Serious tech: How's frame geometry design done?
Frank Krygowski
physics article, and the Soma Smoothie page that Andrew linked to.
"What it is...
The Smoothie road race frame is agile and responsive. Excellent for
stage races and competitive club rides. But also comfortable enough
for daily commutes. ..."
Well, of _course_ it's agile, responsive, comfortable and all the
rest. But seriously: Anyone have inside information on how bike
companies arrive at geometry these days?
As the physics article made clear, it's not possible to do it by pure
physics and math. Seems they must copy other bikes, including their
own previous models, probably making fine adjustments to fork rake,
angles, bottom bracket drop, wheelbase, etc. for various purposes.
But who exactly makes those decisions? What references or resources
do they use? How do they learn that part of the trade? How many
prototypes would be made to dial in a fairly typical bike model?
I suspect that there's such a wide range of acceptable dimensions that
it's actually not a very demanding process. (Otherwise, how would one-
man shops be able to do good custom frames?) But I'd like to learn
about how it's actually done.
- Frank Krygowski
AMuzi
For any given fork (and designers do work around standard
models of available forks now) there's a clearance drop and
head angle set which will get you to desired trail values
for a given wheel size. There's so much 'prior art' that no
new rows need be plowed.
Modern bike models differ in that a racing or raceable bike
will be under a meter wheelbase with 450 calipers and
23~25mm tires while a sport bike [1] will have 5~6cm longer
wheelbase, longer chainstay, clearance for 28~32mm tires and
mudguards hence a 500 caliper.
Specialty models (cross, touring etc) have their own
requirements such as less drop for cross and track, much
more for loaded touring and so on.
A quick perusal and comparison of various geometries at
gunnarbikes.com or at somafab.com may be a useful exercise
for the curious reader. Bicycle frame design is not black
art but the limits and requirements span a very wide range
across riders' intended use.
[1] 'sport' is what we called 'race' 40 years ago. Race
bikes got shorter over time, right to the point of
ridiculous, and then settled around 98cm WB with a 41 CS
(medium size).
--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971
kolldata
right. for a single use frame, here a woods/trekking bike, copy frame
geometry and photo pages into Word for the available model range then
compare.
You will see within the chosen use range, specs toward touring,
agility, climbing, power straight ahead eg double track,...
small differences in mm but significant in performance. I am a clod
rider with the frame and wheels taking me thru it no problem.
Tºm Shermªn °_°
Well, not *all* designers use standard forks. Some even use a single
leg and a one-sided wheel mount.
RANS has built prototype bicycles with adjustable fork rake, so handling
can be more readily optimized by trial and error.
> Modern bike models differ in that a racing or raceable bike will be
> under a meter wheelbase with 450 calipers and 23~25mm tires while a
> sport bike [1] will have 5~6cm longer wheelbase, longer chainstay,
> clearance for 28~32mm tires and mudguards hence a 500 caliper.
>
Some bicycles have adjustible wheelbase length to better fit the rider
and improve front/rear weight distribution.
> Specialty models (cross, touring etc) have their own requirements such
> as less drop for cross and track, much more for loaded touring and so on.
>
> A quick perusal and comparison of various geometries at gunnarbikes.com
> or at somafab.com may be a useful exercise for the curious reader.
> Bicycle frame design is not black art but the limits and requirements
> span a very wide range across riders' intended use.
>
Here is one attempt to provide a mechanistic approach to bicycle
handling: <http://www.calpoly.edu/~wpatters/lords.html>.
> [1] 'sport' is what we called 'race' 40 years ago. Race bikes got
> shorter over time, right to the point of ridiculous, and then settled
> around 98cm WB with a 41 CS (medium size).
Here is another approach to going fast:
<http://farm4.static.flickr.com/3195/3105378175_b2fd47a94c.jpg>.
--
Tºm Shermªn - 42.435731,-83.985007
I am a vehicular cyclist.
Peter Cole
Since none of those characteristics can be quantified, much less
measured in a repeatable way, bike sellers can claim anything with
impunity. Many riders with collections of different geometry bikes
report that when switching between them the new bike may feel "funny" at
first, but they quickly adjust and stop thinking about it. Similar
experiences have been reported by people who deliberately made forks
with a range of trail, the differences were only permanently noticeable
at the extremes.
