Low temp fillet brazing
Bobaxe
bronze alloys. These typically have a liquidus around 1700 degrees. I've
been told that a moderate silver content (25%) alloy can be used for
fillet brazing while providing the benefit of a liquidus around 1250
degrees and a reasonably wide plastic range. While I don't doubt such an
alloy exists and that one could build at least a small fillet using it, I
don't know how to evaluate the strength of the resulting joint. Is the
reduced heating of the tube a virtue with no drawbacks. If so. why do all
the pros use bronze (surely the braze alloy cost is small compared to the
other ingredients in a $1000+ frame)? Thanks for any guidance.
Sincerely, Bob
WJT
>bronze alloys. These typically have a liquidus around 1700 degrees.
>don't know how to evaluate the strength of the resulting joint. Is the
>reduced heating of the tube a virtue with no drawbacks.
Bob,
Fillet brazed frames are pretty. But it kills the tubing. Examine the warranty those
pros offer. Low temp brass is still heated to 1500+F and chrome-moly really, really likes to stay
under 1100F to retain it's normalized strength. Silver soldered lugged frames are easy to weld
at 900F and the tubes love it. Reynolds tubing, other than 753+, like the higher heats that
brazing requires, but not over large areas.
If you must fillet (without a bondo joint under the paint), expect about a years
lifespan under fairly harsh conditions. The tubing will crack on the downtube 2" from the
headtube, and at either end of the seat tube within 1" of the welds.
If you use a JetFluxer with high silver content filler you can keep the heat to 1000F or
so. The frame will be pretty, but the joints will have zero strength. Art has it drawbacks. -Bill
Doug Milliken
In a previous article, bob...@aol.com (Bobaxe) says:
>Can anyone comment on alternatives to the high temp standard low fuming
>bronze alloys. These typically have a liquidus around 1700 degrees. I've
>been told that a moderate silver content (25%) alloy can be used for
>fillet brazing while providing the benefit of a liquidus around 1250
>degrees and a reasonably wide plastic range. While I don't doubt such an
>alloy exists and that one could build at least a small fillet using it, I
>don't know how to evaluate the strength of the resulting joint. Is the
>reduced heating of the tube a virtue with no drawbacks. If so. why do all
>the pros use bronze (surely the braze alloy cost is small compared to the
>other ingredients in a $1000+ frame)? Thanks for any guidance.
I know at least one bike mfr that uses a brazing alloy that is about
one-half silver -- to fillet-braze stainless steel. I believe that this
alloy was chosen for its color match to the stainless as much as for the
low temperature. Results are good, but the one guy in the shop that can do
it is pretty d*** good at his craft.
I tried and didn't get anything like a good looking fillet -- even
though I'm pretty good with a normal brass rod.
--
Hans-Joachim Zierke
Bobaxe writes:
> Can anyone comment on alternatives to the high temp standard low fuming
> bronze alloys. These typically have a liquidus around 1700 degrees. I've
> been told that a moderate silver content (25%) alloy can be used for
> fillet brazing while providing the benefit of a liquidus around 1250
> degrees and a reasonably wide plastic range.
It is possible to braze it at 610 degrees with 40% silver, and at 650
degrees with 56% silver cadmium-free.
Both setups need so tight a tolerance in tube metering, that you might not
want to pay for it. It might get economical with high-precision laser
cutting equipment getting cheaper.
> If so. why do all
> the pros use bronze (surely the braze alloy cost is small compared to the
> other ingredients in a $1000+ frame)? Thanks for any guidance.
