Rigid body modes 3

[isponmo]

Dear All,

I would like to ask something regarding rigid body modes. Maybe it is a trivial thing, but I don't really get it: When I create a new model, I usually run a free-free modes check as part of a set of model checks. Then, I check that there are six rigid modes with very low eigenfrequencies. The thing is that sometimes, when I have a look at the modeshapes, they don't look rigid at all. To put a simple example, if I analyze a cylinder, I would expect to see three rotational and three translation rigid movements, but what I get are some modes in which the cylinder contracts or expands radially.

Is there any explanation for this? Am I getting anything wrong?

Thanks a lot!

Cheers,

Israel.

 

[GregLocock]

If there is deformation in the elastic parts of the structure then there must be actual energy involved in the mode shape and so the frequency would be non zero, hence not an RBM

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[shaun8567]

"To put a simple example, if I analyze a cylinder, I would expect to see three rotational and three translation rigid movements, but what I get are some modes in which the cylinder contracts or expands radially."

These are in fact rotational DOFs, but the expansion/contraction effect that you're seeing is the result of the linear underpinnings of a modal analysis. If you had a fixed-end cylinder (like a cantilever beam) with a torsional load applied on one end, then the results of a linear static analysis will show the cylinder expanding. I should point out that this is just a visual manifestation of the linear assumption; the numeric results calculated by the simulate (strains, stresses, etc) are fine (assuming you don't violate the small displacement approximation). If you were to run a non-linear analysis, then this effect would go away.

 

[GregLocock]

Good one. Also it may just be a visual effect, for instance the animation of the 'roll' mode of a cylinder would show the cylinder's diameter changing, as all the surface points move tangentially, not around the circumference.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[isponmo]

Dear shaun8567 and GregLocock,

Thank you very much for your replies! For newcomers and people with just a few motnsh of experience like me, it is very useful that the community answers to questions like this

I see your points but I don't understand it totally, though. Here my questions:

- "If there is deformation in the elastic parts of the structure then there must be actual energy involved in the mode shape and so the frequency would be non zero, hence not an RBM"

I totally agree, but I never manage to get exactly zero energy and zero frequency in a free-free analysis. In fact, in

http://femci.gsfc.nasa.gov/validitychecks/vc2.html

it is said that the eigenfrequencies should be small, but not zero. I am wondering if the solution is meant to be like that or it has something to do with error introduced by the FE discretization or by the numerical algorithms used to solve the problem.

- "These are in fact rotational DOFs, but the expansion/contraction effect that you're seeing is the result of the linear underpinnings of a modal analysis. If you had a fixed-end cylinder (like a cantilever beam) with a torsional load applied on one end, then the results of a linear static analysis will show the cylinder expanding. I should point out that this is just a visual manifestation of the linear assumption; the numeric results calculated by the simulate (strains, stresses, etc) are fine (assuming you don't violate the small displacement approximation). If you were to run a non-linear analysis, then this effect would go away."

Probably I am getting something wrong from your explanation. I understand that if you have a torsional load the cylinder will expand, but I am not able to see the connection between torsional effects and rigid body mode. If the cylinder is rotating as a rigid body, it is not supposed to be suffering a torsion, right? Wouldn't that be like stating that a cylinder moving in one translational direction as a rigid body suffers, say, compresion?

- "Good one. Also it may just be a visual effect, for instance the animation of the 'roll' mode of a cylinder would show the cylinder's diameter changing, as all the surface points move tangentially, not around the circumference."

What's the difference between moving tangentially and around the circumference?


Again, thanks a lot.

Kind regards,

Israel

 

[GregLocock]

"What's the difference between moving tangentially and around the circumference?"
Draw it up.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[isponmo]

Dear GregLocock, shaun8567 and all,

What do you think about the last things I pointed out? Unfortunately, they are not clear to me yet

I don't know what the meaning of "draw it up" is(I'm not a native English speaker). Do you mean that I take a paper and draw the cylinder? XD In any case, I can see three "directions" in a cylinder: tangential, radial and length direction...

Thanks again,

Cheers!

Israel

 

[GregLocock]

"I take a paper and draw the cylinder?"

yes

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[isponmo]

Ok... but what about the other two questions??

 

[IceBreakerSours]

Wow!

Are you new to this forum? If so, please read these FAQ:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083