Showing results for 
Search instead for 
Did you mean: 

A kind of strange result in Response Analysis


Hi All,


I was trying to do Frequency Response analysis in NX Nastran. Incidentally I compared the results of the same model, that I made in another software. So I used the same FEM file for Freq Response analysis in two diffirent tools, one being NX.


Though the magnitude is different, the other software is not showing any load excitation at the first natural frequency which has a mass participation of 63% in the desired direction. Does anyone can guess why this is happening? Both have the same damping factor too.






Re: A kind of strange result in Response Analysis


The top graph looks pretty darn crude, you're missing a lot of modal content compared to NX Nastran...  I wonder what piece of software that is...


That being said, the 5Hz is entirely different, is it even the same acceleration component and excitation direction you're at?  Can you share a model? 

Re: A kind of strange result in Response Analysis



my observations:

1. on the upper graph there is different order of magnitude for the acceleration than for the lower graph, so first problem may be that the unit system in upper software is not consistent (?)

2. At the first glance it looks that the results from NX have more steps through the frequency. Maybe because of that some of the responses on the upper graph were omitted. Note that they have been calculated for only few points with a step of 5Hz. Compare the values for exact (not approximated!) frequencies and check whether they are identical.

Try to set up smaller steps for 'upper solver'. Most probably default frequency step in NX nastran is smaller (or number of steps is higher) for chosen range of frequencies.


I am curious whether after that the results are convergent...