All the pros use either 40% or 56% silver in _lugged_ frames, at least
those pros whos frames I might buy. In lugless construction, many builders
use high strength alloys with some nickel content. The damage to the tube
is a function of heat and time. Since lugless needs much less time to heat
up for brazing, this can be a good compromise of strength versus costs.
hajo
--
T . E . L . E . K . O . M
We specialize in cross subsidization
Private Customer Shafting Government Staff Grafting
BikeNbear
development work on it and rumor has it that Trek also had a hand in
developing the alloy along with Mike Melton who did some work on some
Olympic bikes. This is proven stuff for fillet brazed frames and is the
only rod I use and I personally fillet brazed about 500 Tesch frames and
Dave Tesch probably brazed another 200 frames with the same type of rod.
If you have cracking problems 2 inches from the head tube- down tube
joint, its not as a result of fillet brazing. We never had any problems
with cracking. The old Fisher mountain bike team frames were all fillet
brazed back in the days when Joe Murray was racing and they never had
frame failures that could be in any way be associated with fillet brazing.
Leo Castellon
BikeNbear
Joshua_Putnam
>I know at least one bike mfr that uses a brazing alloy that is about
>one-half silver -- to fillet-braze stainless steel. I believe that this
>alloy was chosen for its color match to the stainless as much as for the
>low temperature.
Silver brazing is usually recommended for stainless steel.
Brass can cause cracking, but I don't remember why. Henry
James Bicycles recommends brazing their dropouts with
silver solder with gap filling capability, and supplies
BAg24 (50% Ag, 2% Ni) filler with them. I don't fillet braze,
so I don't know if this would work for that.
--
Jo...@WolfeNet.com is Joshua Putnam / P.O. Box 13220 / Burton, WA 98013
"My other bike is a car."
New & used bike parts for sale: finger Joshua...@WolfeNet.com for list.
Suntour PowerFlo 7-sp freewheel, 13x28, new, $26 + postage
WJT
>joint, its not as a result of fillet brazing. We never had any problems
>with cracking. The old Fisher mountain bike team frames were all fillet
>brazed back in the days when Joe Murray was racing and they never had
>frame failures that could be in any way be associated with fillet brazing.
Leo,
We sent back at least 6 brazed frames for that very reason. They were eventually
warrantied. Localized fatigue from the heat band. Very common in aluminum construction too.
-Bill
James W Gourgoutis
BikeNbear <bike...@aol.com> wrote:
>Use the bronze rod that Henry James sells. The Gasflux Company did some
Does anyone know the AWS designation for this type of rod? Or for the
aforementioned "low fuming bronze rod", for that matter?
p.s. glad to see this thread--it's answering some questions I've had for
a long time. Thanks.
-Jim
. . . .
James Gourgoutis <sko...@pitt.edu>[http://www.pitt.edu/~skoop](412.683.7654)
Hans-Joachim Zierke
WJT writes:
> Fillet brazed frames are pretty. But it kills the tubing. Examine
> the wa those
> pros offer. Low temp brass is still heated to 1500+F and chrome-moly really,
> really likes to stay
> under 1100F to retain it's normalized strength. Silver soldered lugged frames
> are easy to weld
> at 900F and the tubes love it. Reynolds tubing, other than 753+, like the
> higher heats that
> brazing requires, but not over large areas.
As soon as you look at strength graphs after brass brazing, you'll see that
Reynolds 531 does not "love" it. Even with hand-held silver brazing, you'll
see the heated area on the graph, though it is very near to invisible.
Bushido
>Reynolds 531 does not "love" it. Even with hand-held silver brazing, you'll
>see the heated area on the graph, though it is very near to invisible.
What I've seen with Reynolds 531 is one season frames no matter what
construction. Silvered 531 has tended to crack at the lug junction (cast lugs)
due to retained hardness in low heat joinery. Brazed 531 with either lug will
go a full hard season and then will be a nice soft ride for the rest of it's life.
Since I no longer use 531it's really academic. Feel free to construct as you
please. -Bill
Greg Lewis
: Bushido writes:
: > What I've seen with Reynolds 531 is one season frames no matter what
: > construction. Silvered 531 has tended to crack at the lug junction (cast
: > lugs) due to retained hardness in low heat joinery.
