3 weeks ago
First I want to say "hi" to the whole community.
I have been following the conversations in this forum for quite some time and found help for many cases. So thank you all for your contributions and help.
After searching the web and the forum for a solution to my current problem, I decided to ask the question here. Please be patient, if I have overlooked a thread where this topic has been discussed already.
Here is the description of my situation:
I want to simulate a contact problem with the solver SOL 401 multi-step nonlinear. The minimodel will be attached via screenshot here. Roughly speaking, I want to move a deformable block (punch) against a deformable support and observe the resulting contact results when different dies are put between these two parts. I have to say, that I use only the GUI of Simcenter 3D and no further test-editing for bulk data or something. All the solutions should be "clickable" in the gui. I hope, that this will be possible. Currently, I use quadratic HEX elements (CHEXA20). If this can cause trouble with the contact, please let me know.
Let's give some more detailed information about the model setup. I am using the software Simcenter 12.0. In the screenshot you will notice both parts without any die in between. The gap between the contacting surfaces is 15mm in the start position. I want to move the punch (block) towards the supporting plate, establish the contact and press the punch a little bit further to evaluate contact pressures and forces. Using the "enforced motion constraint" I defined a travel way of 17mm overall, so 2mm will be the displacement where the pressing will take place.
The contact I use is the "surface-to-surface contact" with a coefficient of static friction of 0.2 and no other parameters to be defined on the corresponding dialog. I tried different settings for the search distance, because I discovered problems with contact and gaps when using SOL101. But this is due to the linear solution characteristics, I suppose.
My very first question is:
- How do I trigger the time stepping? I want to have the results for each increment, and I want to be able to force the size of those increments. My objective is to observe the contact realization in detail. And in this special case, it is mandatory to consider the complete way of travel. So I can't close the gap first and just simulate only the moment of touch.
Note: Following the tutorials concerning "time dependent boundary conditions", I also used a modeling object "TimeStep" where I defined the time increment and the number of time increments (10 steps in 1 second will yield an increment of 0.1s; see screenshot). But when I run the solution, it is obviously not used. The initial increment used is the complete analysis time which is 1s.
I have another simulation with a different problem where the time stepping works. But I am not able to find the point, where the definitions of these simulations differ from each other. I also messed around with some of the "nonlinear control parameters" (see screenshot). But nothing changed. Even when I force the initial time step (DTINIT) to be another value than the default, it will not take place in the simulation. Maybe, I got the purpose of this parameter wrong.
It would be nice to know what I am doing wrong. And I think it is only a small thing. Maybe kind of a lost checkbox.
l
And here is my second question:
Of course, the solution of this problem has a converged result. The solver tries to solve the situation for a time increment of 1s but fails to converge, then it performs a cutback to 0.5s and finds converged solutions for time increments 0.5s and 1s. So far so good. If I used the default values for the search distance in the contact pair definition, I will get a warning in the solution monitor that "no contact elements formed ... solve attempted without contact" for the first try of the increment 1s. That's quite easy to understand. After the cutback there is also no contact formed and for the second try to get from time 0.5s to the endtime 1s the contact will be established and everything is fine.
- Here is the question: What do I control with the search distance? When there is a large gap between the contact partners, it seems to me that there is no need to define any value. Am I right on this?
I have set the parameter for contact stiffness update TSTEPK (I found this in a tutorial or in a web video). Would you also recommend this setting?
Long story to be told here. I come from ABAQUS and have to convert my old workflows into Simcenter.
I really hope, that somebody can help me on this. From my point of view these problems should not be a big deal, but I am struggling to find the proper answers out there.
Any help is much appreciated. And thank you very much in advance.
cheers Nick
3 weeks ago
In the attached images, you appear to be editing the dialogs directly on the navigator. This is just a container of all modeling objects in your part.
To actually use a modeling object, it has to be explicitly selected in a solution. Edit a subcase to select a time step modeling object. Edit the solution to select the global nonlinear control parameters modeling object or edit a subcase to select a local nonlinear control parameter modeling object.
3 weeks ago
Thank you Mr. JimB for your very quick answer. Seems to me, that it took you less time commenting my questions than me for creating the topic ;o)
I tried several things and here is what I got so far:
I did add the constraints to the active subcase "multi-step nonlinear". Editing of this subcase shows the possibility to directly define the number of increments and time period for the subcase. So with this definitions I can deal with fixed time increments. That would be sufficient for my current problem. So I could handle my issue as solved. Thank you very much for your help.
But I want to understand the procedure by selecting the modeling object. In your answer you mentioned three different ways.
I think I didn't get the point exactly. I understand that the definition of the modeling object "TimeStep" is only some kind of an item in a library and it should be used/selected in the corresponding dialogues. You said that one should select this object directly in either the solution or in the subcase. Unfortunately, I can't find the point where to select it. I try to visualize my screen for the edition action, in order to make my point clear.
quote: Edit a subcase to select a time step modeling object.(1)
Edit the solution to select the global nonlinear control parameters modeling object (2)
or edit a subcase to select a local nonlinear control parameter modeling object (3).
(1) Subcase is defined and the edit dialogue shows the following screenshot. I didn't find a way to select a modeling object. By the way, for the use of fixed increments, I took the definition marked by the red circle.
For point (2) and (3) I also can't see a chance to make the proper selection. What did I do wrong? Or am I searching at the wrong place?
Maybe you can go a little bit in more detail where to find the selection possibilities. That would be great.
best regards
Nicksen
2 weeks ago
Hi,
Joining the discussion, I would like to ask how to run the strategy - Automatic Timing Scheme with Initial, max and min parameters?
best regards
Michał