previous page: 4.8 Choosing a ride
page up: Bicycles Misc
next page: 5.2 Frame Building Part 2

5.1 Frame Building Part 1


This article is from the Misc Bicycles FAQs, by various authors.

5.1 Frame Building Part 1

From: terryz@microsoft.com (Terry Zmrhal)

I am currently enrolled in a frame-building class at Ti Cycles in
Seattle. Dave Levy of Ti Cycles offers this class once a year where he
takes 4 students who want to learn about frame building and teaches
them, over four Saturdays what goes into building a frame. We learn
some of the technique of designing and building and do most of the work
on the frame ourselves. Once done, we have a custom frame built mostly
by each individual with professional guidance. We learn enough for a
one-time deal, but we certainly couldn't go off and build at will

Dave Levy has been building frames for 8 years - 4 years with R&E of
Seattle (Rodriguez name) and 4 years on his own at Ti Cycles.

I was very interested in the class, but didn't really want another
steel-frame bike. My dad has wanted a custom bike for awhile. I
mentioned this class and the idea of building a frame for him. After a
little thinking, he was sold on the idea. So I'm building the frame and
getting the experience, but the frame is actually for my dad.

I'll share what I learn during each class about the process of building
a frame. Some of this may be obvious, some not, but I'll explain
everything I can remember anyway.

The first class was this past Saturday. We began at 8 in the morning
and didn't finish up until after 5 in the evening. We didn't waste any
time getting going with the process of building up the frame. We
started with some shop procedures. Safety and caution was of the utmost
importance since many of the machines are large and dangerous. The
other main word of advice was to double and triple check everything
before you do it - once done, there's no undoing it.

After this little shop talk we picked out our lugs, fork crown, and
dropouts from the small selection he had. Ti Cycles doesn't do many
lugged frames so the selection was limited. I chose Shimano vertical
dropouts for the rear and Shimano dropouts for the front, both sets
having eyelets for fenders/racks. We then cleaned the inside and
outside of the tubes and marked them with our name. All the tubes and
parts we would need were in a box about 6 inches square on the end and
about 4 feet long. Quite amazing that everything fits inside such a box
and then we whittle and braze to come up with a final product of a
frame. Inside the box was a head tube (1 1/4"?), a steerer tube for the
fork (1"), double butted top tube (1"), double butted down tube (1
1/4"), single butted seat tube (1 1/8"), ovalized fork blades (7/8"),
seat stays, ovalized and tapered chainstays, and dropouts. All the
tubes are the standard diameter for a traditional British frame. I am
using Reynold 653 tubing. Other were using some Tange and Columbus
tubing. My dad chose the Reynolds set since it was better than a
standard chrome-moly tubeset, but not at the high end of tubesets

The tubes are sent covered with a thin layer of an oily substance to
prevent rusting until they are put together. We also looked down each
tube to find where the butting began, if there was any butting and
marked where the butting began and ended on each tube. On some tubes
it's easier to see than others. I was able to easily see the rings on
the inside of the tube where the butting began and ended on one tube,
not on the other. One of the guys is a little tall and heavy so his
tubes were rifled at each end for extra strength. Buttting is where the
tubes are thicker at one end and thinner in the middle; rifling is
putting extra metal bands or strips inside the tube at the end for
additional support.

The first pieces I worked on were the fork blades and chainstays. These
four pieces each needed to be slotted at the end so the dropout could
fit in. First we rounded off the tips of the dropouts using a belt
sander. This would allow the tips of the dropouts to fit more snugly
against the curvature of the fork blade or chainstay. We also sanded
off some of the paint and chrome so the brass when brazing would hold
better. With the chainstays, the slots are off center, towards the
inside of the frame, imagining it in a full frame. By slotting the
stays off center, there is more room for the freewheel. Actually, this
is only important on the right chainstay where the freewheel is, but it
is done on the left side as well for symmetry and aesthetics. The slots
are cut about a centimeter in and about 20mm wide. It's important to
make sure that the slot is lined up with the major axis of the ovalized
stay, the major axis being the longer axis. The stays are ovalized to
provide more clearance for tires. To slot the stays and blades, we used
a rotating grinder.

With the fork blades, we started by bending the blades providing some of
the fork rake. We started with only about 20mm of fork rack. To bend
the blades, there is a tool where the tip of the blade is held steady
and then using the principle of a lever, we push down on the other end
of the fork blade to bend it. Amazing, but it actually works, I was
actually bending a fork blade this way. I had no clue how fork blades
were bent before this, now I know.

