construction break

school is out of session until the middle of January, so construction has had to cease for a while. at this point the brazing of the fork is 100% complete, having received a bit of touch-up in the last few days before the break.
meanwhile i've found some info on local biking clubs and on endurance cycling. today i did a 10-mile ride, my first in a while, and it's got me thinking about doing longer (much, much longer) rides. after all, this is supposed to be a frame for a randonneur bike, but i've never done a randonée.
from the DC Randonneurs:
about randonneuring
lighting requirements

i especially liked this answer to the question, "What is a randonneur?", from the Randonneurs USA FAQ page:
There is no direct English translation of the French term "randonnée", which loosely means to go on a long trip, tour, outing, or ramble, usually on foot or on a bicycle. A person who goes on a "randonnée" is called a "randonneur". (The correct French term for a female participant is "randonneuse", but such distinctions are often lost in America, where we tend to lump everyone together). In cycling, it means a hard-riding enthusiast who is trying to complete a long randonnée inside a certain time allotment. Note that a randonnée is not a race. Overall, about the only thing being first earns is some bragging rights. It is not uncommon for the last finishers to get as much applause as anyone else. Indeed, there is much camaraderie in randonneuring. One does it to test oneself against the clock, the weather, and a challenging route - but not to beat the other riders.
In comparison to other forms of competitive long-distance cycling, such as at the Race Across America (RAAM), where there are following cars with crews supporting the riders every inch of the way, randonneuring stresses self-sufficiency. Help can only be given at the checkpoints along the route, so support crews (if there are any) must leapfrog the rider. Any rider caught receiving assistance from a support crew in-between checkpoints (or, "contrôles" as they are commonly called) will be subject to a time penalty, or even disqualification. Randonneurs are free to buy food, supplies, or bike repairs at any stores they encounter along the route. Once riders have successfully completed a 200-kilometer "brevet", they are entitled to be called a "randonneur" or "randonneuse".

so now i'm contemplating the logistics of doing a brevet, or more accurately, of doing some training rides to prepare for a brevet. my bike clothing wardrobe is set up for commuting, so i have some gear, but probably not the right stuff for hundreds of miles on end.
and there's the big, expensive issue of lighting to contend with.
for now i'm going to sleep.

braze-on distances index

i've added this to the framebuilding wiki at

cantilever/v-brake bosses

From: "goodrichbikes" 
Subject: Re: [Frame] Canti Boss Location - Std Dimensions Available?
Date: Mon, 20 Dec 2004 13:29:51 -0600

700C(622) = 283mm
26"(559) = 253.5mm
650B(584) = 265mm
A post center to post center measurement of 80mm works best for most canti & v brakes.

Curt Goodrich
Minneapolis, MN


Date: Wed, 26 Jan 2000 01:00:11 -0500 (EST)
From: Jeff DelPapa 
Subject: [Frame] measurements question
The last time I asked about canti spacing, the number I got was 80mm center to center,
apparently from the literature of the big S.  I expect it should depend on the rim width,
but they didn't mention a correction.  As to distance from axle, I was told 25mm less
than the bead seat diameter, which worked fine for the canti's I used. (happened to be
scott self energizing -- pre linear pull style.)


Subject: Re: [Frame] Brake Boss Location
From: "Jon_norstog AKA \"Thursday\"" 
Date: Fri, 14 Oct 2005 20:03:56 -0600

I'm working from the 1999 Shimano book.  The spacing between canti posts
depends a little on rim width.  basically it's 77-85 mm, but on a road
bike you could probably go down to 75 and be fine..

I use 80 for mountain bikes and 85 for BMXers and bikes that are gonna
run V-brakes and wide rims.

Different V-brakes will have different tire clearances.  This may not be
an issue with cantis.

Good luck


DT shifter bosses


ISO disc/disk brake mounts

From: ThomasT41(AT)
Subject: [Frame] ISO mounts and 7005 tubing source
Date: Sun, 13 Jan 2002 12:39:27 EST

For those of you interested here are the iso dimensions for disc brake--can
be found on the net but if your lazy like me.............

mounts- 6mm to 9mm thick, dimensions are plus/minus .1mm

rear ISO brake mounts for disc brake are 51mm(between centers) x79.1mmx39.9mm(axle
center to hole centers), inside flush with inside of drop out

front ISO brake mounts for disc brake are 51mm(between centers) x87.3mmx49.7mm(axle
center to hole centers), inside mount surface set 4mm outboard of the inside drop out face

TT brake cable stops/bosses

From: "Andrew R Stewart" 
Subject: Re: [Frame] TT brake cable BO location - since we're talkin' BO's And
Date: Mon, 24 Oct 2005 20:53:48 -0400