: > Brazed 531 with either
: > lug will go a full hard season and then will be a nice soft ride for the
: > rest of it's life.
I've I've seen just the opposite. Lots and lots of 531 frames
lasting years and years. Sorry I just donot believe it.
Has any one else had this experience?
How about with other bikes.
(and lets talk uncrashed bikes).
--
_____________________________________________________________
Gregory S. Lewis | Phone: (512) 471-3105
Center for Nonlinear Dynamics | FAX: (512) 471-1558
University of Texas at Austin | e-mail:
Austin, TX 78712 USA | le...@chaos.ph.utexas.edu
|
____________________________________________________________
Hans-Joachim Zierke
MO writes:
> Are you sure that you are talking about the same 531 as everyone else??
Sure.
> do
> you think that 531 would still exist if what you are saying was true.
I know that many framebuilders use brass brazing for 531. Nonetheless, the
strength degrades in the heat affected zone, and a material control lab
will show this easily. This bad habit was the reason for TI Reynolds to
require a certification for 753.
I remember a discussion with a MAnnesmann engineer, when Mannesmann began
selling straight 0.6 34CrMo4 tubing to framebuilders. He said that they all
laughed about this "753 certification" of Reynolds. Brazing tubes like
Reynolds 753 is a simple standard task for a qualified metalworker (we have
formalized 3 year education for metalworkers in Germany), and if he can't
do something as simple as this, he gets fired tomorrow. The people at
MAnnesmann stopped laughing when they saw the first graphs of the material
control lab. Most of the well-known framebuilders weren't able to braze at
all. This is not amazing, most of them never attended a course for brazing
technique, but learnt by trial-and-error. This _can_ give good results,
and I have seen good results this way, but most cases are different. The
Mannesmann engineer admitted, that TI Reynolds apparently had a more
realistic impression of real-world framebuilding than they had.
Reynolds 531 is tubing with enough excess strength to take the impact
of brass brazing, and the strength won't get down very fast. Nonetheless,
brass-brazed 531 (or Columbus SL) frames will normally break after 2 years
of pro use. That's why the "soft frame" myth had to be invented.
>It
> was invented in the '30's that is a damn long time for any tubing to still
> be in production.
The "531" label is a trademark. The tubing is not the same as in 1935, it
got stronger.
TBGibb
(Greg Lewis) writes:
>I've I've seen just the opposite. Lots and lots of 531 frames
>lasting years and years. Sorry I just donot believe it.
>Has any one else had this experience?
>How about with other bikes.
>(and lets talk uncrashed bikes).
>
Have a 21 year old 531 with lots of miles, no crashes and no problems.
Tom Gibb (TBG...@aol.com)
>
>: Bushido writes:
>: > What I've seen with Reynolds 531 is one season frames no matter what
>: > construction. Silvered 531 has tended to crack at the lug junction (cast
>: > lugs) due to retained hardness in low heat joinery.
>: > Brazed 531 with either
>: > lug will go a full hard season and then will be a nice soft ride for the
>: > rest of it's life.
>I've I've seen just the opposite. Lots and lots of 531 frames
>lasting years and years. Sorry I just donot believe it.
>Has any one else had this experience?
>How about with other bikes. (and lets talk uncrashed bikes).
One of the first frames I built was a Reynolds 531 touring bike.
That was over 15 years ago and it's still going strong.
-John (j...@edsi.org)
Eric P. Salathe, Jr.
Let's put it this way, how many 15-yr-old bikes have you seen that _aren't_
Reynolds 531?! I guess we are to believe that Carbon-Fruit-and-Fibre is the
material of choice for long-lived frames.
--
,
Eric P. Salathe, Jr. sal...@atmos.washington.edu
Seattle, Washington http://atmos.washington.edu/~salathe
Bruce Jackson
Eric P. Salathe, Jr. <sal...@atmos.washington.edu> wrote:
> Let's put it this way, how many 15-yr-old bikes have you seen that
> _aren't_ Reynolds 531?! I guess we are to believe that
> Carbon-Fruit-and-Fibre is the material of choice for long-lived
> frames.
