Service Bulletin 14-11-03: Wing Skin Doubler

One of the Mothership's higher time RV-12s used as a demonstrator was found to have a few "smoking" rivets on the lower inboard wing skin.  This condition arises when a rivet loosens and allows fretting corrosion (very small scale relative motion between mating surfaces due to vibration) and subsequent oxidation of the resulting aluminum dust which is produced.  The loosening was no

doubt a result of the high stress in the skin at this location.  Putting on my engineer's cap for a moment, I'd say this location has the highest skin stress of the entire airplane.

The fix involved installing a rather robust (0.04 inch thickness rather than 0.03 inches for the wing skin) doubler plate with four different rivet types, eight of which are 5/16th inch rather than the usual 1/8th inch diameter rivets.  I was thinking "Wow, it's great that I received the kit for this service bulletin before I installed the wing skins.  Now I won't have to drill out 20 rivets per side like many builders are having to do because their skins are already on!"  Then I proceeded to install all the rivets in the right wing skin, then had to drill out the aforementioned 20 rivets.  This required the invention of several new, stronger swear words since I have essentially worn out all of the customary ones during previous mistakes in this project.  I won't (I hope) make the same mistake with the left wing.  I'm astonished at the strength and complexity of this fix (four different rivet types!), considering that the fix isn't mandatory.

The plans call for final drilling the holes for the 5/16th rivets with a #22 bit, which I did using a bit from my new #1-through-#60 bit set, obtained from Amazon rather than Spruce or one of the other "name brand" aircraft suppliers.  The rivets would not fit into the holes.  Upon measuring the bit diameter I discovered that the #22 bit was considerably smaller than spec, and that the #21 bit in the set had a smaller diameter than a proper #22.  You get what you pay for!  I had checked the bits with a go/no-go gauge, but that didn't catch it.