How to analyze the stress caused by a Freeze plug

[USACompositeGuy]

I have a fitting that has oversized fastener holes. The upper and lower portions of the fitting (flanges) are approx. .180 thick and have ample edge margin to support a freeze plug from a standard tear out strength vs fastener shear (Each flange can support higher load than the fastener shear in tearout and bearing) . The middle section of the fitting is machined out to lighten the fitting. The remaining thickness on the center section of the wall is approx. 0.030 inch thick (possibly thinner depending on hole oversize). The material of the fitting is 17-4 steel and the total height of the fitting face in question is 1.0 inch. The plug installed will have a dia. .340 and will be made from 17-4 in the H1150 condition and will be installed 0.001 inch larger than the hole prior to installation.

My concern is that the freeze plug will expand when installed in the thin wall section of the fitting and crack out the fitting. Can anyone provide an analysis method for the stress on the thin wall? Is it correct to assume that the stress is uniform in this configuration?

On a side note is it necessary to passivate the fitting and the plug prior to installing?

 

[MNLiaison]

In the past I have analyzed inteference fit bushings in lugs / holes using Lame's equations for thick walled cylinders. With the known radial interference value, hole sizes and material properties you can calculate the stresses in the bushing and fitting. Perform a google search or engineering tips search for lame's equations. I don't know if your geometry will meet the criteria for the calculations but give it a try. Below is a URL for an online calculator, I haven't used it myself.

Regards

 

[rb1957]

the pic isn't really clear (to my older eyes) ...

it looks like you've got two flanges and maybe a boss around the holes, and it's the boss region that's the concern.

if you open out the holes for the bush and get left with 0.03" wall then i think you're right to be worried. it tells me too that the boss wall is quite thin to start with, and i suspect its main function is to stop the flanges being pulled together when you install the fastener; if that's the reason, pretty clever design.

anyways, back to the plug ... if you've got planty of margin for the load in the flanges, could you install the bush in a counter-bore ? this might avoid the very thin wall in the boss. otherwise i'd get more analysis (like you're asking for !) ... maybe the standard solution (Lame, as above, or Timoshenko) for a cyclinder with internal pressure; maybe FE. i don't think the stress is going to be uniform, the flanges have much more restraint (to the plug), the thin walled boss much less ... the middle of the boss may just deflect the 0.001", but what's going to happen at the boundary between the boss and the flange ?

 

[RPstress]

1 thou on diameter oversize - is that max or nominal? Even if it's max, its quite a nasty looking situation. It wouldn't surprise me hugely if you got approaching 100 ksi installation stress. I can see it being 50 ksi easy enough.

That might not be a worry, depending on what the basic part is for, and what its condition is - is it H1150 like the intended plug? SCC might be a slight worry if it's H900 or similar.

Normal practice is to passivate parts like this. Presumably it'll be gunked up with PR1301 or similar as part of installation - not sure what the Boeing spec calls out any more. As long as it's reasonably gunged up and the operating environment is benign you might get away with ignoring passivation with 17-4.

Remember that the final OD of the plug and ID of the plugged hole need measuring carefully after all treatments - tenths of a thou can matter in these situations.