So I've been brewing up some ideas about a more elegant solution to running a modern stem on a threaded fork.

The solutions out there today seem needlessly clunky and make it difficult to get proper slammage etc.

This is my idea:



Essentially it's an oversized locknut. I have about 12mm of thread left over on my fork and could have more if I sourced a headset with lower stack height.

It would replace the locknut, thereby securing and tightening the headset just like normal, but would leave a 1 1/8" column on which to attach a modern stem. The M6 thread is to make it easy to install a topcap for the sake of completion. It would allow you to run the stem flush with the top bearing race. Wouldn't look perfect as it's usually hexagonal, but better than what's out there today.

One could also imagine a solution for a threadless headset where the adapter is cut into two parts. Say 3mm of the threaded part is cut off and threaded onto the fork, the other part is then threaded against it to lock and tighten the headset. This would allow you to run a threadless stem on a 1" threaded fork that is cut too short to do it the conventional way.

Weight? ~30g To be compared with a traditional quill adapter which weighs at least 150g.

Been trying to find a workshop to manufacture a prototype or two. So far no luck (for reasonable money). Since no one has the appropriate tool for the thread, it's down to finding someone with tools small enough to make an internal thread of that diameter on a lathe and it's proving difficult. Though so far I've only been calling local workshops.

Thoughts?

Great idea. Some thoughts:
On the top one, I would ditch the idea of running a top cap. You would have to have the top of the adapter recessed like a bowl for most caps to fit, it would serve zero purpose. I would still have a hole in the top, but hexagonal to fit, say, a 10mm Allen wrench. Then you would have a means of tightening the adapter against the headset race.

You have not catered for how to keep the adaptor / stem from twisting off center in this design it seems ??

Good point about the top cap requiring a recess, I figured the extra mms of clearance from the stack height of the stem should be enough. The measurements are not yet final in this regard. It would be nice to have the adapter end up flush with the top of the stem, not requiring a top cap, but that depends on what stem I (or someone else) might want to run. Will probably have the design adapted to my requirements though since I don't expect anyone else to want one..

In terms of tightening it I've had a few different ideas, the problem with the one you mentioned is that it would require extra machining, the current design can all be done on a lathe = cheaper. One idea was to drill two holes through it, large enough to use a hex key or screwdriver to tighten. That leaves ugly holes though which might be visible depending on the stem design where they end up when threaded on.

EDIT: maxxevv - It wont, since it's tightened against the lower bearing race it's fixed in place, similar to how the lock nut of the headset originally works.

Interesting idea.. What would be the minimum amount of steerer you could thread and stiff be safe enough to resist pulling it off in a sprint or other high stress action.. Seems like a lot or torque on few threads.. What will stop the adapter from unthreading?

What is the actual weight of the current quill adapters? There surely must be room in those to trim the weight down to a more respectable level if weight is the issue..

_________________
Flickr
Tarmac SL3 Project Yellow
Follow on Instagram

That's only in one direction. Clockwise. What about twisting in the opposite direction ?

If you look closely at threaded headsets, the lock washer has an extra indent that sticks into the groove down the center of any typical threaded fork. That's how they prevent it the locknut from unlocking itself. You have not included any feature that caters to preventing it from unlocking itself in your design. Do note that locknuts have a different load requirement versus what you intend to do here.

Note too that your design will not work with most headsets used on threaded forks. Reason being that they require the thread of the fork steerer to terminate below the lock nut / headset lock ring and usually include a seal lip around the quill stem center. Which of course is alright if you want to replace the locknut. But you got to resolve the problem of preventing it from unlocking.

crohnsy - One good thing about the awkward thread used on 1" forks has a really fine pitch (24 threads per inch), meaning it is more resistant to vibration and can be tightened to a higher torque for a given length of engagement. Having said that, 10mm should probably be enough. I'm ~64kg though and not very tough on equipment.

maxxevv - I think you've misunderstood the design. The adapter completely replaces the locknut, meaning it is tightened directly towards the bearing race. The washer to which you refer has nothing to do with keeping the locknut from coming loose, it's just to make it easier to install since you can do it with just one wrench. Upon installation I'd probably use some locktite or similar, but it can be tightened to such a torque it shouldn't really be a problem.

Understand the enthusiasm, but read through what I wrote again, perhaps you'll see what I mean.

Note too that I did mention the intend to replace the locknut.

Go ahead if you think you've got it covered but would strongly suggest you try twisting the steering vigorously in the garage first before rolling that bike onto the road. You'll see what I'm referring to there and then.

