It would certainly be possible to design something that has a higher cutoff than the labhorn. Eg F3 40hz.
What are your requirements?
Here is one already.
http://www.billfitzmaurice.com…5e15c0fce66dc3c1970839bdf
Beiträge von mike.e
-
-
Zitat von "Walt"Alles anzeigen
Hello,
When measuring a real system one usually uses the FFT algorithm to get the frequency response. But your model is not a real system. It will not be faster to do it this way because the computer would have to simulate the whole system en then use the FFT algorithm. But by simulating the system the computer has already solved the equations of motion.
Try to simulate by octave. Use much frequency points in the low frequencies and less in the higher frequencies. Currently I am doing a dynamic analysis on a container crane with the use of Ansys software. A harmonic calculation on 25 frequencies takes about 5 minutes on a good pc. This is with about 400 elements. You could try do reduce the number of elements.
Also there are some techniques to reduce computation time but in your case this will be difficult.
Best regards,
Walt
Perhaps for extensive trial and error computation,LINUX based parallel computing could prove useful? -
-
-
-
Very good- I was waiting to see who would do this. What about 200hz?Im wondering what is the cutoff at the upper end.My pc does not enjoy 5mb gif files! :?
Mike.e -