Good point. My oldest bike is an early 70's Gitaine that is full 531
double butted tubes. Until I repainted it and filed down the excess
brass and crude lugs it looked like it was brazed by a drunken sailor
(no offense meant to drunks or sailors). In spite of probably being
overheated the frame has held up to lots of hard use. The claim that
531 frames only last a season or two is preposterous.
--
Bruce Jackson | P. O. Box 13886-NT | GAB 550E
UNIX Systems Admin. | Denton TX 76203-3886 | (817)565-2279
Computer Sciences | jac...@cs.unt.edu | FAX (817)565-2799
Univ. of North Texas | http://replicant.csci.unt.edu/~jackson/
Rinards
fork, that the longer length of the Kestrel fork alters the frame's angles
too much, and the frame should have some length cut off the bottom of the
head tube to restore the designed head angle.
I measured the lengths of my Kestrel fork and my wife's SR prism fork, and
the difference is about 6 mm or so. Assuming a wheelbase of 1000 mm, this
gives a head angle change of .34 degrees (sin ^ -1 of (6 mm / 1000 mm)).
This third of a degree angle change seems pretty small to me. In fact, I
suspect the angle might vary that much (or more) from simple (perhaps even
unavoidable) errors introduced during construction. I can think of many
sources of error, from a jig that is slightly off, to miters that are not
quite tight, to unavoidable heat distortion during welding, to
innacuracies during machining the head tube bore.
Has anyone measured how much the head angle on a titanium frame can vary
from the designed angle?
Damon Rinard
Damon Rinard
PHILMIX
to last one season. It might not let me ride it tomorrow.
Phil Brown
Greg Lewis
degree (5 min.). However I would not worry too much about a .34 degree
change in angle. If you have the opertunity I would try the Kestrel
fork, and if the bike feels fine go for it. I had a replacement fork
sent to me from Medici for an old Medici. It had a shorter fork rake
and changed the angles by a full degree! (73 instead of 74). It
actually rode better with the new fork. Some builders do have
varriations on the order of 1/3 degree and some will be much better.
For building one bikes one at a time, the builder should have no
trouble compensating for varriations as he is building it.
Greg.
Rinards (rin...@aol.com) wrote:
: Serotta says if you use a Kestrel fork in a frame designed for an aluminum
: Damon Rinard
: Damon Rinard
--
Joshua_Putnam
>Good point. My oldest bike is an early 70's Gitaine that is full 531
>double butted tubes. Until I repainted it and filed down the excess
>brass and crude lugs it looked like it was brazed by a drunken sailor
>(no offense meant to drunks or sailors). In spite of probably being
>overheated the frame has held up to lots of hard use. The claim that
>531 frames only last a season or two is preposterous.
In fact, 531 frames may well be more durable than Columbus
frames when poorly built, simply because the Reynolds steel
can stand overheating better than heat-treated Columbus
can. 531 can be brass brazed at over 1900F without
significant loss of strength, while Colombus says not to go
over 1290F. For a frame put together with a 1100F silver
alloy, that leaves a margin of error of 800F for 531 but
only 190F for the Columbus.
--
Jo...@WolfeNet.com is Joshua Putnam / P.O. Box 13220 / Burton, WA 98013
"My other bike is a car."
New & used bike parts for sale: finger Joshua...@WolfeNet.com for list.
Suntour Superbe Pro pedal cages (also fit XC Pro bodies), $15/pair
Hans-Joachim Zierke
Bruce Jackson writes:
> Good point. My oldest bike is an early 70's Gitaine that is full 531
> double butted tubes. Until I repainted it and filed down the excess
> brass and crude lugs it looked like it was brazed by a drunken sailor
> (no offense meant to drunks or sailors). In spite of probably being
> overheated the frame has held up to lots of hard use. The claim that
> 531 frames only last a season or two is preposterous.
