When post-processing results after an adv. non-linear plastic analysis, I encountered two issues:
The charting function did not allow me to plot Von Mises stress vs time step; I used the Vector vs Output set option but when trying to define the Vector values, I could only select element stress (not nodal) or nodal translation (not stress)... Note; the nodal stress results are there/exist - one can probe them on the contour plot. I realise there is probably a work-around by first defining a data series outside of the charting function which of course a bit more more tedious to follow.
When trying to do animation specifically using the Animate-MultiSet option, the result was that the program seems to do so correctly only for one (fraction of a) cycle, whereafter it then seems to automatically default back to what appears to be standard animation where the contour plot at the final time step is simply scaled down during animation. I.e. the contours corresponding to the initial requested time steps are not shown during anaimation. Note; I am keeping in mind that I requested output at every 4th step.
Just wondered whether I am missing some magic setting...? Thanks for any help.
Solved! Go to Solution.
Regarding output vector selection in charting... I assume you're trying to plot element corner stresses similar to what you would see if you used the entity selector with a contour plot of von Mises stress. Those corner/nodal results are not populated in the drop-down lists for output vector selection by default, as they're not commonly selected vectors and it helps keep the list smaller. If you would like to select those vectors:
Corner data will now appear in the drop-down list
This is going to be true for most drop down lists for output vectors that you encounter.
Regarding the issue with animation - could you provide a little more detail on what you're seeing or perhaps a sample model? I'm unable to reproduce that issue.
Thanks a lot for trying to help, Patrick.
I tried the element corner stress option before; at least this time around (with your help) I actually managed to see the different corner options (C1, C2, etc.) being listed! But, so far I could not find a simple way to know which (3D) element corner is which.... The location to be specified is still an element number. If I could directly select a node then there was of course no issue. Im interested in the result at the (surface) node (or corner then) because that is where the stress or strain is usually a maximum.
As far as the animation is concerned, I think I should just try and describe it differently because, the motion part of the animation seems perfectly normal - it is in the detail of the contours displayed at each animation step, where the problem lies. E.g. the correct contour is only displayed once/initially for say the first animation step, thereafter (as the animation repeats itself), the contour displayed together with that first animation step is wrong...(appears to simply scale down the contour associated with the final animation step). The stress or strain distribution, or point where it is maximum, should shift to different locations throughout the animation. The latter is just briefly visible as the animation is started - thereafter not anymore as the animation keeps repeating itself.
I meanwhile assumed the order in which the node numbers are listed per element (this can be done with relative ease), corresponds to the C-nr's associated with element corner results. So, if one first probe the contour plot at a particular node of concern, it is subsequently possible to then reference that node number against a listed element corner (i.e. to know which C-nr. to use for charting). A remaining issue is that I suspect that such an element corner stress (or strain) is not yet the averaged result (from other neighbouring elements) as would be the case if one were to use a nodal result. Granted, if the mesh is fine enough, the differences amongst corner values, from different surrounding elements, should be small.
I had a look at the animation issue again - it now appears to me that I got the impression that the contours associated with earlier time steps are wrong, because of a contour level effect...(?). When stepping through the contour plots manually, the contour level (or limit values) changed as it should when set to Auto levels. However, during animation, it defaults to, or assumes the min and max level of the final time step and stays there (which sort of makes sense in context of animation).
Thanks for your input. First had to find some time to try and do what you suggested. Does the following work flow appear correct to you? (the result looks fine to me, just want to make sure...) :
Model -> Output -> Process -> Convert > keep Average Values selected > select Output to Process > select all Output Sets > set filter Element > select 60171 for VMstress (or e.g. 60178 for Max Prin Strain) > OK > OK.
When charting, set filter Nodal > at bottom of Vector list, select 9000000..Avg-Converted Vec 60171 > select/enter node nr of interest for Location.