Roman- As I like to be able to run a TT pump I make sure that the stop JUST
clears the pump. Trying for as little of the stop hanging out past the side
of the TT. I like the cast bullet shaped type. The distance from the HT
center is about 10cm. My frames are small and I hate cable loops that are
really tight so I give both the brake and gear cables a bit of room for the
casing to loop away from the bars. ( I run Ergo). Another difference I have
is that I run my brakes righty/front lefty/rear so the TT forward TT stop is
on the TT's right side... the rear stop is on the left side... yes the cable
just grazes the underside of the TT. (Oh my kingdom for five Campy rear
brakes made in a mirror image, the cable on the other side). Again on the
rear stop I place it 10 cm from the ST center. Look at a lot of production
alu or carbon frames and on the smaller sizes the rear stop is so close to
the ST that the cable gets a tight loop, enough to cause friction issues.

water/H2O bottle bosses

From: "Andrew R Stewart" 
Subject: Re: [Frame] TT brake cable BO location - since we're talkin' BO's And
Date: Mon, 24 Oct 2005 20:53:48 -0400

As to water bottle bosses. The bottles overlap. The ST one is as low as
possible, straddling the ft der clamp. The DT set so it clears the ST bottle
by a small bit. About 10cm for the ST and 18.5 for the DT, lower boss to the
BB shell outside surface (add 2cm for BB center).Again small frames and also
46-50 tooth large rings. An interesting story here. Our Co-Motion tandem
also has a set of stoker ST water bottle bosses straddling the der clamp.
they didn't know of our 48 tooth large ring and the der clamp lined up
exactly on the lower boss! So I used a clamp diameter 1/8" larger and a shim
(cut down to clear the boss) and the der mounted right over the boss. Can't
attach a bottle on that tube without some kind of clamping but so far we've
not needed it. Andy S.


Subject: Re: [Frame] H20 boss placement question
Date: Wed, 19 Oct 2005 10:57:01 -0700

I have a drilling jig that I use for this.  It places the bottom hole
for the cage at 16cm above the crotch of the ST/DT.  I use it for both
DT and ST boss locations.  Unlike some, I drill and braze WB's on after
the frame is built.
Omar Khiel


From: "e-RICHIE" 
Date: Wed, 19 Oct 2005 14:44:17 GMT
Subject: Re: [Frame] H20 boss placement question

i like my two opposing sets of bosses to
be "symmetric". when the seat tube cage is
lower than the down tube cage, it bothers
my eyes. my lower cage boss is usually 19cm
from the bb centerline.


Date: Wed, 19 Oct 2005 11:39:32 -0400
Subject: Re: [Frame] H20 boss placement question
From: Doug Fattic 

My sense of proportion is the same as Richie's.  I like the seat tube bottle
to be the same distance from the bottom bracket shell as the down tube
bottle.  Sometimes this has to be altered with small frames.  How low they
can be depends on several factors (bb angle, cage, etc) so when I am
particularly fussy, I will put the actual cage with bottle in place and see
where they comfortably just miss each other and mark the spots.  Of course
they shouldn't be so close that a change in cage will make the bottles bump
each other.  I measure from the top of the bottom bracket shell and this
distance is usually between 6 3/4" and 7/8ths.

Doug Fattic
Niles, Michigan


Subject: Re: [Frame] H20 boss placement question
From: "Jon_norstog AKA \"Thursday\"" 
Date: Wed, 19 Oct 2005 23:02:47 -0600

I've got a Waterford Paramount in for some frame repairs that has the
lower boss 7.5 inches above the crotch.  On mountain bikes with short
seat tubes, I move the bosses down, 5.25" or so above the crotch, which
gives you a little more than just enough clearance for the derailleur



From: Margo Conover 
Date: Wed, 19 Oct 2005 11:41:03 -0600
Subject: [Frame] H2O boss placement

I think this "all depends."  On tiny frames I have to go really
low--like 11cm--in order for the rider to get a big bottle (sometimes
even a small one) in there, particularly if they use a top tube mounted
frame pump.  Also, if the frame uses a clamp-on front derailleur, then
you need to be very aware of where that clamp will sit on the seat tube
when placing the lower boss.

Margo Conover
Luna Cycles, LLC


Date: Wed, 19 Oct 2005 16:48:59 +0200
From: M-gineering 
Subject: Re: [Frame] H20 boss placement question

Tom Palermo wrote:
> Seems like most production bikes are hovering around 15cm.

depends on the size of the frame the angles and the cages. I like to
space them fairly symmetrical and often use 200mm (downtube) & 180 mm
seattube. (c-c). But with small frames and big bottles this won't work


over-/under-BB shifter cable guides

From: "Bill Boston" 
Subject: Re: [Frame] over BB cable guides
Date: Fri, 27 Dec 2002 15:55:02 -0500

I always liked to run the cables over the bottom bracket, but never liked
the Campy guides. I preferred to place 2 cable stops on the downtube, one on
the chain stay and one on the back of the seat tube. this provided good
shifting and it was far easier to paint the frame and was also easier to
clean the frame because toy could move the cable housings. If you go this
route, NEVER use plastic covered or lined cable housing as it will trap
water and dirt and eventually seize up. Bare Stainless cable housing is the
only way to go. Another advantage was that you could bypass the cable stop
on the back of the seat tube and run directly to the cable stop on the front
changer if necessary. This was particularly useful with some of the older
changers that needed a cable housing to function. Fortunately those are no
longer being made.

