The current Dirichlet model allows a scalar 'stepSize' and vector 'factors', where the in-homogeneous Dirichlet values (factor not equal to zero) for dof type 'i' is modified as: total[i] += factor[i] * stepSize. The input file code snippet looks like this:
|
* model = |
|
* { |
|
* type = "Dirichlet"; |
|
* nodeGroups = "TopNodes"; |
|
* dofs = "dy"; |
|
* factors = [ -1.]; |
|
* stepSize = 0.001; |
|
* }; |
Instead, introduce a more generic Dirichlet expression, total[i] += factor[i] * stepSize[i], where 'i' is a dof type associated with a node group. Also, total[i] would assume values from initVals[i]. In this way, one can model homogeneous conditions with factor and initVal set to zero, in-homogeneous but constant with initVal not equal to zero and factor zero, and in-homogeneous and temporally changing values with non-zero factor.
The current Dirichlet model allows a scalar 'stepSize' and vector 'factors', where the in-homogeneous Dirichlet values (factor not equal to zero) for dof type 'i' is modified as: total[i] += factor[i] * stepSize. The input file code snippet looks like this:
falcon/src/constraints/DirichletModel.cpp
Lines 17 to 24 in 778b2dd
Instead, introduce a more generic Dirichlet expression, total[i] += factor[i] * stepSize[i], where 'i' is a dof type associated with a node group. Also, total[i] would assume values from initVals[i]. In this way, one can model homogeneous conditions with factor and initVal set to zero, in-homogeneous but constant with initVal not equal to zero and factor zero, and in-homogeneous and temporally changing values with non-zero factor.