Andrew R Stewart 

I will respectively dissagree. At most the axle flats should have only "skin oil" on them. Too much lube allows the arm to travel up the taper too far before the bolt's torque level is reached. This is a commonly known issue.

Also, my 40 years of shop work has seen many hundreds of crank arms that have "rounded out" because they came loose. This year alone we have probably replaced 3 dozen, and the shop I currentlly am at is not the service focused style that I have spent most of my professional life in. I have followed the same procedure (that i discribed before, proper initial torquing of the retaining bolt and a few checks/retightening in the initial weeks following) with my personal and the client bikes for all my years. The number of crank arms that have been problematic using this procedure, that I can truely track with service records, have been virtually nil. The ones that I have had to replace due to "rounding out" have been orders of magnitude more. What this says to me is that when well torqued during the assembly and checked/retorqued again after some riding the tapered cranks are very reliable. But when torqued only once (and with bikes built/sold by others the amount is a variable) and then never redone the likehood of problems is MUCH HIGHER.

Perhaps some of the misunderstanding of the retightening step is in what/how much this means. In an ideal world the bolt would be loosened then retightened using a torque wrench. In the real world of shop service this means that the bolt is just checked for tightness using the experience of the mechanic's feel as the torque judging. After years and thousands of procedures I have a good feel of what is the reasonable amount of torque. My history has proved this, and doing this over and over has tought me this level. Can i tighten a crank bolt and say it's at X.Y pounds of torque, no. Can I say the bolt is well tightened and within working industry range, yes.

With my personal bikes this retightening produces minimual amount of end of wrench movement, perhaps 1/4"-1/2" for a Park CCW. Over the year (I tend to dissasemble my bikes once a year + or-) my checks and retightening amounts to 1" (or slightly more) of wrench end travel. You do the math as to the torque increase. I have cranks I have done this cycle for over a decade so I am conifident that it works.

I will agree that when a ham fisted "mechanic" goes at it all bets are off. But i teach my coworkers my method and there are dozens of wrenches in shops that learned my methods well and those I keep in touch with don't "correct" me. After so many years of doing the same methods I would think there would be a trail of VERY unhappy shop owners (these are the people that are at the spot that the buck stops, the end of the system we call the bike shop world). Many of my employees and coworkers have stayed in the business for decades.

I have had this discussion before. I agree that whatever method works for you is good. But I will advise my customers and coworkers to my methods untill i see first hand the wrongs of them. Andy.