The following is a sorted version of the discussion which occurred on the list-serv Hard Core Bicycling Science. This is an interesting and pretty sophisticated group of people as will be evident from the discussion.
However, none of these people has the final word on the topic. If you have taken a completely different approach to the design of the suspension, then you need to defend it in terms of some of the discussion here, but do not assume that you are wrong.
The purpose of a design class is to force you to understand the existence of constraints and to learn the process of task definition. This is a combination of "soft skills" which must be supported by hard facts and analysis. The discssion below is a good example of a design related discussion.
You should make particular note of the discussion of the effect of vibration on the human body. This is a very different perspective on suspension design and may be less relevant on a downhill bike, but it is certainly an important factor.
Mick
Resorted in order (9/99 discussion):
*****************************************************************************
I'm new to the list and just recently have started to intensely study the science of the bicycle. I'm currently a student of mechanical engineering and am looking to enter the bicycle design industry upon graduation. At the moment I'm designing a full suspension MTB frame from a concept that I came up with about a year ago, specifically, I'm currently attempting to optimize to design by looking at pivot location and line of travel. What I'm looking for is some quality, up to date, literature which will help me with the design process at hand. I haven't been able to find much and any titles that have proven useful in you're design by looking at pivot location and line of travel. What I'm looking for is some quality, up to date, literature which will help me with the design process at hand. I haven't been able to find much and any titles that have proven useful in you're endeavors would be quite useful. I am looking for both bicycle specific titles as well as other related subjects such as biomechanics, dynamics, and suspension design. Thank you in advance for you're help.
Steve Paterson
Mech. Eng. Student
Queen's University
*****************************************************************************
Steven Paterson asks about references for engineering bicycle suspension pivot locations.
I have seen some serious efforts by Maury Hull and co-authors -- I think one of those was Wang. I believe at least one of their papers appeared in VEHICLE SYSTEM DYNAMICS.
My disagreements with their assumptions and approaches range from mild to severe, but at least they represent one set of approaches.
The big issue, in my mind, is not doing an analysis 'correctly' (though I think Hull's is questionable in spots) -- but in formulating a correct model of the relevant factors and design goals. For example, it is very easy to affect bouncing by very slight upper-body motions. Also dramatic energy absorption is possible through rider motions, far greater than in any suspension. Rider 'properties' can be highly volitional, for example if you ride over a severe bump without seeing it you respond very differently than when it is anticipated.
So, even if Hull's analyses had been unimpeachable, I believe his conclusions would still have been irrelevant to good suspension design.
Jim Papadopoulos
*****************************************************************************
Date: Fri, 24 Sep 1999 09:28:05 -0700
Sender: owner-hardcore-bicycle-science@cycling.org
Precedence: bulk
Reply-To: hardcore-bicycle-science@cycling.org
In talking about reference sources for bicycle suspension, Jim Papadopoulos commented:
>The big issue, in my mind, is not doing an analysis 'correctly' (though I think
Hull's is questionable
>in spots) -- but in formulating a correct model of the relevant factors and design
goals.
To which I wholeheartedly agree, with an emphasis placed on the "design goal". As an example: if we limit the discussion to two basic types of suspension (there are many more applications but these are the big ones), then as a designer you must choose between suspension designed for cross-country riding or suspension designed for downhill riding. The requirements for these two design goals are very different, and which you choose determines most of your design constraints. An example: as Jim mentioned, the up and down nature of the pedal stroke inherently introduces a driving force that causes suspended bicycles to bob up and down. The magnitude of this bobbing is dependent upon many factors, several of which are shock damping, spring pre-load, spring rate, and pivot location. Cross-country suspension designs seek to limit this bobbing while at the same time allowing for enough suspension softness to absorb bump forces that exceed a threshold limit. The bobbing force due to the pedal stroke may be just below this limit, thus having very little affect on the suspension. Downhill suspension on the other hand seeks to absorb all bump forces, and many designers go out of their way to make the suspension as compliant as possible. You immediately notice this when you sit on any of the downhill specific supension bikes: they often sag up to 2-3 inches.
