Another question about 'Creating Field Output by Operating on Frames' 1

[BW94]

Hi Guys,

Yesterday I asked a question about the 'operating on frames' output option. This option would give me the highest value of a particular output variable somewhere during the simulation for each integration point. This all seemed to work the way I expected it to work, but now that I want to request the lowest value of a particular output variable, something strange seems to happen;
When i request the minimum values of max. principal strain over all frames (including the 0 increment), the values are not zero everywhere, although all strains should be zero in the 0 increment, because nothing is happening in that step.

Is there someone who could explain this?

So in short my question is: why is the value LE_min not equal to zero, even when this value is zero in at least one frame that is included in this output variable?

Thanks in advance!

 

[BW94]

To comment on this, also the highest value for max. principal strain that I find in the newly created output variable is higher than the highest value that I can find when I skip through all the frames manually.

Anyone with suggestions or explanations?

 

[FEA way]

Strain results are minimum not when they are 0 (at 0 increment) but when they are below 0 (negative) at some point of the analysis.

 

[BW94]

Thanks for your response.

The problem is that the 'minimum values' I obtain with the 'field output from frames' option are higher than 0. See attached picture. Another simulation, where I requested the 'maximum values' with the 'field output from frames' option, also gives me values which are higher than the highest value I can find when skipping trough the frames manually.

any suggestions?

 

[FEA way]

This might be an error actually (they are still found from time to time in Abaqus). Especially that it happens only with strains.

Just to be 100% sure - check if you selected all frames, including the 0 increment one, when defining field ouput with From Frames option.

Or maybe it’s just showing the absolute value. Unfortunately nothing about it can be found in the documentation.

 

[BW94]

I looked into the documentation already and indeed could not find anything useful about it.

Especially the 'highest value' of max. principal strain in my other simulations seems strange, because it is higher than the strains that would occur in that particular node during the simulation when looking into the XY-data. See attached pictures. The highest strain in that node according to the 'create data from frames' is >58%, while the strain in that same particular node does not even exceed 27.5% in the XY data from the original output (so not from the 'create date from frames'-output).

 

[FEA way]

Actually the difference may be caused by averaging.

 

[BW94]

Hi FEA way,

could you explain how that is possible and which one (the normal XY data or the 'output from frames' data) is likely to be more reliable? I do not really understand why there could be averaging differences since I am looking in the exact same node.

thanks in advance!

kind regards,
BW

 

[FEA way]

When you look at Field Output contour plots they have averaging turned on by default (75%). Turning it off will change your results to more realistic and maximum value will not only have different value but probably also will be located in another place. This is especially important in case of such single-element concentrations like in your case. When you create XY data from field output make sure it has no averaging too.

 

[BW94]

Change the Avg% to either 0% or 100% does not alter the huge difference between the original XY-data and the 'output from frames data', so I really don't know how to explain that difference.

Just for my confirmation: the output in integration points is the most reliable output possible right?

 

[FEA way]

It depends on the case, sometimes averaging causes large differences.

Yes, stresses and strains are calculated at Gauss points and then extrapolated to nodes so integration point output ahould be the most accuarate for them.

 

[BW94]

Thanks a lot!