To avoid stripping the crank-axle nut, I press the crank onto the axle and then tighten the nut.
That’s likely to make no difference - the forces involved when you only have a thread or two engaged are very low (a single thread will take far more load than you can give pushing by hand). Meanwhile mine stripped when the nut was fully engaged and torqued up at the end - at which point the crank has got pushed much further up the taper than you can push it by hand.
I don’t push it by hand but with a mechanical press that measures the force being applied and it applies considerably more than I could by hand.
You clearly have better tools than me! Are you still using a torque wrench on the bolt (and just using a lower torque setting)? Because the crank will squirm back down the taper if you don’t tighten the bolt sufficiently, so the eventual installation is still dependent on the bolt torque.
I still reckon the threads on the axle should easily have been able to take the torque I was using - fairly standard for tightening the cranks on a ST axle. Your method would have helped, but that’s compensating for poor quality kit - to be fair I couldn’t really expect better on a £35 uni.
One thing I try to avoid is making one thing serve two functions, as it usually compromises one or both functions. In this case, the standard procedure asks the nut to perform two functions, press the crank onto the shaft and not come lose itself. The crank and shaft are pressed together to prevent them from wearing each other (or the harder one-- most likely the shaft-- from wearing the softer). Wear occurs only if the two separate at some point as might happen on a hard jolt. Pressing them together puts them both into compression at their interface. To cause wear, a jolt must overcome this precompression. This is also how prestressing concrete enables a beam to essentially take tension.
Once the crank has been appropriately pressed, the job of the nut is to keep the crank from working off the tapered shaft. Relatively soft threads do a better job of keeping a nut from working lose than harder ones but they also will limit how much pressure you can apply without stripping. You could case-harden the shaft threads and use a case-hardened nut to apply as much pressure to the crank as you want but the nut would be more likely to fail in its function of not working lose.
This is one application where separating the two functions is easy. The crank is pressed on and the nut is glued on with Locktight. I use modest torque on the nut to not make the Locktight do all of the work.
You don’t need the kind of equipment I have for pressing on the crank. Since you want both left and right pressed on hard and the axle is very strong, you can just lay the left crank on a firm non-marring surface and hit the right crank with a heavy non-marring club. It won’t hurt to be brutal but make sure you apply pressure directly in line with the axle. Substantial pressure at an angle could hurt it.
Interesting suggestion - thanks for the tip. Doing that does at least make retapping to M9 more feasible, as I don’t have the worry about stripping the thread again. Which just leaves the economic question of spending £15 on a die and some nuts for a £35 uni (which will still have a cracking noise when idling), when I’ve already bought a replacement for my own use, and it would only ever be a loaner (or for use by my kids when they get big enough).
I suppose I could try spending a couple of pounds on bearings first, and then using your suggestion to seat the cranks should give me long enough before they come loose to check whether that’s fixed the problem. I don’t think I damaged the taper when it came loose before, as I stopped riding as soon as I noticed. I do hate to throw things away which are easily salvaged - if only I could find somewhere to borrow the die I need.
I have been able to cut new threads over old using a case-hardened nut slightly smaller than the original. This has been most successful when I change metric to inches or vice versa, enabling a very small downsizing. The threads are ugly and not particularly strong but they lock the nut rather well (I would still use Loctite). They could not support the torque required to press the crank but you aren’t going to use them for that anyway.
Thanks for that suggestion too - unfortunately I’m talking about a large resizing, going from M10 to M9. 9mm being the root diameter of the M10x1 thread, which is what I’m left with after all the threads have gone! Worth a try though I guess - will have to see if I have any old M9x1 axle nuts lying around, as with a bit of care I can probably make my own die by filing a taper and some cutting edges. As you say, with your other suggestion I’m not relying on the nut to press the crank - just need to borrow a lump hammer, as that’s not something I own either (before you get worried, I’d use a bit of wood between the crank and something damaging like that).
Very uneconomic use of my time I suspect, but it’s kind of fun!
One important detail is that the nut needs to be much harder than the shaft (or vice versa). If the two have similar hardness they take turns cutting each other, leaving each with useless discontinuous threads. I make my own case-hardened nuts from ordinary steel nuts so I don’t really know how easy it is to buy them but they shouldn’t be too difficult to find.
