Learning Forum

[matt]

In Mechanical, I'm running a modal analysis followed by a spectrum analysis.  With command snippets, I'm doing a manual comparison of FSUM data at a location on a model.  I'm manually extracting FSUM data from the modal analysis for every mode at a location.  I then manually extract the modal coefficients from the spectrum analysis using *GET.  With some basic APDL math scripting, I manually multiply the modal FSUM data of each mode, by its corresponding modal coefficient.  Then in the last step, I SRSS all of those values to get the actual FSUM data for the location.  It matches what I get if I just issue FSUM from the spectrum analysis.  In this scenario, the spectrum analysis uses the SRSS method.

Next, I want to redo the manual verification of the FSUM data using CQC method instead.  But coding the basic math of CQC is more complicated.  I would like to know if anyone has coded the CQC math before.

[Dlooman]

I doubt it, partly because there doesn't seem to be any reason to.  If you are trying to create a verification test, just base it on two modes so the combination of the fsum values will be easy to do with a hand calc.

[matt]

There is, in fact, a reason to do the CQC calculation in the current project that I'm into.  I didn't want to go into all of those details here though since it would confuse the reader(s) more.  

[Dlooman]

OK.  It's going to be hard to automate because the cross-terms aren't "*gettable".  

[matt]

I can retrieve the mode coefficients with *GET in the spectrum solution.  That allows me to compute R for every mode.  Are you saying that I cannot compute epsilon (image below)?  I assumed if Xi (damping) was known and constant and the lowercase r values were known (i.e., the mode frequencies), a hand calc would be possible.