How-do-i-c​onvert-tim​e-domain-d​ata-into-f​requency-d​omain of an mems accelerometer ?

I have to get modal parameters as output. #excluding fft, stft, packet wavelet transfom

[Walter Roberson]

To check, you cannot use fft or wavelet or stft in the solution? If so then is it that the assignment does not permit you to call upon those routines but you could program them yourself? Or are you not permitted to use those techniques for some reason, such as to make you be creative about other approaches?

[Anshul Sharma]

@Walter_Roberson Sir,Thanks for the reply. There has been already work done using those approaches on same problem so I hereby intend to go for some other different methods to solve the issue. If you kindly suggest some methods, that would be helpful.

[Walter Roberson]

piezoresistive or variable capacitance mems? What model parameters are to be solved for?

[Anshul Sharma]

The sensors used were dual channel microelectromechanical systems (MEMS) accelerometers having natural frequency of 150 Hz and sensitivity of 1 V/g. Accelerometers provide voltage output, which is proportional to the acceleration of the points of placement.

Modal parameters will be- frequencies,mode shapes,modal energy,damping ratios. of the structure.

[Walter Roberson]

How many of these two axis devices are being used on the same structure? If multiple are being used are they all aligned on the same plane or are some of them rotated to give three dimensional data?

To confirm, there are multiple such sensors on the structure and you wish to analyze the movement characteristics of the structure? Movement in a plane? Movement in 3d? Movement for each point in the structure since different points could be reacting differently?

What else is known? It is not easy to talk about measuring reactions if you do not know what you are reacting to.

Like is this a test rig setup where you have a number of different runs where you start with the object at rest in a known position and you know the characteristics of the signal to be applied for a time and you know where it is being applied, and you take measurements over these conditions and later try to derive the parameters you mention? Or is this a situation like handling a kid an odd shaped object and the kud is told to go play with it, and the task is to figure out all about how the various parts are connected together and their elasticity and spring constants and so on just by looking at the measurements but without information about exactly what the kud was doing at the time?

[Anshul Sharma]

@Walter_Roberson

Sir, In order to answer your queries I am attaching url of a rasearch article from which I hope many of them will be clarified.This is the same problem I wish to work on. https://link.springer.com/article/10.1186/2008-6695-5-25

Answer 1

[Abhishek Ballaney]

https://in.mathworks.com/help/matlab/ref/fft.html

[Anshul Sharma]

@Abhishek_Ballaney Sir,Thanks for reply. Can u suggest some other approach excluding the above mentioned,that would be much helpful.