Troul 

just tighten it down until everything is flat. what could go wrong?

john m flores 

It's been a while since I studied, but I did take some structural engineering classes in college, and I've have no concerns about using a couple of brass crush washers to help conform to the non-flat surface and then a standoff to clear the fork and other bits. There will be two forces acting upon the bolt,

A) the bending force caused by the rack mounting point being offset from the fork mounting point. In this case, the lever arm - even at 8mm - is so short that 40# of downward force will not exceed the yield strength of the bolt.

B) the sheer force caused by the downward weight. Again, the forces at play here are so small that the bolt is not likely to sheer off.

I have a Tubus Tara front rack and use those spacers without issue. I don't think that a counterbore is necessary. If anything, if the counterbore has any sharp edges - either at the bottom of the bore or at the top - that becomes a stress concentrator. A couple of soft brass crush washers should be sufficient.

smasha

I know, I know... Just call me Mr Conservative.

smasha

Across the radius of where the crush washer would fit, one side would have about a 2mm gap while the other side comes into contact with the fork. I think that's not going to work.

I'm not worried about shear strength. It's the bending forces, with weight and time, that concern me. A spacer/standoff acting as a sleeve can alleviate that, if I can get both ends of the standoff flush.

Makes sense to me about sharp edges on a counterbore. I'll have to check if that can be done with filleted edges.

smasha

Wondering if I should change this project from a a Giant Cross City (aka Escape) 2 to a Trek FX 2. Seems like very similar geometry, designed for a very similar riding.

The specs for the FX2 say "Max tire size - 38c without fenders". I'm wondering if I can push that to 42, since I already have 700x42 (44-622) tires for this project, but 38c would be fine, if I need.

Or maybe find a fork from a Trek FX 2, if it can fit wider tires, and put that on the Cross City 2 frame? Would that likely work?

1st picture is a Trek FX 2 fork/dropouts, courtesy of a second-hand listing. Looks like a dream, where I'm sitting.

2nd pictures shows colour (close enough to my Cross City 2) and mid-fork eyelets.

john m flores 

A spacer/standoff is not going to alleviate the bending force; it's going to be an additional force applied axially.

40# of force acting upon a 8mm lever arm is not applying much bending force at all, certainly not enough to bend the bolt. Try it yourself - put a similar bolt in a bench vise with 1cm exposed and try to bend it with a pair of pliers. It won't budge.

grumpus

Well it wouldn't be brazed. My experience of frame building doesn't extend to aluminium, but if post weld heat treatment is required you might struggle to find someone to do it. Then you'd have to paint the fork. It might be easier and cheaper to find a nice steel fork, or at least an aluminium one that has functional rack eyes.

smasha

You're the second person today (first person here) who mentioned post-welding heat treatment.

For a city-bike, a can of spray-paint might do for painting. Or I might wind up with a Trek FX2 fork on a Cross City (Escape) 2 frame, and the colours should be close enough.

Anyway, one local shop got back to me with this:
To which I responded:
I'm actually not sure if I'd need a large (or larger) FX2 fork. Are forks and steerer-tubes the same size, for different size bikes? Are forks the same size, but steerer-tube lengths vary?

smasha

if I have a 5mm bolt inside of a standoff with an 8mm OD, and all faces/surfaces are flush, and there's a load perpendicular to the bolt... I'd expect the standoff to transfer some of the leverage force into a stretching force on the bolt.

Metal fatigue, though. Over time, it seems like that's just asking for trouble.

grumpus

Steerer varies by frame size and by model, axle-to-crown only varies by model (except in the case of different wheel size for different sizes of the same model). When you buy an aftermarket fork it usually comes with a longer steerer than you're going to need and you cut it down to suit your stack height, but buying a used fork you need to check steerer length. Have you looked at Surly's touring forks?

smasha

The Surly forks look good, except I'm looking for aluminium.

maddog34

the shear load capacity of a 5mm bolt is a bit below one metric ton... an aluminum bike rack is rated to about, what, 50 Pounds? :-D

i'd suggest custom grinding or filing a standoff out of a tall V-brake pivoting base(the taller aluminum part, they come in two thicknesses), but then you characters couldn't rack up another 85 posts discussing this simple fix.

carry on

grumpus

Yebbut /fatigue/ ... the fact that you'll have to overcome half a ton of clamping force before you stress cycle the fastener is easily overlooked.
I keep a few pieces of round bar in brass and aluminium for making spacers - I can have the spacer made in less time than it would take me to scratch around for something to improvise with.

grumpus

Because?

smasha

That's cool, but shear-strength of a bolt matters when the two surfaces held together by that bolt are (a) perpendicular to the bolt and (b) flush, and (c) the forces are perpendicular to the length of the bolt.

Under a purely shear-strength condition, bending strength is essentially irrelevant.

