Female thread relief and stress in a Titanium sleeve

[cfunke]

I have an interesting problem with a design I am working on for a socket termination for a piece of carbon fiber rigging. The sockets are made of 6AL-4V titanium and have a conical shaped seat which matches the end of the carbon fiber. In order to make attachments to this titanium terminal I have used a female thread. My problem is when we run a tensile fatigue test, the designed stress in the titanium sleeve is (Sleeve OD x Thread Major) is 22-26 ksi for a range of sizes. The parts running at 22 ksi are failing much sooner that we would expect, and appear to be failing at the last thread engaged by the adapter in the sleeve (1 1/2-12). Not too suprising. But other sizes running nearer to 26 ksi are not failing during a tensile fatige test to 100k cycles (3/4-16). There is not any internal relief behind the thread in the sleeve, and I think this is a mistake on my part. But I was wondering your thoughts on the geometry of this relief and whether this tensile stress is too high for titanium used in this manner. Thanks in advance.

 

[Cockroach]

The internal thread relief is simply there for ease of manufacturing, either by tapping or cutting. In your case I would expect a tap since threading is small and impractical to cut on a lathe. But the thread relief also assists in full form threads thus full pin engagement. In other words, the pin would not prematurely bottom out.

But I wonder if the imperfect threads at the back of the box are stress risers for your axial load. Could be something funny going on there, I assume you are using a standard tap and not bottoming tap?

It also depends somewhat on your pre-tap drill operation on the box. I assume you are using standard geometry?

Regards,
Cockroach

 

[Tmoose]

The titanium is failing, but the threaded carbon fiber is not?

 

[dhengr]

A drawing of this detail would be most helpful. I’m not getting a very good picture of exactly what you are doing and exactly how it’s loaded. Given a bolt and a nut or tapped mech. part, both having the same modulus of elasticity and other mechanical properties, the first few threads btwn. the two parts, when the bolt is loaded in tension, will have the highest stresses as they start to yield and bring the rest of the threaded connection truly into play. With these higher stresses and the stress raiser of the thread shape, this could certainly be a fatigue problem area. Then also, the thread fit class and the relative difference btwn. the O.D. and the major thread dia. will influence the stress distribution and magnitude.

 

[GregLocock]

Definitely need a sketch. The design of spike and cone terminations for fibres is very similar and the details are critical there.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Cockroach]

I think you're barking up the wrong tree boys! This is a fatigue issue, not a static loading problem. The problem speaks of cyclic loading of a high strength material.

Twenty bucks riding on stress raiser at the imperfect threads as the result of tapping using something other than a bottoming tap. If buddy taps his female box to the same depth with a bottom tap, the problem goes away. This is a manufacturing issue.

Regards,
Cockroach

 

[cfunke]

Thanks for all the responses. I'll try to get to all of your follow up questions.
Yes, this is a fatigue issue. The socket meets our ultimate breaking strength target, however fails to meet our tensile fatigue minimum requirement.
From the manufacturing side, we do not bottom tap, but single point the threads. I guess that is about the same, maybe not.
The carbon fiber is not failing, but there is not a thread on the carbon. The carbon is a tapered cone shape that seats in the socket. I have attached a few photos of the failed socket and a couple of drawings for the original socket design and the improvement we are going to test in the coming days.
The adapter that is screwed into the socket thread has a modulus of approx. 2x the titanium (Ti = 16.7 Mpsi vs. N-50 = 27 Mpsi). Would this difference in material properties cause an issue?
Thanks in advance for all your help.

 

[hydtools]

Add a thread relief so that the mating part thread engages past the last resulting thread. This would remove the stress riser of the last imperfect female thread and prevent possible bottoming of the male thread in the impertect female thread. Think nut and bolt thread engagement for a better fatigue resistant arrangement. Be generous with the thread relief inside radii and exit to the inside taper.

Ted