new BB shell

following my discovery that my Deda ZeroTre 29.9 x 17 mm oval chainstays don't fit in my Long Shen OS BB shell (see this frameforum thread), i've ordered a new Walter shell from Joe Bringheli.

the most recent changes; R. Sachs story

modified the description on the right there to reflect the way the design has progressed. low-trail bikes aren't especially common these days so it seems right to point out that aspect of this one.

in the latest Rivendell Reader (#38), Grant Petersen mentions that their new A. Homer Hilsen has braze-on mounts for the Nitto Mark's Rack. this struck me as a great idea for my bike, as i'm likely to frequently use it with a handlebar bag, when a full-blown lowrider rack would be overkill. i emailed Grant to find out the correct placement for the braze-on, and his reply came the same day:
The Mark's rack is adjustable. We put the braze-ons about 145mm down the blades from the top of the crown, but you can be off by inches and it'll still work. But 145 looks good, I think. It's a great little rack, really useful..
So it looks like the fork will have two sets of rack braze-ons: one for Mark's Rack at the aforementioned spot and another for a lowrider rack at 165 mm from the bottom eyelet, which appears to be the standard from what i've read in the archives. both will be the somewhat-garish hourglass kind, which one brazes to the front side of the blade.
At least for now, i plan to get an Ostrich handlebar bag from V-O.

finally, this email exchange with Richard Sachs made my day.
I wrote to e-Ritchie one August afternoon:
It's hot like a freshly brazed frame here and I'm working (inside) on
cleaning up my frame parts. Went outside to grab the WD-40 to keep
everything shiny, sanding one of your fork blade stiffeners on the
way. Got to a particularly sticky part, sanded a little too hard, and
that combined with the heat (?) caused the thing to bend in half right
at the club cutout.

Is it possible to get a replacement
for it--or, why not, a pair?

e-Ritchie replied 4 minutes later:

He sent them a week later, and a few days after that the box arrived. inside: not one, not two, but five blade stiffeners. he even took care of postage. hell of a guy.

filing fotos

one perk of being a photojournalism student for a couple semesters is that i've got a bitchin' digital SLR. i took it down to my little shop a few minutes ago after a rigorous filing session. notes on the above:
• most of the pre-brazing cleanup of the lugs is done. the BB shell has proved the trickiest, both because of its greater surface area and because it has several tight valleys that are impossible to get into without rifflers. the rifflers seem coarser than the needle files, which makes it harder to precisely file away a specific area, although it also means fewer individual strokes are requires to remove the same amount of material.
• note that i enlarged the hole in the ST and DT sockets. it was a smaller oval shape...i like the larger pseudo-circular window, both for aesthetics and because it will reduce slightly the mass of steel to heat before brazing. i also rounded off the right chainstay socket point, to reduce the possibility of the shell piercing into the chainstay. not absolutely necessary, but i like the look and i'll sleep better at night. i did the same to the bottom head lug's bottom DT point, based on a piece on the Rivendell site. i may go ahead and round off more points, probably starting with the left CS socket for symmetry. the shell is Long Shen LB106, visible in the raw here.
• most of the heavy-duty reshaping i'm doing with a cheap 6" bastard round from the hardware store, several chainsaw files of various widths, and an 8" Nicholson mill bastard. the 80-grit aluminum oxide abrasive cloth i got from Enco is coming in handy here and there as well. i bought the round file after my Enco order, but i'm actually glad i didn't spluge on a nice one because i'm beating up on it pretty badly with all this reshaping. i have a jeweler's saw as well, but i reckon the file is faster for opeing up holes. haven't tried it yet, though.
• Talbot recommends saving most of the cleanup for after brazing, to avoid wasting time in case of a brazing mistake. i'm disobeying him for several reasons: i have free time now, and i'll have less after the brazing is done. i plan to have enough practice brazing that i won't make any big mistakes on the real thing. and i suspect this kind of intricate filing will be easier with the parts easily accessible from all sides, as they won't be once they're brazed around tubes.
• note as well the ghetto-fabulous soft jaw covers on the vise. they're made by bending pieces of aluminum flashing over the jaws. works OK for now, but not durable or sophisticated. i found the flashing at Home Depot in the roofing aisle, for about 32¢ for a 4 x 6" rectangle. picked up about 10 so i can replace as needed.

also included is a photo of the 3/8" pipe nipple modified for use with the fork bending mandrel. pretty simple to make...i bought a high-tension hacksaw and 32t blades that made the cutting take only a few minutes. i forgot to use oil when drilling the hole--big mistake--so i'll have to drill out the remaining scraps of steel later on.

up late

reverting to "creature-of-the-night mode" as a friend puts it. watched a bit of a--dare i say it--cool program on PBS, about designing more economically-friendly spaces and things. PBS has the program website here and there's a seperate more informative one here.
what i watched was mostly on designing buildings to be more efficient. the gist of it was that cities and skyscrapers are generally quite effective places for people to live and work, when compared to less densely-populated areas. they interviewed architects from a big firm that's designed Bank of America's new headquarters in midtown Manhattan, as well as the usual expert types. Brad Pitt did narration. it was very PBS.
anyway, then they started talking about the way buildings are built in the US as opposed to in Europe or Japan. the companies building American buildings are generally not the ones running the building even a year after construction ends, whereas in other countries there seems to be a great deal of emphasis on how the thing will function 10 or 50 years down the line. initial costs are a concern over there, but they're considered along with annual costs, environmental impacts, community effects, and so on.
it strikes me that this isn't only true of American builders--that they're more concerned with making their products cheap up front than the product's lifetime cost. i can see this happening all around me.
it's a problem both of consumers and of producers: American consumers are interested in the cheapest product that does the job right now, and less so with how long it lasts or how well it works 20 years from now.
these are generalizations, but a good example is aluminum bike frames: they're very popular, have a limited lifetime, and can only be repaired by the manufacturer. when your alu frame breaks, the solution is to call up the maker and have them send you a new one. not that many people ride alu bikes long and hard enough to break them, so the makers don't have to replace a lot of frames. and the consumer doesn't care, outside of the inconvenience of being out of a bike while the new frame is in transit. on the surface, alu frames are good for everyone.
but beneath that, there's all kinds of problems. firstly, this cycle isn't good for cycling as an idea. i reckon the reason more folks don't ride alu frames to the point of breaking is they don't like riding them much, and i reckon a good number of those folks would ride more and ride happier on a steel frame. second, these frames aren't free, even if the consumer doesn't pay for them. there's a cost in labor, resources, and energy that isn't immediately obvious.
i'm interested in the more holistic approach, even if it means remaining a niche producer forever.

