Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Define System Stiffness From Vibration Data 1

Status
Not open for further replies.

jkapcoe

Mechanical
Oct 30, 2003
2
0
0
US
Hello,

I do not know that much about modal analysis results, but I was hoping that someone could help me out.

I have some data showing a components deflection under engine operating rpm. What I would like to do is to have a simple calculation which can give me an idea of the components stiffness in that area based on this data. Does anyone know how to do this?

Thank you in advance!
John
 
Replies continue below

Recommended for you

Dear Sir (Greg Locock),
I am having some doubts. If we have the deflections for the particular force then we can obtain stiffness. Which is that stiffness. in many FEM analysis we use to calculate the stiffness by this method only.
Please guide me !
regards
jo'
 
That's right. In an FEA you know much more about the system than you do in an ODS. An ODS does not have a measured force input, it is just an animated display of the deflections in the system due to real time loading. I got the impression that the original poster had a picture of a deflected shape but did not know what the excitation was.

If you have the results from an experimental modal analysis (or an FEA) then the FRF at point B for the point impedance is Hbb which can be found from the point impedance of the excitation point A and the transfer function between them.

As best I can remember Hbb=Hab*Hab/Haa

using complex maths throughout

This slightly surprising result is due to reciprocity (Hab=Hba). In practice it is rather dangerous, as the errors get out of hand very quickly.


Cheers

Greg Locock
 
If what you want to do is get an approximate analytical model there are some system id techniques which attempt to do this by taking input and output data from an experimental model and build up an state-space model of the system. You can then use this to look at system responses for excitations which are different from your original data. Look for a book/articles by Jernan Juang for examples of this kind of modelling. The system id toolbox in Matlab does this as well.
 
Status
Not open for further replies.
Back
Top