I am studying a +1000 K elements model using transient structural module on ANSYS Workbench. In order to save computation time, I had to use symmetry region, where I split the assembly into two identical halves. The model is composed of a sheet that is loaded vertically and running on a free-to-move cylinder, where the cylinder should be supported from both ends.
Question 1: Does this mean that the reaction forces computed by the software is a result for the complete assembly or just the one half (which is the physical model I have).
Question 2: Can I do further savings in computation time by splitting the model to have 1/8 of its original width? If yes, how can I simulate the bearing support at the ends?
Your help will be highly appreciated!
You can take advantage of simulating the model as half or quarter as long as the boundary conditions, including material and load are symmetry. You will always half the load with one symmetrical plane. But of course you can always idealize your model as 2D plane stress, 2D plane strain or 2D axisymmetry if your model fit into one of these category.
Try to understand how ANSYS interpret symmetrical conditions. Usually the software will 'know' there is another half about the symmetrical plane. Bearing support is normally applied to hole/bore. When you idealize your model in 2D, constraining edge/curve would be appropriate.
My two cents worth.
Regards.
Hi Dear Gerges Youssef
1.When you reach to results from half of a domain solution, the forces are just for half of that. But if you work with fluxes ( like W/m2) or with dimensionless numbers(like drag coefficient) , in this case results are independent with the area so in this case they will same for half of your solution domain or with whole of that.
2. when you solve your problem in 2D it means that your domain is so long in that direction ( in depth ). So it's good for you to illustrate for yourself to solve it in 2D symmetry or in 3D. you could also use reference values to reach to a correct answer if you solve your problem in 2D.
Best Regards.
Dear Youssef
The reaction force is just for half of the model, and what do you think about reducing computation time by using shell elements?
Thank you for replies and suggestions. I may consider modeling the cylinder as a shell.
Can I specify shells as rigid too?
I would be grateful if you checked my other question in the link below..
https://www.researchgate.net/post/I_encountered_model_errors_on_changing_a_flexible_body_to_a_rigid_one_in_ANSYS_Workbench_15_How_can_I_resolve_it
Thanks guys.. I actually got different answers from you. Do you have any supporting links for the information you listed? That will help a lot!
I find your answers so useful, thank you people. In fact, I have another question. I am dealing with a 3D model and I am using 1/4 of the real structure. doing so, I have 2 symmetry planes: One in the length direction and the other in the width direction. Do I need to apply half of the displacement to get the convenient stress/forces?
Achraf Ben Afia , Can you please elaborate the objective of your simulation with images?
- Badges
- Science topic