stirring in the UO-8

i realized late last night that the UO-8 i rode and loved but had been unable to measure is sitting in the basement of my parents' house, where i'm staying nowadays. so i pulled it out and took a tape to it to quantify some of my impressions of it.

turns out it's very close to what i've been getting at. as i've been reading on the archives, it appears to be a low-trail design.

it seems my various influences (see sidebar description) are at odds with one another in a way i'm just starting to really grasp.

but i think i've gotta go with the Peugeot. it really felt great all the time. i hopped on it for a second last night and it just felt right immediately. it's too bad the drivetrain was in such bad shape, so it seems right to build an homage to it.

yet more on trail

Date: Tue, 10 Jan 2006 00:59:40 -0700
From: Mitch Harris
Subject: [BOB] Trail talk

Like Jan said in a previous post, trail is only one design feature that
affects bike handling, and works in concert with the whole bike design.
Since we've been talking a lot about trail, I thought I'd offer this
description of what trail is about and how it affects handling. This is my
take on it that I offer for others to add to and dispute. Most of you
probably know more about this than I do, but here it is. Hopefully this
is standard enough that it will help the trail discussion generally. This
comes from mostly 80s sources like John Forester, corroborated by my own

The gist of the trail issue is that lower trail makes the bike a little less
responsive to the rider's leaning or other body movement and more responsive
to steering with arms/handlebar. Higher trail makes the bike more responsive
to rider lean and subtle body movements and small weight shifts, and less
responsive to arm handlebar movements. This is more true the faster you go.

So a higher trail road bike, like 2.1" or 2.2" of trail, will feel a little
less stable at slow (parking lot) speeds, even twitchy as the bike squirrels
around a bit in response to your body leans and weight shifts at slow
speeds. You get accustomed to this enough that you'd have to ride the two
kinds of bikes one after the other to really notice.) But when you get that
2.2" of trail up to speed (say, 15mph) the bike feels very stable
because the rider unconsiously steers by subtle weight shifts without having
to steer with the bars much. The design challenge is getting this kind of
road race stability at speed while still allowing the bike to be responsive
to tossing into corners by arm steering (and with both high or low trail you
want to avoid wheel flop by getting the right balance). This kind of long
trail has become the road bike standard and people have found it comfortable
for most kinds of distance riding, especially unloaded or lightly loaded
riding (or at least rear loaded riding). I understand Jan's posts on low
trail to be exploring how French low trail is also great for most kinds of
distance riding, and better for some, especially perhaps with cushy 650b
tires, and especially with wieght in a handlebar bag.

In the late 70s and early 80s racing bikes in America started to favor low
trail, 1.8" or 1.9" because most American racing was criterium racing on
short flat tight multi-corner courses where quick arm steering seemed like
an advantage. Achieved with very steep head angles, it was known as crit
geometry and is described as responsive by those who like it, and twitchy by
those who don't. By contrast non-criterium road races tend to favor the high
speed stability of a balanced 2.1"/2.2" trail classic road racing geometry.

In the meantime, designers of touring bikes built for riding with a load
knew to design for lower trail because once you have a heavy load on your
bike you don't want the bike to be responding to every little subtle weight
shift. Also you want a heavily loaded bike to be stable at low speeds. So
you wanted the bike to ignore more subtle weight shifts at low speed and
obey arm steering more. This applies even more to bikes that are going to
carry any significant weight on the front/handlebar because heavy bar bag
weight with long trail will make the steering over-react to subtle weight
shifts. This is unpleasant in my experience with my 2.2" trail long
wheelbase 73/73 degree race bike when I tested it with bar bag and 10 or 15
pounds. That bike tours very well with 25 pounds in a Carradice on the back.

Jan would probably say this is the kind of bike that needs to carry weight
in the back and not the front (as I discovered). He's also pointed out that
the softer wider tires of low trail 650b bikes affects trail. Do a search
for his idea on how tires contribute to trail (he covers this in VBQ more
fully I understand). I wonder how that idea about tires and trail meshes
with the mature designed rigid mtb of the late 80s early 90s that had very
long trail figures more in the 2.8" range plus wide knobby tires. Grant's
MB-1 had the steeper 72 degree head angle (and somewhat less trail than most
mtbs) which he promised in the catalog wouldn't be twitchy because of the
fat tires reducing twitchiness. I've found that to be true with my MB-1.
Interesting how many different takes on this there have been on trail. Tony
Oliver in his very interesting book on bike design has still more
and different ideas on trail, especially for tandems. All of the track bikes
I've had have longish trail in the 2.4" range (even with a steep head angle
the fork offset is minor enough that trail is still high) and are very
stable at speed. This surprised me at first since you think of track racing
as being about maneuvering and not about stability. Over the years of racing
it came to make sense to me because on a good banked track you don't arm
steer much (except for certain maneuvers), and on a very steep banking you
can feel like you are on a continuous straightaway as the track leans you
right over on your side on the turns.

