'Soda can crushing' simulation project by stharani


#1

I created a new simulation project called 'Soda can crushing':

Just trying to visualize the crushing behavior of the can


More of my public projects can be found here.


#2

Hi @stharani!

Do you want to perform a buckling analysis on the can? If so then I can tell you that this is already on the roadmap of SimScale.

Best,

Jousef


#3

Hi @jousefm ,

Thank you very much for your message and sorry for the late reply.
Yes, I was trying to incorporate the non-linear buckling behavior into the soda can structure.

I used the Tet- dominant meshing algorithm with automatic element sizing and 2 coarse mesh finess. The anlysis type that is being used for this simulation is non-linear static analysis - advanced.

A displacement driven loading condition are applied to the top cap surface of the can with auto time stepping simulation control.

Though the simulation did not complete 100%, the qualitative results seem quite convincing. The main reason behind the termination of the simulation was that the specified maximum runtime 3600s was reached before the actual completion of the simulation run.

On the other hand, there seems to be a problem with the Newton iteration resulting in a non-converging solution behavior. This could possibly be due to less number of iterations count used for the calculation or increased time step values considered for this problem.

Also, I would be really interested to hear from you regarding the buckling analysis.

Best regards,
Shashidharan Tharani


#4

Hi @stharani,

sounds like a good approach to me.

You can simply increase the maximum runtime to make the simulation complete. Did you already give that a go?

As mentioned in my previous post, the buckling analysis type is on the roadmap of SimScale :slight_smile:

P.S.: You can add a feature request on this page: Feature Requests

I would also add a vote from my side since this feature has been demanded by some users already but not proposed in the “Vote for Features” section. :+1:

Kind regards,

Jousef


#5

Hi @jousefm,

I did perform the simulations several times again with some possible changes and combinations.

  1. With an increased value of maximum runtime
  2. Then changing the writing time intervals, as the simulation results exceeded its memory allocation.
  3. Tried decreasing the convergence parameters or the time increments values.

Since the CAD geometry was downloaded from grabcad, the file seems to be too huge for the simulations to occur quickly and have an insight of the results.

Maybe, I should try the similar analysis with simpler model again shortly.

Best regards,
Shashi


#6

Hey again @stharani,

that is also what I would suggest. Defeature the geometry and iteratively increasing complexity as soon as you have a successful simulation run.

Keep me up to date with your project. Gave your request a vote because I think that this is a necessary feature that has to be added to the platform. :+1:

Kind regards,

Jousef


#7

Hey @jousefm ,

Ya sure, I shall keep you posted regarding my update on that project.

Thank you very much for the vote :slight_smile:

Best regards,
Shashi


#8

Hi Shashi (@stharani)!

Great idea! But I have noticed that though simulation results doesn’t represents the positive displacement but in your simulation setup the displacement is in +ve z direction. I have tried to perform a simulation case with your model. I merged it in to a single part since its made of same material. Furthermore, by doing so I can get rid of assuring mesh consistency over the contact regions which may result in odd stress distribution.

Please have a look at my project here: https://www.simscale.com/projects/ahmedhussain18/soda_can_crushing_-_non_linear_buckling_analysis/

Though simulation didn’t complete fully but I did manage to go until 65 % of 0.1 meters. The results look something like this:

I hope this helps. For question/s, feel free to ask.

Best,
Ahmed


#9

Hi Ahmed (@ahmedhussain18)!

Thanks a lot for your appreciation and your time dedicated in solving this problem. The results look very nice and I am glad to know my mistakes.

When the first simulation run was made, the displacement was taken care and was assigned to be in the -ve z direction. The screenshots were taken right after the 1st run (25% simulation) was completed. Later, in order to spot and rectify the errors, certain parameters were changed (as mentioned in the previous comments) for further simulations . This process has resulted in an unintentional/unnoticed mistake of updating the displacement to a +ve value in the final saved setup.

Once again thanks for the help.

Best regards,
Shashi