The OpenSees Community

I really need your response Frank;

I checked my model again and found that the problem is related with TwoNodeLink elements and when I use TwoNodeLink elements, the following error will appear:
"FEM_ObjectBrokerAllClasses::getNewElement - - no Element type exists for class tag 86
Fatal error in MPI_Recv: Message truncated, error stack:
....
...
.."
and one other thing, I should perform the analysis in two stages. the first stage is geostatic (impose gravity) and the second stage is transient and I used MultipleSupport pattern in order to imposing earthquake displacement record to the base. but after geostatic step, an error is appear that says there is no node to impose the motion to. if I define some of the nodes again after geostatic step, it works but I don't know what is happening to my model and where is the missing nodes. after this error I'm not sure what parts of my model are missing and what parts are not.
How can I fix it?

Hi Frank;
I'm still waiting for your response.

when I run in SP, the model is built successfully but at the beginning of the analyze stage, some times it just freezes and some times it aborts with this error:
"job aborted:
rank: node: exit code [:error message]
0:Hadi-PC: -1073740940: process 0 exited without calling finalize
1:Hadi-PC: 123
2:Hadi-PC: 123
3:Hadi-PC: 123"
By the way, if I just change the element type to bbarBrick, the model will run and no problem happens.
Thanks.

I create my model with SSPbrick elements and it runs well in OpenSeesMP but I can not run it in OpenSeesSP (v.2.4.10). I want to run my model using any number of processors and configuring my huge models to be run in OpenSeesMP is really difficult, so I want to shift to OpenSeesSP .
Do you have any suggestion?
Regards

@mhscott:
the material is linear and I used both mass and stiffness proportional damping in OpenSees, but it gives different results with the same coefficients in SAP.

@fmk:
the connectivity, element, material properties and every thing are exactly the same but as I said, it gives different results with the same coefficients in SAP.

thanks

Hi every body;
I modeled a SDOF system with a massless column and a lumped mass at top of that. I used Rayleigh damping coefficients in order to model a 5% damping ratio.
I created the same model in SAP2000 and used the same damping Coefficients but after transient analysis, the results are different from those from OpenSees.
So, the question is:
What is the difference between them and how can I fix this problem?

thanks

I have a model and 10 earthquake records. When I want to perform transient analysis with four records of these 10 records, OpenSees freezes at last two steps of the analysis and does nothing till I close the executing window. As I said, I have no problem with other records and the problem is shown in just four of them. I tried to run the model on different PCs and the same problem occurred again. I checked the records and every thing is normal about them.
By the way I deleted the last part of record and run the model again but the problem did not solve.

Regards,

Dear Shiva,
It seems you work on similar projects as mine. It would be very useful if we can handle our experiences to each other.
About your question I should say that based on my knowledge about soil behavior and soil constitutive models implemented in OpenSees, there is no no-linear elastic soil models implemented in OpenSees. So you have two choices, using elastic material or using more complicated plastic material models such as PDMY. I think that OpenSees needs to have more simpler material models like Mohr-Coulomb and non-linear elastic soil models.
I hope it helps.

So what should I do if for example I have nodes 100 and 101 in two processors and I want to tie them together?
I should tie them in just one of these two processors and they will tie together in both two processors?!

Frank,
I run it sequentially and it worked well.
I have some MPConstraints (equalDOF) in my model. when I deactive this constraints every thing is well but when I activate those, the error will appear.
I used Penalty method to impose constraints and there is no problem with the sequential OpenSees but in the parallel version (OpenSeesMP) the error will appear when I use factors greater than 1e10 for penalty method.
Do you have any idea about this problem?

When I run my model in OpenseesMP it prints the following error:
parallelMumpeSolver:: Error -9 returned in substitution dmumps

what is the cause of this error?

Hi Deepak,
I use PDMY material and SSPbrick element. My model is really huge and I can not post it.
I have another problem with gravity step. I do gravity analysis in two steps. there is no problem with the elastic stage. When I update the material to consider the plastic behavior of soil, it converges but the settlement of the soil surface will increase with the time and if I increase the time of analysis the settlement will increase as well. The settlement at the end of the elastic stage is rational but after plastic stage the result is completely unacceptable. Do you have any idea about it?
Thanks

Is it the same for update material stage command?
What is the rule of the DomainChange Command?

thanks Deepak.
I fixed the gravity step but now I have difficulties with the transient step of analysis. It does not converge and I'm really tired working on it. Do you have any advice for me?