I have a 2D plate with different material regions defined on it. When I try to run it, it says "Zero or negative pivot. Check for Rigid body modes". However, when I perform the same analysis on the plate with the same material all through, it gives me perfect results. I'm performing the simulation on Franc 2D. Could someone tell me what I could be doing wrong?
Hi,
I assume you are running a static analysis.
I do not have experience with Franc 2D but Zero or negative pivot usually means that you do not have essential displacement boundary conditions set.
This means that your body (or some part of it) is not in static equilibrium.
You should check that the boundary conditions are set correctly when you have different material zones in place.
This all goes for a static analysis. If you are running a dynamic analysis (explicit or implicit) then check that you have defined the density for all the materials.
Hope this helps.
Elad
Thank you both for your elaborate responses.
However, as for the boundary conditions, the model is the same as the one with uniform material. So, I don't think there is an inconsistency there. Same goes, with the element type or connectivity I believe. If it works for the case with uniform material, I don't see why it wouldn't for the case with different material regions. Moreover, the software doesn't give me an elaborate output, node wise.
I can attach the file in case you want to take a look at it.
Thanks again for the help.
It would be helpful f you could describe how the multi-material plate is constructed in the model. Is it a layered plate? with contact between layers or is it a single body with sections of different materiel types. Zero or negative pivoting is an error you get when there are problems with the inversion of the stiffness matrix. This is usually the result of boundary conditions or material constants as I said before. Maybe the program splits the body into different sub-bodies when you define different material parameters? It could help more if you could describe in more detail the problem you are trying to solve and how you apply your BC and define your material.
I think it usually comes in static analysis when there is a rigid body motion. You can confirm this by adding "*Contact Control Assignment, stabilize". If the analysis runs with this addition then revisit the boundary conditions.
It is a single body with different material regions. It is a 2D plate with different material regions (circles). Left end and the bottom end are fixed and a load is applied on the top edge (tensile). There is an edge crack to the right end.
May be you could provide us with a screen capture of your geometry and meshing to help?
Otherwise, I don't know much about FRANC2D specifically, but in general if the areas where different materials are defined have been specified by different surfaces or by splitting the original surface in multiple surfaces, the boundary often end up being superimposed lines and vertices that are not connected. This would result in rigid body motions such as those you encounter.
Otherwise, you may have ended up with a degenerated element near one of those round material region which becomes too skeewed or folds over itself when the load is applied.
One last thought is that you may have defined stiffness and loads in inconsistent units (e.g. stiffness in GPa, loads in N and distances in m), leading to a negligible stiffness and very large deformations.
I hope this helps.
Best regards,
Laurent
I've attached the snapshot here. I will take a look at units' consistency. Thanks for your suggestions.
You have triangular elements (or quads with superimposed nodes) that have a very high aspect ratio (long edge over short edge) near the crack (among others). Moreover, those appear to deform quite a bit on loading. It is likely that those will provide poor results in any circumstances and could result in singularities (for examples, all three nodes become aligned) if the material stiffness is low. It could also be any of your corners triangular elements that, due to its high initial aspect ratio, gets deformed in a way it can't be solved anymore.
I probably am missing something, but to me it seems that simple plates like that are good candidates for mapped meshing which are much less prone to skewed elements.
*Edited aspect ratio definition, sorry for the initial inversion of short and long edges.
I am using franc2d and I would like to know how can I apply a residual stress field. The manual that I have states that I have to write an ASCII file with the extension *.rsi from the option residual load.
Does anybody know how can I write such file in order to input my residual stress field or have anybody an example file?
- Badges
- Science topic
- Similar topics
- Engineering
- Mechanical Engineering