I probably have a stack of the Campy braze on guides kicking around here
someplace if anyone needs them.


----- Original Message -----
From: "tim paterek" 
To: "cliff mcleroy" 
Sent: Friday, December 27, 2002 3:13 PM
Subject: Re: [Frame] over BB cable guides

> My feeling on under or over is 50-50. They both have their advantages and both have
> their disadvantages. By all means, the simplest to install is the underneath plastic
> fitting that Shimano makes -- it only requires one tapped 5 X .8 hole in the bottom
> of the BB. The easiest to build with is the integrated underneath guides like the
> ones Henry James provides. One hint for topside guides: I have always tacked the
> tail end of the left guide to the BB shell/seat tube with silver after bending it
> in towards the seat tube. This prevents the guide from bending upward due to hard
> shifting.
> Tim Paterek

front pannier/low-rider boss

Subject: RE: [Frame] low-rider and rear rack mount locations
Date: Thu, 18 May 2006 13:17:22 -0500
From: "Dave Gray" 

We use of eyelet to center of lowrider rack boss.

Dave Gray


From: "Omar Khiel" 
Subject: Re: [Frame] low-rider and rear rack mount locations
Date: Thu, 18 May 2006 11:24:27 -0700

I use 165 also.  It works for the rigid mounted Blackburn racks.  I measure
c to c eyelet to boss.
Omar Khiel

fork = brazed (mostly)

some preliminary photos, taken right after i turned the lights on after brazing the blades and stiffeners into the crown. this was with 56% silver filler; the fork was built with 3 different fillers: brass at the steerer, 50N at the dropouts, and 56% at the crown-to-blades.

after soaking in warm water, i discovered i had skipped brazing a couple places. there was a lot of ground to cover in one session and the flux covering everything made it hard to tell what had been done. i'll go back in to touch up later this week, and fill in the vent holes at the same time.

i think what i did braze went generally well. there were a couple tricky spots, where i think i should have maybe switched to a smaller tip or a cooler particular the club cut-out in the 3rd photo above. that area just kept overheating, no matter how delicate i was with the flame. i hope to be able to get some silver to run in there in a few days after plenty of clean up.
in the same photo, you can see some excess silver built up in the concave shoreline of the crown. i was afraid of excess there, because it's so hard to get at. my plan is to try re-heating delicately to convince that silver to flow deeper into the crown. if that's a no-go, i'll try an abrasive stone in the air-powered die grinder.

i use a couple small rare-earth magnets to hold the stiffeners in place at the skinny end while i tacked them at the wide end. this worked well, and i was excited at the thought of using this method again for similarly delicate jigging, which i've have a frequent need for lately. after brazing, i discovered they had lost their magnetism. bummer. i assume it was the heat, but they must be quite sensitive because they were a few inches away while i was tacking.

in other news, i've discovered my school's blasting cabinet over the past couple days. it's a pretty basic one, and the media is really worn-out i think, but it works well for cleaning up areas that would be hard or impossible to clean otherwise. it's faster than files and shop roll a lot of times. but it doesn't remove a lot of material. i'll try to find some fresh blasting media.

how i brazed the front dropouts

first i had to get everything to fit together nicely. as i mentioned before, i had trimmed a little too much from the tips of the blades, so i added a mm or two of rake on the bender, covered in the last post. then i trimmed the dropouts, getting the plug nice and square with the lip. at some point i notched the top of the right blade to avoid mixing them up, from a tip in the Paterek Manual i think.
i discussed the brazing with one of the sculpture faculty who's also a good brazer. i said i didn't think i could easily pin a plug-in drop, so i wanted to braze in the jig, but i was worried about setting fire to the thing. she agreed, and said i could make guards for the jig, but that plywood doesn't tend to burn quickly so it can simply be blown out. she said folks usually wet their plywood jig before brazing/wleding in them, but i wanted to avoid ever wetting the jig so it stays as true and flat as possible.
so i set about shearing and bending scrap steel to fit the area of the jig near where i'd be heating. this was successful in preventing a lot of flame-up, although the maple rake block did catch fire briefly. i blew it out.
so here's the dry fit, checking that everything is aligned and that all the guard pieces are in the right place. the jig is clamped vertically in a vise.

the side view, now with flux. the drops are clamped down, with flux covering all surfaces that contact the drop. this got all the hardware fluxy, but i figure i can easily get new washers and nuts if need be.
note the piece of steel between the dropout and the rake block. this is a sliding guard for the rake block, U-shaped and bent around it, so it can slide to the top or bottom of the block, guarding it or out of the torch's way, like for heating the hub side of the drop.

