I am running analysis of a cross-shaped sample that is prescribed a displacement along x and y-axes. I have run several jobs with this model and had no issues. However, the current set of conditions I am trying requires multiple steps while previous ones have all been one step. I want to apply a displacement along both axes, then remove the condition on one axis, allowing it to unload while increasing the displacement in the other axis. I tried deactivating y-displacement while modifying x in step 2 but every time I try to run, it aborts as soon as it tries to begin step 2. I have tried several variations including having a step for unloading where the only change is deactivating displacement and it still aborts at the start if this step.
Error: "too many attempts made for this increment".
Warning: "The strain increment has exceeded fifty times the strain to cause first yield at 103 points, Excessive distortion at a total of 723 integration points in solid (continuum) elements"
Hi Matthew Eaton,
I'd try to model loading and unloading in the same step using two different amplitudes, one for x-displacement and other for y-displacement.
Victor J. Аdlucky The issue is I don't want to set a specific displacement for y. I want it to unload freely. Is it possible to make this happen with an amplitude? In the physical situation which I am trying to model, the x-and y-arms are clamped and pulled to a specified displacement. I then want to "unclamp" y and then apply a further displacement to x.
Unfortunately, my idea is useless for displacement control. But I have one more idea. In the step 2 (unloading with deactivated y-displacement) try to apply tensile y-loads with values equal to y-reaction forces (in the end of the step 1) and amplitude decreasing from 1 to 0. Mayby it somehow will help to "smooth" transition between the steps.
And above all, minimize initial incrementn in step 2.
Victor J. Аdlucky Thanks for the idea. Unfortunately, I tried a few variations of this and it still seems to have the same issue. I also tried applying the force in step 2, then in step 3 deactivating displacement and ramping down force and it failed at the start of 3. I guess deactivating the displacement while the arm is under stress is too abrupt of a change.
Nils Wagner I've attached files for the original job, as well as two modified versions from the suggestions given earlier.
Matthew Eaton I decided to return to amplitudes approach. After the first step (loading1) I run the second step (unloading), in which x-displacement was fixed (with amplitude= 0,1,1,1) and y-displacement was decreasing to zero (with amplitude= 0,1,1,0). The goal was to find the moment when y-reaction force changes its sign. To save the time I stopped the job when increments become increasing due to active plastic deformation. The reaction force in the second step was really changing very abruptly. This way unloading can be simulated. But the question remains about possibility to deactivate y-displacement on the third step (x-loading). I hope that such deactivation when the arm is not under y-tension will not be so problematic. But it needs to be investigated.
Victor J. Аdlucky This is an interesting approach. I will try investigating this further.
- Badges
- Science topic