Thinking About My Way

A recent visit to Dave Bohm's shop has got me thinking more about what I see myself doing for a living, and what kind of shop I want to have.

Dave has a relatively clean indoor shop so all the dirty stuff happens outside. Indoors it's very quiet and a lot like a jewelry studio. Lots of files and vises. His S.O. has her home office in the same room, and Dave's got a triple-flatscreen CAD workstation running SolidWorks. Outside there's two small buildings, plus a big yard for big jobs. He had a lot of square steel tubing for making a big fence for a custom home nearby.
Attached to the house, there's a small machine shop packed with 2 horizontal mills, one with a vertical head attachment(?...!), a benchtop CNC lathe, and a South Bend 10 Heavy lathe. That room is oddly shaped, and somehow also holds a TIG welder, an O/A rig, a big (and really badass) plasma welder, an Anvil Super Master frame jig with tandem attachment, a slew of smaller Anvil jigs, and the washer/dryer. Next to the house is a paint shop Dave just finished building that also houses a Henry James alignment system. Next to that is a storage shed for all the other crap. And outside there's an abrasive mitering machine and a tandem-sized sandblasting cabinet. I'm definitely forgetting a few things.

So he has an incredibly well-equipped shop for doing all manner of deeply customized work. He has a comfortable space where he can do intricate cutting and filing for hours, but also a lot of precise, high-tech machines. It's taken him 15 years to collect everything. To my mind it's a dream set-up.

His tools allow him to go far beyond assembling purchased parts. He cuts every headbadge by hand from raw sheets of silver. His dropouts are laser-cut locally from stainless plate, then heavily thinned and modified. He CNCs his seatpost binders in-house.

He said a lot customers ask him to CNC custom parts, but they're usually surprised at how time-consuming (and thus costly) the process is: Even after a lot of design work, "the first one is never right."

There are several forces colliding at Dave's shop. The fine traditional craft of building beautiful lugged steel frames and forks is present in force. His 10th frameset is on display (still dirty from yesterday's ride), with abundant ornamentation that would make Hetchins pround. Then again he was working on a very custom trike for a disabled woman, involving plenty of CAD/CAM and even a little carbon. And then there are the almost outlandishly oversized mountain forks, complete with laser-cut plate crowns, steerer turned from solid stock, crimped left blade for disc brake clearance, turned blade caps, and so on. TIG and brazing, hand filing and CNC--making disparate processes work together is what I like most about Dave's work.

We agreed that more folks should have lathes in their shops.

On Selecting Tubes

This is the first in a series where I'm going to try to summarize what I'm learning into concise notes. The idea is to isolate the variables involved in making a choice or performing a task.


Smaller is more flexible and thus better suited for lighter riders and those who don't need/want stiffness.
The 29er MTBs I'm working on lately are 38 mm DT, 36 mm HT, 31.8 mm TT and ST.

Wall Thickness.
Thinner is generally more flexible, but may require better technique to join. See also below.
Chris is able to do S3 frames without using a ST sleeve or an externally-butted ST. I had thought that was impossible without excessive distortion.

Alloy Selection.
Air-hardening alloys (e.g. OX Platinum, S3, 853), though often used for the thinnest tubes, are stronger/stiffer/harder after welding, all else being equal.
We talked a little about the engineering specs that are sometimes provided with tubes, i.e. yield strength, ultimate tensile strength, and elongation. Yield strength is the amount (or range of amounts) of force the tube can recover from without deformation. UTS is the breaking point in a pull-apart test.
The most interesting number is elongation. It's a way of describing how flexible a tube is. Two tubes can be physically identical but will flex to different degrees if they have different elongation.
Pipe Dream Cycles has a great tubing comparison chart listing a lot of these specs. Columbus seems to be the only tube maker publishing all three engineering specs on their website, and even there it's confusing...what do "Rm", "Rp0.2", and "Ap5" stand for?

Joining Method.
Brazing results in a more flexible ride than welding, all else being equal. See also above.
Chris says brazing the air-hardening alloy tubes is a waste. Thus it seems Columbus, Deda, and Tange are better choices for high-end, lightweight, flexible frames.

this week

i've been under the weather for most of three days. maybe it's the water out here or something.
we're building a single speed 29er for Mountain Goat cycles. Chris welded up the front end yesterday. OX Gold DT was a pain to miter in the mill because the 38 mm cutter was a little dull, so Chris did it by hand with aircraft shears and the belt sander. it was a little gappy so he did 2 passes with the TIG, the first to lay in some steel to close up the gaps, the second to make a nice clean bead. it came out prettier than i've seen TIG look.
he had me do the chainstay-dropout brazes (sorta testing my skills) and they came out looking OK. it was tough because the dropouts are stainless Paragon sliders, and all we had on hand for filler was low-fuming bronze. i remember Omar Khiel wrote that brass brazing stainless is possible, but i'd never seen it done. actually i assumed the dropouts were plain steel when i was brazing them, and wondered why the brass didn't want to flow the way i was expecting. Chris said the secrets to brass brazing stainless are higher heat and super-cleanliness. i think the brazing was complicated a little further by using old was Gasflux brand, usually blue-green, but this was the bottom of the bucket and it had turned brown.
that was last night around midnight, and i took most of today off. when i got back into the shop this evening, Chris had unbrazed those single-bend chainstays and brazed in some s-bends and the rest of the rear end. apparently the single-bends didn't permit enough tire clearance.

July 8

er, having trouble remembering how many days have passed here.
today we finished the Ti coupler job. Chris welded the couplers in, then i used a lot of 3M Scotchbrite to remove the welding discoloration. i took the lockring off, scrubbed up all the parts, reassembled, and finally polished the frame to blend the finish. Ti is really cool that the right chainstay had a lot of dents from chain slap, so we just buffed them out with some Scotchbrite on a die grinder. same with a chainstay decal the owner wanted removed. then just lots and lots of polishing with red Scotchbrite

then Chris had me cut a couple miters for a 29er for Mountain Goat Cycles. he's got a nice sturdy mill, a Bridgeport clone (i think) called Millport. hard to describe all the steps involved, and i'd guess the procedure is different depending on the machine and the equipment used. Chris uses a stout machinists' vise with 2 pairs of X-shaped-blocks to hold the tube. on softer alloys, we feed at 3 thousandths/minute. on OX Plat and such, it's 1.5 thousandths/minute. the miter angle is set by the mill head, which is probably accurate to half a degree.

day 3

has been a long one. it's been about as hot as it gets here in Tempe, high of about 115° F today.
so i did a little more TIG until it got really hot, then took a break inside for the hot afternoon. around 7 Chris came out to the shop and we worked until midnight. still working on putting couplers on that Ti Merlin. problem is the ID of Ti tubing apparently isn't controlled very tightly, so the couplers don't slide in the way they should. so i'm spending a lot of time grinding various parts' IDs and ODs to make them fit. Ti is tough shit; i'm using a serious double-cut carbide burr and a lot of elbow grease and it's slow going.

he also assigned me to work on a frame that's in for repair. excessively thin steel was used either for the DT or the HT because there was some kind of failure at that joint. he had previously used a cutting wheel to chop off the whole front end, leaving the rear triangle, ST, BB shell, and some leftover weld bead. my job was to grind off the excess steel, using an air-powered die grinder with abrasive discs and Scotchbrite, leaving the tubes almost like the front end had never been there. the steel was True Temper S3, so it was again slow going. Chris says the stuff i was grinding was harder than Ti.

we did a some tube mitering on the mill for a 29er made of high-end steel (OX Plat ST and a NOVA-branded, Columbus-made 29er DT so far). Chris uses special cutters he had made at a local tool grinder especially for cutting Ti, but they also work well on steel. he showed me how to set the cutter exactly in the middle of the tube to be mitered using a Digital Read-Out (DRO): move the cutter to the far edge of the tube (just so there's no daylight showing between), zero the readout, move the cutter to the close side in the same way, pres the "1/2" button then the "Y" button, and move the cutter to zero the Y readout. sounds more complicated than it is: basically you tell the thing where the front and back are, and it tells you where the middle is.

he demoed fillet brazing at the very end, after a few beers. i assumed he'd use the GasFlux rod Henry James sells, but he explained that type of brass (nickel bronze?) work hardens as you try to grind it, so he uses regular low fuming bronze (LFB). he demoed what he called the Zanotti-Sadoff method: after tacking, flow a tiny fillet all the way around the mitered tube. then, with a cooler flame, focus heat not at the intersection of the tubes but a little away from it on both sides, and proceed to build a big fillet. this came out looking pretty nice; it needed a little filing to be really smooth but i would have been fine painting it.

day 2

pretty tired from working and cooking most of the day.
we got a late start, i did some more TIG practice on square tube. it's especially hard (i think intentionally) because i'm welding pretty thin square tube (1/8"?) to 3/16" plate. chris says this is good training for BB shell and dropout welds. focusing the heat on the thick part is a little harder with TIG because the torch cone gets in the way a bit, whereas with brazing the torch tip is by comparison quite thin. adapting to using TIG as a heat source wasn't as hard as i expected though. it's like a super hot brazing tip.
th most annoying part is the electrode. every time i touch it to the rod or the metal i'm welding, it gets dirty and doesn't work as well. Chris has a supply of maybe 20 electrodes, so i can just switch to a new clean one, but eventually i have to stop and sand down the whole batch. if i'm trying to do nice welds i go through the batch in a day of welding, which is really more like 3 or 4 hours of actual welding.

then we went to the grocery to get food, and i asked Chris a lot more questions about what tubes to use. he said to call Lon at Nova tomorrow to order. OX Plat is offered in thick gauges, but only in mountain-sized diameters (duh) which he recommended for my dad's frame. my mom is about 5'5" and 135 lbs, so she'll get a ~52 cm frame in road OS size (28.6/31.8), probably .7/.4 gauge. Dad gets something bigger and stiffer, but i'll decide on the phone with Nova.
he has a couple Paragon BBs and Salsa stainless rear dropouts that i'll buy.

he said yesterday that there's a local powdercoater who'll do a bike frame and fork for under $80, 1 week turnaround. that's awesome since i was told $200 and 2 weeks turnaround at a place an hour outside DC.

i asked about seat tube angles. he said the following: general rule of thumb is 73º ST angle. good runners and spinners need a steeper angle, 73.5º. big gear pushers and bad runners get a 72.5º. the half degree actually makes a pretty big difference, because it translates to about an inch over 50+ cm of length.

re: chainstays he uses 412 mm on a typical 700c racing bike. sport tourers up to 420mm. after reading a lot of Rivendell propaganda i find this appallingly short. i think i'll go for about 440 mm on these frames.

day one

first i interviewed Chris for a few minutes about business and frame design. i'm planning to build at least a couple frames for family members while i'm here, so i asked some questions about what materials to use. my dad is a big guy, 230 lbs, 5'6", so even though i'm planning a road bike for him, Chris said to use OX Plat tubes in a thickish gauge, like a mountain bike. Henry James doesn't list them, but i'll see if Monica knows anything. or else i'll see about some Heavy Mettle.

got a lesson and practiced some TIG welding. it's much different than MIG, a little like gas welding. more finessed than MIG, much less bulky-looking.

Chris took me to run an errand at a metal store. it was unlike anything i've ever was warehouse-sized, and filled with metal of all kinds in all shapes. no tools, no hardware, just metal. alu, steel, copper, brass, stainless. 4" diameter by 3" long steel cylinders. bars of solid copper--never knew it came like that. i saw a guy cutting down about 30 tubes at once using a super-sized horizontal band saw.
in Maryland there's a store i get sheet steel from. everything they sell is listed on one sheet of paper. this store needed a book.

at the end of the day we worked on a Ti bike that's getting S&S torque couplings. still not sure how all the parts will come together. a couple Ti rings needed their ID enlarged slightly, so i've been using a die grinder on them for a couple hours.

we're also working on a bench/table for the Trek store in Tuscon. welded square steel tube TIGed together. so i spent a lot of time cutting down 10' lengths and deburring. tomorrow i think i'll TIG the thing together.

the jig is an Anvil Journeyman.
tungsten is used for TIG electrodes because it has a melting point thousands of degrees higher than welding is done at.
In and Out Burger is far superior to Jack in the Box.
everything is bigger in Arizona.