Modal Analysis

[politeknik]

Could anyone explain to me, what is the crucial steps in developing modal analysis?

 

[GBor]

Your question is VERY broad. Are you looking to perform the calculations by hand or FEA? Is your system SDOF or MDOF? Are you trying to solve for the fundamental mode or multiple modes? This list goes on. Please be more specific.

 

[politeknik]

Dear sir,

First of all, thanks for your response.
Actually, i'm intended to develop experimental modal analysis of generator's end windings (Both DE & NDE sides).Bump tests already been done & the result shows that there are natural frequencies appeared on 95 - 115 Hz (area of interest: 2x of operating speed). Since this my attempt, hopefully you can guide me precisely....

thanks in advance

 

[GBor]

I am not that familiar with generators and am, therefore, unable to advise about idealization of the end windings.

Typically, the engineer will simplify the system to something that is idealized as a series of masses and springs. For a single degree of freedom system, this is a single mass and a spring. The spring stiffness is determined depending on the system configuration (for instance, a beam stiffness equivalent would be AE/L where A is the cross-sectional area, E is the modulus, and L is the length of the beam). The fundamental frequency is then calculated as the square root of k/m where k is the stiffness and m is the mass.

More complicated systems are more difficult, but all of this is covered in a basic vibrations course. Virtually any book on vibrations would have to cover this simplification methodology.

Someone more familiar with your particular application may be able to provide more specific examples.

 

[Twoballcane]

What are you looking for from a modal analysis? It is tough without test equipment and FEA software. Unless it is a simple shape like a flat plate or rod, anything else will require accelerometers or FEA.

Tobalcane
"If you avoid failure, you also avoid success."

 

[GregLocock]

"you can guide me precisely"

Well, you are in luck. B&K have published a lot on the subject. I suggest you read the vibration and the two structural testing ones to start with, then search the rest of their site for the many articles and so on on modal analysis.

http://www.bksv.com/Library/Primers.aspx

Cheers

Greg Locock

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

 

[politeknik]

Actually, i'm trying to predict the system response & structural dynamic behavior by modal analysis. All measurements(FRF & its phase) already been taken using CSI 2310 with good quality control of the coherence.In addition, i also have no idea what is the mode shape & how to identify it?

 

[GregLocock]

How many measurement points do you have? Is it in all 3 directions?

Cheers

Greg Locock

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

 

[politeknik]

Yes, i'm using triaxial probe for measurement. All data is in all 3 directions. Any advice?

 

[politeknik]

End winding was divided into 17 quadrants where the measurement points were at its inner and cap.

 

[Twoballcane]

You can input that data into ME Scope (

http://www.vibetech.com/go.cfm/en-us/content/mescope)

and get some real good pics of the modes. The other is to use FEA. I use ProM becuase it is easier to run a Modal Analysis.

Tobalcane
"If you avoid failure, you also avoid success."

 

[GregLocock]

There's three parts to the problem.

Analysing the data, and displaying the mode shapes, and then doing some calculations based on the FRFs for whatever it is you want to do with them.

Strictly speaking it sounds to me that you may be able to jump to the third part straight away by using SMURF (structural modification using response functions), receptance theory, or from first principles.





Cheers

Greg Locock

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

 

[GregLocock]

http://www.vibetech.com/go.cfm/en-us/content/techpapers-list#

seems to have many articles that are directly relevant.

Nice link 2BC



Cheers

Greg Locock

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

 

[politeknik]

Dear Mr Greg Locock,

Could you please go in detail about the mode shapes & how to identify from FRF spectrum?

Thanks

 

[GregLocock]

No, not until you have read and understood all the literature that you have been pointed at.

Cheers

Greg Locock

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

 

[politeknik]

Sir Greg,

To obtain mode shape of a structure, a minimum set of FRF measurements must be taken either a single fixed input and many outputs or between a single fixed output and many inputs.Why not using many inputs and many outputs?

 

[GregLocock]

You can, see MIMO analysis.

The problem is that although the resulting model is more robust, it is also noisier.

A SIMO model linearises several sources of non linearity, and is easy to analyse channel by channel. A MIMO analysis is based on partial coherence, and the first step is to look at how to break the signals of interest up.

There's nothing really wrong with MIMO apart from complexity and time, in fact where you have large non linear structures it is likely to prove more reliable and less frustrating. For instance on a whole car, if you take two complete SIMO surveys using different excitation points you can spend weeks trying to reconcile the two resulting modal models. That's why we don't do that.





Cheers

Greg Locock

I rarely exceed 1.79 x 10^12 furlongs per fortnight