As the surfaces held together by the bolt are not flush, then shear-strength becomes less relevant, and bending forces become more relevant.

smasha

Because I like aluminium bikes. For how/where I ride, they appeal to me more than steel or carbon fibre,

If I had $5-10k to spend on a commuter/city bike, I'd seriously consider titanium. Of course, locking up in the city is another disincentive to titanium.

smasha

Maybe this is a viable (safe) solution, using the Escape/ Cross City fork?

This was mentioned previously, but I just now got a chance to mock it up. I'd probably use a nylock nut, and all stainless hardware, if I were to go with this.

A 20mm M5 bolt holds the bottom rack-flange to the fork's eyelet, near the dropout. There are washers directly on both the inside and outside of the rack flange, and a nut on the inside of the rack flange, tightened against the bolt-head.

The tension on the nut prevents the bolt from turning, there's a small gap between the rack-flange and the fork, and there's no pressure against the curved/angled eyelet face on the fork.

This gives me about 7mm between the rack-flange and where the bolt goes into the eyelet.
(Edit: With a nylock washer and a flat washer, this may push the flange to about 8mm from the eyelet)
(Edit 2: Maybe I should use liquid thread-lock, and a normal nut, to minimise the length/leverage between the flange and fork)

Not ideal, but good enough?

What could go wrong? What failure modes should I consider?

Picture shows the drive-side of the fork, picture from the front, facing rear.

rumrunn6

don't overload it, don't get a flat, bring P-clips?

smasha

I intend to stay within the 18kg rated limit of the rack (9kg per side), if that's what you mean by not overloading it.

I ride Schwalbe Marathons, so flats are rare, and I have the gear and know-how to fix a flat in the field.

P-Clips... Maybe not a bad idea.

Trakhak

A warning about P-clip use: I used P-clips to install a cheap rear rack whose lower legs were not quite long enough to reach the corresponding holes in the rear fork ends. In the course of the first ride with lightly loaded panniers, the P-clips slipped down past the end of the seat stay and settled on the fork end, scraping it badly before I could get home to fix it. (Aluminum frame, coincidentally.)

Yan 

This will not work long term. That screw will eventually shear.

The friction force between the mated surfaces is a large portion of the shear resistance. Let me phrase it another way: the threaded portion of the screw is meant for tension. Tightening the screw compresses parallel surfaces into each other. The resulting friction is what keeps the surfaces from moving against each other.

Right now you have zero friction because the mated surfaces are not touching each other at all. The entire shear force is being taking up by the screw itself. The screw is not designed to work this way and will eventually fail. This is not a question of if, it is a question of when.

If I was the owner of this bike, I'd be shopping for a replacement fork. They are not very expensive.

smasha

This mostly makes sense to me.

The part that doesn't make sense is this: Most/all (non-carbon) forks that "properly" support a rack require a standoff/spacer between the lower rack flange and the lower fork eyelet. Assuming that the ends of the spacer are flush with the mating surfaces, that spacer is maintaining a tension force on the bolt, but still creating leverage when there's weight on the rack, and shear forces (as you've described) don't seem to apply. In other words, the way it's typically done precludes shear strength (as you've described) from being a factor; instead, a leverage/tension force on the bolt is the limiting factor. By any measure, this is orders of magnitude weaker than shear strength.

* Does this mean that even those setups will fail; not if but when?
* Compared to a setup like that (requiring a spacer), is this proposal any worse?

Most racks come with spacers for this. The Tara rack comes with 8mm spacers. This implies that I can safely "get away with" 8mm of distance between the rack flange and the fork eyelet, at least with some tension on the bolt and flush mating surfaces on either side of the spacer. Compared to that, it's lacking tension on ⪆2mm of "naked"/"exposed" bolt length, and there's no spacer that may otherwise convert some leverage force into some tension force.

john m flores 

You're operating under the misconception that putting tension on the bolt somehow counteracts the downward force of the rack and whatever is on it. It does not. A simple force diagram will show that. Putting the bolt in tension will not magically change the direction of the downward force of the rack.

You can test this by putting the bike on a scale and putting 30 lbs on the rack. Tightening and loosening the bolt will not change the number on the scale.

Whether or not you add a spacer or how much you tighten the bolt will do other things - increase friction in the threads to deter loosening, increasing the tensile forces acting upon the upper surface of the bolt (the tension caused by the tightened bolt + the tensile forces cause by the moment caused by the load, etc...) but it will not change the downward load and the sheer forces causes by it.

smasha

That's kind of my point; if an 8mm spacer over a bolt is suitable for this task, then is ≈7mm of bolt without a spacer suitable for this task? Does it change the equation if part of that bolt is under a tension that keeps it flush and perpendicular with one of the two mating surfaces?

john m flores 

The spacers used for the Tubus racks are not structural. They are plastic and their primary job is to keep the legs of the rack away from the finish of the bike. They do not carry any load from the rack itself.

That last photo you posted - with the nut on - is ok as long as you tighten the nut against the nut. With that tightened, the geometry of the rack will prevent the bolt from unscrewing itself.