Happy trails,


Date: Thu, 29 Jun 2006 14:53:46 -0500
From: "Kogswell Cycles"
Subject: Re: [BOB] Now: trail, was ew bike from Riv -- the A. Homer Hilsen

On 6/29/06, wrote:

> High trail is kinda imprecise. There's a sort of normal trail for non touring
> bikes the runs in the 58cm area-that's what I use for non touring bikes.
> Here's where low trail comes in: loading the front end. French touring bikes
> were built with long rake/low trail because French touring bikes were loaded
> with front panniers and bar bags and very little weight on the back. In this
> configuration the low trail fork works superbly. Americans and British don't
> load that way, we tend to load the back of the bike. Low trail in this
> circumstance always feels kinda floppy, especially at low speed when

Exactly right.

The French were front-loaders and
English and Americans were rear

We figured that since we were using
French tires, why use an English

The answer is that everyone
has her own way of loading a bike
so why not offer a choice of forks
to accomodate them all rather
than forcing everyone to adapt to
our view of the world?

And don't forget tire pressure.
Lower pressure tires need lower
trail front geometries.



From: Mitch Harris
Date: Jan 11, 2006 1:04 AM
Subject: Re: [BOB] Trail talk
To: Murray Love

This all sounds right to me too, Murray, according to what I know. After
reading your description, it occurs to me that I should have said that low
trail bikes require more arm movement for steering instead of saying that
they are more responsive to arm movement. You're right that they are less
responsive to arm movement which is why they require a bit more of it. I was
thinking of it this way: you steer a high trail bike with lean, and a low
trail bike with arms. Another way to put it, after reading your post, is
that you steer a higher trail bike by leaning and turing your arms, while
you steer a low trail bike by leaning a bit harder and steering with your
arms a bit more emphatically. In practice the difference is subtle and you
can steer either bike with arms or lean, of course, but that is the key

The caster you write about is another way to talk about the high speed
stability that is usually what people are after when they build a higher
trail bike. The last sentence of your wheel flop paragraph is another way to
talk about how you tend to steer a higer trail bike with wieght shifts.
Interesting to think of that as wheel flop since I usually only hear that
term when it means "too much wheel flop," where the wheel turns or kneels
more than you want it to. But it makes sense. The body movement and weight
shift/lean stuff is pretty subtle and anybody who can ride a bike already
does it all the time. It's just that higher trail makes a bike more
responsive to those movements. That's why a higher trail bike rides
no-handed a little more easily than a lower trail bike, because you're
steering with body movement/lean only. And because low trail bikes respond
a little less to lean, they do require subtly more arm steering to turn. I
have race-oriented bikes with both low and higher trail and I like both
kinds of steering geometry a lot. Both descend canyons well at speed which
is where I rely on steering geometry. The only adjustments one needs to make
going from one to the other is that the low trail bike feels a little less
able to ride no-handed at low speed right at first (adjusting to exaggerate
lean a bit takes a couple of seconds), and they feel different climbing out
of the saddle too.

Eddy M.'s bike bike business was getting big in the 80s right when the US
market moved toward higher trail bikes (away from Crit geometry), and I
always thought that might have been part of the reason he kept his front en
geometry a trade secret (he wouldn't publish head angle or fork rake at
all) because he didn't want people to overlook his lower trail bikes if they
saw how his steering figures differed from 80s Pinerellos, etc. He knew
they'd ride great.

One thing about low trail bikes, I think (bike designers please comment), is
that it seems like a bigger chanllenge to design well for low trail. With a
lower trail design you are flirting closer to a no-trail bike, flirting with
the lower limit of what is ridable (at least no-handed), while you try to
find that sweet spot that has a low trail advantage while still having
enough trail to have some caster. That spot would be somewhere between
maybe 38mm and 46mm of trail. By contrast, the advantages of high trail are
available across a wider range. Jan's says higher trail begins about
52mm. Higher trail bikes then would range from 52mm to 75mm if you include
mtbs and track bikes, etc. (Obligatory mention that trail is just one
variable that affects the handling, and not that all high trail bikes would
be nice to ride accross that range.) Think of how many times you've seen
someone out on the street riding a bike with a crash-bent-back fork that no
gives huge trail figures (like 100mm) and they seem fine with it. Bend that
fork forward the same amount and you'd make the bike unridable.



Date: Thu, 3 Aug 2006 15:13:19 -0700 (PDT)
From: alex wetmore
Subject: Re: [RBW] question about Atlantis handling with front load, etc.

On Thu, 3 Aug 2006, dave(AT) wrote:
> I see a number of Atlantis builds with a small Nitto front rack with
> a Bert houd bag mounted on top, or to the bars and the rack. These
> are mounted low on the sides of the wheel, obviously, and so deviate
> from the typical touring load distribution where a lowrider rack
> might be used to counterbalance a heavy load in the rear. These
> builds with the small rack / boxish front bag are more like a
> randonneur set-up.

I had a Quickbeam, not the Atlantis. It is also a high trail bike
though, and I do use it with a handlebar bag. I've also used
handlebar bags with my 57cm Heron Touring, which has a geometry that
is almost identical to the 58cm Atlantis. I also own a Kogswell P/R
and use it with the same handlebar bag (each bike has a Nitto small
front rack and I move the bag between them).

