Swept sine dynamic response

[corus]

To assess the dynamic response of a structure in a particular direction and to determine the participation of each mode in the response it was suggested that a swept sine method be used for frequencies within a certain range.
What force/displacement function is input to get this response from a FE model?

 

[EdR]

corus:

You will probably have to generate the forcing function based on the frequency content desired...We used to do this from specified shock spectra....essentially generated a time history (usually velocity) that had the same frequency content as the specified spectra...

Next question is how.....I don't remember (its been much too long since I did this) so can't help much with that....maybe check a vibrations book

I think the forcing function could be either a force time history or a velocity time history...

Maybe that will point you in some direction...


Ed.R.

 

[brad]

Corus,
It depends on what the physical test is. I admit that my understanding of swept sine is fairly limited, but my understanding of it is that simply devolves to a steady state dynamics analysis. The critical question is how the test was run (applied force, applied acceleration, etc).

For a given applied force in test, my understanding is that the equivalent approach in FEA would be to do a constant-force Steady State Dynamics run, and extract the relevant output.

Don't take my response as authoritative, as it is based on a loose understanding of "swept sine testing". If I know the physical test that you are performing, I can probably give a more informed response as to the FEA equivalent.

Cheers,
Brad

 

[bjpil]

Corus,

What I have done in the past is to do a frequency response analysis with constant acceleration (1 - unity) in one direction over a frequency range (1-3000Hz). This mimics a transmissibility test (A out/A in). From the results I will have a better understanding what to adjust to reduce resonance. You could also use force, displacement or velocity.

BJP

 

[corus]

The problem is one in which there are a number of modes of vibration but none that explicitly give a displacement in the required direction. In order to see which modes had the most effect in that direction someone had suggested a swept sine response analysis, although my impression is that they'd heard the phrase and it sounded good, without knowing what it meant. No physical testing will be done of the part at this stage
bjpil's comment sounds good in that it sounds like a spectrum response analysis with acceleration against frequency, and is a steady state analysis. I'm not sure if that is what brad is referring to by a steady state constant force dynamics analysis. Presumably a constant force implies a constant acceleration as bjpil suggests?
Either way this sounds the easiest way of obtaining the required results.
Many thanks

corus

http://www.corusresearch.com