A 531 frame with severely overheated tubes will last two seasons in _pro_
use. This means: for the guys who can output 450 W constantly. Since my
output is in the 150 W and below range, a broken downtube is much less
likely.
Hans-Joachim Zierke
Joshua_Putnam writes:
> In fact, 531 frames may well be more durable than Columbus
> frames when poorly built, simply because the Reynolds steel
> can stand overheating better than heat-treated Columbus
> can. 531 can be brass brazed at over 1900F without
> significant loss of strength,
Hmmh... this must be a wonder material. I guess it wasn't created by
engineers, but by salespeople.
> while Colombus says not to go
> over 1290F.
Sounds like honest info. Any high-strength steel shouldn't. Otherwise, you
will always have recristallization. Some alloys might still rather strong
after that, but you always change material properties.
> For a frame put together with a 1100F silver
> alloy, that leaves a margin of error of 800F for 531 but
> only 190F for the Columbus.
No. If you overheat the silver brazing stuff, it disappears faster than you
can look at it. That's why beginning brazers hate silver. The tolerance for
"working temperature" is quite low.
Rinards
(Greg Lewis) writes:
>A good frame builder should be able to hold the angles to about 1/12
>degree (5 min.). [...] Some builders do have
>variations on the order of 1/3 degree and some will be much better.
>
>Greg.
Thanks for posting, Greg. Could you please explain how you know this, or
what steps the builder must take to hold these tolerances? I would like
to get all the info on this subject that there is to be had <G>, but I
need to be convinced of its accuracy. As you probably know, some
framebuilders have been known to pull numbers out of thin air...
I already recieved one excellent reply from Bikefixr which is very
reliable sounding:
"I measure every frame I sell to check for accuracy. I have a frame
alignment
table and micrometers and various protractors. I find head angle to vary
by
as much as 1.5 degrees from published spec."
Damon Rinard
Joseph Newman
: In article <4g3a9u$c...@geraldo.cc.utexas.edu>, le...@chaos.utexas.edu
: (Greg Lewis) writes:
: >A good frame builder should be able to hold the angles to about 1/12
: >degree (5 min.). [...] Some builders do have
: >variations on the order of 1/3 degree and some will be much better.
: >
: >Greg.
: Thanks for posting, Greg. Could you please explain how you know this, or
: what steps the builder must take to hold these tolerances? I would like
: to get all the info on this subject that there is to be had <G>, but I
: need to be convinced of its accuracy. As you probably know, some
: framebuilders have been known to pull numbers out of thin air...
: I already received one excellent reply from Bikefixr which is very
: reliable sounding:
: "I measure every frame I sell to check for accuracy. I have a frame
: alignment
: table and micrometers and various protractors. I find head angle to vary
: by
: as much as 1.5 degrees from published spec."
: Damon Rinard
I certainly haven't measured large numbers of frames. The number I
came up with are based on what my experience of what one can
reasonable build given standard tools and techniques. I built one
frame my self and talked to a number of builders. The 1/12 of a
degree figure is the limit of measuring accuracy with a standard
machinist bevel protractor. Paterek's (sp?) frame manual instructs
the builder to cold set each joint after brazing (if necessary) to
within this tolerance, and it is the accuracy a couple of other frame
builders have told me to achieve. There is not much reason to be more
accurate than this given the lengths involved in a bicycle. If one
makes adjustments during the building process to tube lengths and
angles, one can get a way with less accuracy. (A bike will really not
ride significantly differently if the angle are off by a quarter--we
could discuss how much of an error is significant but I don't think
there will be too much argument with the 1/4 degree figure).
