a perplexing frame design question

[Greg]

Does anybody out there know what is the historical origin of the 72 to 74 degree seat tube angle
used on most modern road (not tri) bikes? I believe I understand the reason for the head tube angle
and how it has gotten a bit steeper (now towards 74-75 deg.) over the years due to the general
improvement in roads which has reduced the need for highly raked forks which has in turn reduced the
need for a lot of 'fork trail' which has reduced the need for highly slack head tube angles (but
correct me if I am wrong in any of this).

Other than the aesthetic desire to have the seat tube approximately match the angle of the head
tube, I have not read or heard of any other explanation for the choice of 72 to 74 deg. SEAT TUBE
angle. Do you know of any? Since the American bike craze of the 1890s, has someone studied and
quantified the effect of the seat tube angle on the human body or on frame design enough to have
made this angle selection for the entire bike industry? Because if one looks at photos of American
"safety bicycles" from the late 1890s, is interesting to note how head tube angles are quite slack,
and that they are typically matched by a very similiar and very slack seat tube angle. Sometimes
these "slack-angled" bikes from the 1890s have a very interesting seatpost: one that has a FORWARD
elbow bend that then clamps the seat rails well forward of the end of the seat post, like some tri
seatposts of today (perhaps to keep the rider's knee over the pedal spindle when the cranks are 3
and 9 o'clock with such a slack seat tube?). I have read biomechanical studies on the relationship
between seat tube angle (and hip angle) and human power output, but all the studies that I have seen
have been recent (within the last 25 years or so). But obviously frame makers were deciding on seat
tube angles on "safety bicycles" well before that. What guided their choice?

If anybody knows where I can find out more about this, please let me know by posting here or writing
to [email protected]!

 

[Marten Gerritse]

Greg wrote:

> Does anybody out there know what is the historical origin of the 72 to 74 degree seat tube angle
> used on most modern road (not tri) bikes?

framebuilders prefer to build frames with the same seat and headtube angle. Seatpost offset takes
care of the required biomechanical position /Marten

 

[Sheldon Brown]

A shy "Greg" asked a good question:

> Does anybody out there know what is the historical origin of the 72 to 74 degree seat tube angle
> used on most modern road (not tri) bikes? I believe I understand the reason for the head tube
> angle and how it has gotten a bit steeper (now towards 74-75 deg.) over the years due to the
> general improvement in roads which has reduced the need for highly raked forks which has in turn
> reduced the need for a lot of 'fork trail' which has reduced the need for highly slack head tube
> angles (but correct me if I am wrong in any of this).

There's some truth to this, but the trail is not just a function of head angle; wheel radius and
fork rake also enter into it. Even for the same wheel size, you can get any desired amount of trail
with any desired head angle by varying the rake appropriately.

A friend of mine built a recumbent with a head tube angle something over 90 degrees, with reverse
raked fork. It looks very strange (he calls it his "Frankenbike") but rides fine, because the trail
is still in the normal range.

> Other than the aesthetic desire to have the seat tube approximately match the angle of the head
> tube, I have not read or heard of any other explanation for the choice of 72 to 74 deg. SEAT TUBE
> angle. Do you know of any?

Actually, I think you've got it backward. The seat angle is the more basic one, and head angles are
chosen relative to it.

> Since the American bike craze of the 1890s,

That was not an "American" phenomenon, it happened throughout the industrialized world.

> has someone studied and quantified the effect of the seat tube angle on the human body or on frame
> design enough to have made this angle selection for the entire bike industry? Because if one looks
> at photos of American "safety bicycles" from the late 1890s, is interesting to note how head tube
> angles are quite slack, and that they are typically matched by a very similiar and very slack seat
> tube angle. Sometimes these "slack-angled" bikes from the 1890s have a very interesting seatpost:
> one that has a FORWARD elbow bend that then clamps the seat rails well forward of the end of the
> seat post, like some tri seatposts of today (perhaps to keep the rider's knee over the pedal
> spindle when the cranks are 3 and 9 o'clock with such a slack seat tube?). I have read
> biomechanical studies on the relationship between seat tube angle (and hip angle) and human power
> output, but all the studies that I have seen have been recent (within the last 25 years or so).
> But obviously frame makers were deciding on seat tube angles on "safety bicycles" well before
> that. What guided their choice?

