You're supposed to install quill stems so the wedge is below the threaded part of the fork, right?
I just bought a bike with 100mm of threads, so most stems won't fit correctly, even slammed.
How important is that rule? Has anyone actually seen a failure from this? I've seen a bunch of stems setup too high, but never a steerer failure.
If I slam the stem I want to use, the wedge would be about half on half off the threaded part of the steerer.

Yes, the expanders have caused a cracks in the threaded section. To be safe you need a long stem like a Nitto Technomic.

In this situation, I’ve seen steerers crack, on the keyed portion of the threads, when the stem is tightened.

There's nothing wrong with quill stem expanders.

The problem is folks always overtorque these things.

Especially if it's a big box store bike with a 13mm wrench bolt allowing the end user with more grip with that hand tool.

That stem bolt doesn't have to be socked down tight. All it needs to do is make it so a little effort is required to twist the handlebars. All the tough twisting jobs a stem ever sees happen while you are not riding. (Knock your steering that hard riding and I promise you, you've got bigger issues than the stem twisting.)

If you really had your stem bolt too loose, it could slide down on a hard bump but it isn't going anywhere. This isn't like the bolt(s) at the handlebar clamp where failure can have really bad consequences. Should this happen, you just ride home with your bars a little (or maybe a lot!) lower, get out the wrench, raise them and tighten slightly tighter. No big deal.

The problem isn't the wedge, or splitting the fork by overtightening.

The issue is the risk of "notch failure", or metal fatigue over time.

The threads cut about 1/3rd of the way into the steerer tube, weakening it significantly. This isn't a problem if the stem bridges the weak zones and carries loads beyond them.

Having the stem end in the threads is a double whammy. Not only are you losing it's support, you're also concentrating stresses where the fork it's weakest.

When you ride hard, like climbing hills, you pull on the bars. The stem deflects, and the upper H/S bearing acts as a fulcrum pushing the lower stem and steerer in the opposite direction. So the steerer is constantly flexed back and forth with the most stress where the stem ends.

It will break eventually, most likely on a hard climb or sprint, or possibly at a pothole if you have weight on it.

So, it's a time bomb, with the fuse length dependent on how you ride and luck.

Years ago, I broke a threaded steerer tube like that. I returned it to the builder, who acknowledged that he had threaded it way too far and replaced it, grateful that it hadn't resulted in an injury. At the very least, I'd second easyupbug's suggestion of using a Technomic stem.

Ok thanks all. I will re-use the original tall stem and re-think bars to get the reach, or fork out for a technomic

I don't agree with this at all. If your stem got knocked loose, it is absolutely a dangerous situation. I find that steel quill stems are especially problematic in this way. Once it gets knocked loose- it is loose. Aluminum stems tend to give more warning, and have a longer way to go between slightly loose and dangerously loose.
Whereas if your handlebars is slightly loose, it might slip and rotate downwards, but you are still firmly in control of the bike.

Apparently you missed my second sentence.

Do that and your scenario doesn't happen.

And I agree with you about the handlebars spinning in the stem clamp - if you are in the drops. But many of the common perches on the hoods? I've heard too many stories of that going wrong.

So I had a look at some other forks. I have several with this problem. Below is off my 57cm Ken Evans, a very nice race bike, which came with a cinelli stem that would have been on the threads even slammed.
So for this fork a normal road racing quill will not work. You need either a quite long stem, slammed or an extremely long stem.
I have a few frames like this.

I realize best practice here, and it makes sense, but what is that actually going to mean? If someone brings me a bike to overhaul that is like this, does that mean I have to talk them into spending $100 on a technomic which only might fit them? Or even just for a $15 headset service. If someone came in with a tear on their front sidewall i'd certainly strongly suggest new tires.
I did some more reading and found Grant Peterson discussing the issue
https://www.rivbike.com/blogs/grant-...-thwack-update
He posts this picture:

This is a relief. If it is not a critical failure which would cause complete loss of control, and it's something that might take decades of riding to occur, I think it might make sense to treat as best practice 'within reason'.

That's usually what I do, just follow the line.

Also, the steerer in question on that Worldrider is not keyed

It does sound pretty scary. Glad that didn't mess you up.
What stem were you using? And what happened/ how did you notice it?

This whole issue doesn't make any sense, WHY do so many companies spec forks with a tonne of threads, it's just insane.

Thd steering forces involved are miniscule compared to the stem to steerer connection.

The "hold wheel between bars and twist" commonly used tends to have folks over rather than under tighten.

BITD we used to have a concept called, "race tight" which I don't hear much of these days.

Our race tight standard for stems was that the stem should twist before the fork blade would flex. This was important because the freedom to twist reduced the consequences in crashes.

FWIW I've been around for a long while and have never heard of a crash caused by stem twist, and I doubt anyone here ever has either, except maybe 3rd hand. It just doesn't happen.

It's an example of an ignorant designer having a bright idea.

The design of forks with short threads is about a century old, and very well suited to the needs of builders and OEMs who know the final steerer length. Since it's been standard so long, nobody gives it any thought.

So, eventually someone selling generic replacement forks figured they could reduce SKUs by making "universal" forks with tall steerers and long threads, which could be cut to size as needed.

BTW- this isn't a problem unique to bicycles. It's an example of what happens because of the loss of institutional knowledge over time. Flawed changes are made because the WHY of things is forgotten.

The stem was custom made by the framebuilder; the quill length was very ordinary.

I was on the return leg of a mountain bike ride when the steering suddenly went all mushy. I think that was the initial cracking of the tube; next thing I knew, the bars just turned completely independently of the fork. I was going slow and so was able to stop, and was luckily only a mile from the trailhead, where my truck was waiting.

I really have no idea how common it is -- in my case, the framebuilder was profusely apologetic, and didn't understand how or why it had gotten threaded so far down the tube. Must've been made on a Friday.

Well this does not work with steel stems. When a steel stem is loose, it is loose. There is very little gradation between very tight and scary loose.
I don't know if you've worked on cars much. On an aluminum engine, you have to turn the head bolts many turns to gradually compress the aluminum and tighten the head to spec. Whereas on an iron engine, like the 67 MGB I had, the head bolts went from loose to tightened to spec within half a turn. That iron just doesn't compress much.

FB and others, this talk reminded me of an incident well over 20 years ago when I was helping my sister organize a move with hired movers involved.

One of the burly guys grabbed her bike to turn the stem to put in the truck, and I said, just a second, I'll grab the allen key to loosen the stem for you.
Burly, thick as a bucket of nails guy just says, "oh no, no need, I always just grab and bars and twist them really, really hard"

I put on my best diplomatic face, and not wanting big burly guy to get offended and perhaps be rough with other stuff he was packing, so I said, oh I have it right here, did it myself, and that was that.
So be aware of burly movers and your beloved bike handlebars.

re "racing tight" , I'm glad to have read that here years ago, and keep it in mind the few times I put on shifters or whatever. Only once have I benefited from only tightening "racing tight" with a small crash, but appreciate having learned about it.

I wouldn't. I just do what is quicker and get paid so I can work on the next bike.

I don't talk owners into anything.

I briefly make the better suggestion, but if I see they don't care, I stop tallking. My security cameras record audio, so if they try to stupidly sue me for anything later, I have proof that I mentioned the problem.

I like money, fast.