Drop Axle Guni project shelved

WARNING: Those of you for whom tech talk makes your eyes glaze over, read no further!

A few months ago there was discussion of designing a guni with crank axles located below the hub axle, like on those century-old geared penny farthing bikes. But with new materials and tolerances of course. Part of my interest in such a machine, other than the gearing, would be the riding effect of such lowered cranks.

Among the parts needed for a prototype was a hub with secure but swappable cogs on both sides (i.e. anything but standard screw-on track cogs, which Harper has proven to be useless in our world). So I found a Level Components fixie hub, which has bolt-on cogs on both sides.

One of many design issues was how to get the cranks exactly 180 degrees apart. In order to do this, the teeth on all four cogs need to be aligned exactly the same from left to right. I assumed that this hub would be machined in mirror image, such that the teeth of each the left and right side cogs would be positioned exactly the same. Not so. :angry:

The only way to build such a guni is to make a custom hub, so that the bolt-on cogs are perfectly aligned with each other. But I am not willing to invest that much in this project. Therefore another good concept shelved.

Bummer about that. It is a unique challenge to try and retrofit bicycle parts for unicycle use. Sometimes they don’t work out as planned.

But, that is a neat fixie bike hub. I like it when people in the bike industry come up with new ideas to improve things. Too bad they didn’t think to make it symmetric. Maybe they’ll think about that in a future revision. Did you contact the designers and tell them about that problem?

If they made it so the cogs are lined up on both sides it would also be usable as a double chain standard giraffe hub. Though a $180 hub is probably out of the price range for most people wanting a giraffe. And they’d have to consider making a slightly larger cog for giraffe use.

wow, Pete

I was really impressed before with your splendid guni. I really felt bad bashing it’s excess friction, with the up and down power train to the jack shaft and all.
I was so sure of the 6 % power loss per set, based on my knowledge of motorcycle power trains. Of course, I was wrong, not understanding that the motorcycle chain transfered 200x as much power as a bike chain while weighing 10x as much, by using a roller chain. And the bike chain is much more efficient.
I did not understand the difference in the chains at the time I was such a know it all.:o
Now, if I am understanding your post correctly, you are trying the “chain on each side with a sprocket below the axle” design.
Surely, this should reduce drive line friction compared to the jackshaft design.
So you are having alignment trouble, as I understand it. Like the cranks aren’t 180 with the parts you can find now.
All I can say is you are a true uni craftsman hero Pete!:smiley: I don’t know what else to say. I can’t believe this design concept will not see it’s day.
I wish I could help more. You are the man pushing the uni speed record.
Thank you for your efforts. It is an epic endeavor!:slight_smile:

Not to beg into an argument, feel, but you keep saying that the “drive line friction” of my jackshaft design is greater than on a dropped crank design.
Each design employs 2 chains and 4 sprockets. So why would there be any frictional difference of any measurable degree?

And I am not the man “pushing the uni speed record”. I believe that would be the brave riders themselves. There’s a fellow in northern Europe preparing for an attempt to break the current hour record of 16.8 miles.

The chainset on each side design

Moves power through a pair of sprockets and chain at a friction loss of x. The power loss alternates from pedal to pedal, as each takes it’s turn as the driven set. So power loss in the drive train would be x, plus the friction of the sprocket set on the other side that is coasting, y. x will be much greater then y, as it is a percentage of the power needed to drive the whole vehicle, y is a percentage of the power needed to rotate the pedal and unloaded chainset, a much smaller value.
With the jackshaft design, the power on each stroke goes up to the shaft with a power loss of x, and back down to the wheel through a second driven set, at another loss of x. The coasting side pedal does the same. So the friction loss of the jackshaft design will be 2x plus 2y.
We can ignore the minor friction loss of the jackshaft bearings, as it should be the same as the loss from the pedal stub shaft bearings in the other design.
With motorcycle roller chain, x would equal at least 6 % of the drive power. Bike chains I don’t know. Probably more like 2-3 %.
As far as what is the fastest design ?, that may come down to what pedal placement is better for the rider. If having the pedal axis below the wheel axle is OK, then that design has a slight power edge. If it impairs the ride ability at all, that may suck.
Wouldn’t it be great if the lower pedal setup improved stability ! Best case scenario , improved rider power and reduced drive line loss .
Who knows ? One thing I remember from that bike chain friction article, was that a tight chain had lower power loss. That is fortunate, as the setup with the solidest feel is also the most efficient.

what if that hub was machined to accept the same cog, but in a different manner??
the situation i am thinking of is that you could drill and tap the hub in several different places, custom fit to your needs, and have a good way to do it. if you want a new cog, use the same kind, and use a pattern to drill them out within a reasonable tolerance.

i hope that makes sense

Regarding your comparative analysis: 2 short chains on each design = the exact same friction parameter. It matters not which side the chains are located.