High trail bikes work well with handlebar bags. The high stability
given to you by the high trail makes the bike very stable even with a
heavy front load. The Quickbeam does require more muscle than the P/R
when turning, and I can't ride it no hands while turning, but these
are more minor issues.

> Is it important to have rear weight as well, or does your Atlantis
> handle fine with only a front load.

I've been riding the Quickbeam with only a front handlebar bag.

I prefer the handling of the low trail bikes, but they are more picky
about how the load is distributed on the bike. My P/R requires that
the majority of the load be carried up front if there is a moderate to
high load on the bike. The Quickbeam doesn't care and handles well no
matter where the load is located. I expect that the Atlantis would be

> Additionally, I'm interested in how large 45 mm+ tires affect the
> handling of an Atlantis, as well. I have read that one of the
> reasons randonneurs are oft en designed with low trail is to
> complement the widish 36-38 mm tires a 650 B- bike generally has.

Wide tires help a low trail bike by adding some of the stability that
the low trail gives up. A low trail bike with narrow tires can feel
twitchy and like it wants to turn too easily. The wider tires
resist these minor turns and make the bike feel more stable.

A high trail bike like the Atlantis already resists the turns due to
having a high trail. The extra resistance of a wider tire isn't
necessary (but you still may prefer the bike's handling with one size
of tire or the other).

Try a handlebar bag on your Atlantis, I think you'll be happy with
the results.



Date: Tue, 10 Jan 2006 22:28:36 -0800
From: Murray Love
Subject: Re: [BOB] Trail talk

Thanks, Mitch. My understanding is a little different--though not by
much--and I'll lay it out below. I'll leave aside issues of weight
shifts and body english since I'm not sure I understand them all that
well just yet.

My take on the trail issue is as follows:

When you increase geometric trail, you also increase two factors that
counteract each other to some extent: Caster and wheel flop.

CASTER refers to the tendency of the front tire's contact patch to
want to fall in line behind the steering axis, and resist deviations
from that plane. Any bike with positive trail will exhibit
self-correction in some proportion to its speed: the faster you go,
the stronger the self-correcting (or "righting") moment. High-trail
geometry increases caster, but even high-trail bikes have little
caster at low speeds--the bike can easily be steered quite
aggressively with little resistance.

WHEEL FLOP refers to the tendency of the front end of the bike to
"kneel" under steering input. When you turn the handlebars, the front
of the bike pivots about an axis passing through the contact patch and
parallel to the steering axis, and the axle height drops. My
understanding is that wheel flop is what we use to initiate lean via
countersteer--the greater the trail, the higher the wheel flop, and
the easier it is to initiate a lean by very small movements of the

So. If I have this right (and I make no guarantees), we can say the follow

HIGH-TRAIL BIKES are juggling the effects of both higher caster and
greater wheel flop. At speed, caster makes the bike want to continue
in whatever straight line it's in, but relatively minor steering
inputs can cause the bike to begin a lean, and to veer as a result.
At lower speeds, caster is diminished, while wheel flop is unchanged
(and steering angles are higher), leading to lower stability. The
upshot is that high-trail bikes require LESS steering input to
initiate a lean at speed, despite the fact that they have greater
straight-line stability.

LOW-TRAIL BIKES have inherently lower caster and therefore less
straight-line stability, but most bikes with positive geometric trail
will have enough caster at speed to inspire confidence. They also
have less wheel flop, so are not as prone to begin a lean if the
handlebars are nudged. One implication of this is that you need to
turn the handlebars MORE to initiate a lean on a low-trail bike (can
anyone confirm this?).

Last month, Dan Goldenberg wrote on KOG:

"I recall that Merckx designed his frames (and presumably the ones he
rode also in
competition) with low trail, because he liked how a low trail frame handled
on the cobbles and poor roads."

This makes sense when you consider that pave is one of those rare
on-road situations where most of the steering input is coming from the
road surface, as the wheel tries to find its way around and over the
cobbles. If it's correct that a high-trail bike is more sensitive to
steering-axis inputs, it would follow that the bike would be falling
from lean to lean in this situation. A low-trail bike, which is less
sensitive to steering inputs, would be less affected.




Here's what Tom Kellogg has to say about trail on the Spectrum Cycles
web site:

First the definitions;
Fork rake; The distance that the front axle leads (usually) the
imaginary line drawn through the center of the head tube (or steering axis).
Head angle; Is the angle described between the ground and the steering axis.
Trail; The distance between the front wheel contact patch with the road
and the imaginary point where the steering axis meets the road.

As a general rule when dealing with 700-C wheels, a rake of about 5.6mm
will give a frame set "neutral" handling. My use of the term "neutral"
here refers to two things. First, neutral handling means that a frame
set will respond to steering input in the same manner no matter what
speed the bicycle is traveling. Second, while cornering, a neutral
handling bike will have neither a tendency to climb out of a turn nor
have a tendency to dive into the turn, it will simply hold the line that
the rider sets up unless further rider input is applied.

Decreasing trail below the neutral range has a couple of effects as you
might expect. The first thing a rider will notice about a low trial bike
is that it appears to resist attitude changes (or lean angles). It
requires more physical effort to get the bike to lean into a corner and
more effort to get it to straighten up. The second thing that you will
notice is that while cornering at higher speeds, the bike will have a
tendency to climb out of the turn on its own. Finally, you will find
that the way the bike responds to rider input is effected by the speed
of the bike. As you might have guessed by now, at lower speeds, a low
trail bike will have a tendency to want to go straight and do so pretty
much on its own. What you will find at higher speeds (like over 30mph)
is that a low trail bike will become quite vague in the front end. The
front wheel will feel as though it is wandering a bit and the contact
patch feel will go away.