If angles are varying by 1.5 degree from published specs, I would
think that most likely the builder has changed the design or that the
published specs are wrong (note I am not questioning that the figure is
accurate). If bikes of the same size are varying by
this much within a given batch of bikes the builder is very sloppy,
and should not be sell bikes! One cannot build good frames with this
kind of variation unless one is building them one at a time. Mass
production does not lend itself to this kind of variation.
Think about what a 1.5 degree variation would do to the bike. In the
seat angle it would put the seat cluster back by about 1.5 cm! That
is quite significant. In the head angle it would change the wheel base
by a similar amount and the trail by quite a bit (I don't have
a calculator in front of me to give you number). The handling
characteristics would be completely different!
One must change the fork rake or top tube length or something
to make sure the bike handles well. Just arbitrarily change the
seat or head tube angle by this much without concern for
its effect on the whole bike would make for some very bad bikes
I do know of one moderately popular Italian builder who`s frames of
a given size have a noticeable variation in head tube length.
I am a long way from an expert on these matters but I don't think
I've said anything unreasonable. Hopefully the experienced builders on the
net will add to (or correct) this.
Joseph Newman
: In fact, 531 frames may well be more durable than Columbus
: frames when poorly built, simply because the Reynolds steel
: can stand overheating better than heat-treated Columbus
: can. 531 can be brass brazed at over 1900F without
: over 1290F. For a frame put together with a 1100F silver
: alloy, that leaves a margin of error of 800F for 531 but
: only 190F for the Columbus.
I don't have the liturature in front of me but that 1900F
figure sounds high. Also I know that Columbus says not
to go over 1290F but that would preclude using brass.
The vast majority of bikes out of Italy are made
using brass not silver with Columbus tubing!
BTW, I am not sure what they are using now but Masi used 531 rather
than Colombus for years and years. (and Merckx raced on them). I'm
not sure of tubing wall thickness now, but back in the days when 531,
and Colombus SL were used in almost all road racing bikes (also some
Colombus SP which is thicker walled was used for larger frames),
Reynolds used slightly thicker walls than Colombus. This was the
reason Masi used 531.
The geometry of the bike, and the geometry of the tubing is more important
wheather Colombus or Reynolds (or true temper or prestige or ...) is used.
Joshua_Putnam
>Joshua_Putnam writes:
>> In fact, 531 frames may well be more durable than Columbus
>> frames when poorly built, simply because the Reynolds steel
>> can stand overheating better than heat-treated Columbus
>> can. 531 can be brass brazed at over 1900F without
>> significant loss of strength,
>Hmmh... this must be a wonder material. I guess it wasn't created by
>engineers, but by salespeople.
>> while Colombus says not to go
>> over 1290F.
>Sounds like honest info.
It's not that Reynolds encourages going over normal brazing
temperatures, but that 531 isn't a heat treated alloy, so you're
not losing the added strength of heat treating. There definitely
is a loss of strength from overheating, just not as large a loss
as with heat-treated CrMo.
For 531 tubes, Reynolds suggests a maximum liquidus of 1700F.
For 753, they say not to go over 1290F, the same as for Columbus
heat-treated steels. They're ven more conservative on the rear
stays of 731, 653, and 708 tube sets, saying not to go over about
1200F.
On the other hand, for their 525 CrMo tube set, they say
temperatures of 2000F are acceptable.
Greg Lewis
[snip]
: For 531 tubes, Reynolds suggests a maximum liquidus of 1700F.
: For 753, they say not to go over 1290F, the same as for Columbus
: heat-treated steels. They're ven more conservative on the rear
: stays of 731, 653, and 708 tube sets, saying not to go over about
: 1200F.
the 731, 653,and 708 rear stays are the same material as the 753.
in fact the 653 rear stays are exactly the same as 753. The 731
has slightly larger diameter and tapers slightly. (don't remember
what the 708 is but I'm pretty sure it is the same as 753 and
653--don't have the specs with me.)
: On the other hand, for their 525 CrMo tube set, they say
: temperatures of 2000F are acceptable.
