Hi,
I am just wondering if there is any contact type in Nastran In-CAD similar to "No Penetration" contact set in Solidworks? I have read this topic as well: https://knowledge.autodesk.com/support/nastran-in-cad
Thanks,
Solved! Go to Solution.
Solved by John_Holtz. Go to Solution.
Hi maxhDL7VT,
I am not sure what the "no penetration" contact is in Solidworks, but I would guess that it allows parts to separate but not pass through (or penetrate) each other. If that is the case, it is the same as In-CAD's "Separation" contact.
Hi John,
Thanks for your reply. I have tried "Separation" contact type before. The issue is when I change the bonded to separation, I get endless iterations from Nastran solver during simulation. It goes on and on and never ends like an endless calculation cycle.
No Penetration in solidworks means selected components/bodies cannot penetrate each other during simulation.
Thanks,
Hi,
I think we need a complete description of what "No penetration" simulates in Solidworks.
Think of a flat plate in contact with another flat plate. Different types of contact are used to allow or prevent motion in three general directions:
Hi john,
Here is the definition of "No Penetration which I took from solidworks help;
No Penetration: Available for static, drop test, and nonlinear studies. This contact type prevents interference between Set 1 and Set 2 entities but allows gaps to form. This is the most time-consuming option to solve.
Thanks,
Hi John,
Also, shall I change any settings or parameters? When I apply "Separation" contact to parts, they penetrate each other because obviously contact is not initiated in In-CAD. I guess I should change some settings/parameters?
Thanks John,
Yes, I suggest that you change something if the results are wrong . We cannot say what needs to be changed because we know nothing about the analysis or the results. If you want to provide some images or attach the model so that we can understand the analysis, then someone can provide some ideas for the next step.
Note that there WILL be some penetration in most analyses. When the separation is zero, the reaction force is zero. The only way a reaction force is created at the contact surfaces is if the "invisible springs" in between the two contact surfaces are compressed, and that only occurs if the parts penetrate. In normal circumstances, the amount of penetration is insignificant, but there are situations in which the parameters can be changed to minimize the penetration.
Also, the deformed shape can exaggerate the penetration. In some cases, you need to edit the contour and set the "Deform Options" to an "Actual" value of 1 to get a correct deformed shape. Of course, the displacement results will indicate whether the parts have penetrated or not regardless of the deformed shape.
Good day John,
Thank you very much for your reply. I'll attach my model here so you can understand the analysis I am working on.
Regarding change of settings/parameters, I am technically addressing In-CAD parameters. please see here.( please Note that I am doing Linear analysis)
Thanks,
Hi John,
How can I send you a simulation file (90 mb size)?
Hi Max,
If 90 MB is too large to attach to the forum (I am not sure what the size limit is these days), you can do one of these:
Hi @Max.H.Pour
I think the problem with the penetration is related to the difference in mesh size between the two faces. WIth the automatic contact, it is trying to prevent the nodes on the wheel (with the large mesh) from penetrating the surface on the shaft (with the fine mesh). That leaves a lot of nodes on the shaft that can penetrate the wheel.
The easiest solution is to edit the contact pair between the shaft and wheel (H004004:1 <-> H002142:1) and change the "Penetration Type" from "Unsymmetric" to "Symmetric".
Give it a try and let me know what you think.
By the way, I tried other changing other parameters, but so far none have really made a difference on your original setup.
Hi @Max.H.Pour
I wanted to follow-up with you on this contact problem. Have you gotten better results? Do you still have any questions?
Hi John,
I have gotten better results by changing the "Penetration Type" from "Unsymmetric" to "Symmetric". It only works when the data type is set to centroidal. It is still very high on corner data type even though I refine the meshing to very small sizes.