The upshot of this post is: if you are interested in designing a downhill suspension bike, I suspect the motocross industry may provide you with the most relevant information. If you are designing a cross-country bike, you should look at more references specific to bicycles, since the pedal forces are of primary concern and do not exist on other forms of suspended transportation. I don't know of any good motocross references, but there have been many senior projects submitted at universities throughout the nation dealing specifically with this topic. One university I know of that has copies of these projects available is Cal Poly San Luis Obispo, CA. Their web address is www.calpoly.edu and you can search the library for senior projects dealing with bicycle suspension. I'm sure they would make copies of the projects and send them to you if you asked (although they may charge a fee).
One additional comment I would like to make regarding suspension design is that the shock makes it or breaks it. Pick the shock you want to use and design your suspenion around that shock (most shocks can also be tuned when built to match specific suspension requirements). You would be amazed at how big a difference the shock can make in the way the bicycle rides. I'm not just talking about picking a "good" shock over a "non-good" shock, I'm saying the shock and the suspension must work together, so design your pivot locations (swingarm rate and leverage) around your shock (or vice versa if you have the resources). A good shock company should be able to supply you with spring rate and damping characteristics that will allow you to model the system together.
Garrett Smith
Truvativ International
*****************************************************************************
At 02:25 AM 9/23/1999 -0400, Steven Patterson wrote:
> At the moment I'm designing a full suspension MTB frame from
> a concept that I came up with about a year ago, specifically,
> I'm currently attempting to optimize to design by looking
> at pivot location and line of travel. What I'm looking for
> is some quality, up to date, literature which will help me
> with the design process at hand.
As Jim P. mentioned, there was a good deal of research performed at UC Davis under Professor Maury Hull. I would look for research under his name as well as his students - the ones I remember were: Eric Wang, Stan Needle, and Mike DeLorenzo. Many of Hull's papers were published in Cycling Science. I don't think the information will really help you though. Some of Jim's articles in Cycling Science are better for looking at the effects of pivot placement. Also search the HBS archives at www.bsn.com/cycling. We've had a few discussions already about this subject.
I recommend ignoring anything you have read about suspension designs in magazines - the information is almost always incorrect. Try looking for books on vehicle dynamics too. Pay special attention to dive and squat (load transfer) and how to negate it. Since the C.G. height to wheelbase ratio for bicycles is large, the effects can be significant.
Marc Pfister
*****************************************************************************
> Steven Paterson asks about references for engineering
> bicycle suspension pivot locations.
Patents are also a source of information.
/Marten
*****************************************************************************
Jim Papadopoulos reponded to my post regarding suspension by writing:
>>as a designer you must choose between suspension
>>designed for cross-country riding or suspension
>>designed for downhill riding. The requirements
>>for these two design goals are very different,
>>and which you choose determines most of your design
>>constraints.
>I'm not going to argue against what you say, but
>I think the issues are sometimes much more subtle
>than 'cross country vs. downhill'.
To which I agree. My use of cross-crountry vs. downhill was only to serve a point: that the design goals and constraints can be very different depending upon the use of the product. For example I'm sure recumbent suspension opens up a whole new arena of design goals and constraints.
>It seems likely to me that a rider learns to perform
>either slight torso motions to reduce bouncing,
>or gets used to it in some fashion, so that it becomes
>less of an issue. If that's true, what is the
>proper design goal then????
Again, I agree, with the key phrase being "less of an issue". I don't think it makes the issue non-existent, or even that much less of an issue. From my experience, a well designed full suspension cross-country bike will require the least amount of adaption by the rider. Even with biological adaption and a full range of gears, I'm fairly confident that I would out-ride myself on
a cross-country bike vs. a downhill bike (for cross-country riding that is). The opposite is true for a downhill race. In both these cases the design of the machine has a larger impact on performance (and hence perception) than biological adaption (at least in the short term ~ less than a million years or so). So, I would say the design goals remain the same.
