Time dependent transient deformed shape

[ZegersSander]

Hi,

I'm performing some modal analysis. I use the mode superposition method. I can retrieve the results for a single node, but I want to watch the deformed shape over time. The model is damped, so it should vibrate a bit before comming to a standstill.

How is this to be achieved in Ansys?

 

[cbrn]

Hi,
provided that you already performed the expansion of results, or that you have included it in the solution phase, then you should be able to achieve what you want by Animating Over Time (or Over Timesteps).

Regards

 

[ZegersSander]

Thank you for your reply.

I've found this function you mention. I've expanded the first five modes. But when issuing the command, Ansys asks for a mode number to be issued with the set command?.

It would be helpfull if there was an example of how to do this. I tried searching the verification manual, but it didn't turn up anything so far.

Is it correct that this has to be done in /POST1 instead of /POST26 which I would expect it to be?

Any further help will be greatly appreciated

 

[cbrn]

Hi,
sorry, I realize that I may have mixed up things... In fact, animating over time is meaningful in transient analysis. In modal analysis, the deformed shapes you see are the eigenforms of the solution. They remain constant though their amplitudes decrease with time, if you have damping. In this case, the animation gets another meaning... And I can understand why ANSYS wants to know which mode you want to animate.
Remember that in ANSYS, in a modal analysis each mode is stored in a set, that's why you use the *SET command to select the mode (or the equivalent GUI path).
Anyway, yes, all this pertains to /POST1 (because you don't deal with a "time history" in the strict meaning of the term).
Hope that gives you some idea...

Regards

 

[cbrn]

Hi,
disregard my previous post. I should have drunk more coffee this morning...
Mode superposition is for calculating transient responses starting from previously-calculated natural modes. So you're OK.
As for how to correctly handle the results from a modal superposition, I think the right place to look at is in the Transient Analysis part of the Help.
I'm sorry I can't help you precisely, because the transient analyses I've made were all solved with the direct method.

Regards

 

[ZegersSander]

Hi,

Thanks for the replies.

I'm getting more and more familiar with all the issues. Maybe it's just not possible?

I can review the time-history of the superposition modes in Post26, per node. Ansys can draw a graph there where the displacement is plotted against the time.

It's also possible to export all the time data to a .csv file. This I can read into Excel, and plot the data for a specific point in time.

This all has to be done with the .rsdp file.

However I cannot plot a deformed shape from this file in Ansys. Ansys tells me that I need to set a mode number. This has to be done from the .rst file. This file doesn't contain the time data, only the mode-shape data. Once I load this file, I can only choose the 5 modes I expanded, so it's frequency based, not time based.

The helpfile is of not much help up to now. I can calculate it alright, but not display the results.

So maybe it's not possible from the mode-superposition method, but I find this hard to believe as all the data is there.

Any further help is greatly appreciated.

 

[doatis]

I'm sorry if I'm repeating stuff that's already been said before...

When you were looking at your results in /post1, did you expand your mode-superposition results? You need to do this in order to view results in /post1.

There is an example of a transient analysis using mode-superposition in the ANSYS help manual.

Structural Guide > Chapter 5. Transient Dynamic Analysis

Section 5.4 Performing a Mode-Superposition

At the bottom of the page there is an example script showing the basic APDL to do all of this.

Hope this helps,
Doug

 

[cbrn]

Hi,
another hint: the mode superposition may have been stored under the "999999" set of the results file.

Regards

 

[ZegersSander]

I'm happy to report that I've got it working.

It appears I didn't read the manual close enough and the replies here made me do it better. It turned out I needed to expand the results, which I thought I did, but did at the wrong point. I expanded the results before the modesuperposition calculation. At that point there were only the modeshapes available.

Sorry to the first couple of reply-ers that said so, but I didn't read carefully enough.

Now to get more than 40 frames worth of animation.......