Hi Pavan

For the basics, take a look here

http://qringtech.com/learnmore/recommended-reading/

For OpenSource EMA software, take a look here

http://www.openmodal.com/

http://hosting.umons.ac.be/html/mecara/EasyMod/index.html

http://www.noisestructure.com/products/products.php

Hope this helps

Claes

In addition to the software mentioned by Claes, you will also need two accelerometers. One at a fixed location and the other to move over the object. Using a two-channel FFT analyser, frequency response functions (FRF) are obtained from the fixed and roving accelerometers. These data are used by the relevant software to generate the mode shapes, frequencies and associated damping. It sounds complicated but is relatively straightforward in practice.

Hi Eric

I believe you forgot the hammer?

EMA usually uses force as reference, i.e. the minimum set up is 1x accelerometer (response) and 1x force instrumented impulse hammer.

With one accelerometer as reference, you otherwise get an ODS or Operating Modal Analysis.

I see from your profile that you deal with large scale structure so I am guessing OMA is your regular choice, while EMA is the more common for smaller scale objects.

Sincerely

Claes

with of help of accelerator or fft analyzer can give you natural frequency

It might be useful for you to consider a simpler approach (simpler than those mentioned in previous responses) if the following conditions apply:

That simpler approach is described at:

https://www.researchgate.net/publication/304827861_MATLABR_Curve-fitting_for_Estimation_of_Structural_Dynamic_Parameters

Claes,

You are correct. I did leave out the hammer. However, it is still possible to do true EMA using the two accelerometer technique with an external excitation source such as a shaker.

Regards,

Eric

Hi Eric

You can even to it with a single accelerometer when using Cepstrum as demonstrated by Bob Randall, e.g.

https://link.springer.com/chapter/10.1007/978-1-4614-2425-3_18

Conference Paper New cepstral techniques for operational modal analysis

Out of curiosity - how do you get the mass scaling for the modes without force? I mean, OMA gives you natural frequency, damping and mode shape - but it is not scaled with respect to system weight.

That said, you need only scaled modes when using them in combination with a simulation model.

I find that OMA (SSI) tends to provide more consistent results than LSCF (or LSCE).

In the days of old, one measured the shaker current and used it to compute the driving force. In my book, this would simply be a version of force reference EMA albeit without the force sensor.

I am guessing you do something more clever? Feel free to provide a suitable reference. Also, would you expect it to work for in operation vibration of, say, a motor or a case with random excitation?

Sincerely

Claes

Some very good information is available from - Ewins D.J., “Modal Testing: Theory, Practice & Application (2nd Edition)”, Research Studies Press, Hertfordshire, UK, 1999. There may be later editions but mine is from 1999.

Natural frequency can be found with a single accelerometer and a hammer/shaker. But to know the mode shape, one accelerometer is not enough. Two or more sensors may be required to find the elastic bending modes/shapes

Hi What is the key difference between commercial Software and open modal?

Hi, you could simply try hammer testing. Depending on what sensors and tools you have at your disposal, you could go about it in several ways.

Hi Pavan,

Besides all of the mentioned methods, you can use the signal processing technique. Accordingly, first of all, you must be recorded the signals of the structural response via the recorder sensors against ambient vibration or any seismic source/generator. Then, you can implement the operational modal analysis on the obtained experimental vibration data. Finally, you will achieve rather than the inherent characteristics of your drill body.

Good Luck, Milad.

Is there a major difference between commercial and open source python-based software? Shall I replace my commercial software with the openmodal software code? My requirements are FRF, FFT, Modal analysis, ODS, etc.

- Let use FFT function, let write your own code.

- From time history data, natural frequencies can be obtained and the corresponding mode shapes are determined using the imaginary parts from FFT.