Issue with the PDE toolbox (vibration modes of a plate with rods)

Hello,

I seeking to simulate the mode of vibration of a plate topped with rods (cf figure)

I used the PDE toolbox and the following commands :

model = createpde('structural', 'modal-solid');

importGeometry(model,'plate.stl');

E = 210E9;
nu = 0.3;
rho = 8000;
structuralProperties(model,'YoungsModulus',E, 'PoissonsRatio',nu, 'MassDensity',rho);
                       
generateMesh(model)
pdeplot3D(model)
RF = solve(model,'FrequencyRange',[70,1000]*2*pi)

However I get way too much vibration modes and frequencies in the NaturalFrequencies variable

What have I done wrong ?

Thanks !

[Jan]

@Terry Jin: In this forum it is not liked, if somebody bumps a question without providing additional information. If you have a good reason, wait at least more than 24 hours. Thanks.

I've cleaned the comment and the flag.

Getting no answer is usually a strong hint, that the question contains too few information to be answered. You do not mention, why you assume this: "I get way too much vibration modes". Maybe only your assumption is the problem.

[Terry Jin]

Because I calculated the frequencies of the flexion and compression modes and there is not 20563 different frequencies between 70 and 1000 Hz... So either the calculations are very wrong either I made a mistake in the code, which I'm trying to understand.

What kind of informations should I add ?

Thanks for the answer and sorry for the flag.

[Jan]

Sorry, I'm not working in this field. Looking in the documentation does not reveal, which units are used for the MassDensity and the YoungsModulus, but the values E = 210E9; rho = 8000; are lloking really strange for a physicist. Is the material extremely stiff and dense?

[Terry Jin]

The Young Modulus is in Pascals. For steel E ≈ 200 GPa = 200E9 Pa. And the density of steel is 8000 . The SI units are used, and I don't think that the problem comes from here. Maybe from the geometry ?

[Terry Jin]

Actually, I'm not sure that I'm using this tool correctly. What I'm trying to do is to simulate an experiement that I did to compare the frequencies of flexion and compression. The idea of the experiment was to have a plate with rods (cf figure) and to emit certain frequencies on the left side of the rods to record that signal on the right of the rods. (Then a fast Fourier transform is used to compute the spectrum).

Do you think it's possible to do the same thing numerically on matlab (finie elements method) ?

[Terry Jin]

Something like that but a simpler version

[Terry Jin]

I'm getting started with the MATLAB PDE toolbox. I'm not sure of what I can do with it.

Answer 1

[Terry Jin]

No answer on this forum.

[Torsten]

Too specific for the forum - maybe a question for experts with physical background knowledge from the support.