another view, showing the sliding guard a little better:

i wouldn't say it was a hard braze, but it was a little nerve-wracking, because i had no way to confirm that i had done a good job. i tried to compensate by basically pouring the silver into the joint, adding more than i figured i would need. this resulted in a few clumps of 50n on the blades, but it comes off easily enough.
i mostly brazed in this configuration, with the jig vertical, so the silver would flow down into the blade, but i also briefly fed silver into the 3/32" vent holes i had drilled, tipping the jig so the drops pointed diagonally down, persuading the silver to flow the opposite direction. how well this worked, i dunno. it did make the vent holes fill up with silver, so i drilled them out again afterward.

the fork, about two-thirds done. the blades-to-crown will have to wait until next week.

still a little silver left to file off of the right blade.

obviously the jig guards weren't totally successful in keeping the jig burn-free, although the plywood pieces never actually flamed up.

so far as i can tell, everything is straight. i'm slowly building a set of wheels to use as alignment gauges.

new bending rig

today i modified parts of my bending setup, and now it's working well.
the big problem before had been that the tip of the blade kept slipping off the hold-down bit and popping up. kinda scary the first few times it happened, although i always had a hand on the blade and a cheater in it so it never went anywhere. but not good for bending.

see this post for a breakdown of the hold-down design, including the original bit:

the problem, i realized, was that the bit was tapered gradually. the blade had nothing to hold it position securely. i did some filing and hacksawing, ending up with this:

this better emulates the R. Sachs method of using a sacrificial front plug-in drop to secure the blade tip.
also, since the original photo, i flattened off a corner of the hexagonal lower portion to better position it in the V-groove.

i don't think i photographed the bender after i modified it to keep the plywood from splitting. as silly as this looks, it works well (especially for a bender that cost maybe $10, excepting the optional $20 router bit). i can hear the plywood creaking apart a little when i'm cranking down, but i can report no kinks in tweaking .9 mm Deda Tre blades.

this is a 90° V-groove; i'd like to try something with a shallower angle, which would probably keep the plywood intact more easily and still support the blade nicely.

the bolts are 1/4"-20 round heads, holes made with a cordless drill, spaced just far enough to allow a washer on either side. i had them about as tight as i could with a big crescent and screwdriver, i might loosen them up for the next bending sesh.

birth of a fork, breaking ground on a frameset

the fork is begun. the 1st real braze went well.

crown is an R. Sachs Newvex, brass brazed to a True Temper Verus 1" threaded steerer.

i used my biggest tip for brazing and got the flame pretty hot (for brazing at least) to get the crown hot enough. it was a pretty long braze, but not an especially hard one. i brazed it upsde-down, flowing brass from the crown bottom to the race. the trickiest part was pulling brass down the sides, i.e. heating up the area around the blade sockets, where there's lots of mass and it's hard to directly heat the desired area. i heated up the adjacent areas, sent the flame into the steerer, and eventually the heat conducted itself to where it needed to go. the brass flowed. and it was good.

i got the flux off with a quick dip in boiling water, provided by the photo dept.'s hot pot. that worked wonders on what would've otherwise been a long, arduous job. i think i overheated a little because the flux was baked on and looked really glassy, almost like enamel.

the joint is now almost completely cleaned up, so photos at this point are kinda unfair. after cleanup i spent a few hours trimming my blades and test-fitting everything in the fork jig. i was tired and got over-zealous in trimming, so i'll need to increase the bend slightly, to make up a ~1/4" between blade tip and dropout ledge.

this is the front, right? i've never installed anything but cantis.

at top middle here you can see a little gap in the brass. i went back later and filled this, as well as a couple other little voids.

back at last, and eve of the fork

a lot has been happening.

the frame:
i plan to begin construction proper tomorrow afternoon, following my furniture class. the first braze is the steerer-to-crown. i'm planning to do it in brass, because i'm worried about the second braze that area will see, blades/stiffeners to crown.
i filed down the stainless plug-in front drops to fit the fork blades, with some difficulty. good clamping of the drop and a good 2nd cut flat file were the keys. they're considerably oversized, or at least they felt that way after a couple hours' work on them.
i still need to trim the blade tops to length, probably using the abrasive shop saw. then i have to wait for an MSC order to arrive with some wire drill bits to replace the ones i lost out of Freddy Parr's pinning kit. the first drill bit (into the lug pre-fluxing) is a #41 wire gauge, the second bit is a #43. these, of course, don't correspond to any fractional size; i tired that.

i had cut my TT and DT too short for the current design iteration, but i just got around to ordering replacements, as well as a new set of rear drops, all from Kirk at Pacenti Cycle Design. the new drops are socket verticals with one set of eyelets. i hadn't realized before i did research that a rack has to be attached to a cast-in eyelet, not a brazed-on one. next project is a rear triangle jig to get those verticals lined up.