: --
: Jo...@WolfeNet.com is Joshua Putnam / P.O. Box 13220 / Burton, WA 98013
: "My other bike is a car."
: New & used bike parts for sale: finger Joshua...@WolfeNet.com for list.
--
Hans-Joachim Zierke
Joshua_Putnam writes:
> It's not that Reynolds encourages going over normal brazing
> temperatures, but that 531 isn't a heat treated alloy, so you're
> not losing the added strength of heat treating. There definitely
> is a loss of strength from overheating, just not as large a loss
> as with heat-treated CrMo.
And...:
> On the other hand, for their 525 CrMo tube set, they say
> temperatures of 2000F are acceptable.
And now - put the two pieces together. Is there a remote possibility that
at least one part of the Reynolds PR isn't commercially grade pure
nonsense?
Joshua_Putnam
>Joshua_Putnam writes:
>> It's not that Reynolds encourages going over normal brazing
>> temperatures, but that 531 isn't a heat treated alloy, so you're
>> not losing the added strength of heat treating. There definitely
>> is a loss of strength from overheating, just not as large a loss
>> as with heat-treated CrMo.
>And...:
>> On the other hand, for their 525 CrMo tube set, they say
>> temperatures of 2000F are acceptable.
>And now - put the two pieces together. Is there a remote possibility that
>at least one part of the Reynolds PR isn't commercially grade pure
>nonsense?
I'm not sure if any of this info is in their PR material
anyway. It doesn't show up in any of the glossies I have,
only in their brazing specs.
525 isn't heat treated or thin-walled CrMo -- it's a set
specifically designed for machine brazing on cheap bikes.
The data on strength losses in 531 vs. heat treated CrMo
are independently verified in at least one of my frame
brazing books, though I don't have them handy at the moment
so I'm not sure which it is. I know from discussions with
various frame builders that brass-brazing 531 at 1600F is
quite common and produces durable frames.
--
Jo...@WolfeNet.com is Joshua Putnam / P.O. Box 13220 / Burton, WA 98013
"My other bike is a car."
New & used bike parts for sale: finger Joshua...@WolfeNet.com for list.
Gary Helfrich
: If angles are varying by 1.5 degree from published specs, I would
: think that most likely the builder has changed the design or that the
: published specs are wrong (note I am not questioning that the figure is
: accurate). If bikes of the same size are varying by
: this much within a given batch of bikes the builder is very sloppy,
: and should not be sell bikes! One cannot build good frames with this
: kind of variation unless one is building them one at a time. Mass
: production does not lend itself to this kind of variation.
I would be very suprised if production bikes from any manufacturer are
built to a tighter tolerance than +- 1/2 degree. This is not so much an
indictment of low manufacturing tolerances as it is an indication of the
riders' relative insensitivity to small changes in geometry.
The head angle can vary by as much as 1/4 degree just by differences in
welding distortion. This is based on my experience with steel and
titanium frame manufacture. I would think that aluminum would have
similar tolerances. Mass production lends itself quite well to this
"sloppy" tolerance. Weld a couple of thousand bikes, and I think that
you will agree. People tend to forget that these are bikes, not hard
drive head assemblies.
: Think about what a 1.5 degree variation would do to the bike. In the
: seat angle it would put the seat cluster back by about 1.5 cm! That
: is quite significant. In the head angle it would change the wheel base
: by a similar amount and the trail by quite a bit (I don't have
: a calculator in front of me to give you number). The handling
: characteristics would be completely different!
You might like to believe this, and so would the bike magazines. I've
done an interesting experiement a few years back that convinced me that
the gepmetry geeks were full of it. I built four identical steel frames
and put identical parts on them. Two of the bikes had 73 degree head
angles, one had a 72 degree head, and the last one was 74. The forks
were identical. I would ask people to identify the bikes in the order of
increasing head angle by riding them. No one was told that two of the
bikes were the same. I never met anyone who could get them correctly
every time. A interesting variation that I would do was to lower the
tire pressure in the front tire of one of the bikes. People would always
think that the one with the lower tire pressure had the slackest head angle.
Gary Helfrich
Arctos Machine
Rinards
fork would change a frame's geometry enough to require shaving the bottom
of the head tube to correct for it, as Serotta claims. I calculated the
change in head angle from installing the Kestrel fork to be .34 degrees.
My
thought was that there must be some unavoidable variation in head angle
just
from welding, etc., and that this variation might even be greater than
that
caused by installing the longer fork!
The consensus appears to be that all frames vary somewhat, typically in
the
same range of variation caused by installing a Kestrel fork (or more), and
that this amount is insignificant anyway, as illustrated by Gary
Helfrich's
"four bikes, four head angles" blind test ride experiments. Below is an
edited compilation of responses. Thanks to all who contributed!
Sincerely, Damon Rinard
-------------------------------------------------
From: Bike...@aol.com
I measure every frame I sell to check for accuracy. I have a frame
alignment
table and micrometers and various protractors. I find head angle to vary
by
as much as 1.5 degrees from published spec. The lateral alignment can be
out
as much as 3 or 4 degrees-this much needs serious realignment. But again,
fork rake, rim and tire choice can make this much a difference on a
perfect
frame.
And from le...@chaos.utexas.edu (Greg Lewis):
A good frame builder should be able to hold the angles to about 1/12
degree (5 min.). However I would not worry too much about a .34 degree
change in angle. [...] Some builders do have
variations on the order of 1/3 degree and some will be much better.
I certainly haven't measured large numbers of frames. The number I
came up with are based on what my experience of what one can
reasonable build given standard tools and techniques. I built one
frame myself and talked to a number of builders. The 1/12 of a
degree figure is the limit of measuring accuracy with a standard
machinist bevel protractor. Paterek's frame manual instructs
the builder to cold set each joint after brazing (if necessary) to
within this tolerance, and it is the accuracy a couple of other frame
builders have told me to achieve. There is not much reason to be more
accurate than this given the lengths involved in a bicycle. If one
makes adjustments during the building process to tube lengths and
angles, one can get away with less accuracy. (A bike will really not
ride significantly differently if the angle are off by a quarter--we
could discuss how much of an error is significant but I don't think
there will be too much argument with the 1/4 degree figure).
If angles are varying by 1.5 degree from published specs, I would
think that most likely the builder has changed the design or that the
published specs are wrong (note I am not questioning that the figure is
accurate). If bikes of the same size are varying by
this much within a given batch of bikes the builder is very sloppy,
and should not be selling bikes! One cannot build good frames with this
kind of variation unless one is building them one at a time. Mass
production does not lend itself to this kind of variation.
Think about what a 1.5 degree variation would do to the bike. In the
seat angle it would put the seat cluster back by about 1.5 cm! That
is quite significant. In the head angle it would change the wheel base
by a similar amount and the trail by quite a bit (I don't have
a calculator in front of me to give you number). The handling
characteristics would be completely different!
One must change the fork rake or top tube length or something as well,
to make sure the bike handles well. Just arbitrarily change the
seat or head tube angle by this much without concern for
its effect on the whole bike would make for some very bad bikes.
And finally, from lap...@wco.com (Gary Helfrich):
I would be very surprised if production bikes from any manufacturer are
built to a tighter tolerance than
Damon Rinard
Melissa Kepner
>A interesting variation that I would do was to lower the
>tire pressure in the front tire of one of the bikes. People would always
>think that the one with the lower tire pressure had the slackest head angle.
Ooooo, that's nasty! Wish I'd thought of it.
Jim
---------------------------------------------------------------------
Melissa Kepner Jim Adney
mgke...@facstaff.wisc.edu jra...@njackn.com
Laura Kepner-Adney
Madison, Wisconsin
---------------------------------------------------------------------