Hi all,
I'm simulating residual stress generated in a metal casting. I'm using Ansys Fluent to obtain the temperature history and I'm using Ansys mechanical to obtain the stress generated due to the temperature gradient.
When I use linear material properties the simulation converges without any issue. However, if I use multilinear hardening material model for different temperatures there's difficulty in converting.
Any idea how to solve a similar kind of issue?
Thank you.
Cheers,
Rashiga
Hello Dr. Claudio Pedrazzi
It’s so nice to hear from you again and thank you for your reply.
Regarding your suggestions,
-Yes, large displacement in turned on.
-It’s a transient simulation where the temperature profile is from solidus to room temperature. I have used high time step number 10E8 to deal with the nonlinearity and automatic time stepping is turned on.
-Mesh seems to me is in reasonable quality lowest element quality is 0.3.
-From reading the solver output, element distortion is the reason for non convergence.
-I did not check the “Newton-Raphson-residuals” I will check that.
One thing I’m having an issue with is how to deal with when we have relatively high initial temperatures where there is significant deformation due to thermal expansion. Futhermore due to rapid initial cooling, it will result in highly distorted elements.
Rashiga Walallawita Hi, @Rashiga-Walallawita In simple way you can just try small reduce or small increament of the posion's ratio in material properties in ANSYS. Also, watch the below video and hopefully you will get some idea to resolve this issue, if not then just ping me back!!
https://www.youtube.com/watch?v=OZbKnxKTd_0
https://www.youtube.com/watch?v=OZbKnxKTd_0&t=115s
Good luck!
&
Regards,
Hi Abdul Gaphur ,
Thank you for your reply. Can you please explain how the changing poison ratio help the convergence? Also, in the video, they say to add plastic strain data. But what if we only have data up to a certain point? Do I have to interpolate the points where there are no practical data?
Thank you.
Cheers,
Rashiga
Rashiga Walallawita By the way, i dont knw which hardening exactly u r using in multilinear. Bcoz kinematic or isotropic multi materials are both act different...but as of my studies, if you vary the posions ratio with or wothout refining mesh, that give u converged state. As u knw, in plasticity, the material is normally deformed under tensile stresses to the point of fracture, until get a desired shape , so if you reduce poisions ratio the effect of deformaing is vary so..i knw it will new to u but u can try and see.
regrds
Rashiga Walallawita hi can you pls screen shot ur mesh model, materials properties table and spread sheet and error u r getting.
In this case ,u might be using engineering stress so can u change that into true stres-strain data. Also check, the plastic strain whatever u have given whther exceeded the max value u have assigned.
secondly, increase the maximum substeps.
But, more easier to comment after u screen shot above things.
regards,
Rashiga Walallawita hi,
I had used bilinear material data in my thesis. But in that case, you should use LS-Dyna as solver. Maybe it will give you some idea, you can find it in my publications.
Best regards,
- Badges
- Science topic