Learning Forum

[ReneSe]

For my bachelorthesis, I am doing an modal analsys. The analysis itself is no problem with workbench. I get the frequencies and the animation of the form/shape.nBut is there any possibility to get the modal shapes (vectors) out to make a MAC-Matrix with them? Like standardized to [1 yy zz] or smth like that?nI heard that smth like that is possible with APDL, but also in workbench?n

[peteroznewman]

You will find the answer to your question in this thread.n/forum/discussion/2569/comparing-stresses-from-modal-analysis-and-static-analysisnn

[ReneSe]

Hey peteroznewmannSadly I won't really understand what to do.nI copied the modopt,,,12,,on thing to my ds in the modal-folder and re-run the analysis.nWhere do I find the eigenvectoren then? I still only get the frequencies.nSorry if it takes a bit longer, I'm no native and my workbench is in german :Sn

[peteroznewman]

Please insert some images to show your mesh and the mode shapes. Did you use a Command Object in Workbench to normalize the maximum deformation to a unit value? Did you plot the Deformation, Deformation 2, Deformation 3 for Modes 1, 2 and 3? If you did, you can export those Deformed shapes.n

[ReneSe]

nnHere are some pictures. I hope they help.nIn the second picture you can see the additional command I set. Also tried it in the ds-file.nnHow can I use a command to normalize the deformation to a unit value? (I think you mean like for the thid picture where there is shown the deformation of the first mode? Isn't this in milimeters?)n

[peteroznewman]

Move the Command Object up to the Modal branch, you put it in the Solution branch which doesn't help.nn

[ReneSe]

Thank you, that definetly changed the outcome.nBut where do I find the eigenvector-matrix now?n

[peteroznewman]

Total Deformation plot is the eigenvector for mode 1 frequency.nTotal Deformation 2 plot is the eigenvector for mode 2 frequency, etc.nYou can export the X,Y,Z deformation of every node from the plot, that is your matrix.nYou will also want to export the X,Y,Z coordinates of every node also.nn

[ReneSe]

So if I export the total deformation data of mode1, the excel/txt I get is the eigenvector?nIt has nearly 22000 entries. n

[peteroznewman]

Yes, that is correct.nIf you replaced the circular tubes that make up the frame with centerlines of the tube, and you meshed the centerlines with beam elements that had a tube cross-section of the appropriate dimensions on each beam, you could get the number of nodes down to around 100 or so.n

[ReneSe]

Thank you very much! nI already tried my MAC-Matrix in matlab and it works perfectly fine.nnA last quick question tho:nIf I export all my nodes somehow, I could plot them and the deformation after in matlab right? (Would be better with less nodes then)n

[peteroznewman]

Yes you can export all that data and plot it in matlab or you could just plot it in ANSYS Mechanical : )n

[ReneSe]

I wan't to develop a matlab app that compares two models and their modes and shapes. Therefor I need to export to matlab.nnIs ist possible to make an external measuring point in modal analysis and export the deformation at that point? Like having 5 points on every beam of the frame that I get the deformation from?nIt is possible in mechanical, that I know.nIt kinda shouldn't be, cause a modal analysis isn't actually a deformation with force, am I right?n

[peteroznewman]

Yes, you can create a result in Mechanical, such as Deformation at a Named Selection of Nodes, and select a desired Mode number as part of the definition and export that. n

[ReneSe]

So how would I do that?nI already found an extension and other ways to show me the node numbers. But how do I export my Modes to mechanical and get the deformation coordinates?nI also got the hint from my professor to use things like plot elements and rigid body elements that would give a double node, follow the existing nodes. Is this somehow possible?n

[ReneSe]

I already found a better version.nI simply export all the node coordinates, plot them in a scatter plot and for the deformation I add the Eigenvector and plot again.nThanks again for the help so far. n