]
  [u]
  []
[]

[AuxVariables]
  [w]
  []
[]

[AuxKernels]
  [w_kernel]
    type = ParsedAux
    variable = w
    expression = 'sol'
    functor_names = 'u'
    functor_symbols = 'sol'
    execute_on = 'NONLINEAR_CONVERGENCE'
  []
[]

[Kernels]
  [diff]
    type = Diffusion
    variable = u
  []
[]

[BCs]
  [left]
    type = DirichletBC
    variable = u
    boundary = left
    value = 0
  []
  [right]
    type = DirichletBC
    variable = u
    boundary = right
    value = 1
  []
[]

[Postprocessors]
  [w_change]
    type = AverageVariableChange
    variable = w
    change_over = nonlinear_iteration
    norm = L1
    execute_on = 'NONLINEAR_CONVERGENCE'
  []
  [num_nl_iterations]
    type = NumNonlinearIterations
    execute_on = 'TIMESTEP_END'
  []
[]

[Convergence]
  [test_conv]
    type = PostprocessorConvergence
    postprocessor = w_change
    tolerance = 1e-7
  []
[]

[Problem]
  previous_nl_solution_required = true
[]

[Executioner]
  type = Steady
  solve_type = 'PJFNK'
  petsc_options_iname = '-pc_type'
  petsc_options_value = 'hypre'
  nonlinear_convergence = test_conv
[]

[Outputs]
  csv = true
  show = 'num_nl_iterations'
  execute_on = 'FINAL'
[]

Input Parameters
Input Files