Discussion Forum

March 12, 2010 8:24pm UTC in response to hattie liu

Re: how to analysis deformation of cantilever beam under a certain frequency???

Hi

there are a couple of ways, but first you must inderstand that if you have a beam vibrating, around any of its natural the resonance frequency, the amplitude is governed by the Q=1/(damping factor) * the excitation you define (or the "numerical damping" used to keep the solver stay stable).

If you do an eigenvalue analysis you do not define any excitation value, so COMSOL normalises the displacemenst, you can compare modes between each other but not in absolute values, the displacement are often expressed in meters, even for mm large parts, this is simply a normalisation issue, normally you should not look at the units for displacements in an eigenvalue simulaton.

If you do an harmonic analysis you must define an excitation i.e. F[N] force somewhere (I use typically 1[N] or 1[kN]) then I know that my values calculated are to be considered as "per [N], respectively per [kN]".

You can see the displacement if you ask for it (you must turn plot deformed "on" and select surface/boundary - displacement plots). The values observed are very strongly depenendent on the damping of your material or other physicsal phenomena you have defined. For harmonic sweep you define a "parameter" as freq and let it go/scan over the desired frequency range, just as for a parametric solving case.

Finally you can define, in transient mode, a time dependent Force excitation as a sinus function and do a time scan, transients are typically used for non harmonic excitations, but its possible to run a sinus, its simply less efficient than the harmonic analysis for pure sinus excitations.

Hope this helps you on the way
Good luck
Ivar

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse