Frame Building: Day 7

To be clear, a lot of what I write during this class is basically notes for my future reference.

I had my first measurement mishap today. Luckily it was only in the drawing rather than the frame. I’d drawn in the tire centered over the diameter of the tire rather than the radius. This threw off the position of the chain stay dimples. But like I said, it was only in the drawing. The actual position of the dimples were marked with the actual wheel mounted to the free hanging chain stays.

Although I noticed the problem, I didn’t figure out the cause. Joseph did, and that allowed me to continue with brazing them in place. It went, meh, okay. The fit was rather rough at the bottom bracket. The ports are straight but the chain stays are tapered, so every time I inserted the tubes, the points on the bracket points would spread apart again. Long story short, it caused some large gaps that needed to be pinched while hot (a good two-person job) and a truck load of silver to fill. Next time I’ll take more off the drop out end so the bracket end will be less into the tapered section when I trim.

Most of us also had time to align the stays. I must’ve had some good karma because it was nearly perfect. The drive side dropout was a millimeter low and the bracket spacing was 127 mm instead of 130 mm. Both were quickly fixed.

Next was the seat stays, which have been cut, test fitted, and I’ve prepared the points. I’ll be using a traditional points, spooned, and they’re ready to braze when I get in tomorrow. A lot of today was demos so I didn’t have the time to finish them. Also a contributing factor was I wanted a long scallop here, so that meant more time to file and book-match them.

I wasn’t able to use my first choice of seat stays because the end diameter of the .7 mm tubes was too small to fit the plug-in style rear drop outs. That would mean I’d have to cut off a good portion of the small end of these tapered tubes, and that would make them too short to reach the seat tube lug. So I’m using .8 mm. It’ll have a stiff rear triangle. Jan Heine seems to prefer this design but I have not enough personal experience for an opinion.

Regardless, the entire bike is now .8 walled tubes, butted to .5 on the main triangle, except for the seat tube, which is .9/.6 as that is the only diameter we had available.

Besides the seat stays tomorrow, we start adding the brake posts and bridges. As a special bonus, we’ll have the honor of visiting Joseph’s own shop after class. I’m looking forward to that. Not so great is that Ron, the owner and UBI’s chief engineer is heading back to Ashland tonight for the remainder. I’m going to missing his encyclopedic knowledge and technical input. On the bright side, maybe the biggest need for that is gone as there isn’t any more structural work to be done and the jigs have been put away.

Here’s where it stands:

My little corner of the shop. In the vice is the partially completed frame with a Grand Bois Hetre EL on a Velocity A23 rim I made a few months ago. This is the actual tire and wheel I intend to use, which is helpful to have on hand when designing and measuring the frame.

Did I mention it was hot? Today was the hottest day of the year, I believe. Like most shops I’ve been in, it gets hot inside, especially when working. I thoroughly enjoyed the ride home though, despite the heat.

Frame Building: Day 5 & 6

On Friday, I left with the head tube and top tube brazed together. By the end of the day Monday, the whole front triangle was brazed and the chain stays ready form mitering. I needed to drill a vent in the seat tube under the lug. Like most things, front and back orientation is important but easy to reverse.

By the end of today the chain stays will be brazed on, the frame aligned, and I’ll begin on the seat stays, which take the longest.

Here’s where we stand:


Frame Building: Day 4

For those of us fillet brazing most of the day was spent practicing that technique. The lug people began working on the forks in earnest. As a group we went through the process of drawing the rear of the bike to scale in order to select and optionally bend chain stays.

I was able to finish the fork, let it cool and get it into the heat tank before class was over. I’m excited to see the results without flux all over it.

The 20 mile commute is getting easier but I’ve noticed what I’m calling “competitive commuters”, so far only men. More on that later. Time for class.


Frame Building: Day 2 & 3

By the end of Day 2 we’d completed mitering tubes, silver brazing lugs, and copper brazing dropouts. For theory we talked about bike design and geometries, and fitting. It’s the only day we’re given homework, which was to calculate some frame dimensions based on the bike we’re building. Our instructor, Joseph Ahearne, and I are the only two building 650b wheeled bikes, although his is for touring. One fellow is building a purely road bike, another a mountain bike, and another guy is building sort of a roadish cross bike. After test riding Joseph’s personal bike, very posh I’ll add, I’m convinced that a shorter top tube might be the best fit for me.

Ron, the owner and chief engineer, gave me some ideas that caused me rethink the order I’ve been designing frames. Adjusting my process caused me to come up with angles that are closer to available lug angles and tube measurements that come out nice and round. Some of this is beginner’s luck, I’m convinced because the order I’m still using is still different than the one he described (which is bottom bracket drop, seat tube angle and length, top tube, head tube angle, and then down tube.

Day 3: Today we continued practice brazing but also drafted full-size paper drawings of our frames, measured them and had them verified. We also began working on our forks. Mine has a 68 mm rake, which is at the limit of the bender, within 2 mm. I’ve chosen the Pacenti Paris-Brest fork crown for its capacity to accept wide tires, and Henry James insert fork dropouts, mostly because I’m feeling more comfortable with my silver brazing skill than my brass. I feel like I should have the fork completed by the end of the day 4.

Freshly bent fork blades. The dropouts I’m using are beside them (the careful reader will notice they’re both non-drive side). All of that is sitting atop the full-scale diagram of the frame.

More photos are forthcoming. I’m having trouble accessing them from my phone at the moment.

Frame Building: Day One

I’m writing this from a coffee shop, people are dresses in either cycling garb or semi-retro hipster attire, two middle age women are making out on the sofa, my barista is talking describing her favorite form of contraception (male contraception), and I am sipping what is arguably the best mocha I’ve ever had.

Yes, it can only be Portland.

Monday was packed, leaving the house by 6 am to make to the school by 7:45, then all day in class until 5, and then home by 6:30. Although the 100 minute commute is beautiful one, I’m not used to it and it’s wearing on me. I figure that will go away, but it doesn’t leave much time in the day for much else. Working from home spoils a person.

So far we’ve gone over the equipment learned to get the torches going and tune the flames well. We’ve brazed a couple of water bottle bosses, mitered a top tube, filed lugs, and brazed them. That’s just the first day. We haven’t started working on our actual bikes yet but by Friday we should have our forks done, that is for the people that aren’t fillet brazing.

We’re at lunch now. Photos and more to come later when I have time at a computer.

Making Good in My S&S Investment

Last night I put the Soma Grand Randonneur into the S&S case, in preparation for my flight to Portland.

To pack an S&S traveler coupler equipped bike, one much remove the following items: rack, fenders and bottle cases obviously, but also the drive side crank, seat with stem, and handlebars with stem.

Once done, the bike splits in two and fits into the case quickly, perhaps five minutes. There aren’t a lot of ways it fits, somewhere around 1, but how many ways do you need?

I threw in quite a few other parts for the new bike, but as you can probably tell from the photo, there isn’t room for the rack and fenders. Those are stowaway in my suitcase.


If you read the current Bicycle Quarterly, you might be of the opinion this is a pointless exercise. In The article is about Japanese “rinko bukuro” travel bikes They make unsupported claims about couplers on a bike’s performance. The essay continues with a series of photos where Natsuko demonstrates how she readies her bike for travel. (I’m on a plane and writing from memory, so I might have the names wrong.)

It all seems elegant, refined, as if an ancient Japanese tradition, not unlike a tea ceromony. Couplers are for Neanderthals.

I take issue with a couple of things.

First, I have noticed no appreciable differences in the way the bikes rides or perform with couplers, either of them. If this were true, and negative, I wouldn’t have installed them on my Soma.

Secondly, Natsuko’s bike appears to be maybe a 49 cm frame size. A smaller frame is obviously going to fit into a sack much easier than the 61 cm frame size I ride.

Lastly, this method of packing is only valid for carry-on, and only on trains, and only in Japan. Heavy qualifications. As an American island dweller who must fly and required to relinquish his bike to TSA, none of those apply to me.

While I’d rather all forms of mass transit had internal bike racks like Amtrak, that’s not the world we live in. Most of the world doesn’t live in Natsuko’s world either.

For the rest of us, we have S&S couplers and bike shippers. For my efforts, I saved $100 in airline fees. Two and a half more round-trips with this bike and they’ll have paid for themselves.

Update: Explaining why TSA didn’t shut the lid is another topic.

And also this….

By way of BSNYC, there is this, proving:

  • Some solutions don’t need problems (they come with their own)
  • There’s a niche market for cyclists who have had an assectomy
  • Bike design is easy — anyone can do it
  • You can get a patent for just about anything

Where to start? Well… it’s orange. How weird is that?!

Some thoughts: There should be a law against abusing an old Schwinn like that. Where does one get a brake cable that long? From the photos, it looks like ‘riding’ this bike results in quite a lot of pain. Note that this seatless bike still requires a seat tube… I would have moved the front mech to the down tube.  Yeah, that would make this all better.

As the pictures hint, apparently you can’t reach the pedals with your feet slung behind you. WHO WOULD HAVE THOUGHT?!  That’s probably why this guy just left the pedals off:

And so did this guy, and why he added a engine:

Of the three designs here, this the only one where laying your bike down won’t snap your head off.


Frame Design: Final Draft

I went back to the drawing board, so to speak, this time starting with the fork trail and head tube angle I want. Then I worked backward along the frame with the wheelbase and seat tube length as parameters, arriving at this design.

Caption goes here
This design is built around the fork rake. Top tube is shorter than the seat tube now. The seat tube angle is tighter and the bottom bracket angle is more relaxed.

Armed with a 64 degree seat tube/down tube angle bottom bracket shell, I could shave half a degree off the seat tube angle pushing the bottom bracket slightly forward. This change seems more typical of many of the modern rando designs I’ve looked at. They tend to have a top tube a centimeter shorter than the seat tube and an angle there of 72.5 or 73 degrees.

Doing those things let me  lengthen the seat tube  to have a reasonably sized head tube with shorter extensions than were previously required. Modifying the head tube length below the down tube joint was still necessary. That prevents the down tube from interfering with the fork crown as the handlebars are turned. There is ample room there now, so I could lower the entire bike by shaving the bottom of the head tube and adjusting the chainstay angles accordingly but the bottom bracket height is fine.

That’s as far as I’m going with the design. Even this much is partially theoretical in that I’m not sure what lug angle combinations are going to be available. Compromises will have to be made, no doubt. But after letting it cool the last two days, I’m still pretty happy with this last design. Good thing too, the CAD started crashing and now that file won’t even open.

I listed out what I wanted to get out of the course: brazing technique, bending lugs, bending chainstays, and bending fork blades. Three out of four are torturing metal. Also, they confirmed my spot in the 2-day TIG seminar immediately starting the day after the frame course.

Frame Design: Version 2

I realized last night that class starts in just two weeks.

While far from being second nature, manipulating the design in CAD getting a little easier. One key point I was missing was constraints. In Visio you can add rulers to show the length and angle of aspects of your drawing. In AutoDesk Inventor Fusion these measurements (the green lines) actually do something. Specifically they lock in the aspect that it is measuring. You can change the measured value and it will try to adjust the related values to compensate, if it can.

So moving an object without moving a bunch of other things can take a while, but at least now it’s deterministic whereas before it seemed random to me.

Pretty close to final version.
Pretty close to final version. Low bottom bracket, low trail, long wheelbase.

This latest version makes several changes.

  • The head tube height has been corrected and is pretty close to the final version I think. The extensions at the top and bottom account for the headset stack, fork crown, and the reinforcement rings I’ll be adding.
  • The top tube has been lengthened and the seat tube shortened, both to 60 cm. This matches the seat tube length of the Soma Grand Randonneur but gives me extra length on the top tube, where I’ve usually felt slightly cramped.
  • The front wheel and fork have been added to the drawing and I’m mostly happy with the placement and geometry. Toe overlap is clearly not a problem, and both the rake and trail are where I want them.
  • The wheelbase has been reduced 3cm but I may increase it from 106 to 107 cm by lengthening the stays. This would improve stability for no-handed riding and provide room for a frame pump mounted behind the seat tube. Right now the wheelbase is somewhere in the middle of, say, the Soma and MAP randonneur geometries. Pushing it out a little more would be closer to the MAP, but still shorter than it and longer than the Soma.

One thing this drawing has taught me is that I’ll probably need one of these between the front fender and fork crown because of the tight clearance and angles between the two surfaces.

Placing braze ons, studs and other accoutrement in the diagram seems pointless. The actual positioning will be done with the rims in place and using jigs, so I don’t see a purpose in the exercise. The goal of the drawing was to get the tube lengths and angles pretty close, and they are that now I think.

Now I’m excited to ride it!

Don't follow me, I'm lost… (and Italian)