Thread Strip out

[STVV]

Hello,

I am working on a compression seal joint that consists of two hollow shafts threaded together. The shaft with the female inner threads will be turning which will transmit torque to the other shaft (which has the male threads). The way I see it is I need to size both shafts to handle the torque (not a problem) but I also need to size the threads to prevent strip out. I am seeking analysis formulas and methods for calculating the thread strip out. Any help would be appreciated. I think I need to calculate the stripping shear area of the threads and use that to get my max allowable shear but I am not sure how to relate the known torque value to the shear stress in the threads.

Thanks

 

[Tmoose]

For bolts threaded into material with similar E the first few threads work awful hard. Something like this.

http://bonrouge.com/test/test_files/thread_stress.gif

For my first cut I'd probably assume only 2 or 3 threads were engaged.

But, before I worried about thread stength issues, I'd spend time ensuring the threaded/sealing joint scheme could be made commercially with good-enough centering (based on rpm) and butting faces with such good accuracy they won't make the shafts form a kink at the joint. That makes a gasketed face out of the question. Asking threads to provide centering or location is generally a mistake. A well controlled pilot diameter and a pilot face should be provided. See pages 46 and following here -

http://www.deublin.com/pdf/RU073 US.pdf

 

[desertfox]

Hi STVV

I agree with the previous poster about centering on screw threads, but assuming you can live with error I see the following problems:-

1. your shaft must go metal to metal as well as compressing the seal otherwise the thread will not fully tighten and when you apply the torque it will compress the seal further
or eventually damage it.

2. your shaft can only apply torque in a clockwise direction
assuming a right hand thread.

3. If the materials are of the same properties then I would
use the rule of 1.5*thread diameter for thread length
engagement and that should be suffiecient to prevent the
threads stripping.

4. Analysing the stresses the threaded portion will act like
the shaft and will in fact be your weakest link.
I would look for the principle tensile stress which will
run at 45 degrees to the shear stress due to torsion and
if the material is brittle you will need to consider
stress concentration in the thread also.

regards

desertfox

 

[STVV]

Thanks For the input. We are using NPT threads on a current test model and the joint does not leak. The alignment is an issue I am addressing. The RPM value is very low. No exotic material needs to be used so I do not think brittleness will be a factor. I do like the Idea of a conical lead in. I think what both of you are saying is the standard 1.5*thread dia for thread engagement will work.

 

[Tmoose]

I think the common "rule" is 1.5 diameters of thread engagement very generally develops full fastener strength in weaker materials.

Depending on how hollow the shafts are I'd anticipate the wedging of the tapered threads might do something exciting long before the threads "strip." In order for the threads to "strip" they have pull against something, like butting shoulders.