BEST WAY TO DO A MULTI LOAD ANALYSIS

[Xprivate]

Hi All , i have two loads for one Boundary condition, for a same BC i have an initial loadcase and after my component is deformed with the initial case, a second load case follows. What is the best way to perform this analysis? i explored some
1)defining two subcases
2)running the first case and extracting the displacements , using this displacement as the initial load to run the second load case
3) or directly giving the first load case as initial conditions and second case in the loads of Boundary conditions card.

what is the best and most accurate way to go? thank you

 

[SWComposites]

Are you running a linear analysis? If so, just apply all of the loads in one case.

 

[Xprivate]

Hi, its a non linear analysis

 

[rb1957]

1) probably best. Separate the two cases so you can factor one (what you wanted to do in another thread).
2) this probably would work but would be such a pain in the a$$ (editing every node co-ords) ... but then you'd lose the internal stresses that these displacements create, so it won't work either.
3) don't understand.



"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[Xprivate]

thanks rb1957, i think like you say option 1) is the best to proceed . i used the subcases to run the analysis.

 

[ThomasH]

I assume that this is a continuation of the question in the Nastran forum. And now you want to set this up in Femap. There are more than one way to set this up but I will try to show you the main steps.

I assume that you have one set for BC's and two sets for the loads. The first load set should include the base pressure and the second should include the pressure and the point load.

In Femap select "10 Nonlinear Static", then select "Global Requests and Conditions"-"Nonlinear Options". For "Increments or Time Steps", input 10 and for "Output Control" - "Intermediate", select "1 Yes". This creates the non-linera analysis option for both your subcases.

Now move down a bit in the options to "Boundary Conditions", select that option and click "Edit". Select your Constrants and click "OK". These constraints will apply for both subcases.

Now I will simplify a bit and define two subcases.
The bottom line is your "Analysis Set Manager" probably reads "No Cases Defined". Click on that line and Click the "Edit" botton. Now you get "Analysis Case". I suggest for "Case ID" input "1" and Subtible "Pressure". Then click the "Next"-botton, now click the box that says "Skip NLPARM". For this load case /Sub case you will use the parameters set in the previously described setup. Again, click the "Next" button and for Loads select the pressure load.

If you now select (click on) "Case 1" and click "New" you will get "Case 2". Selects a subtitle, "Skip NLPARM", Click "Next" and select your second load set.

Now I think you should have a working setup for the analysis.

If you use several subcases in a nonlinear analysis the results will be continously added. So case 2 will have the completed case 1 as starting point, and so on.

 

[Xprivate]

THanks Thomas for detailed info, this is what i was talking about when I said defining two subcases [1) in my question]. I thought there were some better ways to do this. Because on of the problem i have is that my model is big(takes 10 hours to finish on case) and convergence is an issue. so before i finish my two subcases i will be spending weeks to get the solution. And to perform many iteration is hell lot of time. So yeah.

 

[ThomasH]

One way that may speed things up is to define the NLPARM for both subcases. If I assume that you have convergence issues for the second subcase you can use fewer "Increments" for the first subcase. That should be faster.

How big is your model (in dof's)? There can be a number of reasons for the long analysis time. The obvious is the model size but there can also be different issues with convergence that can be handled with the settings. Also, I assumed that you use NX Nastran but you may use another "flavor" . I would ask in the Nastran forum regarding specific settings for the solver if that is the problem.

Edit: The reason I recommend the Nastran forum is that there are a lot of similarities between NX and MSC. And I am not sure that the users of MSC Nastran follow the Femap forum .

 

[Xprivate]

it mostly because of the different BCs i am simulating , but my model has approximately 250000 nodes and elements . Yeah i am using FEMAP for this but solver offcourse NX. I am okay with the solver setting its just that the cases i run are bit tricky and takes a lot of time which i cannot disclose more .

THanks for your help

 

[rb1957]

"mostly because of the different BCs i am simulating" ... do you meant the two load cases have different BCs ? Then you can't run them as one model, no??

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[Xprivate]

@rb1957, i meant to say that the non linear parameters have to be changed everytime because i have 20-30 scenarios. but this is one scenario where i have same BC for two different cases.