Multiple subcases

[Anonymous]

I'm trying to produce random vibe results for X, Y, and Z directions.  I set up three subcases for the 3 inputs but I can't figure out if they were all run together or separately.  If this is correct, how do I load results for each subcase?  If not, then how do I properly set this up for 3 separate inputs and outputs?

[Andrew.Sartorelli]

Hi Carlson,

Each subcase will run independently, at the end an RMS result will be calculated that combines all of the subcases. When review results you should see multiple sets within the same FNO file, each one corresponds to the different subcase. I hope this helps.

Andrew

[Anonymous]

It does and doesn't help.

1st, in the "Subcases" dropdown under Results, there is no mention of which Subcase you are retrieving results for. Have to keep track of 1st, 2nd, and 3rd. Would be nice to have these labeled and that to be included in the displayed results. Also, all 3 inputs are shown on the plot for all 3 results sets. Each should only show its coresponding input. Below.

2nd, if I look at all 3 cases X, Y, Z at 767, I have roughly 1492 psi, 10 psi, 10 psi. But if I pick PSD at 767 Hz there is a magnitude of 3.7 e5 psi. What are PSD results? There is no 4th load case. Are they simply peak accelerations in in/sec2?

I've attached the model. Not sure if it will get through at 30 Mb.

thanks,
Lee

Below image is showing results for X axis input although it displays all 3 subcases.

[Andrew.Sartorelli]

Hi  Carlson,

Sorry about the late response here, things always get a little crazy around the Holidays. I certain agree that we could do a better job with the post-processing there for a random response analysis. I'll see about generating a enhancement request based on your feedback.

Regarding the PSD result, I'm working on putting together some documentation for you it's just taking me a little bit of time.

Regards,

Andrew

[Anonymous]

Thank you for the response.

[Andrew.Sartorelli · Accepted solution]

Hi Carlson,

I apologize about the delay, but I think we've got the complete answer for you now!

If you'd like to have each PSD analyzed seperately, then you should create three seperate analyses instead of using subcases. You can duplicate an analysis by right-clicking and selecting Copy/Duplicate. The reason for doing this is there is only one RANDOM card written per an analysis, rather than one being written at a subcase level, this should also help alleviate some of the confusion with the results labeling.

The PSD output is the PSD times the transfer functions squared, with the transfer function being the frequency response result.

The RMS output is area under the curve of the PSD output plotted versus frequency. If you input your PSD in units of g^2/Hz, then you would need to multiply this by the gravitation constant.

I hope this helps, and again sorry about the long delay here. Please let me know if you have any further questions.

Regards,

Andrew
______________________________________________________________

If my post answers your question, please click the "Accept as Solution" button. This helps everyone find answers more quickly!

[Anonymous]

Unfortunately, the cable tension issue is preventing me from performing the necessary random vibration analysis. Until I can get tensioned elements there, any random vibe would be meaningless. Thank you for clearing up the subcases, but I really do not understand the purpose of "subcase" if you actually have to create separate analysis for each axis.

[Anonymous]

Was there documentation created?

The problem I run into is that I always have 3 axes to run.  This is done before we take our hardware to an actual shaker table where we shake all 3 axes.

Unless I have all the kinks worked out of my FEA model, creating 3 diffent simulations might start as simple as duplicating and changing the loads, but additional changes (point masses, materials, etc) require modifying all three analyses.

Also, when I set up 3 different analyses and want to switch between them to see the results, I have to resolve each time as only the last result set is available.  Hopefully, this can be fixed in the settings.

thanks,

Lee

[Andrew.Sartorelli]

Hi Lee,

With regard to the results only saving for the last set, is this occurring in 2016 or 2017? We made a change in 2017 so that every analysis will have a unique filename that is automatically generated. With 2016 you had the option to save the .NAS and .FNO based on the analysis name or the Inventor file name, which could produce some less than ideal behavior as you've mentioned.

Regards,
Andrew

[Anonymous]

2017 has taken care of this.  Thanks.

[Anonymous]

Fixed, but I'm still wasting a ton of time editing 3 analyses to run 3 load cases. I wanted to change some connectors but I have to do each change 3 times. Point mass, 3 times, etc.

[Anonymous]

Also, when I reopened the analysis, there were no plots where there once were freq vs mode number and mass participation plots. There are now none.

[Andrew.Sartorelli]

Hi Lee,

You will need to re-load the results once you reopen the file. This should then load in the FNO data as well as all the information for the XY plots.

Regards,
Andrew