Initializing the _hist material property for the phase field fracture model from the input file #16991
-
|
Hello all, I have a question regarding the history variable approach used in the phase field fracture model in MOOSE, when I have a distribution of preexisting cracks. The question is about the initialization of the _hist material property. According to this post (https://groups.google.com/g/moose-users/c/Bf_nkvs-j3k/m/yXib3yPMBwAJ), I have to hard code ComputeIsotropicLinearElasticPFFractureStress::initQpStatefulProperties() function, if I'm not interested in creating the cracks in the mesh itself (real cracks). I was wondering if it has become possible somehow to initialize this history variable in the input time, e.g., using an Auxillary variable or a material property. Any insights/suggestions are really appreciated. Thank you, |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
Do you have to use a history field? I recommend against it. Instead, it's better to apply initial condition on c, and enforce the irreversibility using Now, if you insist on using history to initialize the crack, you can take a look at this PR: #15102 which tried to do what you are describing here. It seems to me that PR was really close to be merged, I don't know why it got closed in the end. Like I said there, I'm not a big fan of the history approach, but I also have no objection in us supporting that capability. |
Beta Was this translation helpful? Give feedback.
-
|
This is very helpful. Thank you so much. |
Beta Was this translation helpful? Give feedback.
Do you have to use a history field? I recommend against it. Instead, it's better to apply initial condition on c, and enforce the irreversibility using
VariableOldValueBoundsAux.Now, if you insist on using history to initialize the crack, you can take a look at this PR: #15102 which tried to do what you are describing here. It seems to me that PR was really close to be merged, I don't know why it got closed in the end. Like I said there, I'm not a big fan of the history approach, but I also have no objection in us supporting that capability.