The upright bicycle as we know it is a very mature design, which has grown by a sort-of evolutionary
process. Over the last century and more, hundreds of thousands of very smart, clever folks have
spent millions of hours riding their bikes and thinking about ways that they might be improved. One
of 'em will make the down tube a little lighter, whaddaya know, the bike is lighter and works well.
Another one will make it a little lighter still, but then it turns out to be to easily damaged, so
they go back to the first guy's improved version. All of the dimensions of the bicycle frame have
been arrived at by similar processes.

The basic saddle position has to do with weight distribution. If the saddle is too far forward, you
put too much weight on your hands (with tri-bars, though how much weight is too much changes, since
it isn't carried by the hands.) If it's too far back, you can't pedal as powerfully.

If the seat tube angle is sub-optimal, riders will move the saddle back or forward to compensate,
even if an oddball seatpost is required to do it.

I personally believe that the optimal seat angle for most folks is around 72 degrees. Bill Farrell
of FitKit fame developed a concept called "Q Factor" which takes into account the rider's femur
length, and this concept states that a rider with proportionally long femurs will want a shallower
seat angle.

In the early '70s, a fad for steeper seat angles developed, first in Italy. This fad was primarily
caused by a desire to shorten the chainstays, which some people believed would make bikes faster or
more maneuverable. This fad has not entirely run its course.

Head angles do and should relate to seat angles...or perhaps more specifically to chainstay length.
If a bike has short chainstays, you also want a relatively short front-center dimension, to maintain
a reasonable weight distribution between the front and rear wheels.

A more vertical head angle makes a bike slightly more maneuverable, but tends to also make it a bit
harsher riding.

The upper limit on how steep the head angle can be depends on top tube length and the issue of
avoiding excessive interference between the rider's toes and the front wheel.

Sheldon "Geometry Was My Favorite School Subject" Brown
+---------------------------------------------------------+
| It is good to learn from your mistakes; | It is better to learn from the mistakes of others. |
+---------------------------------------------------------+ Harris Cyclery, West Newton,
Massachusetts Phone 617-244-9772 FAX 617-244-1041 http://harriscyclery.com Hard-to-find parts
shipped Worldwide http://captainbike.com http://sheldonbrown.com

 

[Greg]

marten wrote:

> framebuilders prefer to build frames with the same seat and headtube angle. Seatpost offset takes
> care of the required biomechanical position

Thanks for the reply.

Yes, I do see that. But WHY is that "framebuilders prefer to build frames with the same seat and
headtube angles"? Is it a desire for an aesthetically pleasing look? Or for some other reason(s)?

 

[Greg]

Sheldon wrote,

> There's some truth to this, but the trail is not just a function of head angle; wheel radius and
> fork rake also enter into it. Even for the same wheel size, you can get any desired amount of
> trail with any desired head angle by varying the rake appropriately.

Yes, but assuming wheel radius is fixed at the ICU mandated size, and for a historically steel fork,
a given amount of rake is necessary to get a reasonably compliant ride over unsmooth roads. This
requires a narrow range of head tube angles to get a reasonable amount of trail to get decent bike
handling, right?

> Actually, I think you've got it backward. The seat angle is the more basic one, and head angles
> are chosen relative to it.

This is confusing. I thought it was the head angle that was limited to a narrow range given the
limitations I listed above. ??

> Since the American bike craze of the 1890s, That was not an "American" phenomenon, it happened
> throughout the industrialized world.

I wrote this because the craze was far stronger in the US (at the time) than it was in other parts
of the world (see, "A Social History of the Bicycle" by Smith).