Increasing trial above the neutral range will cause opposite effects for
the most part. At lower speeds, handling response will be light and
consequently, attitude changes will be much easier. During cornering,
the bike will have a tendency to drop into a tighter arc than the rider
might have intended. Finally, speed's effect on handling is reversed.
While low speeds give a light feel during handling maneuvers, high speed
sets up a very solid front end feel.

Although high trail frame sets give safer (more inherently stable)
handling than low trail frame sets do, high trail frame sets are still
inconsistent in the way they respond to rider input. Interpreting from
the basics above you can see why we usually aim for neutral trail. It
does not require the rider to consciously hold a bike down during hard
cornering, nor does it require different rider input depending on
changing speeds.

For some frame designers though, it is not always that simple. For
example, look at the way Eddy Merckx designs most of his frames. He
usually uses less trail than the "ideal" as he did much of his racing on
the pave and likes the way a low trail frame tracks under really
horrible conditions. Granted, they do not act as consistently under a
variety of speeds on good roads, but they really work on northern
Europe's country tracks.

two finds

first, according to this shop owner, Sun has made their CR-18 rim in 650B. coincidentally, i emailed Sun yesterday to ask when and if they'd do exactly that. this means 650B is doable for about $70 less in rims, although tires would still be pretty steep; Harris has NiftySwiftys for $40 apiece, as opposed to about $40 total for a set of Panaracer Tourguards.

i stumbled across this browsing the KOG message boards for info on trail...there's a link to a very interesting .pdf of a Vintage Bicycling Quarterly article comparing the Kogswell P/R with three different forks.

high trail vs low trail

regrettably, it seems the design i decided on and had plotted won't be quite final. at issue now is how much trail i want.

In the blue shorts: Jan Heine and the iBobbers, advocating trail around 45 mm for general-purpose riding, especially if a handlebar bag is to be used (and I plan to use one sometimes), and especially with 650b wheels (which I'm not planning to use at this point).

In the red shorts: almost everyone else, including Rivendell.

the other factor complicating things is that more rake/less trail would push the front wheel forward, allowing a shorter top tube with even better toe clearance. for those of you just tuning in, i've got a short torso, so a TT c-c of 57 cm is a bit of a compromise (necessitating a shorter-than-desirable stem). so i'm kinda biased towards the iBob side.

another factor is that i could still switch to 650b. the drawbacks would be limited rim/tire/tube availability (and probably higher prices for those parts) and slightly decreased efficiency (wider tries, smaller wheels). upshots are better toe clearance and possibly handling.

see also this on the P/R, and this on crazy Porteur races in Paris

now to hear arguements from the iBobs:

From: "Tom T"
Date: Wed, 30 Mar 2005 12:16:23 -0800
Subject: [BOB] VBQ frame geometry- pneumatic trail & geometric trail an

To follow up on the thread of the subject title a little while back,
here's something you might find interesting. I must say that I was more
than intrigued when Rivendell ( Grant) had that adjustable rake fork
made to test the handling and wrote the article about it in the Reader
# 31?.

I've ridden that fork with with different settings around the parking
lot but it wasn't long enuf to to make an impression. Then I got the
recent issue VBQ and the frontend geometry article really piqued my
interest so I thought why not?

It just so happens that my Rivendell fixie has identical geometry as the
1954 Alex Singer tested ( bike # 7)minus the the 61 mm raked fork. (Mine
is a 56cm with a 57cm top tube, 73 head angle, 44 rake). And, I also
have a spare fork with 48 mm rake made recently by Roland Della Santa
with the Heron forkcrown and 24mm round blades. (Thanks Todd K.)

I gave Roland a range of 58 min to 61 max. since he didn't know if it
could go that far without rippling the blades ( sometimes it happens).
He has a jig that can rake to max 70, never been used up until now.
The one he normally uses can only go up to 55 max. as that is the most
he ever needed rake a fork for most frames built.

The fork came back reraked to 60 mm with the paint intact, was installed
and ridden around the neighborhood hillsides. The initial impressions
are very good to me and a number of shopguys. They were curious about
this latest project of mine, and they did let me use their tools. 2 of
them commented that it corners so smooth and surefooted , so much
better than the
$3k+ latest Trek carbon Mardone bikes they're selling. Hmm, It was a
good thing it was just outside of earshot range of their buying

Initial impressions seem to bear out Jan's claims of better handling in
the VBQ article.

FYI, currently the wheels are 700c fixed, single speed, Pasela 32c
(actual width, same as the stock Quickbeam tires), inflated to about
75psi. dia. of wheel: 693mm measured. From the Trail-0-matic calculator I get 43 mm of trail. Front
center is now measured 603.
The fork is now shortened by about 2mm so it isn't a problem with
clearance. That's about what you'd lose for the typical 72.5- 73 degree
head angle for every 10 mm of additional rake. Can someone figure this
with some simple or not simple math ?
BTW, this reraking wouldn't work if your brake pads are already at the
top of the slot.
The TCO I had with the 32c tires is solved with room for fenders should
I choose to install them later.

Bike looks great.

Addtional test riding will be done in the next few weeks with Rolly
Polly tires (700x27c, 684 mm dia. gives 42mm trail), with loaded
handlebar bag, and on Mt diablo with twisty fast descent and plenty of
decreasing radius curves. If I like the handling with the RP tires I
would have a few more forks reraked the same- with the Saluki being
the only exception, all my bikes have 73 or 72.5 head angle and 45mm to
50 mm rake.