i've been collecting new and used parts for this new frame for a while now. last Friday i took receipt of the last of the wheel parts, which came from, Harris Cyclery in Mass., and various eBay auctions and local Craigslist purchasers. so now i'm also working on a set of wheels, including my first non-singlespeed rear wheel.
wheelset specs:
hubs: Deore XT 9-spd 36h rear, Deore LX 32h front
rims: Salsa Delgado Cross rear, Sun CR-18 front
spokes: all DT Alpine III, triple butted, 2.34-1.8-2.0 x 292 rear and 300 mm front
the triple-butted spokes fit nicely into the XL holes of Deore hubs (2.6 instead of the usual 2.4 mm).

my life:
i'm between college visits for a few days now. RISD in providence and WIT in boston this past weekend, Parsons in NYC next. RISD has several amazing shops, but no CNC in the industrial design shop, and it looked like their machine tools don't get a lot of use (i'd like to be pretty handy with a mill and lathe in 3 years). Parsons, i know from a phone call to them, has CNC machines and a 3D printer in their product design shop, plus an interesting program call "integrated design" that seems to consist of designing/making/studying whatever one wants, plus a few required classes.
i should have an internship for this spring lined up by Christmas, by a wonderfully fortunate series of events. specifically, the fact that i can braze is a big reason i'm being considered. i've been busy working on the application amongst everything else.

the semester is wrapping up, and that means finishing the pieces i've been making for my classes. framebuilding skills have really come in handy in working on the piece for my metal sculpture class, which i'll unveil here after its completion.
i underestimated how nuanced filing is. the proper direction and angle of the stroke in relation to both the vise and the workpiece is something learned by doing a lot of filing. mitering practice was a big help there.
at almost the last minute, i started to experiment with oxy-acetylene welding areas i had previously MIG welded. the plan was to smooth out the contours of the joints, which is impossible with MIG. it's typically done by grinding or filing, but puddling the steel and allowing it to assume a rather natural shape at the joints proved faster and more in line with the aesthetics of the rest of the piece. wow, i seriously go to art school.
i also kinda brass-plated one small area by brazing a thin layer of brass over the bare steel. my sculpture teacher, who had seen me unsuccessfully try to brass braze months ago, watched. afterwards, i was talking to him about learning to braze. the part that took me forever to learn, that i think had been the source of much of my frustration, was adjusting the flame at the torch handle; the key is to get the gases slowly flowing out of the tip, and use a barely carburizing flame. the acetylene has to be at nearly as low pressure as is possible, with just a touch of oxygen mixed in. the gases should be faster/hotter for brass, of course, i'm especially talking about silver brazing here.

cut list

these are based on my own AutoCAD drawings, and differ from the BG101 numbers by a couple mm here and there, even after all the tweaking with it i could stand.
the first cut is the steerer length, likely tomorrow evening.
i did a dropout braze today into a scrap fork blade with 56% silver, went quite well. did some destructive testing with a big cheater bar. the dropout bent rather easily, although some small cracks eventually formed on the sides, on the chain side and the opposite.

[in mm]

max headset stack 40
HT length 163.20
crown height 18.5
=221.70 of 240

fork blades (along steerer axis)
axle-crown bottom 370.62
axle-crown top 389.12

head tube
overall 163.20
DT hole, center from HT bottom 23.20
TT hole, center from HT top 35.00
between holes c-c 105.00
between TT/DT edges 72.00
to TT top edge 20.00
to DT bottom edge 5.00

top tube
c-c 562.56
top edge 531.10
bottom edge 530.77
front miter to 31.7 @ 74°
back miter to 28.6 @ 74.5°

down tube
c-c 628.86
bottom edge 594.35
top edge w/o ST miter 612.52
top edge w/ ST miter 593.83
front miter to 31.7 @ 60.19°
back miter to 34.79 @ 90°

seat tube (to TT top)
c-t 570.09
back edge 560.19
front edge w/o DT miter 560.17
front edge w/ DT miter 543.34
bottom miter to 34.79 @ 90°
secondary miter to 31.7 @ 59.69°

cooking vs. brazing [rant]

besides working with metal and wood to make stuff, my passion is cooking.
i learned to cook the only way i can imagine: from watching and cooking with my mom. then i started to experiment on my own when i got my first real apartment a couple years ago. and i worked as a waiter before that, which taught me about how to combine ingredients like a good restaurant does.
a mistake a lot of people make when cooking is that they're scared of undercooking meat, especially chicken. i make chicken so it's undercooked when i take it out of the pan. there's two factors in play: the chicken continues to cook after the heat is turned off (not a lot, but enough that it's noticeable) and chicken tastes best when it's just barely cooked all the way--a little overcooked and it can get tough. so a little pink in the pan is a good thing. it takes a few tries to see how much pink is just enough.
the best way to get a well-cooked thick slab of meat is to brown the outside with high heat (like on a grill) for a just a couple minutes, then turn the heat down to about a medium-low (like in an oven) to fully cook the insides while keeping it tender and juicy. this is how restaurants do it, the good ones at least. i used to always wonder why the cooks always ran between the grill and the oven until i figured this out.