The blades were part of the 653 tubeset. As I was bending one of them,
I heard a large pop. Dave said it was probably just the bolt holding
the end of the blade in place moving. As it turns out, I had cracked
the tube at the end along the major axis of the ovalized blade! They
had never cracked a blade this way before, but Dave also said he has
heard of having problems with the 653 fork blades. They replaced it
with True Temper tubing which they've never had problems with.

The fork blades are bent somewhat before slotting because they cut off
about 2 cm at the bend end. The very end can't be entirely bent and so
doesn't provide a smooth curve. By cutting off some at the end, a
smoother curve remains. This was mostly for aesthetics again. With the
fork blades, the slots need to be centered, not off center as on the

Once the slots are cut, they are ground down using the belt sander again
to provide an angled surface. This was to provide an easier method of
brazing the dropout and blade/stay as well as for aesthetics. Each was
lighted sanded on the inside, outside, and along the edges.

Now we were ready to braze these. We started by applying a flux - which
was a Boric Acid aqueous solution. The flux would help with the brazing
by keeping the brazing area clean, preventing oxidation, and a few other
things which I can't remember. It was a good thing to use essentially.
To braze we were using an oxy-acetylene torch. I've never used one of
these before so I asked lots of questions, but still don't remember all
Dave said. The basic idea is that two gases are mixed, one which will
burn, the other which will promote the burning. The acetylene will
burn, and oxygen will promote the burning. You can control the torch by
varying the mixture of these two gases. The idea was to get a hot flame
at about a medium setting with a small inner cone, which was very hot.

Once we had a flame going, Dave showed us how to braze. We started by
heating up the whole area slowly by holding the torch a little bit away
from the dropout and blade/stay. When the flux stopped bubbling and
started looking glassy then we moved the torch in closer to get it
hotter. We would fill each side between the dropout and the wall of the
blade/stay with brass. We would apply heat and the narrow brass rod to
the area, the brass would melt and flow into the tube. To move the
brass, we would heat where we wanted the brass to be and then let
gravity do it's work and the brass would flow to that area. This was
how we would get the brass farther into the tube. There were a lot of
things to consider here - making sure you applied brass in the right
place, in the right amount, making sure the area didn't get too hot,
letting it cool at the right time, moving brass around. Here's an area
where I most likely would not be able to reproduce myself.

We did the two stays and blades this way.

And moving on. Next was prepping the lugs. By using a very narrow belt
sander and a round sander, both small hand-held tools hooked to an air
compressor, we would sand the edges of the lugs, round the inner edges
of the lugs and sand the inside of the lugs. Sanding the edges of the
lugs would provide definition and some aesthetics. Rounding the inner
edges of the lugs would promote the flow of silver when it came time to
braze the lug and tube together. Sanding the inside of the lugs was to
make sure that there was enough clearance for the tube. The tube should
hold, but it shouldn't be difficult to remove or move either. This
process took awhile since the tools are small and the areas are even

The next steps we started on but didn't finish entirely. I started
working on finishing the areas around the dropouts. This would mean
sanding and filing using a variety of tools - the belt sander, the hand
held sanders, files of varying sizes, and sandpaper - to make the
dropouts looks good. Getting rid of the flux that had hardened, getting
rid of excess brass, providing smooth curves, and definitive edges. I
was only able to start because we had a bottleneck around the tools
necessary to do this. This could take hours depending on how meticulous
you want to be, but this is also the areas that people will notice the
most. I wanted to make these look good.

The other step was beginning the frame design and mitering of the tubes.
The frame design program which Dave was using was working, so we
couldn't actually see the frame design. We did miter the seat tube to
fit into the bottom bracket shell and then cut the relief for where the
down tube would fit in. To do the mitering, they had a large drill
machine. Everything on the machine was extremely precise. Everything
was checked for tolerances, levelness and exactness. I thought this was
pretty impressive.

I think that's all there was to this first class. The next class isn't
until December 4th. I'll post another report then.

Terry Zmrhal
'The Horizon is but a line to be crossed, not a limit to be reached.'


Continue to:

previous page: 4.8 Choosing a ride
page up: Bicycles Misc
next page: 5.2 Frame Building Part 2