Tom T


Date: Mon, 21 Mar 2005 14:37:58 -0800
From: Jan Heine
Subject: [BOB] Re: VBQ article freaked me out

>That was probably Jones with his URB (UnRideable Bike) research. He
>tried to establish that stability was dependent on "mechanical
>trail," what Jan called "geometric trail" in his article.

Mechanical trail, according to Bicycling Science, is the vertical
component of geometric trail. Geometric trail (usually only called
"trail") simply is how far the steerer axis intersects the ground
ahead of the contact patch.

"Pneumatic" trail is a phenomenon that works like geometric trail to
a degree, but is caused by tire deformation. (The tire act as a
rubber band. At the front, it is laid down without deformation, no
matter the side loads. But if there is a sideload, the tire deforms
increasingly toward the back, with the maximum deformation at the
rear. This pulls the steering straight, because there is more
deformation behind the steerer axis than in front.)

I hope Jim Papadopoulos doesn't read this over-simplified
explanation. The VBQ article was the result of numerous discussions
with him and many others...
Jan Heine, Seattle
Vintage Bicycle Quarterly
c/o Il Vecchio Bicycles
140 Lakeside Ave, Ste. C
Seattle WA 98122


Subject: Re: [BOB] Rake/Trail vs. Head Angle
From: mlove(AT)
Date: Mon, 12 Dec 2005 10:59:27 -0800

Forbes wrote on 12-12-2005 10:21:36 AM:

> Is there an accepted formula/best design practice
> related to a frame's head angle versus the amount of
> fork rake/trail? I was taught that the steeper the
> head angle, the less one should rake the fork. Is
> that commonly accepted practice? Or is there debate
> on the matter?
> Thanks!
> - Forbes Bagatelle-Black
> Santa Clarita, CA

Forbes, this has become a bit of a recurring issue across the BOB/RBW/KOG lists,
thanks to Jan Heine's writing in VBQ. There wasn't much debate on the matter until
Jan began to point out that French cyclotouring bikes of the mid-20th century have
excellent handling characteristics despite the fact that the geometric trail on
these bikes was often much lower than the commonly accepted 50-60mm found on most
modern bikes. One of the main reasons the old cyclotouring bikes can afford such
low geometric trails is that they used fat, low-pressure 650B tires with a large
contact patch providing a large degree of stability--what Jan calls "pneumatic

My further readings on the issue seem to have converged on the following heuristics,
and I'm sure that the more knowledgeable among us will correct me if I say anything
blatantly stupid:

- Design for the trail you want, taking into account things like toe-clip overlap,
aesthetics, what have you. If TCO is a concern, go with a slacker head angle and
larger fork rake, because...
- Within the typical head angle range of 72°-74°, changes in trail will predominate
over changes in head angle.
- If you'll be using 650A/B or 26" tires run at 50psi or so on the road, you can
reduce the geometric trail somewhat--even more so if you'll be loading the front
down with a rack or large handlebar bag. The old French bikes had geometric trails
as low as 11 mm, but most were up around 30 mm, if I recall correctly.
- Conversely, bikes with skinny 700C high-pressure tires may benefit from higher
trail--the J.P. Weigle randonneur tested in the latest VBQ had 52 mm of trail with
the fork designed for lightly loaded front ends, and 42 mm with the fork for loaded
front ends. The Rambouillet and Model P have trail figures in the high 50s, the
Atlantis and Surly LHT in the low 60s.

Dublin, Ca
*** the definitive post from J. Heine
Date: Mon, 12 Dec 2005 17:43:01 -0800
From: Jan Heine
Subject: Re: [BOB] Rake/Trail vs. Head Angle

>>An interesting thing is that a mountain bike with a skinny slick often
>>feels more twitchy than a road bike with stock size tires.

Wheel size plays a role. That is why 650B is so popular - it feels
different from 26"

>>For low trail lovers, my Bike Friday has 33mm of trail (from the small
>>wheel with normal fork design of 73deg, 40mm offset). Some riders
>>have found that they prefer to reduce the fork often to get more trail
>>and more stable handling. 15mm of offset would give it 59mm of trail,
>>closer to a road bike:

Another example that shows that simply looking at trail is
meaningless. A 650B bike with fat tires and 33 mm trail will be
lovely. Yet I never could ride my Bike Friday (no idea about that
bike's geometry, but probably was similar) no-handed at less than 30
mph. And I rarely go that fast. On the other hand, as a teenager, I
had a Peugeot folding bike - awful contraption - with the same wheel
size, yet riding no-hands at moderate speeds was not that difficult.

Tony Oliver, in his book "Touring Bikes," has a diagram that
indicates that bike with MORE trail are LESS stable - exactly the
opposite of what most Americans believe. At first, I thought it was a
misprint... but it isn't.

However, it isn't a matter of either Americans or Oliver being right.
The thing few people understand is that two factors play a role: More
trail increases stability, especially at high speeds (to the point
where it is difficult to change line in mid-turn), yet it also
increases wheel flop (which persists into moderate speeds). So you
have two counteracting phenomena, which is why so different bikes can
handle similarly.

When you add a front bag, you increase wheel flop, and you really
need a low-trail geometry. Tire size and pressure also plays a role -
no point rehashing the VBQ geometry article here (Vol. 3, No. 3).