Go on then - how do you case harden a nut? Is heating and then quenching sufficient? Though presumably I need to heat it to a higher temperature than is easy using standard domestic kit.
I suspect I might not have too big a problem with the nut being harder than the shaft I’m trying to cut though - it seems to be pretty low quality.
I saw this thread, and I myself have developed a click, so I thought I’d post about it.
I’ve been riding my KH24 since Christmas, but in the past week I’ve noticed a clicking sound. It usually happens after I static hop, rolling hop, drop, etc. In the next revolution or so of the cranks, I hear a click.
I put lube on the places where the spokes touch, and it still happened, so I don’t think it’s the spokes. I’ve tightened my seatpost, cranks and pedals, and it keeps happening. The harder I listen, the more I think that the sound is coming from the right crank, though at first it may have come from the left, so I don’t know 
So, any advice? Do you think anything’s broken?
Thanks.
You could try the tightness of the bearing caps ???
Oh yeah, I also tightened them 
I was hoping you would ask. Case-hardening is one of the most interesting and surprising simple things you can do to steel. As you probably know, if you get steel hot to just below the point where it begins to flow, the rate at which it subsequently cools, called quenching, can greatly vary its properties. Fast quenching makes it hard but brittle. Slow makes it tough but soft. There are many applications, such as swords or a nut used to cut threads on a shaft, where the ideal would be a very hard outer shell with a tough interior. In fact, an ancient form of sword making forges together, rather like an onion, progressively harder shells of steel. Case-hardening does the same thing to any shape and is incredibly easy. You just throw the red-hot steel into a container of case-hardening powder. The powder initially has very high heat capacity, which makes it quench the outer layer of metal rapidly. The first few micrometers thus are very hard. Now for the intriguing part. As the power absorbs heat it is transformed, losing its heat capacity. The result is a quenching profile ranging from very fast on the outer skin to very slow at the core. Even for something as small as a nut you may have to wait several hours or even a day before it comes back down to room temperature.
I was hoping you would ask. Case-hardening is one of the most interesting and surprising simple things you can do to steel. As you probably know, if you get steel hot to just below the point where it begins to flow, the rate at which it subsequently cools, called quenching, can greatly vary its properties. Fast quenching makes it hard but brittle. Slow makes it tough but soft. There are many applications, such as swords or a nut used to cut threads on a shaft, where the ideal would be a very hard outer shell with a tough interior. In fact, an ancient form of sword making forges together, rather like an onion, progressively harder shells of steel. Case-hardening does the same thing to any shape and is incredibly easy. You just throw the red-hot steel into a container of case-hardening powder. The powder initially has very high heat capacity, which makes it quench the outer layer of metal rapidly. The first few micrometers thus are very hard. Now for the intriguing part. As the power absorbs heat it is transformed, losing its heat capacity. The result is a quenching profile ranging from very fast on the outer skin to very slow at the core. Even for something as small as a nut you may have to wait several hours or even a day before it comes back down to room temperature.
This is an answer to aracer’s #30.
Case-hardening is one of the most interesting and surprising simple things you can do to steel. As you probably know, if you get steel hot to just below the point where it begins to flow, the rate at which it subsequently cools, called quenching, can greatly vary its properties. Fast quenching makes it hard but brittle. Slow makes it tough but soft. There are many applications, such as swords or a nut used to cut threads on a shaft, where the ideal would be a very hard outer shell with a tough interior. In fact, an ancient form of sword making forges together, rather like an onion, progressively harder shells of steel. Case-hardening does the same thing to any shape and is incredibly easy. You just throw the red-hot steel into a container of case-hardening powder. The powder initially has very high heat capacity, which makes it quench the outer layer of metal rapidly. The first few micrometers thus are very hard. Now for the intriguing part. As the power absorbs heat it is transformed, losing its heat capacity. The result is a quenching profile ranging from very fast on the outer skin to very slow at the core. Even for something as small as a nut you may have to wait several hours or even a day before it comes back down to room temperature.
Hmm - so heating with a blowtorch and quenching in water wouldn’t do any good? Had a funny feeling it involved stuff I don’t have.
Thanks for all the help anyway - engineering advice at a level I didn’t expect to find on here.