> The upright bicycle as we know it is a very mature design, which has grown by a sort-of
> evolutionary process. Over the last century and more, hundreds of thousands of very smart, clever
> folks have spent millions of hours riding their bikes and thinking about ways that they might be
> improved. One of 'em will make the down tube a little lighter, whaddaya know, the bike is lighter
> and works well. Another one will make it a little lighter still, but then it turns out to be to
> easily damaged, so they go back to the first guy's improved version. All of the dimensions of the
> bicycle frame have been arrived at by similar processes.

So, today, when a person or company is designing a road frame, the best they have to go on, is that,
"Joe down the street did it this way, so we'll do it that way too"? That seems amazing. One would
think someone would have applied more scientific scrutiny to the whole thing...

> The basic saddle position has to do with weight distribution. If the saddle is too far forward,
> you put too much weight on your hands (with tri-bars, though how much weight is too much changes,
> since it isn't carried by the hands.) If it's too far back, you can't pedal as powerfully.

Yes, the weight distibuiton of the rider's C.G. is critical. The position of his rear or hands is
less so, but plays into the equation, of course. So you are saying that the seat angle has not much
to do with biomechanical efficiency, but mostly with weight distribution and handling?

> If the seat tube angle is sub-optimal, riders will move the saddle back or forward to compensate,
> even if an oddball seatpost is required to do it.

In my experience, I see far more riders adapting to some crazy positions ("that's just the way
the bike came", or "the pro riders/triathletes now ride this way", etc.) than tinkering on their
own with their position to find better efficiency or power. Some bike shops are just as bad in
this area.

> I personally believe that the optimal seat angle for most folks is around 72 degrees. Bill Farrell
> of FitKit fame developed a concept called "Q Factor" which takes into account the rider's femur
> length, and this concept states that a rider with proportionally long femurs will want a shallower
> seat angle.

Yes, I know about the FitKit. But wouldn't a rider with such long femurs (and resultant shallow seat
angle) then be uncomfortably "folded over" at the waist if he were to use this shallow seat angle
and still keep his upper body low and aerodynamic?

--------------

Someone else led me to a link to the article "the myth of KOPS and bike fit" (linked to your
website?) What do you think of this article? I also have never found ANY real substantiation as to
why in the heck having your knee over the pedal spindle at the 3 o'clock position makes any
difference whatsoever in pedalling power or efficiency. What seems far more important is seat height
(easily affected by sliding one's rear fore and aft across an angled saddle) and hip angle (the
amount the rider is 'folded-over' at the waist). But even these have been pitifully examined by
folks in the exercise science field.

Thanks for all the interesting input. Any more would be welcome.

 

[Sheldon Brown]

I wrote:

>>There's some truth to this, but the trail is not just a function of head angle; wheel radius and
>>fork rake also enter into it. Even for the same wheel size, you can get any desired amount of
>>trail with any desired head angle by varying the rake appropriately.

A shy Greg wrote:

> Yes, but assuming wheel radius is fixed at the ICU mandated size,

Why would one assume that? Was this supposed to be a discussion related to racing bikes only?

> and for a historically steel fork, a given amount of rake is necessary to get a reasonably
> compliant ride over unsmooth roads. This requires a narrow range of head tube angles to get a
> reasonable amount of trail to get decent bike handling, right?

I would assume the frame builder would also build the fork, so any desired amount of rake could be
accommodated. I've ridden very nice riding bikes with head angles as shallow as 68 degrees, in fact
I even own one. I've also got a '30s German bike with a head angle on the order of 64-65 degrees,
but haven't much ridden it for other reasons.

>>Actually, I think you've got it backward. The seat angle is the more basic one, and head angles
>>are chosen relative to it.
>
> This is confusing. I thought it was the head angle that was limited to a narrow range given the
> limitations I listed above. ??

Nope.

