Originally Posted by
Jean Daspry
Filling it with solder would be under the brazing method, an interesting idea. I've yet to see an example of a brazed freewheel, but some reportedly have done it.
Now that‘s a constructive answer, thank you so much.
The radial forces on the inward and outward section will be bigger with the pawls jammed backwards, due to the wedge profile? Can you explain how please, I don‘t quite see it.
Do you know how the forces and fatigue of the needle rollers/balls against the outer ring would compare with those on a roller clutch?
I don‘t have an inner threaded portion though, and locking the hub is the goal, the failing would be for it to unlock.
What do you think about adding an identical, flipped freewheel? Surely the load and margins would stay in spec, then?
By the way, what about replacing the freewheel with a locking keyed bushing? Nobody prefers it over the other options?
I endeavor to be constructive.
You're right that a roller clutch exerts high radial forces on the assembly, necessary, because when locked, it is relying on friction between smooth metal parts, and it would be designed for those loads. Ratchets do not exert nearly the same radial loads; If you draw a line tangent to the diameter at the pawl pivot points, then join that with the angle of the pawls when locked, and a third line perpindicular to that first tangent line, you have a right triangle which describes the forces involved; Let us say the pawl angle is 30 degrees, probably less, but that's a good worst case, when engaged with the tooth, the "adjacent" side (across from the hypotenuse/pawl-line), is your circumferential driving force, which is (pawl compression load)cos(30 deg) = 0.87 or 87% of the pawl compression load, pretty good. The radial load is the "opposite" side to the 30 degree angle, so (pawl compression load)sin(30 deg) = 0.5 or 50% of the pawl load, so-so, making the pawl locked angle as 20 degrees will cut that down to 34%, better. 10 degree pawls provides 17% radial and 98% drive loads, now we're talking.
With a roller clutch, it's the opposite in concept; A higher wedge angle will exert less radial load, but may slip. A lower wedge angle is more likely to generate sufficient friction to lock, but at the expense of higher outward radial load, so those inner and outer rings need to be stronger and stiffer (two distinctly different properties, by the way) than with a ratchet. That is the advantage of a ratchet, the assembly can be lighter and manage the loads, at the expense of clicking sound when overunning, and needing strong and hard pawls, pivot pins, and teeth, and very durable pawl springs which are tiny but may go through millions of cycles in the life of the hub.
Trying to lock the pawls outward is even worse than a roller clutch; The rollers are solid and hard so difficult to compress; The pawls, at a shallow angle and the ratchet run backwards, the compression force trying to push the pawls in is great; Possibly something solid steel and very hard and high strength metal (like a roller or ball bearing) behind the pawl might stand up to the load, however, a steel ball or pin behind the ratchet will be pushing on the center of the pawl, and the unsupported end of the pawl will probably snap off, as well as the outer ratchet ring, under outward radial loads well above design, may "burst" circumferentially, as its hardness will probably prevent much plastic stretching before ultimate tensile failure.
Two coaxial opposed ratchets could work. However, bike hub ratchets are designed for human drive loads. If you have a motor that exceeds that load, or, under hard braking you exert greater torque on the ratchet than normal human power, it may overload whichever ratchet is handling braking.
Let's see... did I forget any of your questions? I think that will do for now.
Like I said, I don't understand exactly your intended implementation for the locked hub, if I recall, I think you may have said a motor is involved. But I hope the above helps.