Basics: Problem with contacts

[filiptheking]

http://oi58.tinypic.com/2ezmgra.jpg

Take a look at the picture above. It is a hexagonal bolt inside a hexagonal sleeve .

♦ It is 8 mm between two flat parallel planes on the bolt.
♦ The gap between the bolt and the sleeve is 0.35 mm .
♦ The sleeve is fixed and there is a torque on the bolt to 20 Nm

1. Why is the maximum deformation over 4 kilometers? Where can I have done wrong?

2. Why is it tension in the bolt before it touches the sleeve?:

http://oi61.tinypic.com/34skyfs.jpg

3. Must the sleeve and the bolt be in contact with each other in the assembly, before i import them in to ANSYS?

 

[flash3780]

Obviously your result is not real. The contact elements are blowing right through the target elements and your model is probably only converging because you have weak springs turned on.

To solve your problem, I'd recommend adjusting your model so that the parts start out in contact. From there, you can solve the untorqued case to make sure everything is working properly. Finally, you can add your torque load (with auto load step bisection turned on, of course). If you're still pulling through the contact elements, try starting with a smaller time-step.

You might also consider adding break-edges to your contact surfaces and fillets to the inside angles if stresses are important to you. This looks like a test case, so I'm assuming you have a more complicated model in mind that you're trying to solve.

Hope that helps. Happy Holidays!

-Chris

 

[filiptheking]

Hello Chris and thanks for your reply.

I managed to get a succesful analysis when the model started in contact and when I used these settings:

http://oi62.tinypic.com/2qbhhu1.jpg

The problem is that I really need to do the analysis when it starts without contact. As you guessed, this is only a test case, and for the more complicated model I need to do it that way.

In the final model that i want the analyse, the parts are not in contact initially. They get in contact first after one of the parts deform a little. So I want to test it on this test case first. Any ideas how to do that?

Happy Holidays to you too!

 

[filiptheking]

I want it to work when I change the moment to the opposite. But it fails brutally:

http://oi60.tinypic.com/mux36c.jpg

Except that the moment is 20 Nm instead of -20 Nm. It's the same settings as the one that was succesful.