>>The upright bicycle as we know it is a very mature design, which has grown by a sort-of
>>evolutionary process. Over the last century and more, hundreds of thousands of very smart, clever
>>folks have spent millions of hours riding their bikes and thinking about ways that they might be
>>improved. One of 'em will make the down tube a little lighter, whaddaya know, the bike is lighter
>>and works well. Another one will make it a little lighter still, but then it turns out to be to
>>easily damaged, so they go back to the first guy's improved version. All of the dimensions of the
>>bicycle frame have been arrived at by similar processes.
>
>
> So, today, when a person or company is designing a road frame, the best they have to go on, is
> that, "Joe down the street did it this way, so we'll do it that way too"? That seems amazing. One
> would think someone would have applied more scientific scrutiny to the whole thing...

Most bikes are designed by marketing/styling experts, not engineers or scientists.

Also, the liability crisis makes it scary for any big company to do something that is outside of
the ordinary.

>>The basic saddle position has to do with weight distribution. If the saddle is too far forward,
>>you put too much weight on your hands (with tri-bars, though how much weight is too much changes,
>>since it isn't carried by the hands.) If it's too far back, you can't pedal as powerfully.

> Yes, the weight distibuiton of the rider's C.G. is critical. The position of his rear or hands is
> less so, but plays into the equation, of course. So you are saying that the seat angle has not
> much to do with biomechanical efficiency, but mostly with weight distribution and handling?

Right.

>>I personally believe that the optimal seat angle for most folks is around 72 degrees. Bill Farrell
>>of FitKit fame developed a concept called "Q Factor" which takes into account the rider's femur
>>length, and this concept states that a rider with proportionally long femurs will want a shallower
>>seat angle.
>
>
> Yes, I know about the FitKit. But wouldn't a rider with such long femurs (and resultant shallow
> seat angle) then be uncomfortably "folded over" at the waist if he were to use this shallow seat
> angle and still keep his upper body low and aerodynamic?

This is where parallel design comes in, because matching a shallow head angle to the shallow seat
tube angle allows the top tube to be kept in reason without excessive toe-overlap.

In any case, most cyclists don't want to be "low and aerodynamic."

> Someone else led me to a link to the article "the myth of KOPS and bike fit" (linked to your
> website?) What do you think of this article?

I find it quite persuasive, that's why I put it on my Website

http://sheldonbrown.com/kops.html

> I also have never found ANY real substantiation as to why in the heck having your knee over the
> pedal spindle at the 3 o'clock position makes any difference whatsoever in pedalling power or
> efficiency.

Same here. I prefer to be a bit farther back than that.

Sheldon "Not Much Interested In Racing" Brown +--------------------------------------------------+
| For every complex problem, there is a solution | that is simple, neat, and wrong. | --H. L.
| Mencken |
+--------------------------------------------------+ Harris Cyclery, West Newton, Massachusetts
Phone 617-244-9772 FAX 617-244-1041 http://harriscyclery.com Hard-to-find parts shipped Worldwide
http://captainbike.com http://sheldonbrown.com

 

[Jobst Brandt]

Greg <[email protected]> writes:

> Does anybody out there know what is the historical origin of the 72 to 74 degree seat tube angle
> used on most modern road (not tri) bikes? I believe I understand the reason for the head tube
> angle and how it has gotten a bit steeper (now towards 74-75 deg.) over the years due to the
> general improvement in roads which has reduced the need for highly raked forks which has in turn
> reduced the need for a lot of 'fork trail' which has reduced the need for highly slack head tube
> angles (but correct me if I am wrong in any of this).

The angle is designed to optimize axial fork loading from road shock. A three dimensional
probability curve will show that with a common road roughness there is an optimum angle to prevent
the fork from being loaded in bending both in severity and load magnitude. In olden days roads were
unpaved and rougher, requiring a lower angle of attack for this condition.

You can test this by riding no hands on some average roads with the head bearing adjusted loosely
enough to make clunking noises on bumps. I think you'll be surprised how small an amplitude of road
roughness will cause reverse loads (clunks) in the head bearing. In addition, frame builders can
tell you which way fork blades break in fatigue (at the fork crown). I think you'll find that with
72 degrees they break forward more often, and at 75 degrees, rearward.

Jobst Brandt [email protected] Palo Alto CA