If you're tired at the end of a long ride, it's easy to imagine the bike
is "fatiguing", or if you fail to negotiate a fallen log on the trail
that your bike isn't "agile" enough, and most bike sellers will be happy
to reinforce those suspicions.
kolldata
>
> - Show quoted text -
well, that was my point.
fabricators publish geometry and image. The format is kindof standard,
even.
Comparing stats in a standarized format allows the beyond monkey a
quantified overview of what he's riding into.
Then you can ROITS, exspeariencing what the geometry stats said to
you, wether you have an improvement or not.
Dan O
> Here's a serious bike tech question.
(Qualify up front my naivette - but you guys know that :-)
> It riffs off the bike stability
> physics article, and the Soma Smoothie page that Andrew linked to.
>
> "What it is...
> The Smoothie road race frame is agile and responsive. Excellent for
> stage races and competitive club rides. But also comfortable enough
> for daily commutes. ..."
>
Marketing blather.
> Well, of _course_ it's agile, responsive, comfortable and all the
> rest. But seriously: Anyone have inside information on how bike
> companies arrive at geometry these days?
>
> As the physics article made clear, it's not possible to do it by pure
> physics and math. Seems they must copy other bikes, including their
> own previous models, probably making fine adjustments to fork rake,
> angles, bottom bracket drop, wheelbase, etc. for various purposes.
>
> But who exactly makes those decisions? What references or resources
> do they use? How do they learn that part of the trade? How many
> prototypes would be made to dial in a fairly typical bike model?
>
> I suspect that there's such a wide range of acceptable dimensions that
> it's actually not a very demanding process. (Otherwise, how would one-
> man shops be able to do good custom frames?) But I'd like to learn
> about how it's actually done.
>
The good ones use what is prove to work, and try incremental changes
that they theorize might bring improvement. Racing in all it's
various forms is a good proving ground.
(Also, to wrap up my initial qyualifier, let me take this opportunity
to apologize for polluting this *awesome* group with my personal
problems and what not, and to thank *all* of you - my friends - for
keeping tech alive in a context of enabling life to the fullest.)
Dan
> Here's a serious bike tech question.
(Prefacing qualifier of my naivette.)
> It riffs off the bike stability
> physics article, and the Soma Smoothie page that Andrew linked to.
>
> "What it is...
> The Smoothie road race frame is agile and responsive. Excellent for
> stage races and competitive club rides. But also comfortable enough
> for daily commutes. ..."
>
Hyperbolic marketing blather.
> Well, of _course_ it's agile, responsive, comfortable and all the
> rest. But seriously: Anyone have inside information on how bike
> companies arrive at geometry these days?
>
> As the physics article made clear, it's not possible to do it by pure
> physics and math. Seems they must copy other bikes, including their
> own previous models, probably making fine adjustments to fork rake,
> angles, bottom bracket drop, wheelbase, etc. for various purposes.
>
> But who exactly makes those decisions? What references or resources
> do they use? How do they learn that part of the trade? How many
> prototypes would be made to dial in a fairly typical bike model?
>
> I suspect that there's such a wide range of acceptable dimensions that
> it's actually not a very demanding process. (Otherwise, how would one-
> man shops be able to do good custom frames?) But I'd like to learn
> about how it's actually done.
>
The good ones start with what's proven to work, and make incremental
changes based on their theory of what might advance the state of the
art and science and joy. Racing in all it's various forms is an
ideal proving ground for this.
(To followup the preface: I apologize for polluting this fine group
with my personal problems and mental issues. Thank you, my friends,
for keeping tech alive in the context of helping others live life to
the fullest.)
landotter
Laterally stiff and vertically compliant, anyone? Or from a steel
obsessed website I was reading the other day: steel's slight
compliance is "liveliness" and "dancing" while CFRP's same
characteristics are "isolating" and "dead".
Tºm Shermªn °_°
> Laterally stiff and vertically compliant, anyone?[...]
The above can of course be accomplished with suspension. Or by proper
design on a recumbent.
Not so much on a DF upright.
--
Tºm Shermªn - 42.435731°N, 83.985007°W
landotter
$southslope.net"> wrote:
> On 7/10/2011 12:02 PM, landotter wrote:
>
> > Laterally stiff and vertically compliant, anyone?[...]
>
> The above can of course be accomplished with suspension. Or by proper
> design on a recumbent.
