Titanium Putty

[Dougt115]

I have a titanium bolt threaded into aluminum and sealed with a titanium putty. Now the parts are fatigue failing in the threads when the bending load is applied. I am wondering if this is a bad design combination?

 

[EdStainless]

Two low modulus materials, a lot of flexing, and materials with low fatigue strength.
Sounds like a plan for disaster.
You may need more bolts or larger bolts.

Are you contolling the pre-load? Are they holding tight?

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[Tmoose]

I kind of doubt the putty is at fault.

Subjecting a bolt to cyclic bending is looking for trouble.
Poor Geometry beats (and kills) excellent materials every time.

"In applications where screws are subject to fatigue from dynamic
loading, the importance of proper preloading during assembly
cannot be over-emphasized. The proper preload is especially
important for rigid-type (metal-to-metal) joints, where it has been found
that the use of a preload greater than the external load will usually
eliminate the possibility of fatigue failure. "

http://www.holo-krome.com/pdf/Consolidated_Tech_Manual.pdf

 

[Dougt115]

Thanks guys I now have a path forward.

There are three areas to correct, preload, radii, and depth of the threaded hole.

 

[tbuelna]

Dougt115-

Ti and Al are a poor galvanic match, so galvanic corrosion might be something to consider. A metal filled epoxy is not the best choice for sealing a thread interface. It would be better to use a sealant material that remains flexible when cured.

Your description of how the bolt is loaded was unclear. Is the bolt itself actually subject to cyclic bending loads? Or is the bolt securing a component that is subject to bending moments? A bending moment applied to a bolt threaded into a tapped hole is a bad situation, since there is a big stress concentration created at the root fillet of the bolt thread pitch just before the start of the tapped hole threads. Normally you try to avoid putting the threaded portion of a bolt in shear or bending.

It would also be worthwhile to evaluate coatings for the threads to prevent galling. If the threads begin to gall/seize at installation and you're only relying on wrenching torque to determine preload, it would be quite easy to install the bolt without any clamping force created on the component being attached.

 

[Dougt115]

tbuelna,

All of the worst things you mentioned is what the previous engineer did.

Even worse the part it is attaching are co-axial with the bolt so it gets an even bigger lever arm.

Now I am trying to fix it.

Best option so far is to go deeper and take the bending load off of the threads. Also the part has more material just below where the current threads end so it is substantially stronger.