sequence of rigid body modes and signifiance

[enthu1980]

Hello All,

I have started learning frequency or modal analysis using FEA. I have one basic question.

Will the sequence of first six rigid modes during frequency or modal analysis will always remain same? Also, what is the significance of these rigid body modes?

Please bear with me for this basic question

 

[236804]

As per my knowledge the sequence is not at all important. The significance of rigid modes is that you can see any unknown constraints etc. If there no constraints model is complete free, then your rigid modes zero or nearly zero value. This rigid body mode frequencis gives you a idea about the model.

Hope this answers your question.

236804

 

[cbrn]

Hi,
You may even not see these rbm at all, if the FE program you use filter them out automatically.
The significance of the rbm in a modal analysis is null, because they are only the trivial solution of the eigenproblem where all eigenvalues are zero.
The sequence will remain the same for the simple reason that they all have a frequency of 0 Hz and thus they will be ordered not by frequency but by degree of freedom (UX, UY, UZ, ROTX, ROTY, ROTZ).

Regards

 

[GBor]

I would suggest a text on vibrations may be useful for you at this point. Rigid Body Modes are the result of infinite deflection along a particular degree of freedom, and are related to the constraints in your FEM. For instance, if you have a 6 degree of freedom beam element oriented perpendicular to your 5 degree of freedom plate element, then the beam will want to rotate about its longitudinal axis and the plate won't be able to stop it (no constraint from the plate element). If you tie a couple of beam elements to the end of your original beam, but orient it in the plane of your plate element, the new beams will constrain the rotation of your original beam and eliminate the RBM.

The significance of an RBM depends entirely on your modeling intent. If you have a static object with no moving parts, an RBM will tell you that you haven't properly constrained your model. If you have a rotating shaft on a piece of machinery and your FEM calculates 1 RBM, you may have some confidence that the model is performing as expected.

I don't think most FE programs necessarily filter RBM, but they do allow you to idenify how many you expect. They will also usually identify excessive deflections and that your model isn't tied down (constrained) sufficiently.

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group

 

[GregLocock]

In the funny old world of vehicle dynamics we worry long and hard about the sequency of rigid body modes.

For instance, when we mount an engine into the car, the frequencies of the bounce pitch and roll modes are fundamental to the performance of the system.

Note that just because they are rigid body modes they are not necessarily zero frequency - it is well worth differentiating between the two phenomena.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GBor]

Greg,

I'm not familiar with vehicle dynamics, but what you are describing in "bounce pitch and roll" doesn't sound like what I think of as a classic "rigid body mode". It sounds more like the motion of a rigid body on an elastic foundation...a single degree or multidegree of freedom system?

I think of a "rigid body mode" as the result of zero restraining stiffness, so the body is able to deflect infinitely. Since, for a single DOF system, omega = SQRT(k/m) and k (stiffness) is zero, omega would be zero. Am I missing something from vibrations?

I'd have to think about what this means in a multi DOF system with a forcing function. Guess I've been away from the books too long!

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group

 

[johnhors]

Greg,

I think you have muddied the waters just a bit ! In your "funny old world of vehicle dynamics" the term rigid body mode must mean something different to what the rest of the world thinks it means (as defined by Garland)!

 

[enthu1980]

I got the answer similar to what Greg had said by an Analyst working in an automobile company.

 

[enthu1980]

Hello Greg,

Can you please explain, if possible, somewhat about "the frequencies of the bounce pitch and roll modes" so that we can have better understanding about topic,

 

[GregLocock]

I think it is just a terminology thing.

If you analyse an engine then the first 6 modes are rigid body modes, in which the engine itself is rigid, but it moves in various ways on the engine mounts, typically at frequencies between 6 and 35 Hz.

At a frequency of (say) 80 Hz we get the first flexural mode of the engine, where it bends in its own right.

So in our terminology a RBM is merely a non-flexural mode whereas a zero frequency mode is an unconstrained motion.

The same terminology is used with smaller components on the engine, so the alternator, for example, will have 6 rigid body modes where its mounting brackets are flexing, but the alternator itself is just a mass, and then at say 1000- 1500Hz it will have its first flexural mode, where the casing starts to flex.

Typically the alternator's rbms are important for engine noise, and the flexural modes are responsible for whines and so on.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[cbrn]

Hi,
Greg, it seems to me that in automotive (which is definitely not my field) you use a very specific terminology which is not adherent with the common definition of "rigid body motion". It's not a problem, it's just a matter of terminology as you say. When you say that the first freqs of an engine are "rbm", you don't want to say that the eigenvector is null for a particular dof, rather that "the engine moves as an infinitely-rigid body upon an elastic foundation", as Garland said.
Garland, I fully agree with your remarks. I also used a terminology which could sound unclear: the "significance" of a rbm in a modal analysis is null in that sense that it is not a vibration; nevertheless, the information given by a rbm CAN be extremely significant! your definition of rbm, by the way, exactly matches mine in mathematical terms: when you solve the eigenproblem, one trivial solution of the system is to have null eigenvector (so that anything multiplied by zero equals to zero by definition!!!), but of course this is not a vibrational shape.
Btw: a solver of Cosmos for example automatically recognizes RBMs and doesn't display them, so for a free/free beam the first shown eigenvalue (and eigenform) is in reality the "seventh" (the first 6 being RBMs for the fundamental 6 DOFs of the system). On the countrary, Ansys treats the RBMs as "real" eigenvectors.

Regards

 

[GregLocock]

Well, just to add to the confusion, we certainly have zero frequency RBMs on a car with a definite mode shape. For instance, the car on its suspension has (ideally) yaw pitch roll, bounce and lateral modes. These have positive frequencies It also has a rolling along on the wheels mode, which is zero frequency, but has a definite direction, and residual.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[cbrn]

Hi,
Greg, this is interesting and, I think, not really surprising: if we consider the whole car "turning around" the wheel axis, then this is a RBM in the common definition of this term: ROTZ (being "Z" the wheel axis) is the DOF, the eigenform is a constant unit for any point of the car (or the entire car mass value if we normalize to mass, or also a constant zero because this makes no difference, 'cause the value of the eigenvector only states the initial position w.r.t. the axes), and the frequency is zero (i.e. it's a constant movement, not a "true" vibration). It's the same thing I encounter in my world made of turbine rotors, where the first torsional eigenform is the null one around the machine axis (same thing also pointed out by Garland).
Am I correctly interpreting all this?

Regards

 

[johnhors]

..... and if driver of the car uses a humpback bridge to get airborne and simultaneously puts the gear box into neutral, you have then at least ten rigid body modes (6 for the car as a whole 3 translations + 3 rotations plus 1 for each wheel plus possibly several more if you include the drive train, but that's probably going a bit too far!)