>
> Not so much on a DF upright.
>
I get it. All those double squishy mtb frames are a figment of my
imagination.
Tom Ace
> Laterally stiff and vertically compliant, anyone?
Achieving (the illusion of) that with oval seat stays
was patented: http://tinyurl.com/p4513985
Tom Ace
Tºm Shermªn °_°
Er, I *did* mention suspension.
James
The Merida O.NINE is a hard tail MTB that has, IIRC. 9.4mm of
"suspension" travel with 100kg hanging from the seat post. I asked for
the test method from Merida. Details of the reply are here;
http://www.merida-bikes.com/en_int/cms/63/o-nine
--
JS.
James
For the custom frame I had made, I made most of the decisions based on
what I had before and what I wanted to change.
I wanted the front wheel a little further from my toes, and the rear
wheel to have a little more clearance to the frame. I also know what
position I like the seat and bars relative to the BB, so we adjusted the
head tube length and seat post angle and length to suit. I was quite
happy with the steering on the old bike, so (almost) matched the head angle.
We iterated the design drawings and remeasured my old bike until I was
satified that the modifications were just what I wanted. I have no
complaints with the result. The Columbus tubes provide plenty of EH&M too.
A good custom frame design is in achieving the right fit for the individual.
The combinations and permutations of seat posts, seats, fork steerer
length and head stems make it possible to mass produce a small set of
frame geometries and have them fit most people, but the result isn't
always optimal. You might end up with the seat all the way back on the
rails, or an uber long head stem that adds flex and unnecessary weight
to the front of the bike. (The bending moment on the steerer would be
increased by a long stem I think).
--
JS.
Michael Press
<85730357-d38d-44e1...@r2g2000vbj.googlegroups.com>,
landotter <land...@gmail.com> wrote:
> Laterally stiff and vertically compliant, anyone? Or from a steel
> obsessed website I was reading the other day: steel's slight
> compliance is "liveliness" and "dancing" while CFRP's same
> characteristics are "isolating" and "dead".
Pure poetry.
--
Michael Press
Frank Krygowski
In other words, you made adjustments based on what you'd ridden
before. And this is what I suspect goes on in almost all bike
companies, from builders of one-off custom frames to, say, Trek or
Cannondale.
In the larger companies, I'm still curious about who it is that does
the work of specifying those adjustments (i.e. the frame geometry) and
especially, how they learn their craft.
To put it another way: Say at some mid-sized bike company, all five
of the guys designing frame geometry were killed in a bus crash. Say
five complete novices had to take on the job. Would anybody riding
the next model bikes even notice?
BTW, I'm convinced that the "poetry" writers Michael Press referred to
have as much influence on the process as the designers. I'm betting
"Stiff and responsive" measures just the same as "Harsh and
unforgiving," but influences the feedback loop in different ways.
Over the decades, the adjustment process has resulted in a gradual
change in bike geometry. We have something interesting now with Jan
Heine and _Bicycle Quarterly_. His particular "poetry" is claiming
that lots of trends have been in the wrong direction - that certain
features of 1950s racing bikes are actually better for many riders
than the current high-tech stuff.
I wonder how often such naysaying has been part of the "poetry." And
will this affect the fashions five or ten years from now?
- Frank Krygowski
Dan O
Did you even read my suggestion?
Frank Krygowski
You mean "The good ones use what is prove to work, and try incremental
changes that they theorize might bring improvement"?
Um... yeah. I read that.
Did you read the original post? Do you have any answers to the
questions I asked?
- Frank Krygowski
Dan O
Yeah (and qualified my naivette in initial response). And sorry if I
kind of dismissing the breadth of your last restatement of the
question (as is evident, "I can feel... one of my spells coming on")
From that other thread:
"I hate to sound like I take anything on sheer faith. I appreciate
that the good...
... [racing]...
... mechanics / builders / tuners may have a closer grasp to
the physics, but I'm still sure they don't believe they have it all
figured out."
It's an art and science with infinite variables depending on
individual humans - no two of which are exactly the same - but on the
edge, I imagine experienced feelings, gut sense, intuition, and daring
all play a part along with a decent understanding of the physics; and
experience (such as racing) is the proving ground.
James
Probably not quite. Stiff hopefully is a measure of torsional
stiffness, but is not necessarily directly proportional to vertical
compliance (which may be affected by seat post length and material for
example).
