[fogl]

Hello everybody,


I would like to use the solidDisplacementFoam to simulate a press fit of two tubes. The OD of small tube is 2um larger than ID of large tube.



My idea was to first separately simulate a compression of the small tube so that it fits inside the large tube perfectly (with fixed displacement BC). Than map the displacement filed of the first solution to initial field for the "press fit" case (with zero traction BC).



I did all as described, but the solution is not logical - the solution is that displacement filed is zero. It seems like initial condition was not taken into account. Please see the two images below.


Does anybody have an idea why such solution?
Is there some other way to simulate a press fit?


Regards
Klemen

[fogl]

I am still struggling with my test case to model the press fit and i am not sure if my case was set correctly (if the physics was modeled correctly) . I did not add any boundary condition between the two cylinders. Is there somebody who could comment on that.



Currently my model was set as there was only one cylinder with initial displacement field. Would it be more suitable to set some kind of boundary condition between the two cylinders which allows the transfer of normal pressure?


Regards
Klemen

[pete20r2]

Can you please share the case? I'd be keen to try this myself.
Of course you need to specify a boundary condition, OF is not smart, computers are not smart, you need to tell it exactly what to do.

[fogl]

Thank you for your reply. I understand that i have to set the BC - I have them, but i am not sure i should add them also between the two cylinders.


What BC should i add between the cylinders?



Please see the case file attached.
To run it run ./meshAll and then run solidDisplacementFoam.


Regards
Klemen

[pete20r2]

Hi,
You have a typo in meshAll, I think you need to slow down when you type "fields".
If you fix that you will start to see results.

The two cylinders are part of the same mesh so no boundary is necessary.
I'm not sure what the issue is but the displacement/stress fields look wrong to me, they should be uniform around the pipe.

[fogl]

Thank you for your reply. There is a typo, but it does not affect the simulation, because the rho, nu and E fields were already set at the case i provided.


You are right, the two cylinders are part of the same mesh and boundary is not necessary, but i am not sure if there should be one - should there be some kind of BC that allows the transfer of normal pressure (i am not shore of the physics of this case).


As you see from results they are the same as shown on the two images in my firs post - there is some initial displacement/stress of inner cylinder, which should be transferred to the outer cylinder. But the result of simulation is that the displacement/stress filed is zero (which is not what it should be, so i assume there is something wrong).


I am not sure which displacement/stress fields looks wrong to you, the initial filed is uniform around the pipe? Please check the D magnitude or draw the Glyph to display gradient. The final filed looks wrong to me too which is what is the problem


Regards
Klemen

[pete20r2]

Did you rescale to the new data?

[fogl]

Yes.



I was expecting part of the displacement to be transferred to the outer ring. I assume the solution for the displacement should be in range of one half of initial displacement - so it should be easily seen in the same range setting.


Regards
Klemen