Truvativ International
*****************************************************************************
Reply-To: hardcore-bicycle-science@cycling.org
The biggest hurdle to get over is the tight coupling of a deterministic system with limited degrees of freedon (the bicycle) and the non-linear, real time, adaptive control element (the rider). This is not to deter anyone from trying to model the bicycle half of the system to analyze performance. Rather it is to recognize the challenge of developing meaningful criteria for design evaluation in a case where the control element is a major factor in performance and the system is (almost) impossible to rationally test without the control element.
*****************************************************************************
*****************************************************************************
Reply-To: hardcore-bicycle-science@cycling.org
Bruce Vermeychuk writes:
> The biggest hurdle to get over is the tight coupling of a
> deterministic system with limited degrees of freedom (the bicycle)
> and the non-linear, real time, adaptive control element (the rider).
I think that just scratches the surface because people like world champion Frischknecht rides a rigid bicycle for most of his MTB races and uses the suspension fork of his sponsors mainly because they are the sponsors. You won't get that comment from him on the record but it is perceptible when you ride with him.
That means that on a rigid bicycle, suspension is furnished, has been done in the past, by the legs and arms and the most important part of this is the arms, surprisingly. When the hands are shaken sufficiently, vision is blurred and this limits speed because the rider is riding into a fog of his eyes making. This is why the suspension stem has been pursued as an alternative. Unfortunately, the stem has no other benefits and is also a less accommodating linkage in the direction needed.
So it boils down to modeling the abilities of the human body, the part of the system that is most often ignored by the designers of off road bicycles who think they are modeling the behavior of the system. I am more convinced of the pragmatic approach that has served bicycling
well up to now. Much of the motorcycle suspension designs we see are are largely unsuited to bicycling because they are derived from a system in which the vehicle outweighs the rider and in which the vehicle has power to make up for losses of the damped system. It also does not have a rider who is lunging on the pedals to propel it.
Jobst Brandt <jbrandt@hpl.hp.com>
*****************************************************************************
Reply-To: hardcore-bicycle-science@cycling.org
Jobst Brandt writes:
> When the hands are shaken
> sufficiently, vision is blurred and this limits speed because the
> rider is riding into a fog of his eyes making.
>
> So it boils down to modeling the abilities of the human body, the part
> of the system that is most often ignored by the designers of off road
> bicycles who think they are modeling the behavior of the system.
Several years ago, while working to create a "passenger comfort model" to objectively evaluate automatic transmission shift quality, I came across some fascinating work done by the U. S. Army Tank Command. The focus of the work was to determine exactly how much shock and vibration (and the corresponding power spectra) the tank crew could tolerate before their operational performance was significantly degraded.
I'm still hunting for the original sources - this may not be available on line, I think the work was done between 1948 and 1955. What I remember was that relatively low amplitude vibration in a frequency range of 0.1 to 10 Hz had the greatest negative effect on crew performance. From experience, this seems to be in about the same range as off road bikers may experience if rigidly coupled to their ride - so they don't sit.
If I am able to relocate the sources, I'll post a pointer.
Bruce Vermeychuk
*****************************************************************************
Reply-To: hardcore-bicycle-science@cycling.org
Bruce Vermeychuk wrote about work published from the US Army Tank Command. References that I had found from the US Army Tank Automotive Center might be similar. The main stuff was a bunch of papers authored by Fred Pradko and others, on Human Vibration Response and on Comfort Criteria.
For example ASME papers 65-WA/HUF-19 and 66-WA/BHF-15 and SAE papers 650426 and 660139 (660138 by others also seems relevant)
Jim Papadopoulos
*****************************************************************************
Aside on 9/99 discussion:
*****************************************************************************
Reply-To: hardcore-bicycle-science@cycling.org
Hello everybody!. I did not write to the list for the last year or so.
Now that I see you are in the middle of an interesting discussion on bike suspension, I find it attractive to reenter the list.