Responsive is a term used to describe how agile the rider feels in
controlling the bike. Is it easy to turn quickly? Does it feel like
you're dragging a sack of potatoes when you stand on the pedals, or does
it feel light and lively? These terms are not necessarily related to
stiffness, harshness or compliance.
> Over the decades, the adjustment process has resulted in a gradual
> change in bike geometry. We have something interesting now with Jan
> Heine and _Bicycle Quarterly_. His particular "poetry" is claiming
> that lots of trends have been in the wrong direction - that certain
> features of 1950s racing bikes are actually better for many riders
> than the current high-tech stuff.
Quite possibly. It is possibly true to say that most race bred machines
are not suitable for the majority. Being designed for a specific
purpose makes them less suitable for general purpose.
Fitting a rack and full mudguards to my racing bike would not be
practical, for example.
--
JS.
Dan O
True especially of optimization for what most people probably think of
as bicycle racing, but racing in all it's myriad forms is an ideal
proving ground for performance characterisitics in any direction.
> Fitting a rack and full mudguards to my racing bike would not be
> practical, for example.
>
Surely people somewhere must be racing such bikes. Now, whether any
frame designers with a mainstream market are paying attention...
David Scheidt
:To put it another way: Say at some mid-sized bike company, all five
:of the guys designing frame geometry were killed in a bus crash. Say
:five complete novices had to take on the job. Would anybody riding
:the next model bikes even notice?
The next model, maybe not. They're likely to be cautious, and not
change anything more than the paint color. but they're likely to do
something stupid in two or three model iterations, where an
experienced designer would say "you nimrod, if you do that, nasal
demons will kill the rider".
:Over the decades, the adjustment process has resulted in a gradual
:change in bike geometry. We have something interesting now with Jan
:Heine and _Bicycle Quarterly_. His particular "poetry" is claiming
:that lots of trends have been in the wrong direction - that certain
:features of 1950s racing bikes are actually better for many riders
:than the current high-tech stuff.
Well, 50's racing bikes were only a small step away from practical
bikes. Modern racing bikes aren't very useful for anything but
racing, and pretending your a racer. Few non-racers are well served
by that.
--
What's the rule on that?
Tºm Shermªn °_°
>> Fitting a rack and full mudguards to my racing bike would not be
>> practical, for example.
>>
>
> Surely people somewhere must be racing such bikes. Now, whether any
> frame designers with a mainstream market are paying attention...
With a full hard-shell fairing, one does not need mudguards and a rack.
And yes, people race such things:
<https://www.youtube.com/watch?v=PTx7DKSlbPg>.
thirty-six
$southslope.net"> wrote:
> On 7/11/2011 9:46 PM, Dan O wrote:
>
> > On Jul 11, 4:00 pm, James<james.e.stew...@gmail.com> wrote:
> >> [...]
> >> Fitting a rack and full mudguards to my racing bike would not be
> >> practical, for example.
>
> > Surely people somewhere must be racing such bikes. Now, whether any
> > frame designers with a mainstream market are paying attention...
>
> With a full hard-shell fairing, one does not need mudguards and a rack.
> And yes, people race such things:
> <https://www.youtube.com/watch?v=PTx7DKSlbPg>.
>
"On the weekend of the 5th and 6th June was the Cyclevision event in
the Netherlands. I took part in the six hour race on the Sunday,
riding the same Sinner Mango that I ride to work, with Harry's camera
mounted on the back of my Mango. These are a few bits of video from
the first hour and a half."
NETHERLANDS - mostly flat and the course had exceptionally large
radius bends. I wonder why such a course was selected. Almost
totally impractical in England, there are corners in cycle friendly
roads, and steep hills. The only practical roads would be those
recommended for HGVs, a very poor mix of vehicle types.
thirty-six
>
> > On Jul 11, 4:00 pm, James<james.e.stew...@gmail.com> wrote:
> >> [...]
> >> Fitting a rack and full mudguards to my racing bike would not be
> >> practical, for example.
>
> > Surely people somewhere must be racing such bikes. Now, whether any
> > frame designers with a mainstream market are paying attention...
>
> With a full hard-shell fairing, one does not need mudguards and a rack.
> And yes, people race such things:
> <https://www.youtube.com/watch?v=PTx7DKSlbPg>.
