Threaded connection's initial contact modelling using ANSYS workbench

[Pakhoz]

Hi everyone! I am doing a simulation of threaded connection in ANSYS workbench and have some problem.

In this 2D model, the initial make-up is modeled by setting an "initial overlap" (shown in figure 1).

This is how I did:
- The initial contact between pin and box as "frictional" with coefficient of friction 0.03. Formulation used is Augmented Lagrange. With interface treatment: offset (ramped) 0.08 inch.

The other constraints/ loadings:
- The upper part (box) is restricted by "frictionless support" on its edge on the right to ensure it can't move along x direction

- The lower part (pin) is restricted by "frictionless support" on its bottom edge to ensure it moves only along x direction.

- A pressure load is added on the left edge of the pin.

My question:
Is the way I address the initial contact right ? How can I get the stress induced by this initial contact ? And why this model is always not convergent ? Can anybody give me some advice?

Thank you for your suggestions.

- Pakho Cheng

 

[Mech_Engineer]

I don't think this is a good representation of a threaded connection; in reality the model should have an initial gap between the threads, and then you should move one of the parts axially to create a connection on one side of the thread faces.

Your current model basically assumes an oversized thread being forced into an underiszed one, but the initial geometry interference you have cannot be solved because you need to start with no interference and then slowly introduce the interference over some time steps.

 

[Pakhoz]

I don't think this is a good representation of a threaded connection; in reality the model should have an initial gap between the threads, and then you should move one of the parts axially to create a connection on one side of the thread faces.

Your current model basically assumes an oversized thread being forced into an underiszed one, but the initial geometry interference you have cannot be solved because you need to start with no interference and then slowly introduce the interference over some time steps.

Thank you for your reply.

Instead of making an initial gap between the threads, can I develop these two surface bodies initially contact with each other with no interference, and then give the pin a y-direction displacement equal to the "initial gap" to model the interference?

Because I think when the two parts are at make-up situation, both the flanks of each thread will have contact. If you move one of the parts axially to make them contact, the contact region is probably on only one flank of the thread?

If what I say makes sense, how can I introduce the interference slowly over some time steps as you say?

Thanks.

P.C