The proposition that there are two classes of suspension: for cross country and for descent, might be not totally correct, simply because it is based on the assumption that the suspension is of the trailing swing arm type. Now assume one bike that has the front steering hub connected to the rear wheel through one rigid frame, say a fat tube. The rear swing arm is pivoted just at the rear wheel axle (You read it right). The BB with pedals and chainring at the front of this arm is suspended from the rigid frame by a (For example) rubber band, from the BB to the frame on top of it. This is a simplified description of the bike I have designed.
The suspension travel can be bigger than in standard bikes, because you can design the rigid frame with a convex shape, looking to the ground, so as the BB can raise much higher than in standard bikes. The bottom of the travel is when the pedal or the chainring hits the ground. This ridiculous bike. can be operated (And was), with no seat, for research purposes, just to eliminate the effect of seat linkages. I have not been able to detect any jumping or pogoing of this suspension even with the softest suspension. Then,there is a bar (You can cal it seatbar) that goes, pivoted at both ends from BB to near the seat, and is connected to the steering hub with another doubly pivoted bar ( You can call it top bar). This top bar goes from the steering hub to a pivot near the seat where it connects wit the seatbar.
These bars, swingarm, seatbar topbar and the bike's frame, form a typical hinged quadrilateral, so fancy nowadays in Mechanics. This configuration produces a nearly straight line motion in the seatbar, up and down, so you can install a pair of spring loaded friction pads anywhere in the mainframe. To allow for a good operation of these pads, the seat bar itself is of rectangular crossection or has two plates welded to the sides.
The real bike I have been driving for over one year now, has the top bar hinged not at the steering hub but to a sleeve, that slides up and down along a very long handlebar stem-This sleeve, which is connected to the handlebar is suspended by a very long coil spring.
In my probably biased opinion this is a wonderful bike.
First: front and rear wheels are joined by a very rigid frame and the gravity forces applied to it are tamed by a very soft suspension. So its behavior with the ground is as good as in the best nonsuspended bike, or better. Second: the very soft front and rear suspensions makes an "oldman" comfortable bike. Third: the great lateral rigidity of the hinged quadrilateral (Specially with the addition of the friction pads, that also act as guides) makes a "=B7young man" excellent sports bike, making it possible precision handling in spite of the softness.
Anybody wishing to see the bike please go to
http://www.losalcores.com/melchorduran
*****************************************************************************
> "Melchor Duran" wrote
>
> The rear swing arm is pivoted just at the rear wheel axle (You read it
> right).
>
1889 whippet? (http://www.bikeculture.com/bcq/bcq8/ar8-1.html)
/marten
*****************************************************************************
Reply-To: hardcore-bicycle-science@cycling.org
Melchor Duran wrote:
> The proposition that there are two classes of suspension: for cross
> country and for descent, might be not totally correct, simply because it
> is based on the assumption that the suspension is of the trailing swing
> arm type.
Also there may be a third type - one which is designed to give lower rolling resistance - the new softer ones tend to be less efficient
> This ridiculous bike. can be operated (And was), with no seat, for
Just as a side note - the latest trials motorcycles have no seat at all.
+=====================================================+
Resorted in order (‘98 discussion):
Craig Good (craig@real.uwaterloo.ca)
Thu, 26 Feb 1998 17:39:33 -0500 (EST)
I have recently gone through the academic exercise of developing a computer algorithm
that will derive the equations of motion for any planar mechanical system. The only
stipulation is that the planar mechanical system must be made up of elements found in a
library of modelling components. Now that the code is done, I am looking for
applications to use my program with. As a demonstration, I would like to develop the
equations of motion to predict PEDALLING INTERACTIONS in a rear suspension mountain bike.
I have created a simple model and I would like to invite everyone to comment on my model.
As this will be a demonstration of my algorithm, please keep in mind I want to keep
the model as simple as possible.
MODEL DESCRIPTION
-----------------
I have created a 3 mass model of a bicycle and rider which isolates the motion of the rear
suspension. One can view a picture of my model at:
http://real.uwaterloo.ca:80/~craig/research/research.html
My model constists of 3 masses, the bike frame and forks, the crankset, and the rear
wheel. The frame is attached to ground at the front axle by a pin joint. The
frame is also connected to the crankset by a pin joint. The crankset is then attached to
the crankset by a chain drive element. The rear axel of the bike is constrained by a
linear slider which is parallel to the bicycles direction of travel (kinematically
constrains rear tire to remain on the ground).