>
Like the video shows, riding around in big circles on level ground
away from general traffic are all they are good for. Why are there
no chicanes or corners? Regular bike races usually have at least one
tight corner, a bottleneck and a hill. HPV races are boring in
comparison. Does a real bike stimulate an adrenaline rush that you
are unable to cope with?
Tºm Shermªn °_°
> On Jul 12, 8:40 am, Tºm Shermªn °_°<""twshermanREMOVE\"@THI
> $southslope.net"> wrote:
>> On 7/11/2011 9:46 PM, Dan O wrote:
>>
>>> On Jul 11, 4:00 pm, James<james.e.stew...@gmail.com> wrote:
>>>> [...]
>>>> Fitting a rack and full mudguards to my racing bike would not be
>>>> practical, for example.
>>
>>> Surely people somewhere must be racing such bikes. Now, whether any
>>> frame designers with a mainstream market are paying attention...
>>
>> With a full hard-shell fairing, one does not need mudguards and a rack.
>> And yes, people race such things:
>> <https://www.youtube.com/watch?v=PTx7DKSlbPg>.
>>
>
>
> "On the weekend of the 5th and 6th June was the Cyclevision event in
> the Netherlands. I took part in the six hour race on the Sunday,
> riding the same Sinner Mango that I ride to work, with Harry's camera
> mounted on the back of my Mango. These are a few bits of video from
> the first hour and a half."
>
> NETHERLANDS - mostly flat and the course had exceptionally large
> radius bends. I wonder why such a course was selected. Almost
> totally impractical in England, there are corners in cycle friendly
> roads, and steep hills. The only practical roads would be those
> recommended for HGVs, a very poor mix of vehicle types.
Tight radius bends are not a problem with a roll cage, but just add to
the fun:
<https://www.youtube.com/watch?v=Iwn_davpik8&feature=player_embedded>.
Tºm Shermªn °_°
Tight corners, heavy traffic, and plenty of excitement:
<https://www.youtube.com/watch?v=Iwn_davpik8&feature=player_embedded>.
(Same video as on other recent post.)
thirty-six
$southslope.net"> wrote:
> On 7/12/2011 4:23 AM, thirty-six wrote:
>
> > On Jul 12, 8:40 am, Tºm Shermªn °_°<""twshermanREMOVE\"@THI
> > $southslope.net"> wrote:
> >> On 7/11/2011 9:46 PM, Dan O wrote:
>
> >>> On Jul 11, 4:00 pm, James<james.e.stew...@gmail.com> wrote:
> >>>> [...]
> >>>> Fitting a rack and full mudguards to my racing bike would not be
> >>>> practical, for example.
>
> >>> Surely people somewhere must be racing such bikes. Now, whether any
> >>> frame designers with a mainstream market are paying attention...
>
> >> With a full hard-shell fairing, one does not need mudguards and a rack.
> >> And yes, people race such things:
> >> <https://www.youtube.com/watch?v=PTx7DKSlbPg>.
>
> > "On the weekend of the 5th and 6th June was the Cyclevision event in
> > the Netherlands. I took part in the six hour race on the Sunday,
> > riding the same Sinner Mango that I ride to work, with Harry's camera
> > mounted on the back of my Mango. These are a few bits of video from
> > the first hour and a half."
>
> Who else races their all-weather commuting and shopping bike?
I did. I used a cycle cape and mudguards, usually zip-tied.
thirty-six
You are a joke, those curves have a 10m radius, not 3m ( and 165deg)
like could be found in a road race circuit or criterium. It wouldn't
of course be a fair race with upright cyclists as they sweep through
the bends bouncing off the bubbles and turning the occupants over like
beetles on their backs. I'd have great fun in a combined race if I
had the energy.
Tºm Shermªn
> like could be found in a road race circuit or criterium.
3m radius of curvature? YHGTBSM!
See Exhibit 4.9 (page 18):
<http://www.mhd.state.ma.us/downloads/designGuide/CH_4.pdf>.
> It wouldn't
> of course be a fair race with upright cyclists as they sweep through
> the bends bouncing off the bubbles and turning the occupants over like
> beetles on their backs. I'd have great fun in a combined race if I
> had the energy.
I think you would find the upright cyclists would be the ones getting
the worst of the collisions here on Planet Earth.
Planet Trevor, where you live, may be different.