strength of out-of-round cold worked bush

huangyhg

strength of out-of-round cold worked bush
hi there, i'm currently facing an out-of-round reamed up cold worked flange bush (fti). the remaining wall thickness after further reaming up for the repair bearing will be about 0.055 inch (1.4mm), about half of the "design" thickness of min. 0.1 inch. my question is what the strength (static/fatigue) effect is on the lug and the flange bush. from my engineering judgement i would guess that the lug has already been cold worked when installing the original fti bush and the compressive stresses at the lug hole will remain. the reduction in wall thickness of the fti bush to 0.055 inch seems to me still acceptable.
however i cannot find any literature that substantiates this, except for some vague quotes on internet.
anybody out there with more experience on this?
i suspect that you should replace the bush.  an fti bush means to me that the design was pushed fairly hard to achieve an acceptable fatigue life, and any change would need to be carfeully substantiated.
i think losing 1/2 the thickness of the bush wall would affect the residual stresses in the lug, 'cause the bush is now more compliant than originally.  a short-cut would be to ask fti how effective the bush would be.  alternatively, you could calc the numbers, from roark or timoshenko.
rb1957 is right, there is an issue with the residual stresses in the lug.  i suspect that whatever is inserted in the bush will affect its compliance depending on the fit, etc.
if it's a clearance fit, maybe you could argue that since the bush is now more compliant, the residual compressive stress in the lug material, at the surface of the bore, is higher that it would otherwise be.  in isolation, this is good for fatigue, but a less rigid bush might mean that fretting is introduced or that the hole behaves more like an open hole under load, rather than a rigidly reinforced hole. depends on the structure.
the are a few solutions that calculate the residual compressive stress at the hole periphery for an unfilled, cold expanded hole. (but since i don't have references for them, i guess that this is not really helpful.)  anyway, maybe these are a starting point if substantiation is required.
if it's an interference fit in the bush, maybe you could argue the compliance is unchanged.  however, i guess if your using a small forcemate bush or something like that, it's probably a clearance fit in the bush for a removable fastener or similar.

on the other hand, this sounds like a production concession, so maybe time is tight.  if so, i would definitely go with rb1957.  just replace the bush, paying appropriate heed to the possibility of damage to the bore of the hole in the parent material when you remove the bush.
thank you rb & fastmouse,
it was indeed a production concession on a part for which we do not have repair authorization ourselves. indeed time is tight and apparently this particular bush was not fatigue critical. for this case our proposed repair of reaming up and installing oversize bearing was accepted. meanwhile i have gotten access to fti product specifications and technical library articles on their internet site and i will investigate these bushings a bit more. note that for this concession replacing the bush with the same one per b/p would have resulted in the same defect because of the production set-up we are bound to.
sorry, but odd and odder ...
i surprised that an (expensive) fti bush isn't fatigue critical ... if fatigue isn't a problem, why use fti bushes ?
then to say "replacing the bush with the same one per b/p would have resulted in the same defect because of the production set-up we are bound to."  says to me that your production process is wrong, and this problem will recurr.  i had assumed that the problem was due to in-service wear, but this seems to me to say that it was due to an installation issue.
rb1957,
i know and agree with your remark on the fti bush, my idea exactly. unfortunately the design is fixed, but i've made the same comment too. with production set-up i meant the same as you; installation issue.