I have not yet included the elements which make up the suspension system. However,
since my algorithm can handle any of elements in the component library, this is as simple
as adding the appropriate links and springs in the right places. In the
drawing I have added a simple suspension which would add 2 more masses and a spring to the
system.
Applied Forces.
The mass of the rider is not directly included in the three mass model. However, it is
added by applying appropriate D'Alembert forces (Fmx,Fmy) at the rider's center of mass
(xm,ym). The pedal, seat and handlebar loads are then applied to the model. These
interface forces were measured for both standing and seated climbing by Stone et al.
(1990).
A retarding torque is also applied to the rear wheel. This appropriately simulates
the air resistance, inertial and gravitational forces a rider would encounter. This
retarding torque is a function of the velocity and acceleration of the rear wheel
(derivative of theta wrt t and 2nd derivative of theta wrt t) My model is now
complete and I am ready to generate the equations of
motion which when integrated will predict suspension 'bob' as a function of my input
loads.
I would appreciate any constructive critism of my actual model.
Thanks,
Craig
Note to Moderator: I have tried to keep this explaination as simple as possible.
However, I have had to skip over some stuff at the potential expense of running on for
pages and pages. Please let me know if there is any thing that needs to be rewritten
or redrawn. Thanks a million.
###Most engineering models contain significant simplifications, which play out as
limitations in quantitative or qualitative applicability. If anyone digs into your
work I'm sure you will get further opportunities to explain details!
Moderator####
-----------------------------------------------------------------------
Craig Good
EMAIL: craig@real.uwaterloo.ca
WEB: http://real.uwaterloo.ca/~craig
262 Anita Court, Graduate Student, MOtion
Research Group (MORG)
Waterloo, Ontario University of Waterloo, Office: CIM
2709
Canada N2K 2R4 885-1211 Extension
3646
(519) 888-9249 40 Knots + Steep Port
Ramps = Day Off School
Challenge the Known ---> Embrace the Unknown
-----------------------------------------------------------------------
Jobst Brandt (jbrandt@hplabsz.hpl.hp.com)
Thu, 26 Feb 1998 18:38:01 PST
Craig Good writes:
> I have recently gone through the academic exercise of developing a
> computer algorithm that will derive the equations of motion for any planar
> mechanical system. The only stipulation is that the planar mechanical
> system must be made up of elements found in a library of modeling
> components.
This is not the first time someone has come up with a model for thevarious aspects of a
bicycle, however, it is the rider that is not practical to model. If you follow the
threads on wreck.bike, there are endless discussions why one frame rides better than
another. Some of this is pure fantasy but I believe part of it is real. I mean
that there is great disagreement on whether one bicycle works better than another.
The reason is that riders have greatly different techniques. I had it easy with
modeling the bicycle wheel because I didn't use it as a design mechanism, but rather an
analysis of what the wheel can support. The wheel got to where it is
pragmatically. Ultimately,bicycles are relegated to do that as well, regardless of
people's desires to include its peculiarities in design forecasts. The reason for
failures is often not anticipated but the school of trial and error weeds them out.
> Now that the code is done, I am looking for applications to use my
> program with. As a demonstration, I would like to develop the equations of
> motion to predict PEDALING INTERACTIONS in a rear suspension mountain
> bike.
I'm not sure the rear suspension has an engineering solution. A springy rear end,
and even a front one, interferes with the basic
operation of propelling the bicycle. I have had the opportunity to talk with Ritchey
and Frischknecht about this and for most races the rigid bike is preferred. Rock
shocks is a sponsor and furnishes needed money for the team but without anyone directly
saying so, it appears that the suspension is not an advantage. Downhill racing is
another matter entirely. For this the best vehicle is a motor-less dirt motorcycle
with more than one foot of front and rear travel.
