I'm trying to use the Model Order Reduction technique to reduce the stiffness matrix in AceFEM/AceGen. But there is a problem with the manipulation of the global stiffness before solving it by SMTNewtonIteration[]. Has anyone met some problem like this?

  • 3
    $\begingroup$ "there is a problem" - what problem? Can you include enough code to reproduce the problem? $\endgroup$
    – MarcoB
    Commented Oct 17, 2022 at 17:45
  • $\begingroup$ hello MarcoB, because using SMTNewtonIteration[] will automatically assemble the stiffness matrix and solve it. But I want to modify or manipulate the global stiffness matrix before solving it. Is it possible to do that in AceFEM/AceGen? Thanks a lot. $\endgroup$ Commented Oct 18, 2022 at 8:25
  • $\begingroup$ Have you seen this? $\endgroup$
    – user21
    Commented Oct 19, 2022 at 4:22
  • $\begingroup$ I don't think that what you are looking for is possible. link1 and link2 $\endgroup$
    – Max
    Commented Oct 19, 2022 at 10:20
  • $\begingroup$ The direct answer is No, due to reasons you mentioned. However there are some ways one could circumvent this issue, but It depends what you want to do. You can introduce 2 domains and make 2 element codes. Then deactivate domain 2, assemble the FULL tangent of domain1, and then do some operations on it and export it to elementData to elements 2, activate them and deactivate elements 1 and you can deactivate required nodes if needed. Then the SMTNewtonIteration will assemble Tangent of elements 2. Or you can contact author, prof. Korelc and ask him for the c source code. $\endgroup$ Commented Oct 19, 2022 at 10:28


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.