I am trying to deal with simple static structural sim, but oftenly I get the
RUN TERMINATED DUE TO EXCESSIVE PIVOT RATIOS IN MATRIX KLL. error message.
Well, from what I understand, this can mean that something is just not fixed well in the model and "flowing", (excesive free DOFs) and I should run something similar to "modal analysis" to find out what is wrong, suggested is to set the BAILOUT parameter to value -1.
^^^ USER ACTION: CONSTRAIN MECHANISMS WITH SPCI OR SUPORTI ENTRIES OR SPECIFY PARAM,BAILOUT,-1 TO
^^^ CONTINUE THE RUN WITH MECHANISMS.
The only problem is, that modal analysis as it is ends with fatal error, so I still know nothing about the model and when I try the BAILOUT parameter, I can't set it... I open the table Solution -> Edit -> Parameters tab -> Parameters -> Edit and I set the BAILOUT parameter, but when I close this tab and open the Edit again, the value is still 0...
So, what can I do now? Am I missing something when trying to set up the BAILOUT parameter? Is this the correct way how to know if the model is correctly connected?
Solved! Go to Solution.
When you edit a parameter, you need to click the Add button
Then click OK at the bottom of the dialog. You should see your PARAM in the list of specified properties if you re-open the parameters modeling object:
Also, Nastran will provide nodes and degrees of freedom, in the f06 file, that correspond to the excessive ratios. Sometimes these can help you identify the problem, which can be undefined stiffnesses in connection elements like springs, 1D elements spinning around their X axis due to lack of rotational connectivity, etc.
Thanks a lot, this is exactly the information I needed to hear.
See, I saw there the button "Add" all the time, but once I clicked it it changet to "Remove" and I had no idea what happend... I onlz knew, that when I opened the "Parameters" second time, the BAILOUT parameter was gone and I didn't noticed the new cathegory "Specified Properties".
So now, when I knew what to look for, I finally understand, so thanks again.
I have additional question, related enough to this topic:
Now, the BAILOUT is -1, the simulation ended succesfully (In this case, it is a very simple simulation, only Shell, fixed on some edges, loaded with gravity, no other elements, one body, It even succesfully finished the modal analysis with no visible errors in the mesh) what are the results I can look for, what the BAILOUT gives me to identify, where was the problem?
Thanks a lot for your help,
I would start by looking at the displacements. It's likely they don't make sense, and hopefully the part of your model that is not well connected is moving relative to the rest. From this you will need to figure out what in your fem is causing this. Maybe you have multiple meshes and the junction nodes are not coincident. You can see this with the Element Edges check.
You can also look at reaction forces, to see if they occur where they should, and if the sum makes sense. Since you applied a gravity load, the sum of the forces should equal the mass times the acceleration in the loading direction, and 0 in the others. A possible error is that your model has no mass. You can check the mass of your model using the Finite Element Model Summary.
I hope it is not a bad thing to reopen this topic, when I have another relevant question:
This is a simmilar model. It is one shell object, connected with the StitchEdge tool, it's calculation ends with the KLL error and when i use the BAILOUT -1, I have following results. As you can see, nothing is "floating", but individual nodes are "buckling" in rotation (the second image). Do you have any idea what to do with it?
I am thinking about finer mesh, or linear quads instead of parabolic, but I don't know. and I am still kind of lost in it.
Thanks for all the help,
The drilling degrees of freedom (rotation around the element normal) are not restrained for parabolic shells. You can increase PARAM,MAXRATIO or you can switch to linear shells, for which PARAM,K6ROT takes care of the drilling dofs.
thanks for the explanation, now it is clear for me. Just to finish this task:
To increase the MAXRATIO (to 1e15) did help to get the results without further KLL errors or warnings, but it did not repaired the illogical rotation results.
So the more usefull answer is use the linear elements. That fixed both.
Thanks again and by this I consider this topic finally truly closed.