The Berthoud tested in VBQ Vol. 4, No. 1, had about 65 mm of trail,
and it worked OK at moderate speeds, but not well at low speeds (<10
mph), where it suddenly would sverve. This especially with a
handlebar bag. (This lack of stability is due to wheel flop.) At very
high speeds, it was rock-steady (high trail taking over), but once
you picked a line, you couldn't change it. If there was a pothole in
mid-turn, you would hit it! Also, the Bertoud was sensitive to
cross-winds at low and moderate speeds, but not at all at very high

The Weigle tested in VBQ Vol. 4, No. 2., had about 41 mm of trail. It
was wonderful at low speeds even with a full bag (riding no-hands
climbing at 7 mph! / very little wheel flop), it was great at
moderate speeds, and it was good at very high speeds. I liked being
able to change line in mid-turn, but in a full aero tuck at 50+ mph,
it lacked the confidence that the Berthoud had (less trail and
"stability"). Putting my hands to the drops made it totally fine. The
bike was not sensitive to cross-winds at low and moderate speeds, but
was more sensitive to them at very high speeds.

So which is "more stable?"
Jan Heine, Seattle
Vintage Bicycle Quarterly
c/o Il Vecchio Bicycles
140 Lakeside Ave, Ste. C
Seattle WA 98122

Date: Fri, 19 May 2006 11:10:42 +0100
Subject: [CR]Peugeot frame design.

It was asked:
Does anyone know why Peugeot would have spec their bikes like this? To be
different in a French way? A trend at the time? A famous racer had a
bike with these angles?

The July 1971 issue of Le Cycle shows "The plan of service des courses
Peugeot of a racing frame" It may be the frame used by Delisle in the 1970
Tour but I am not certain of that. The details are:-
Head angle:73 Seat angle 72 Top and seat tube 56cm
Front centre 588mm Rear centre 430mm
BB drop 80mm BB height 262mm
Fork rake 55mm Trail 46mm
An interesting dimention is measured horizontally from seat lug line to
rear dropout line and is 248mm
The angle between the seat and chainstay is 90
All this seems entirely conventional, if perhaps a little old fashioned
for 1970. For example Eddy merckx had a fork rake of 40mm and a rear
triangle of 415mm on his 1970 Tour bike.
I can only think that if Peugeot did sell bikes with a 76 degree head
angle and a long fork rake it was so they felt "lively" during a quick
test spin round the block.

Ray Green, Brighton, England
*** and last my post and a response
Subject: Re: [BOB] What do with a slack HT angle?
From: mlove(AT)
Date: Wed, 9 Aug 2006 10:42:56 -0700

Ethan wrote on 09-08-2006 10:24:51 AM:

> HowdyBobs
> This is usually the type of thing that gets hashed out on the
> Framebuilders' list, but I'm searching for Bob insight because the
> frame in question is quite Bobbish.
> Here's the situation: I've got long legs and a short torso. I'm set

> on using 700c wheels front and back, mostly because the lugs and tube
> I have will only work for that (I think). So to avoid toe overlap an
> let the top tube be as short as possible, I made the HT angle
> relatively slack (71.5°).
> I had originally planned to go with ~59mm trail, meaning 52mm fork
> offset with a 30mm tire. After reading what I could find on low trai
> designs in the archives, I'm a convert and ready to go low. But how
> to acheive this with a slack HT angle? To get low trail, I'd need to

> increase the rake/offset, which means yet more wheel flop, which so
> far as I know is undesirable.


Wheel flop is a function of only HT angle and trail. Increasing fork
offset will decrease both trail and wheel flop, all other things being

WF = Trail * sin(HT) * cos(HT)


Ken Janes, lug cutter

this got relayed to the FBs list. great photos of a great shop. I've been doing lots of filing lately so it's fun to see other folks' tools and setups. tonight it was the right drop; time-consuming becuase it's in 3 planes that make it hard to remove steel quickly.

my Enco order came in, so now i've got 2 brand new Nicholson half-rounds, 8 and 10", a needle file set, and a riffler set. i'm wishing i'd ordered 2 needle file sets. and the link above suggests a wire wheel for the contoured parts. and i need more dremel bits. and chalk for the files.

shopping for schools

This is going to be slightly outside the official topic of this blog, but it's definitely connected.
I'm about to enter my second year at the Corcoran College of Art + Design, which at times feels as pretentious as it sounds. I was originally planning to major in Photojournalism, but partly because of the project chronicled herein, my interests have shifted towards Industrial Design. Problem is, Corcoran doesn't offer said major, and lacks a few things I'd need to study it on my own.

So I've set out to look for a place to transfer to, without much success. I've looked at community colleges, trade schools, engineering-focused universities, and other art colleges, but can't find anything with the right combination of small, practical, and technical. I like the artsy-ness of Corcoran, but I'd like to have access to a broader range of facilities and courses.

This project has really changed my way of thinking about my future. It's the thing I've worked the hardest on in a long time. I hope my passion for it is evident in the words preserved here.

I've re-discovered a love of working with my hands, of making stuff, and making it my way. I used to do this kind of thing as a kid--finding wood to whittle into little boats and such--but it feels so much different and so much greater now that I'm doing it as a project, with research, organization, paperwork, all over the course of months, rather than a Sunday afternoon spent whittling. I hunger to learn more skills like these, skills I might not even know exist yet. That's just it: I don't want to study...I want to learn to do things.

If this rings a bell, or if you know of a place like I describe, or whatever, shoot me a note: ethan theatsymbol

So far I've looked at:
Rochester Institute of Technology (looked great but too big; also, lake effect snow)
Rhode Island School of Design (current favorite, but I'd have to wait a year to transfer)
Wentworth Institute of Technology (looks promising; website not so hot)
Parsons The New School for Design
The Pratt Institute (don't know much yet about these two)

Henry James fork blade bending mandrel/jig

as requested by Siren


back with instructions

the scans aren't great...the plan was kicking around in my Timbuk2 for a couple weeks. the original is life-size and quite accurate, so maybe these are too.

jig backboard w/pics

these photos kinda suck 'cuz i took them with a cheap camera...after the jig is finished i'll go to it with the 20D and do it up right.
this is just the backboard with the full-size plot spray-adhesived on (after an unsuccessful 1st attempt with Elmer's..bad idea). clamping blocks are to be glued next, then screwed in from the back.
unlike Talbot, i plan to only use this for pinning it all together, then free braze in a repair stand. i made the cutouts for the lugs with this in mind, trying to strike a balance between space for a drill to pin and space for as much clamping block for best alignment.
i used a circular saw for the straight cuts and a thin jigsaw blade for the curvy parts. went a lot more smoothly than anything else has so far, other than having to get a new circ saw blade.

