Hi Braj,

Where did you come by this statement? If we would be considering a free structure, the first two modes would be rigid body modes and thus, would tell you little about dynamics.

Hence, the dynamic behaviour of a structure can in general not be understood by known the first two eigenvectors/mode shapes. It highly depends on the excitation of the structure, i.e., which frequency range is contained in the exciting signal. Therefore, there is no general rule on how many modes are sufficient. This must be judged on a case by case basis.

To find out which modes are dominant in a certain transient deformation case, you can compute the modal participation factors. But to do this you must have an idea of the real deformation that will arise under operational conditions.

This is definitely a topic you can discuss in more details with Elmar Woschke

and his group.

Cheers,

Sascha

Dear Samy,

while I agree to the fact that the modal participation factors are a decisive, I have to add that neither symmetry of a structure nor linear elasticity have mich impact on the fact of two modes shapes are sufficient to describe the dynamic behaviour. Just think about high-frequency dynamics and wave propagation. In these cases two modes shapes are not nearly enough to describe the dynamics. I would say that two mode shapes are almost never enough. These are very rare cases where this would hold. So I would recommend to forget about two mode shapes and use the modal participation factors to judge the validity.

I must reiterate at this point that the external excitation is the key factor determining the frequency range, which in turn determines the modes that are excited.

Kind regards

Sascha

Very nice discussions about this subject.