How to model a cable in OptiStruct and apply a pretension to it?

JVR

Hi, firstly, I would like to model a steel cable structure with OptiStruct. The cable must work only under stretch/tension. None compression, bending or torsional reactions.

 

I have already worked with CRODs, but these elements provide all the before mentioned reactions at the same time and I do not know how to apply a pretension load to it or if it is possible for linear static load steps.

 

Recently, I watched a video tutorial of CGAPs which can be used for the purpose of simulating a cable structure while also applying a pretension load. The problem is that I was not able to set them up properly for my example model (a separated testing environment that I use to test if any of these elements work as I need). For your reference, this is the video that I watched.

 

I would like to mention that I prefer to be able to use these elements and the pretension in linear static load steps if it is possible.

 

Thank you beforehand for the help.

loistf

Hi JVR. 

I can recommend this article. 

Tutorials, How To, Troubleshooting - Cable modelling in Optistruct (altair.com)

Does it solves your doubts? 

JVR

loistf said:

Hi JVR. 

I can recommend this article. 

Tutorials, How To, Troubleshooting - Cable modelling in Optistruct (altair.com)

Does it solves your doubts? 

Hello loistf,

 

I followed the tutorial, nevertheless, I was not able to achieve the same results and behavior of the CGAP shown.

 

Yours sincerely,

loistf

JVR said:

Hello loistf,

 

I followed the tutorial, nevertheless, I was not able to achieve the same results and behavior of the CGAP shown.

 

Yours sincerely,

Hi JVR. In which sense you didn't achieve? If you are more specific I might have some suggestion. Ideally you show a reproducible example. 

JVR

loistf said:

Hi JVR. In which sense you didn't achieve? If you are more specific I might have some suggestion. Ideally you show a reproducible example. 

Hi loistf, thanks for your response:

Okay, so basically if I set up the CGAP element with the linear static load step with the FLIP option marked to have a simple cable element behavior, it gives similar results as a CROD element. All is fine until this point. Unfortunately, if I set up a preload (F0), it is ignored or shows very weird results without common sense.

In the other hand, I already tried setting up a non-linear static load step with the exact same model, boundary conditions and loads, and I was not even able to get a convergent solution. After opening the .h3d in HyperView, the result has no sense, shows a massive deformation on the elements. I also tried setting this model with a preload (F0) but, as I already commented, the original model with the non-linear static load step, does not converge.

If you need it, I can upload you these models to have a better understanding on the problem.

Yours sincerely,

loistf

JVR said:

Hi loistf, thanks for your response:

Okay, so basically if I set up the CGAP element with the linear static load step with the FLIP option marked to have a simple cable element behavior, it gives similar results as a CROD element. All is fine until this point. Unfortunately, if I set up a preload (F0), it is ignored or shows very weird results without common sense.

In the other hand, I already tried setting up a non-linear static load step with the exact same model, boundary conditions and loads, and I was not even able to get a convergent solution. After opening the .h3d in HyperView, the result has no sense, shows a massive deformation on the elements. I also tried setting this model with a preload (F0) but, as I already commented, the original model with the non-linear static load step, does not converge.

If you need it, I can upload you these models to have a better understanding on the problem.

Yours sincerely,

sure, do not hesitate to upload the files if you can. Myself or somebody else in the forum might have suggestions. 

 

JVR

loistf said:

sure, do not hesitate to upload the files if you can. Myself or somebody else in the forum might have suggestions. 

 

Hi loistf!

I have attached the .hm file that I have been using to test the behaviour of the CGAP element in this reply.

The PGAP is still not preloaded with the "F0" input because the model is not even able to run the non-linear static analysis and I have not found a solution for it.

The non-linear static analysis is mandatory in case of trying to apply the preload, otherwise, the reference guide states that this "F0" input would be ignored.

Yours sincerely,

loistf

JVR said:

Hi loistf!

I have attached the .hm file that I have been using to test the behaviour of the CGAP element in this reply.

The PGAP is still not preloaded with the "F0" input because the model is not even able to run the non-linear static analysis and I have not found a solution for it.

The non-linear static analysis is mandatory in case of trying to apply the preload, otherwise, the reference guide states that this "F0" input would be ignored.

Yours sincerely,

Hi again. Took a look at your model. I believe that the problem of nonconvergence in your model is not related to the cable itself, but to the fact that the model is not supported enough for a static analys. I recommend you add constraints on the plate below to test the cgap and f0 behaviour. Then, when you have seen your model working properly, you can consider removing the bottom-plate constraints you added as much as you can.

You can probably see rigid body modes if you ran a normal modes analysis, which would prove that your model is not proper, yet. 

JVR

loistf said:

Hi again. Took a look at your model. I believe that the problem of nonconvergence in your model is not related to the cable itself, but to the fact that the model is not supported enough for a static analys. I recommend you add constraints on the plate below to test the cgap and f0 behaviour. Then, when you have seen your model working properly, you can consider removing the bottom-plate constraints you added as much as you can.

You can probably see rigid body modes if you ran a normal modes analysis, which would prove that your model is not proper, yet. 

Hi once again loistf,

As you indicated on the previous reply, the model was not properly constrained instead of a wrong setup on the PGAP property, explaining why the analysis failed every time.

I created a random SPC on the bottom plate and the analysis worked perfectly.

By the moment, I think this will be enough for me to keep testing these CGAP elements to validate them as a substitute for the CRODs in the real models. I still need to validate that they do not work under compression, nevertheless, right now I have to prioritize a different project.

Thank you so much for all the help provided.

Yours sincerely,