Long Tail 2

a post to the iBob list linked to this post by the owner of Heron. this got me thinking about the Long Tail again, because Todd talks about saturation of the "Bobish" niche, what with Rivendell, Heron, and Velo-Orange all making similar lugged steel frames for similar purposes.

How to define "Bobish"; it's an ephemeral idea. It means taking the scenic route, not wearing a lot of polyester, riding a steel bike, taking a picnic with you. It's the type of riding most Rivendell owners do, the kind they advocate in their literature.

As with a lot of things, there's a wide spectrum of Bobishness. I'm bobish, but I'm also young and I like to go fast. Most of my riding is on the streets of DC. All the other Bobs ride a little differently from one another in different places. 50 years ago none of the Bobs would know about any other, and now we're all able to talk to one another almost effortlessly, trading parts and stories and recommendations.

The result is going to be more Bobs, as the gospel spreads--Bobs have kids and talk to their "real life" friends. Three framemakers in our one arena thus isn't necessarily a lot. The past few years have seen the explosion of the Bob niche, and I expect it will continue to grow, at least in part because it makes sense. It's a type of riding centered on common sense, fun, and exercise. And it's addictive, as can be seen from looking over a day's iBob emails.

Talbot jig scans

Recognizing the copyright law disaster I might be about to create, I offer 3 scanned pages from Richard Talbot's classic Designing and Building Your Own Frameset.

page 43: Miscellaneous Aids. Tube blocks (with note for oversized tubing modification), clamping plates, and wedges for chainstay alignment in BB shell.

page 44: Fork Jig. Drawing and construction notes.

page 45: Main Triangle Jig. Drawing and construction notes. Using oversized tubing, the 1/8" shim is unnecessary.

the long tail of bikes

I'm currently reading an excellent book by Chris Anderson, editor of Wired. It's about how the economy is increasingly based not on blockbuster hits but on an enormous number of niche markets, each serving a small community but adding up to some big numbers. It also talks a lot about the act of blogging as a catalyst to this increase in niche sales, so in keeping with the spirit, I offer this link to the Long Tail website.
In the part I'm reading now, Anderson discusses great examples of businesses and groups that operate in the Long Tail: Wikipedia, Amazon, Alibris, SETI@home, and all manner of feel-good internet stuff.
So naturally I've got to consider the Long Tail of Bikes. But I haven't fully considered it yet, so I don't want to write about it. In the meantime, Velo Orange has released preliminary specs for their semi-custom randonneur.

mandrel v2 pics

A little more about this thing:
the pics should hopefully be a good supplement to the text of the previous post on how to make one of these. i also added anothe rlittle flourish during the final filing of the groove. at the tight end (~7 to ~9" radii) the groove is decidedly semicicular to fit the 12.5 mm end of the blade...then it gradually becomes a wider and more V-shaped groove starting a little left of center, so by the time it reaches 12" radius it's able to fit the 20 mm width of the blade near it top.
Just for looks i sanded the surface of the groove up to 220 grit and waxed the whole thing.
If you're interested in buying something like this we should talk. It's not something I'd want to do for a living, but I've got the procedure down pat now.

mandrel #2 done

so a couple days ago i finished a 12" radius fork blade bending mandrel from the H. James plans. upon further research, it became clear that radius would only be suitable for the recently-more-popular Continental or Italian camber, in which the blades curve rather gradually from near the top to the dropout. I decided that in keeping with the style of the bike, I wanted a more old-school British camber, in which the blades arc tightly only near the dropout, with perhaps the top two thirds straight and the bottom third curved.

despite having just finished the 12" mandrel, i undertook the construction of a new mandrel, based on the designs of John Clay of Tallahassee. this one has a varying radius, starting at 7" near the end where the dropout will be secured and changing to 12" at the other end. following are the steps i undertook to make the rough cut:

• determine a center point at least roughly 8" from the left edge and 12" from the top edge of a piece of 3/4" thick cabinet/high-grade plywood
• use a carpenter's square to draw lines radiating at 90° angles from the centerpoint: one 7" long to the left and one 12" long to the top
• use a compass to draw arcs 8", 9", 10", and 11" from the centerpoint within the area described by the right angle lines
• use a protractor to mark 18° increments in the relevant quadrant
• use a ruler to sequentially connect the centerpoint, an 18° mark, and an arc, starting with the 8" arc and the 18° mark closest to horizontal, progressing towards the 11" arc and the 18° mark closest to vertical
• now sketch the curve, attempting to connect the intersections just marked. the curve will have a slight teardrop shape.

now rough cut with a jig saw (after clamping down the plywood, donning goggles, and practicing), file to a smooth curve (a 4-in-hand rasp is perfect), and file an even semi-circular trough along the center of the curve (triangular, round, half round, chainsaw, and knife files are useful). figure out a way to hold the dropout end of the blade in place while bending (still working on that myself).

frenchy frame by Brian Beaudette

my favorite is the fork crown here

this is very much the look i'm shooting for

this came from this post on an interesting blog by a former aircraft his fixturing. gotta get some safety wire.