in the same way that a brazed joint can be cut open to reveal how complete the braze was, food can always be tasted or cut open a little to determine both how completely it's cooked and how it tastes. i taste constantly while i'm cooking, and i've learned to associate specific tastes with the way the food looks and cooks. this verifiability is important to me; it makes both processes less daunting.
learning to cook is mostly developing a sense of what's going on inside the food, by extensive trial and error. i sense that the same thing is going on while i practice brazing. i'm at a gross disadvantage because i have nobody to watch and take mental notes from; i would have taken many more years learning to cook without having my mom to emulate. i still call her up for advice sometimes. she's cooked everything, and she gave me a Joy of Cooking so i can look things up on my own. the framebuilding equivalent is the Paterek Manual, although even that doesn't come close to the exhaustiveness of the JoC. here's a blog about a woman cooking every recipe in the JoC

my favorite thing to do is to bring one cuisine in and incorporate it into an unrelated cuisine. this weekend i tried blackened salmon sushi rolls, which was pretty successful. sorta like a cajun version of spicy tuna roll.
i'm very much influenced by other cooks. my grandparents' friend Jimmy Chu used to be a cook at a nice restaurant in Taiwan. he went with me to buy my wok. he showed me how to slice meat thinly, marinade in special soy sauce with sugar and black pepper, then throw it in the wok with garlic, ginger, and spring onion, all sliced super thin. cook on high for about 3 minutes, then throw in some chopped vegetables, and serve over rice. i've been riffing on this structure for a couple years now. i've tried every kind of meat i can get, changed the marinade, used other vegetables, etc.
i still come back to the original quite often. it's an artful way to cook and eat--simple, healthy, pretty cheap, delicious, with plenty of room for variation. and the wok is an incredible tool to use. it requires great care in storage and use, because it's very thin non-stainless steel. a properly heated wok comes to cooking temperature in about 15 seconds. removing it from the heat for 10 seconds to flip things around causes the heat inside to drop significantly--the polar opposite of another of my favorite pans, the cast iron skillet.

anyways, this kind of riffing and influencing is what i find myself doing when i work in the shop as well. a big part is learning the technical side of things, how to use each tool well, learning to cook. but there's always the question of _what_ to make, and for that i always try to pick up ideas from where i work and live, from what surrounds me, or from what i wish surrounded me.

i discovered about a year ago what an amazing culture Ethiopians share. we have a lot of Ethiopians and Ethiopian restaurants here in the Shaw neighborhood in DC. i've asked friends for recipes, figured out the staples of Ethiopian cuisine. it actually has several European influences, Italian and Portugese among them. i've found the markets near me that cater to local Ethiopians. i come in and ask for berbere, the orange-red spice mix, and kibbeh, the flavored butter. i always get funny looks because i'm white but i know a few Ethiopian words, and because i'm usually dressed like a biker.

fork jigged

all the ingredients are ready for the fork. i had been waiting on the blades to arrive from Florida, where John Clay generously raked them--John gets a big thank you.
John's also been bit by the French bike bug. after seeing the photos of his frames with the lovely tight curve starting right at the dropout, i knew he was the guy for the job. they're way better than the pair i screwed up; the bend is straight along the oval axis and the curve is smooth.
the blades are too long, but i took some photos of everything jigged up anyways. it was good to look at the setup and think about any problems ahead of any assembly.

top view, or as close as i could get. the jig is hanging off the edge of the welding table, with the top left edge weighed down by a big hunk of iron.
blades are Deda Tre .9 x 425 mm, steerer is True Temper 240 mm, crown is R. Sachs Newvex, dropouts are cast stainless with two eyelets, braze-ons are big honkin' hourglass things from Ceeway held in place by welding rod.

from the steerer's POV. it was harder than i thought to get the steerer clamped down right. when i clamped one end, the other end popped up the opposite direction. once i shorten the blades i'll have more room for clamps. there's a block of MDF along the steerer to pad it a little. maybe i'll make a couple concave wood clamping pads.

side view. the position of the braze-ons isn't exactly the way they'll ultimately look--i didn't have the numbers handy. the bottom will be 165 mm from one eyelet, the top will be 145 mm from the crown top.

at center bottom is the maple laminate rake block. it's 65.5 mm tall x 80 mm wide. the steerer centerline sits .5" above the backboard, and the axle centerline sits .5" above the rake block, so the rake block height is the same as the desired rake. i should have made it about 10 mm narrower to allow more room for nuts to position the dropouts.
above the rake block is the axle chassis made of 3/16" angle, drilled for 1/4-20 bolts on the bottom and for the 5/16" dummy axle on the sides. these were hard to make accurately, and because of them, the axle is ever so slightly out of parallel with the backboard. i could only notice it with a 12" dummy axle though, and then it was a mm off from one to the other. this is the fourth time i've made those angle bits, and this is the best iteration, so i decided it was close enough.
the idea is that everything but the rake block is reusable on a different fork, as long as it has a 1" steerer, at least 100 mm spacing, and roughly the same length.

looking up the fork from the axle.

what's wrong with this picture?

machining school

i returned Friday night from a 2-day trip deep into Virginia. i stopped in Charlottesville to visit family and a few favorite bars, then toured Central Virginia Community College Friday morning.
I had thought of getting an Associate's in Machine Tool Technology or somesuch from CVCC before transferring for an Industrial Design BFA degree, and i wanted to tour it before deciding. i'm glad i went, but now i'm leaning toward transferring directly to RISD. CVCC's grads seem to come from and/or go directly into industry, and i guess i want a more flexible degree and varied experience. i would like to take some classes at a purely technical school, maybe this spring. Applied Technical Math, Heat Treating, that typea stuff.
this is, again, only tangentally related. and some people reading this see this sort of thing every day, while some will barely know what i'm talking about.