I have been to plenty of bike shows and the suspensions I see offered there are exercises
in how complex a linkage one can make to approximate a swing arm without actually having
one. There are many theories where the pivot point must be located, most of which
are predicated on countering the adverse effects of suspension while pedaling hard.
I have not yet seen one that a racer would use in Paris-Roubaix.
> I have created a simple model and I would like to invite everyone to
> comment on my model. As this will be a demonstration of my
> algorithm, please keep in mind I want to keep the model as simple as
> possible.
At this point you have come upon the difficult part. What is the input? As you
know, garbage in = garbage out.
Jobst Brandt <jbrandt@hpl.hp.com>
tullio@TheRamp.net
Sat, 28 Feb 1998 10:20:21 -0600
Specialized has been publishing the results of a rear suspension study conducted by the
Technical University of Hamburg and reported by Mountain Bike of Germany. I am told
that the study utilized computer modeling . I have not read the report from the
University nor the magazine article so I cannot comment on the particulars. Is
anyone familiar with this study?
By the way, the summarized results shown by Specialized rank their FSR as the best in the
study for both downhill and cross country use (out of 10 bikes tested). This may
explain their interest in promoting the study.
Todd Kuzma
********************************
* Tullio's Big Dog Cyclery *
* LaSalle, IL
*
* 815-223-1776
*
* e-mail: tullio@TheRamp.net *
* Raleigh-Specialized-Bianchi *
* Waterford-Torelli-GT/Dyno *
* Burley-Co-Motion
*
********************************
Mon, 2 Mar 98 9:34:13 PST
I have a copy of the study in German. Sadly, I don't read German, except for a
few key words like "auspuffen". I am suspicious of the methodology
they used. However, I am riding a Specialized FSR, and it is an excellent rear suspension
design. What bothers me is that they also gave high ratings to a Trek or Fisher
Y-bike, a design that in real life is demonstrably inefficient. ###I assume you DON'T
mean: 'lower energy output for a given energy input'. If you do, please clarifyhow that
concept applies to suspensions. Moderator###
A simulation of a rigid rider riding up a smooth hill on a variety of rear suspension
designs would be illuminating, and possible.
John Olsen
-------------
Original Text
From: tullio@TheRamp.net, on 2/28/98 10:20 AM:
>Specialized has been publishing the results of a rear suspension study
>conducted by the Technical University of Hamburg and reported by Mountain
>Bike of Germany. I am told that the study utilized computer modeling.
Mon, 2 Mar 98 13:53:06 PST
I wrote:
>What bothers me is that they also gave high
>ratings to a Trek or Fisher Y-bike, a design that in real life is
>demonstrably inefficient.
Then the Moderator interjected:
>###I assume you DON'T mean: 'lower energy
>output for a given energy input'. If you do, please clarify
>how that concept applies to suspensions. Moderator###
Yes, a trigger word, isn't it? Some suspension bikes couple pedaling forces into
suspension forces enough that the suspension is forced to work noticeably when a rider
pedals, even if he does it smoothly. Low pivot URT designs, unless set up with
enough preload that there is no static sag, get exercized on every pedal stroke, using
some of your pedaling energy to warm up the damper. They are slower than a more
neutral full suspension bike.
There. That should get some fun discussion going.
John Olsen
Mon, 2 Mar 1998 19:49:34 EST
In a message dated 98-02-26 22:02:05 EST, Jobst Brandt writes:
> I'm not sure the rear suspension has an engineering solution. A
> springy rear end, and even a front one, interferes with the basic
> operation of propelling the bicycle.
You're ignoring a couple of factors here. The first is the human factor. Bumps in off-road
riding interfere with human performance. I will not even attempt to offer any
discussion of size of bumps, isolating continuous vibrations versus discrete impacts, etc.
But if you agree that a rider will encounter even one type of bump that will force them to
stop pedaling and take their butt off the seat on a rigid, then you must consider the
human factor before discounting it. Or if you will agree that even one vibration type will
force a rider to use their arms less than they would like (tired wrists), then you must
also consider the human engine factor.
