how to make high penetration for contact elements in ansys apdl

[C.Amini]

Hello everyone
I am modeling FSW ( friction stir welding). I already simulate it using the tool that had no pin only including shoulder.
Lately I wanna simulate the process with tool including shoulder and pin which should penetrate 5.6mm but my simulation became divergent after penetration of 2.1mm . and some times these errors arise during solution:

1- Some of the element based thermal loads may be inappropriate. Check
applied element loads carefully. Ensure that film coefficients and
bulk temperatures are always positive.
2-Element 11310 (type = 1, SOLID226) (and maybe other elements) has
become highly distorted. Excessive distortion of elements is usually
a symptom indicating the need for corrective action elsewhere. Try
incrementing the load more slowly

how can i adjust options of contact element or some options of solution to achieve high penetration (5.6mm) and avoid divergence
any comment would be helpful and appreciated
regards
cyrus

 

[karachun]

If solid mesh becoe distorted then you can use mesh nonlinear adaptation NLAD.
Link
Link
Link
Link

 

[C.Amini]

hello Karachun
thank you for your response, those links you sent, were included in handy informations. I am on it hopefully to simulate the process completely. used element is solid226
I am using ansys apdl ( not familiar with ansys workbench) meanwhile the analysis type for me is transient ( because of this apparently I have limitation to use meshing adaptiving) ) but I feel that I can use these informations to achieve the goal.
at the end I am so grateful
regards

 

[C.Amini]

Hi karachun
I studied nonlinear adaptive meshing. unfortunately all these methods are used in structural static analysis. is there a way to implement this method in transient analysis. in the other hand my analysis is coupled analysis and I use solid226 to do it ( not including special elements that were mentioned usable in NLADM).
looking forward your response

 

[karachun]

I use this method only to NL static problems. Try to search in documentation.
Link
Link
There is a list of supported elements at second link and there are no solid226 at this list.

 

[AnPham]

You can refer this tutorial:

https://www.sharcnet.ca/Software/Ansys/16.2.3/en-us/help/ans_tec/tecfricstir.html

and find the file named: fsw.dat in ANSYS install dictionary.

 

[C.A]

Hi AnPham
thank you for your answer.
I already modeled this sample that you mentioned above. it was so useful. configuration of my simulation is based on this sample. but this sample ignored the pin. that is why I developed the sample with pin. but at the plunge stage , due to high deformation and element highly distortion, the problem did not became converge.my problem is how to converge the problem ( transient analysis with element solid 226).
anyway thank you for your response

 

[AnPham]

If you want to model the Pin, it will be very difficult problem to perform in ANSYS. It's not recommend.
You can try with other software to model full Tool geometry.

 

[C.A]

the idea came to my mind is that modeling the plunge stage statically and then use of solid 186 ( the element is applicable to perform rezoning). for ding that I wanna run an example. it has to do with an example in help ansys " Rezoning Using a Generic New Mesh"

https://www.sharcnet.ca/Software/Ansys/16.2.3/en-us/help/ans_adv/Hlp_G_ADVRZSMP.html

.then changing element and key options with solid 226 and since then I go on the process in accordance with your example in previous post. but unfortunately I did exactly the commands in this example " Rezoning Using a Generic New Mesh" but as I wrote in my post before

https://www.eng-tips.com/viewthread.cfm?qid=450428

the example doesn't work to me.I do not know why?

 

[AnPham]

Nonlinear Adaptive Region only available for Static Structural.
Your problem (Friction Heating) is Transient Structural. So you can not apply Nonlinear Adaptive Region for this problem.

 

[C.A]

I think non linear adaptive meshing is different from than Rezoning.meanwhile I told I wanna to model fsw statically till I arrive high penetration (at the plunge stage initially).
in the NLADM for 3d problem I have to use solid 285 which tetrahedral shape and four nodes is available for it that I will have problem to change this element with solid 226 at the dwell stage.and multiframe restart ability is not available in NLADM and we know I need restart analysis after plunge stage.so I decided to use other Technic named rezoning with solid 186 because when I want to change this element with solid 226 I will not run into any problem. maybe I am wrong.

 

[AnPham]

In this problem, I think you can have other way to model.
That is using Moving Heat instead Tool model. That mean, calculate Heat Power by Pin and Shoulder, then apply and make it moving.
With this way, solving time much smaller than Tool model like ansys help.

p/s: it's not like your initial idea

 

[C.A]

thank you AnPham
yes this idea is known in FSW, If I couldn't simulate with pin maybe I use this method
Regards