Today, as I was putting more miles on the original epicyclic unicycle hub, I was torquing up a steep hill like mad in 1.5 mode with Phil’s (pdc) triple drilled 170mm Kooka cranks and the axle slipped. Needless to say, I UPDed and looked like a fool which was beneficial to those close enough to enjoy it. Also, cranking uphill at low speed I escaped uninjured. I was close to the machine shop at my lab so I walked back and I tried anchoring it in direct drive to see if I could ride it home in 29" mode. It behaved as if a weld broke. As I was changing gears, I realized it wouldn’t matter which weld broke it wasn’t going to work. It didn’t. I took the bus.
The hub has countless (rather, uncounted) miles on it but I have put well over 1000 miles on it by myself. Many other people have used it for extended periods of time so I can’t say with any confidence how many total miles are on this hub. The hub, in its present rendition, was assembled and has remained assembled since November 14, 2002 so it’s been a long time. I have recently taken to riding curbs at high speeds on my commute as well as hopping in place and riding off curbs. I have been stressing it. Picture riding a curb at a speed higher than you can run out. Stupid…fun. Actually it is in some ways easier than the Coker because the wheel is light and nimble and the corrections can be made more quickly.
I will have to disassemble the wheel on this prototype version to get to the hub and dissect it. I will reweld the broken part and reassemble the whole unicycle. I will try to take photos of the guts after cleaning. Those with Schlumpf hubs should not worry, as I have said before, Florian is a much better mechanical engineer than me as well as a better machinist. He also uses case hardened parts. I will ask him to comment on this particular failure point when I get to it and perhaps he can shed some light on how his design is (hopefully) beefier in that region.
I hope to do this soon because I leave for Thailand in a month. I would like to have one of the SARs riders abuse it while I’m gone which will be three weeks. Steve DeKoekkoek has a NanoRaptor tire of mine I would like to switch to because the Schwalb Big Apple has enough crown clearance at 50 psig but not at 60 psig. It’s a tight fit. Then Steve gets the Big Apple for awhile.
You can have the Nano back any time. I may have to go back to a smaller tire though, I have a sneaking suspicion my 29er rim is too narrow and not enough crown clearance for the BA.
No, I was too lazy to go out in my garage to look at the current clearance before I posted. I just looked, the Nano is not at full pressure and there is 1/4 inch clearance.
I’m weeping…what a sad day. Seems like only yesterday that 'Shifty and I were out hunting eagles together, and now this bad news.
Harper, I hope you’re able to get her fixed up and running again, as I’d hate to have to piece-part her out across the club. But just in case, I’d ask the rest of the SARs to PM me with your bids for the components. I’m thinking I’ll keep the Steve Howard frame, but the tri-drilled Kookas are available, as is the seat and pedal set.
I’ll tear it down while you figure what you can get for the parts independently. I rode the Big Boy in to work today. I was amazed at how slow it was. I also botched the first freemount on the short 6" cranks. I couldn’t stop abruptly and go into a hop either. It was an alien craft.
A sad end to a dream? Or can you get her running again?
Here’s me on the far side of the world, and I knew your unicycle by name and reputation. Think about that - because even if geared unis become commonplace, Blueshift will always have a unique place in history, like the Wright Bros. Flyer, Stephenson’s Rocket, and Campbell’s Bluebird.
Harper, you don’t give yourself enough credit. You had the vision AND the ability to make the geared hub happen …
After seeing this thread I wondered what weld might have broken. I was having trouble remebering the details of the guts of the hub so I looked at a 3D model of the hub I created some time ago using Autodesk Inventor software. The model is of Harper’s second generation hub that can be taken apart without dis-assembling the whole wheel.
A couple evening ago I converted the Inventor model to Solidworks (my current 3D modeling software of choice) and decided to make it available for viewing. I created a couple Adobe Acrobat .pdf files, an .avi file that plays in Microsoft Media Player and an “eDrawing”.
