Axial only constraint to simulate a thread in a pressurized pipe.

[acrmnsm]

So I am not an abaqus expert, I have asked our local support but they have not been that great.
As a more old school guy who has been running simulations since the 90's (I know that doesn't make me that old school as there are a good few users on here who started way before me..), I am always looking for decent ways to approximate answers that reduce model build and solve time.

I will try and simplfy my question:

Imagine a series of threaded pipe connections which are subject to internal and external pressure and axial tension/compression loads. The pipes have variable thickness along the lengths, are modeled in axisymmetric and we want to estimate the forces at the threads to compare against the hand calcs.
Thread ratings are easy to calculate by hand, so I do not wish to get involved in modelling threads and having lots of contact etc. These are repeated calcs so I wish to develop a protocol to solve this.

If I use tie contacts at the threads then this does not allow for the pressure to separate the threads, which we would like to get a feel for, ie if separation is greater than x, then we can expect the thread rating to drop, or even unzip.

So is there a constraint or connector I can use to axially fix the two thread faces together, but still allow radial separation.
It works with rough contact, but as there are some non linear things going on elsewhere we are trying to reduce solution time. Abaqus keeps complaining that rough contact leads to longer solution times.
I looked at cartesian connectors but cannot seem to work out how they work?

Any tips would be greatly appreciated.

 

[FEA way]

It should be possible with connectors utilizing connector failure behavior. However you can also use mesh-independent fasteners which are convenient way of defining various connections. This capability is described in details in Abaqus documentation.

 

[Mustaine3]

I would recommend to also test if an explicit analysis can reduce the run time. The contact is no problem there and with little movement in the model, you can add a lot of mass scaling.