As far as tight chains go, however, that is the ONLY real-world issue which you can discern while riding. There is a delicate balance in chain tension necessary to eliminate crank slop, while not inducing chain friction. Ask any giraffe mechanic.

I get what you’re saying. But the hub & cogs are machined with 3 bolts and 3 “lugs”, hence only 3 possible circumferential positions of the cog on the hub. No alteration of their proprietary hub design is possible. Nevertheless, you have solved the alignment issue in theory…nice work!

Pete, you didn’t understand

Lets try again, with just one pedal. When the power from the pedal goes to the wheel through one set of sprockets and one chain, this will have a friction loss of x. If we make the power travel from pedal to wheel through 2 sets of sprockets and 2 chains in series, the friction loss is 2x. OK ? Remember that only one pedal is driven at a time.
Now about chain tension and power loss. I gave you a link to this paper from John Hopkins U before. You did not read it or you have forgotten. It is about power loss in the bike chain sprocket system. One of their discoveries was the counter intuitive measurement of lower power loss the tighter the chain. A most helpful discovery don’t you think !:slight_smile:

I’d like to end your debate by suggesting that the actual riding of chain-driven gunis, as well as giraffes, presents the final proof. You may enjoy your theories, but you can never account for all of the actual variables. Bottom line: street sense (riding impressions) trumps book smarts (formulae) in this kind of inquiry.

To me, this project just wasn’t worth hundreds of dollars to machine a custom spec hub for the application. Seems to be most folks’ roadblock on the street of dreams.

I am not debating the value of experiance

Nor can I debate that resistance adds in series, not in parallel ! I can only point it out. That, and how with your aversion to book learning, I am glad you are not building an airplane. Lets change the subject.
During my time of employment as a welder in a fabrication shop, I saw lots of one off parts spun up quickly on a medal lathe. Since the hub you desire is round, I believe you will be very pleasantly surprised at the ease and low cost of the machining required. Any round shaped aluminum object you can dream up and will fit in the chuck can be shaped in the metal lathe easily, in my experience.:wink:

if you tighten the crap out of a bike chain it puts a huge stress on the bearings, a sideways load on bearings means slower/stiffer bearings. if a chain is loose it has more? power loss because of slop and stretch in the chain. if you tighten it, you increase the friction, but the ratio is in favor of a tightened chain.

I would have thought that a tight chain would have more friction

The article I linked to above is about the only research I could find on this subject. They measured less friction with the tight chain so it’s something to consider. On a giraffe, many comment on how their sprockets are a bit out of round or off centered. That may sku the results a bit. I got the impression that the university coats were using new precision stuff.
They reference a paper being available at the end of the article. Might be worth while to learn some other stuff. Like how much % larger sprockets help etc.

You could remove or machine off the current lugs, epoxy in the original holes, and re-machine. This could be done in a a day of work (obviously the epoxy would need a while to set).

Alternately, just make a custom cog or two. if you can model it in solidworks, making it on a cnc mill could take as little as 15 minutes, and programming it could take less than an hour. Just an idea. Since the cog is a wearing part, I would run off three or four at once, just so you have replacements.

I’d offer to do this for you, but after the KH geared hub came out, I really don’t see why making a jackshaft gearing system is worth the trouble. Honestly, how many of us really want more than 1:1.5? Which is not to say I look down on your previous work at all. Purple Phaze (the name of it, no?) was amazing.

Go make friends with someone with a cnc mill. You’d be amazed how easy it is to make cogs with one.

Probably the wrong place for this comment, but I’d actually like LESS than 1:1.5. I think I’d like my coker geared more like 1:1.35

Answer to expected question. I like the stability of the 36" wheel (vs. 29er)

You’d be surprised how much benefit a 1.22 ratio can have on a 36".

As for machining some custom cogs, between finding a good CNC machinist, and giving some compensation, and the actual time it really takes to make even the simplest parts, it’s a very big job.