Frank Stewart runs the program. he showed me around the shop and gave me a demo of their Hass Mini Mill by making a set of T-nuts.
Frank's been at CVCC since the early 70s. he's turning 62 next month and plans to retire soon, when he'll start working on his pet project, a motorcycle frame. he had a jig set up on I-beams in a corner of the shop; i regret not taking pictures of it.

here's one of their older mills, a manual machine with 2-axis CNC added--that is, the machinist controls the Z axis, the up-down of the cutting tool.
on a newer 3-axis mill he showed me how it's possible to program the tool graphically on-screen, with touchscreen controls and an animated "dry run" of the program.

here Frank's double-checking something on the Haas. Frank double-checked a lot. he also had to reposition the coolant hose (blue and orange thingy) a lot to keep it pointed at the tool.
they have just one Haas. it's about the size of a couple refrigerators and almost totally self-contained--there's a full guard so Frank stays clean and the chips just fall down into a big bucket; coolant gets recycled constantly. it's not a big stretch to imagine one in a garage, like any other household appliance. in fact he mentioned a guy who works nearby at a big manufacturer, then goes home and makes parts on his Mini Mill on the side.

Frank, lending some perspective, about to start the program, which is transferred to the mill from a PC via 3.5" diskette.

the tool spinning, about to make another pass. it took down the sides in several passes; Frank said cutting away that much at once wouldn't work.

now it's midway through a pass.
after cutting the sides away, the machine autmatically switched tools, drilled 3 holes, switched tools again, and finally tapped the holes. Frank paused the program often to reach in and make adjustments or add tapping fluid.

the finished part. this would be cut into thirds to make three individual T-nuts.

i thought this was about the coolest thing i had ever seen, and when i said so, Frank said, "You know, some folks come in here and see this and just like you they say it's cool. One time I had an English teacher in here and showed him this and he told me he would just hate to stand in front of a machine and do that all day."

so about that black flux...

don't believe everything you hear about it. it's way better than white (for me to learn with at least), but it's not a complete cure-all. witness the following bits.

the BB shell is the Long Shen i filed but couldn't use due to chainstay fit problems discussed in earlier posts. the head lug is a $2 Rivendell reject. i did the shell Sunday and the lug earlier today. on the lug i tried all the silver i have now: Sif 55%, Wolverine 45%, and Harris 56%. the higher-Ag pair felt about the same...maybe the Harris was a little faster to flow, hard to say.

first, BB shells are frickin'hard. there's pins in the way, there's a lot of surface to cover, and the thing is thick. i thought a bigger tip would help with the extra bulk. i tried a #15 in my PurOx/OxWeld rig, compared to the #6 i used for the previous sleeve joint. (the only intermediary step is a #9.) this was so big that i couldn't direct heat where i wanted it when doing more focused heating of the area to be brazed. broad convex surfaces worked fine, though.
also, i find my brazing improves dramatically over the course of a joint. when i start, i tend to overheat things badly, but after a few minutes it's like muscle memory kicks in a little and i'm better. so there's a few spots that are nicely done, and those are the last spots i did. i think before i do any "real" frame brazing, i'll warm up on mild steel.
the damn curly things on the Riv lugs are really tough, they're so small and right at the edge. i burned the hell out of one.
for the head lug i had the shop to myself so i and set up 2 torch rigs. i set a #9 for preheating and a #6 for brazing. i had tried to change tips mid-braze before, but that took so long that it almost wasn't worth preheating with a larger tip. the 2-torch way worked well enough that i'll likely use it again, maybe with a #9 and #4 instead.

it's pretty much a perfect Z shape.

here i it got way too hot and i could tell i'd done it.

this i like. i broke off the pin and filed fluch to try that part out.

no silver penetration around the DT at the BB shell á la Little Fish's first Audax. i'm glad i tried a BB beforehand.

worship at the temple of the black flux

tried it today and loved it. gave me lots more flexibility with heating. i did overheat one part trying to bring the filler down with the flame, but i felt the difference between "filler filling" and "cooked like a biscuit" was far greater. even the part that i cooked had decent penetration.
i had expected more trouble getting the flux off afterwards than with white flux, but it wasn't noticeably harder.
this was another split-tube practice joint, which isn't really the best way to practice i think; sometimes the gap isn't very even so the silver won't flow evenly. i'm gonna try a bottom bracket shell this weekend.

