Wanted to perform FRF with test load (freq vs magnitude) as input

Hello All,

I wanted to perform FRF with test input which are in Freq (Hz) vs magnitude (G2/Hz). I added test input as curve and assigned it through RLOAD1 Tc option. Rest of parameters are same as FRF. Now analysis is done but stresses are very high like 2000 MPa which are doubtful as mounting part is sufficiently thick.

I want to check if my process is correct. If not, then what is correct process.

With my current approach stress are very high so which are the areas i have to check to get stresses in plausible range.

Note: While adding magnitude (G2/Hz) i have multiplied these values with 9810

Request your support. Thank you.

Find more posts tagged with

English

OptiStruct

v2022.2

Accepted answers

All comments

Adriano_Koga

this might be useful

https://insider.altairhyperworks.com/wp-content/uploads/2018/01/T-T-1285-OptiStruct-Units-to-be-maintained-for-Random-response-analysis.pdf

if your PSD is in G2/Hz, then your unit load should be 1.0 G, and them your output will be G for RMS acceleration.

N_Pat

Hello Adriano,

Thanks for input. So with consistent unit is it possible to implement PSD input in FRF directly without RANDPS card and TABRND1?

Xiaolei

Hello Adriano,

Thanks for input. So with consistent unit is it possible to implement PSD input in FRF directly without RANDPS card and TABRND1?

Random analysis can be considered as an extension of FRF analysis. Let's say you have a single PSD input, then PSDouput = (FRF^2) * PSDinput.

Theoretically it's possible to factor everything of PSD into FRF analysis input, but I don't think it's a good idea. Firstly, setting up a PSD analysis is relatively straightforward and can be easily performed using OptiStruct. Secondly, a PSD analysis provides additional insights, such as the Root Mean Square (RMS) output, which is directly available in PSD analysis but not in FRF analysis.