Ahhhhhh a whole different can of Mil Spec
As for Shot Peening...well lets just say I agree about the crack propagating thing (only due to the crack running into a rounded stress reliever, the shot impact area), but not the "compressive" part. It is a percussive residual, it can only be compression if there is containment...and on this semi-ductile surface we are talking about metal is neither added or subtracted, just moved around so there can't be containment. we are talking about a process that AT BEST only "disrupts" to a depth of .00005, as we are talking about a air hardening heat treated steel. Now this works great for "surface cracks", but does nothing for stress risers created by sharp machining angles and carbon/chromium inclusions in poorly amalgamated material (a "bad pour" in the metallurgy business) so it only deals with surface cracks ( yes they can grow ) and the MPI can't tell you about inclusions...It ain't X Ray.. so we are talking about a superficial treatment and a test that while making one feel good, don't really tell you about your bolt. (These are the exact reasons Reed and John question this process)
I'll admit that I don't have any experience with shot peening in industry, so I'll have to take the easy way out and drop some quotes from a couple of my textbooks...
Juvinall & Marshek - Machine Component Design said:The most common and versatile of the cold-working treatments is shot peening. It is...blah blah blah...since the area is resisted by subsurface material, the skin is placed in residual compression. The thickness of the compressive layer is usually less than a millimeter. The highest compressive stresses occur slightly below the surface and are commonly of the order of half the yield strength.
and...
Kalpakjian - Manufacturing Processes said:....they [shot] make overlapping indentations on the surface, causing plastic deformation of the surface to depths of up to 1.25 mm. Because plastic deformation is not uniform throughout a part's thickness, the process imparts compressive residual stresses on the surface, thus improving the fatigue life of the component.
Since a bolt is hardened steel, I would expect the depth of compression to be quite a bit smaller, like you said.
The only limit to residual stresses that I'm aware of is that they have to sum to zero, otherwise the part'll be flying around and causing mayhem. As long as you have all compressive stresses in the radial direction, they sum to zero, and you're good.
I will agree with you about MPI though...it probably is more of a feel good test than anything. And certainly a rough bolt or one with inclusions is a definite no bueno.
KurtM said:Now since we are talking about failure, metallurgy, and stress risers of which I have a bit of passing knowledge, and since you seem to have a good grasp on this subject....where do bolts usually fail?, the least common is to shear a locking lug, although it does happen , usually on full auto guns that see lots of mag dumps.. Where is it?
If a bolt's gonna let go, it'll usually do it at the cam pin hole. There, obviously, you've got the least amount of material, highest stresses, and a big freakin' stress riser.
KurtM said:For extra credit, What Bolt part fails the most out of all of them...not to include firing pins being cooked by pierced primers? and consumables like gas rings and extractors.
Well, since you've taken most of my choices away , and since firing pin retainers usually stay together (as long as you have a shrouded carrier and you don't put one of those dumb solid retainers in), I'm gonna have to say it's the bolt that goes most often, with the cam pin as second.