How can I get an interative simulation?

[lebasi]

Hi all,
I´m new with Ansys and it is becoming my worse personal nightmare¡. I m working with transient thermal analysis (tig welding torche). I need to run a file with various simulations but before solving Ansys says "your model has 1 warning, do you still to solve?". I would like to eliminate this question to run all file without interruption. is that possible?

Regards,
thanks, isabel (mechanical engineering, Spain)

 

[Stringmaker]

Isabel,
Chances are you probably have something defined improperly or something of that sort. Many times warnings in Ansys are very trivial. For instance, if you have a few elements which give shape warnings you will get a warning when you solve. Check the output window and/or log file to see what the warnings actually are. If they're trivial just continue on with your solve and you should be fine. If you get a significant number of them and are certain that all are irrelevant and will not have a negative impact to your solution try typing /NERR,0 into the command window before you solve or past that in your input file.
If you're having trouble understanding the warnings you're getting post them here verbatim. Chances are they're easily fixed. Make certain that you investigate them before completely dismissing them.

Good luck,
-Brian

 

[lebasi]

My warning is about shape limit: "Element shape checking is currently using one or more non-default
limits. Issue SHPP,DEFAULT to reset, if desired" . I had changed this limit to try to avoid the warning before solving, because I have 15 elements not so good in my model.
I have used the comand /nerr,0 and I still have warning, before solving appers in the window : "a check of your model data produced 1 warning should the solv comand be executed?"
do you know some way to avoid this question and solve directly?

 

[Stringmaker]

As long as the poorly shaped elements are not in an area of high heat flux or have boundary conditions applied directly to them I wouldn't worry all that much about the error if you have elements of good shape surrounding these. An option you may want to persue is to set SHPP back to default as it instructs you and refine the mesh in that area and see if that will get rid of your warnings. Do you have poorly shaped hexahedral elements? Do you have tetrahedral elements that are not able to properly transistion? Have you tried using a different mesher? Have you tried repairing the geometry? In 99% of situations poorly shaped elements can be corrected. You have to decide whether it's worth you time to do so or not. Unless they meet the criteria above or are in an area of interest I wouldn't worry about it.

 

[cbrn]

Hi,
as Stringmaker says, it's not always necessary to have "perfect" mesh with all-fairly-shaped elements: as long as none has gone into error, and as long as the poorly-shaped elems are in zones where the results are not relevant for your analysis, you can accept them.
If you are launching several solutions from an APDL routine, then the message you mention should only popup the first time but the program will not wait for your answer: every message will be prompted in the Output Window, no input from the user should be needed, because the execution is under the control of the APDL.
If you still have troubles, and you have already checked and "accepted" your mesh, you can issue SHPP,OFF or SHPP,SILENT to avoid messages popup.
A further control is given by /UIS,MSGPOP,3 (only errors displayed).

Regards

 

[lebasi]

hi one more time
Thanks both of you for your replies, it helped me so much.
Now I have another problem. How I said in my first email I design in Ansys a Tig welding torch to study how temperatures and voltage are. I put convection loads on surfaces through argon pass "sfa,all,,conv,h,Ta", current flow With this command "F,all,AMPS,i/node" I didnt know how to do in different way, I thought if my current flow is 300 A I can apply 300/ nodes number to each node. it is right?
Also I used cero voltage at electrode point and I applied temperature of 3000 K.
I simulated for 1 second and the results were good, but if I simulate for 3.5 min, temperatures are so low, around 300 k, it doesnt have any sense, because torch must increase its temperature. something I m doing bad but I dont know what?

isabel

 

[cbrn]

Hi,
the only thing I can suggest is to check how the loads are applied throughout time. It seems that the "heat take-away", i.e. the convection load, is effectively applied at every timestep, while the "heat generation", i.e. the 3000K source, is only active at the first timestep.
In fact, I suppose that for your 1s simulation you ran a single timestep TIME=1, NSUB=1, while for 3.5mn you ran something like TIME=210 DELTIM=<timestep>.
I think you have a mistake in the setting of your transient analysis. It's difficult for me to suggest more details about the settings because I don't know which ones you have already done.

Regards