i took these right after i turned off the torch. the metal was still hot enough to need holding with pliers.

here you can see the part i mentioned that i had cooked. after clean-up with hot water, this black spot was still mostly there. other than that blotch, the joint cleaned up to about what it looked like pre-braze.

i'm awful proud of that line of silver all the way round the edges.


black flux, more silver, cotton gloves, one temp. crayon shipping tomorrow from Airgas Reading. the black flux is what i'm really waiting on.

i tried another couple brazes today with a technique recommended by Chris at Circle A Cycles in Providence. i found out about them from a messenger buddy at work; i'm planning to visit when i go up to tour RISD.
he told me to split a tube lengthwise and wrap it around a length of same-size tube, and braze it like a split lug. this worked pretty well, except that there was a part where i think the gap got too big and silver didn't flow. my furniture teacher (who's one of a couple good brazers at school, both otherwise wood people) helped, directing a propane-air torch up through the tube, giving some nice even heat while i provided more focused heating with a pretty big, wide oxy-acetylene tip. i had the fan blowing from the same direction as the propane torch for ventilation, which made the tube into a little blast-furnace typea thing.
i tried to do the same furnace thing later on, but i had the lights on and the torch adjusted poorly so it overheated quickly and the flux burned off quickly.

bending the other day was not successful. the plywood split along the plys with the 90° v-groove cut into it, so i reinforced with 1/4-20 bolts just under the groove. that took a while, and even then i wasn't able to get the bends i wanted. i got one blade curved nicely, but in doing so i kinked it with the solid round cheater bar. that one i bent first on the 8" radius bender, then on the 6"; i tried to start the second blade on the 6", but then it kept popping out of the hold-down bit when i tried to switch to the 8". the second time that happened i called it a day, figuring things were getting a little unsafe.
i'm working on getting new blades and outsourcing their bending. Bringheli's Deda ZeroTres will likely be the winners, but they may be a bit short for the brake reach of the Dia-Compe centerpulls (70 mm). so maybe the pads will sit in the top of the slots, or i'll dig up some shorter centerpulls.

more on the school front as well: i'm planning to take a tour next week of the shop at CVCC, a comm. college in SW Virginia offering an Associate's in Machine Technology. i found out about it through this list of schools offering CNC instruction. i'd been looking for a list like that for a while--it's given me some hope that there are schools out there (actually, nearby) that have programs i'm interested in.

bending, finally

later today i'll (hopefully) bend my fork blades. i've got an 8" and a 6" plywood bender ready to groove with the v-groove router bit i just got. the hold-down/bit assembly worked well on a test bend with brake line. the plan is to first put an 8" bend in both blades, then switch to the 6" if need be to add a tighter curve to the dropout ends.
the only missing piece is a cheater bar. last time i used a partially ovalized piece of black pipe, a la the Henry James instructions, with poor results--the blade de-ovalized somewhat. this time i'll scrounge around for some solid bar that will fit inside the blade. i might even forge a blade-specific cheater bar if i can't find something suitable.
i had really wanted to use sand to help prevent kinks while bending. the problem is, i can't think of a way to plug the blade at the dropout end while still allowing the hold-down bit to fit in. maybe melted wax dripped in from the crown end? i'd need some kind of stop to prevent it from dripping out the dropout end.


i tried a brass joint yesterday with (i think) pretty good results. the joint definitely helped me relax and take my time with heating. also i turned out all the lights for the first time, which helped as well. there was still some ambient light coming from other parts of the shop, and plenty from the torch. i'm _starting_ to get the hang of this.
this was the first test of the aforepictured tube holder, and it worked well. there's a fair amount of slop between the pipe attached to the pliers and the vise, which was only bad when i wanted to precisely position a tube.
the reason i wanted to do that was kinda silly. i'm running low on lugs, only 3 Rivendells left, so i tried brazing two tubes together without a lug, just letting one rest on top of the other. i tried to pin it, by drilling down through the top of the mitered tube, but the nail wouldn't go in right and the flux was drying so i just let one tube rest on the other. this required positioning the bottom tube horizontally, as i discovered when the top tube slid off when the flux heated up. i re-positioned and tried again, and the brass seemed to flow pretty well, except in one spot where i think the miter was off and the tubes were too close.
today i cleaned the flux off a bit and tried some destructive tesing. i held the unmitered tube in the vise and stuck a bar of solid stock into the mitered tube. it pried off without a whole lot of effort. i guess that's the reason for a fillet.

things seem good here.

that big black blotch there isn't so nice. i could tell when i was heating that i overdid it in one spot; i think that was it.

note the hole in the upper tube where i had tried to pin.

you can see a little line of brass around the inside there.

the weird red coloration appeared after soaking in diluted toilet bowl cleaner for a few minutes. today i tried soaking in pure T.B.C. for about an hour with better results. the flux came off quickly with some convincing from a file, although i had to emery off a lot of rust.