Algor and threads 1

[smalld]

Hi,
I Have a rod with a 3/4"-16unc thread on each end that are 1" long , overall length about 13" with a 1.5" shoulder and fillets in the corners. My Question is about applying boundary conditions to it. One end is screwed into another parts and is "fixed" the other end has a 17,500lbf load applied in tension. the model is imported from Inventor 09 so the od's are threadless with the surfaces split-up. should the BC,"Fixed" be placed on the OD where the thread would be or on the very bottom face? or is there a better way to model a fixed threaded joint?

 

[GBor]

What you have looks OK to me depending on what information you are trying to get out of it. Is this just a test because it seems as though it would be easier to hand calc this one?

 

[smalld]

well, a quick hand calc 17500lbf/.441in^2(3/4" dia) = ~39.6 ksi, with algor, with a b.c on the very bottom face perp to the axis, 41 ksi Von Mises stress..... Ok all is good... but once I change the B.C to fix the OD where the threads are and you get around 80.5 ksi.... so that is why I was looking at the best way to model a threaded joint.

 

[GBor]

Where is the peak stress for each calculation? Does it relocate with the boundary conditions? For instance, is 81ksi right at the end of the BC?

My thoughts are that it may have something to do with poisson. If you pin the last row at the top of the threads, do you still get that high stress?

It sounds like you may have your own answer...if fixing the face is closer to a hand calculation, that may the right approach for your application.

You may also want to check in to 3-2-1 support methodology. The best paper I've seen on this is at

http://www.roshaz.com,

but the gist is that you use balanced loading and apply minimal boundary conditions to address rigid body motion and computing round-off differences. For your model, you would apply an equivalent load on the other end of your threaded rod and then pick three points where you would apply Tx, Txy, and Txyz at individual nodes.

The thought behind this is that boundary conditions are theoretically infinitely stiff, so load is going to go through them, but they aren't going to be able to "cushion the blow". You will get elevated stress levels around boundary conditions that may not be realistic.

 

[smalld]

"Where is the peak stress for each calculation? Does it relocate with the boundary conditions? For instance, is 81ksi right at the end of the BC?" Yes, you nailed it on the head, max stress is right at the end of the BC. I will try a pin and go from there.
Thanks.

 

[GBor]

I started thinking a little more about your problem this morning. At the threads, I would think an axial constraint would be more realistic. You have to pin something in the planar directions to prevent rigid body motions, but I would suggest a "pinned" node at the center of the end face and axial constraints on the face for the threads.

To put some geometry to it: Txyz at the center of the end face. Tz in the thread area. This will allow Poisson's effect to displace the threads in the xy plane, but will move your stress to the proper location at the top of the threads.