I certainly don't want to see some "tarmac polishing" from an inadequately thought out design.

Maxxevv has a point. The stem and handlebar combination is a pretty big lever. There's no telling what might accidentally hold the front wheel in place when moving the bars to the left, you could easily unscrew it. It's possible that with enough speed or a heavy front wheel, gyroscopic centering forces might be strong enough to resist the turn, you'd have no steering at 50mph!

Thanks again for your feedback. This forced me to take a closer look at the requirements of the component and this is my attempt at an analysis.

The problem with threaded fasteners in general is that the problems that can arise are compound in nature. So before determining the force required to loosen the component once assembled, lets have a look at modes of failure for threaded components.

In general threaded fasteners are designed in such a way that the bolt should fail before the threads do. This leads to a situation where the length of thread engagement is recommended as twice the nominal diameter of the bolt. In this situation this is unlikely to be the case, especially since if it was, there would be better alternatives to solve the problem. This recommendation is however for a solid bolt and the steerer tube is hollow! So maybe we can get within this safety margin anyway. Let's have a look. In order to determine a maximum allowable fastening torque for the adapter, lets find the clamping force at which the steerer tube would yield. Assuming a wall thickness of 0.9 mm and an effective outer diameter of 25 mm the stress area is ~68 mm^2. Yield strength for basic CrMo is ~480MPa. Giving a maximum clamping force of 32,7 kN. Under normal conditions we would need to apply 166 Nm of torque to achieve this.

So to be on the safe side we can apply 100Nm of torque to the assembly without risking the steerer tube. This is also safely within the margins of stripping the threads from both the converter and the headset bearing race.

So what is the relation between fastening torque and loosening torque you ask? Well, this is also a difficult question to answer, especially without a lot of test data. But since the friction between the components during assembly is when the parts are in motion, we can get a few clues. Dynamic friction is generally lower than static friction. This means that when the component is in place, it will take more force to unfasten, than it did to fasten it. However, to be on the safe side we can assume that the force required to loosen the adapter will be the same as it was to fasten it in the first place.

This means applying 100 Nm of torque. That is quite a lot. I would guess that 100 Nm would be enough to destroy a lightweight cf steerer. I also put some thought into ways of mounting it and came to the conclusion that using a stem is probably the best option. I have an old, heavy and strong stem I can use for this purpouse. This will also give an indication of the amount of force necessary to dislodge it.

And after all this we can add a threadlocker of choice to the assembly. This would increase the torque required to unfasten the cockpit, and also safeguard to some degree against thermal expansion and vibrations.

I think that about covers it, it's been a while since I worked with threaded components at this level of detail so please let me know if you have objections to any of the above.

In other news I found a workshop that will make the part for me. It will be machined from a 2024-T351. It's really strong, but not very corrosion resistant so will have to be careful in that regard, and make sure to use a compund to keep it from seizing. 80€ for one, 150€ for three and 160€ for eight pieces. Not bad.

Thread strength/stress calculations are not done that way. Do a Google .. the proper answers/methods can be found on the web.

Example: http://www.roymech.co.uk/Useful_Tables/ ... Calcs.html

And to put the stress loads at the thread into perspective, assuming a typical 80kg rider, under hard braking, or in the worst case scenario, encounters a bump on the road, the load is magnified typically 4x.
Assuming, all the weight is tilted forward onto the hoods, the typical distance of the hoods to the steerer joint in question here is approximately 0.25m.

In such a totally plausible scenario, the total moment at the threads is 80x4x9.81x0.25m = 784Nm.

You sure your numbers have a similar basis ???

Nobody is trying to put you down, but randomly plucking/plugging ideas and figures doesn't put things right. There are serious flaws in the idea that need to be worked out to prevent a catastrophic accident on the bike.

I don't want to have to say " I told you so... " later.

Wow!

You clearly have an answer for everybody and money to burn... Look forward to seeing the results..

_________________
Flickr
Tarmac SL3 Project Yellow
Follow on Instagram

There is no reason you couldn't have a small expander inside the steerer tube, that the external adapter screws into. More complicated, heavier, but probably still lighter than what is out there and still "slammable"

Sent from my EVO using Tapatalk 2

Perhaps I'm confused as to precisely how this is to work, but if the adapter is in fact acting as the lockring, no way I'd want to attach my bars to it.

Think I might have to swallow my pride on this one and admit that it's not a very good idea.. Oh well, back to the drawing board.