Multipoint constraints

[JASPERFENG]

Dear professor Dodds

If I want to release the multipoint constraint, is that possible to be realized, This is mainly used to simulate the crack growth under low-cycle fatigue actions and the crack plane is not at the symmetry plane. I tried to modify the MPCs between the steps. It reports the errors: "the determinant of the Jacobian matrix for gauss point of the element is non-positive". "The element" denotes the element with MPCs

Thank you very much
Best regards

Max

[rhdodds]

Hi Max, We can communicate directly on this. Please try running a linear-elastic model with released multi-point constraints — does that work ? I suspect the issue is the sudden release of restraining nodal forces (from the MPC) on already deformed elements. Are you using large displacement formulation ? For symmetry plane constraint releases, those forces are decreased gradually over a number of time steps. This issue has been on my list for a while…. Bob

…

On Mar 6, 2018, at 2:31 AM, GithubMarkMike ***@***.***> wrote: Dear professor Dodds If I want to release the multipoint constraint, is that possible to be realized, This is mainly used to simulate the crack growth under low-cycle fatigue actions and the crack plane is not at the symmetry plane. I tried to modify the MPCs between the steps. It reports the errors: "the determinant of the Jacobian matrix for gauss point of the element is non-positive". "The element" denotes the element with MPCs Thank you very much Best regards Max — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#3>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AL4g_hpgUpQ7rT5wyY0wu-onHiYhw78fks5tbjtDgaJpZM4SeOrI>.

[JASPERFENG]

Hi professor

I tried running the linear-elastic model with releases MPC, still cannot, the same error occurs. I used large displacement formulation before, now I tried small displacement, also cannot, the same error occurs.

I checked the Warp3D manual, Warp3d tackles with MPCs by: " the incremental equilibrium equations are modified to eliminate the dependent displacements, the equations solved and the dependent terms recovered immediately from the constraint equations" does that mean no nodal force between the tied nodes, therefore we cannot release the MPCs like that we release the absolute constraints in Warp3D ( can release the nodal force gradually).

It will be grateful and appreciative if you can elaborate on that. and the barriers of realizing the release MPCs in warp3D.

Thank you very much
Best regards
Liuyang

[rhdodds]

Hi Liuyang, I’ll run some problems this weekend. I’m giving mid-term exams this week. I suspect you have discovered the problem. When MPCs “disappear” between time steps, the code does *not* capture the corresponding nodal forces and release them over a few subsequent steps. That procedure is followed by the release of simple constraints as you have found. Best, Bob

…

On Mar 7, 2018, at 1:14 AM, GithubMarkMike ***@***.***> wrote: Hi professor I tried running the linear-elastic model with releases MPC, still cannot, the same error occurs. I used large displacement formulation before, now I tried small displacement, also cannot, the same error occurs. I checked the Warp3D manual, Warp3d tackles with MPCs by: " the incremental equilibrium equations are modified to eliminate the dependent displacements, the equations solved and the dependent terms recovered immediately from the constraint equations" does that mean no nodal force between the tied nodes, therefore we cannot release the MPCs like that we release the absolute constraints in Warp3D ( can release the nodal force gradually). It will be grateful and appreciative if you can elaborate on that. and the barriers of realizing the release MPCs in warp3D. Thank you very much Best regards Liuyang — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#3 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AL4g_rSXZIet3RVyKjjUPGM-69Ofc5fGks5tb3rNgaJpZM4SeOrI>.

[JASPERFENG]

Thank you very much, professor

Best regards
Liuyang