Working with Rigids

Hello everyone.

I'm working on a project which consists of a rod-trussed structure .I'm having some difficulties working with rigids. I have followed the Altair's student frame linear static analysis.

Since I need my SPC to be in a rigid, the configuration is this one:

<?xml version="1.0" encoding="UTF-8"?>

I am using two rigids: one for the SPC and another one for applying a force (brown line). Whenever I use the configuration above, OptiStruct says there's dependency between rigids (because the two rigids are applied to the same node). So, in order to fix that, I have tried this configuration:

<?xml version="1.0" encoding="UTF-8"?>

Which yields this result:

<?xml version="1.0" encoding="UTF-8"?>

But if I try to run a different configuration (for example, linking the SPC rigid to 3 nodes, except the one where the force rigid connects), I get this result:

<?xml version="1.0" encoding="UTF-8"?>

Which basically zeroes my displacements and stresses in the model. I have no ideia which one is giving me more realistic results and I'd could use some help trying to figure it out!

Thanks

Unable to find an attachment - read this blog

Find more posts tagged with

English

OptiStruct

Accepted answers

All comments

Adriano_Koga

your rigid elements (yellow, and brown) are composed of only 1 rbe each, or are they made of 2 small elements each?

If you are using 2 rigids for creating each one, make sure that the independent node is the middle one for both of them. The independent node should be the one with SPC.

For the force it is ok. Alhogh maybe a RBE3 would be better to avoid adding stiffness.

LFSS

your rigid elements (yellow, and brown) are composed of only 1 rbe each, or are they made of 2 small elements each?

If you are using 2 rigids for creating each one, make sure that the independent node is the middle one for both of them. The independent node should be the one with SPC.

For the force it is ok. Alhogh maybe a RBE3 would be better to avoid adding stiffness.

They are composed of two rigids, one for yellow and one for brown. I've made sure the middle one is the independent node for both of them, as well as where the SPC and the force is applied. What's the difference for an RBE3? is it better for my case?

Adriano_Koga

RBE2 creates a kinematic condition between nodes, thus artifical stiffness is added to your model. So the dependent DOFs will follow exactly what the independent DOF does. That's why it is called rigids.

RBE3 is used for load and mass distribution, and does not add stiffness to your nodes. Dependent node will be calculated from 'an average' of the independent nodes. Actually it depends on the weighting factors.

Maybe sharing your model would make it easier to check what is going on.

LFSS

RBE2 creates a kinematic condition between nodes, thus artifical stiffness is added to your model. So the dependent DOFs will follow exactly what the independent DOF does. That's why it is called rigids.

RBE3 is used for load and mass distribution, and does not add stiffness to your nodes. Dependent node will be calculated from 'an average' of the independent nodes. Actually it depends on the weighting factors.

Maybe sharing your model would make it easier to check what is going on.

I understand. I'll try it later today.

I've attached the model file to the original post.

Adriano_Koga

I think I didn't get what you see as a problem.

Running your model as it is, this is what I got (just adjusting the legend a little bit, and adding a couple 'measures').

By adding a RBE2 you're essentially imposing that the same ZERO displacement (SPC 123456) of the central node will the carried to the 3 nodes connected to it.

Which is exactly what you got here.

For sure you should look again at what are the BCs in your real structure and try to bring them to your model.

There is some stress in the whole structure, although it is low compared to the stress in the red regions.

For buckling you requested only the 1st mode, which is ok, as it is the lowest one, but you would probably want to look at others too.

<?xml version="1.0" encoding="UTF-8"?>

LFSS

That's interesting.

So the difference between my first model (with the SPC rigid linked to 2 nodes) to my second one (SPC rigid linked to three nodes) is that in the second one I get a overstiffened structure relative to the first one, did I get it right?

As for the buckling, thanks for the advice!

Just out of curiosity, how do you display the tubes like that? I only get the line contour, but this one you showed here has a better visualization!

Adriano_Koga

correct.

2nd model was stiffer thatn the 1st one.

This image was just a display, not the real section.

I went to Preferences and turned on visualization for 1D elements as a cylinder.

image.png.108bd2f8da9b30c9bf98243efecb46a2.png

LFSS

correct.

2nd model was stiffer thatn the 1st one.

This image was just a display, not the real section.

I went to Preferences and turned on visualization for 1D elements as a cylinder.

<?xml version="1.0" encoding="UTF-8"?>

Thanks a lot man!