Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2021-02-22 18:13:50

LouisB
Member
Registered: 2020-12-21
Posts: 4

Pretension Bolted Connection

Hello,

I am trying to simulate a bolted connection with a pretension with the following steps:

1. I pull on one side of the bolt with a traction force and fix the other side from t= 0s to 100s (resnonlin0 in my .mess file)

2. Then I want to put a contact between the nut and the bolt and a tie connection at t=100s until the end of the simulation (200s in my example) to simulate the tensioner release. (resnonlin1 in my .mess file)

The problem is that the first step is working well when I run it independently but when I want to add my contact and my tie connection with the nuts I am getting a "NonConvergenceError" which is giving me some advise that I tried to follow (like increasing the ITER_GLOB_MAXI ...) but I still have this error.

Could someone help me to debug my problem ? give me some advise and/or send me an example that I can study ?

Unclosed a .zip file with the .hdf, .comm and .mess of my simulation.

Regards,
Louis


Attachments:
BoltedCo 2.zip, Size: 1.23 MiB, Downloads: 1

Offline

#2 Yesterday 07:52:38

mf
Member
Registered: 2019-06-18
Posts: 170

Re: Pretension Bolted Connection

Hello,

try the attached example for a start.
1) the TIE must be a DIDI, as it refers to the already tensioned state, not the zero state (untensioned)
2) no need for a contact yet. It would be necessary when you put 2 pieces to connect in between your nuts.
3) I created a partition of the nut on the bolt.
4) it could be even better with just fixing the lower ring-shaped part of the nut--> you'll also get the forces in the nut, now it is rigid because you fix the whole nut.
5) your time frame of 100s or whatever does not make sense because everything is linear (material..). There is no time dependence in the material, also this 'time' in static-nonlinear is just a way of ordering forces etc. It is an artificial simulation time that has nothing to do with real time in the most cases.
Just a little advice for non-convergence: if everything is linear and convergence is hard to reach, there is for sure something wrong in the definition of the model (conflicting BCs, moving bodies, or 'rattling' contacts in non_linear etc.). If convergence is not reached within 10-20 iterations, it probably never will. It is easy to see with RESI_GLOB_RELA, if it starts at ~10^⁻12-->good, if it starts at 1-->bad,

Mario.

Last edited by mf (Today 07:58:22)


Attachments:
StudyF.zip, Size: 1.42 MiB, Downloads: 3

Offline

#3 Yesterday 13:06:28

mf
Member
Registered: 2019-06-18
Posts: 170

Re: Pretension Bolted Connection

Hello Louis,

the above posted .hdf is SM2020.

Here are .med (in .med 3.3 not 4.0 as in SM2020) and .comm. Let me know if you can import it in SM2019. Both files need to be imported separately,

Mario.

Last edited by mf (Yesterday 13:11:19)


Attachments:
CurrentCase.zip, Size: 686.17 KiB, Downloads: 5

Offline

#4 Yesterday 17:48:47

LouisB
Member
Registered: 2020-12-21
Posts: 4

Re: Pretension Bolted Connection

Hello Mario,

Everything is working with the .comm and the .hdf file.
Thank you for your help.

Louis

Offline