mkoijn9
Mechanical
- Nov 3, 2007
- 10
Hi, I've just written (in Fortran) an implementation of the Polymax modal analysis algorithm, and I was wondering if anyone has a set of multireference frf's (with the correct results also) I could use to test my algorithm to see if it is working correctly as I don't have access to any comercial modal analysis software.
I have already synthesised a set of frf's(two references) 25 outputs with two modeshapes at exactly the same frequency but 90 degrees out of phase with each other. It identifies the modal frequencies, and that there are two of them at the same frequency, but the modal participation factors seem to be not correct, they indicate participation of both modes in all frfs(both references) but I arranged the data set to have one reference on the node line of one mode shape and vice versa, so one participation factor for one mode shape should should be showing a value and the other participation factor for the reference on that modes node line should be zero, and vice versa.
I also created a second data set with the two mode shapes slightly apart in frequency( a few hertz) and then it works as expected. Should the algorithm be able to detect modes with absolutely identical modal frequencies?
Any ideas?
I have already synthesised a set of frf's(two references) 25 outputs with two modeshapes at exactly the same frequency but 90 degrees out of phase with each other. It identifies the modal frequencies, and that there are two of them at the same frequency, but the modal participation factors seem to be not correct, they indicate participation of both modes in all frfs(both references) but I arranged the data set to have one reference on the node line of one mode shape and vice versa, so one participation factor for one mode shape should should be showing a value and the other participation factor for the reference on that modes node line should be zero, and vice versa.
I also created a second data set with the two mode shapes slightly apart in frequency( a few hertz) and then it works as expected. Should the algorithm be able to detect modes with absolutely identical modal frequencies?
Any ideas?