If you are really interested in this stuff, the eDrawing is what you want to see. However, it requires that the eDrawing viewer be installed on your PC. It’s cool though because it allows you to explode the assembly, rotate, zoom, pan, make parts invisible, etc. You can download the free version of the eDrawings viewer at solidworks.com. The eDrawing viewer will also work with AutoCad files.
My guess is the sungear to sungear bearing assembly weld. It’s the weaker of the two welds because it has wall thickness limits imposed by a 17mm shaft and the minimum available diameter off-the-shelf spur gear for the application. I’ll know tomorrow when I take it apart. I haven’t taken it apart for so long I had to look at the drawings to figure out how to do it. I’m still not sure what the best way is.
I hope you get this baby back up and running. I remember, when I first heared about Gunis, I serached and found your plans and develpoment, then reading through the process of making it to its first rides.
I took apart the unicycle and then the hub today. The weld that failed was the one that I suspected. All of my photos came out so poorly I won’t post them but rather link to all of them in a folder on my website and then throw them away later. The photos are here.
The first five, 404-409, show the sungear and sungear bearing assembly removed from the hub. The weld between the two parts is still visible and the sungear is pressed on so tightly it took some doing to get it off. The next three shots, 410-412, show the two parts after they have been separated. The final six, 413-418, show the two parts skimmed, cleaned up, and pressed back together. I machined a small valley between the two parts to get a slightly deeper penetration weld.
The larger, tapped six hole pattern attaches the torque arm to the sungear bearing assembly. The smaller six hole pattern is there for a 1/8" drift punch in case I ever needed to get the bearings out of there. I was glad I at least planned ahead for that. I had also included jackscrew holes in the hub to press it apart with 1/4-20 screws. I will probably have it welded by the end of the day and assemble it later this week. Then the wheel has to be rebuilt.
Sadly, I also noticed a rip in my airseat that I will have to deal with eventually. When it rains it pours.
Here is an e-mail from Florian Schlumpf after I described the failure mode to him. Understandably, with the poor photographs, it is difficult to comment on the failure mode. From his description, his design in this area is entirely different and, from what I can tell, stronger from a construction standpoint and more efficient from an assembly standpoint. My guess is that Florian cuts and case hardens his gears to his own specifications.
Spring steel is exrtremely hard. My torque arms on my prototypes are made of spring steel for that very reason. It is difficult to deal with but hard, tough, and extremely strong for a given thickness.
Greg,
I don’t know your design, so I don’t feel competent to comment on a failure of your hub. We do not weld any parts on our hub. The torque lever, made of spring steel, is pressed directly on the sunwheel, held by three noses, similar to those of an internal hub of a bicycle. Well, as you may have heard, we too had some unexpected surprises, that made necessary an exchange of the first batch of hubs, that were shipped in 2005. I think, a development of this kind always is a long process of improvement. I only learnt after shipping the first hubs, what people are doing with them!
Hope to see you again sometime, too! You won’t be at Unicon this summer?
And here is another e-mail describing the attachment of the sunwheel to the torque lever. This scheme requires no welds and has the potential to use a large contact surface area
Greg-
I made as many components similar or equal to our bicycle gearing system, in order to profit from our experience in that field. All gear and clutch parts are case hardened, to withstand the high torques in a bicycle bottom bracket (our test torque for the bike gearing system is 250Nm / 180 lbft). The sunwheel of our hub has three grooves, the torque lever has three noses, that engage into these grooves.
BlueShift is back together and has been ridden. Everything seems hunky-dory. If my memory is correct, which I doubt more daily, there was more backlash before I repaired it. This is what I think not what I know. Perhaps the weld was cracked for some time and allowed some additional slipping. I wouldn’t think that would last long before it failed completely, though.
I may switch to a different saddle arrangement. I have not yet repaired the tear in the seam of my nylon airseat cover. The airseat I had on it was good for distance. I wonder if nylon covers are still available?
I don’t think the old-school ballistics nylon seat covers are still available. Bedford may still have the Rocks seat cover. The Rocks cover is still in his price list. Otherwise the KH Fusion cover works. Not as much volume as the ballistics nylon seat covers, but works well enough.