The second issue is terrain levelling. Now you will probably first tell me that a bicycle
doesn't have enough sprung weight to give sufficient preload to level terrain in the way a
on-road car or off-road motorcycle does. But the "bouncing" action which turns
forward momentum into vertical momentum on bumps takes away from your horizontal speed and
therefore your efficiency in getting to your destination. So suspension could forseeably
help enough in rounding out the bumps to make up for it's bio-inefficiencies (and I assume
bio-feedback was what you were referring to in your quote above, Jobst.) Also by turning
bumps into more sinusoidal affairs (less bouncing, less energy lost into arms and legs
than with sudden "sharp" bumps).
So that's why I disagree with your quote above, and imagine that a computer simulation
could tell us a GREAT deal about what performance envelope a suspended bike can help
improve (range of bump size, speeds, length of ride). I doubt that Suzuki or Honda have
considered propulsion efficiency (related to engine output--not traction efficiency) as a
#1 design criterion.
Luigi Petrighi
Tue, 3 Mar 1998 00:32:48 -0600 (CST)
John Olsen wrote about SUSPENSION 'EFFICIENCY':
>Some suspension bikes couple
>pedaling forces into suspension forces enough that the
>suspension is forced to work noticeably when a rider pedals,
>even if he does it smoothly. Low pivot URT designs, unless
>set up with enough preload that there is no static sag, get
>exercized on every pedal stroke, using some of your pedaling
>energy to warm up the damper. They are slower than a more
>neutral full suspension bike.
Your position is that pedalling forces can cause suspension motion, which therefore
transforms mechanical energy into heat, and makes the bicycle slower?
I don't find the argument compelling..... why do you think it's true? (Unfortunately, this
is another area where rider awareness of the damper could easily affect perceptions or
behaviour.)
There are two reasons I'm dubious. One is an issue of magnitudes -- small damper motions
really dissipate very little energy (especially when damper force is due to orifice flow
i.e. proportional to V^2, rather than viscous force i.e. proportional to V). I think that
Wang and Hull measured oscillations and estimated that they dissipated about 5 W. In steep
uphills, with a base power of 250 W, this might mean a speed change of 2%. If your speed
is low, your speedometer might not even resolve the difference. (While at higher speeds,
the percentage change in speed will be less.)
Going further along these lines, it's possible to reduce bouncing by properly timing
upper-body countermoves. What happens to energy dissipation then (within the rider's
body) is anyone's guess.
Now, it IS credible to me that suspension upsets pedalling. But this would be true even
without damping, i.e. without any energy lost to the bicycle. You could lose your rhythm,
or your freedom to generate substantial inertial reactions from slight body motions. In
other words, you could be prevented from pedalling effectively. Or, you could end up
moving your body in such a way that your muscles absorb energy (like stopping a swing). So
here is a key question: is a suspended, but undamped bicycle, noticeably faster on smooth
terrain than a damped one?
When bumps are present, there's also a question of energy loss while traversing them (for
example, coasting down a bumpy trail). It's not obvious to me that the addition of a
damper would necessarily reduce coasting speed. (To resolve this analytically could
require an accurate prediction of how a person's body shakes around due to a bump.)
My perspective is that it's impossible to make a good theoretical argument, and pretty
difficult to prove by testing, whether damping reduces smooth-surface speed.
Jim Papadopoulos
Tue, 3 Mar 98 12:08:06 PST
Item: 2 riders, one on a hardtail mountain bike, the other on a
comparably-equiped Full Suspension mountain bike, are riding on the road. Mr.
Suspension is working hard to keep up with Mr. Hardtail. They switch bikes, and the
new rider on the fs bike is working hard to keep up. The FS bike is strongly
coupled, that is, pedaling forces deflect the rear suspension
with no other input.
Item: a strong, experienced rider borrows a FS bike from me, and rides his daily
commute on it. He mounts his cycling computer on the bike, which has similar
tires and equipment to his own, and rides his level, 20 mile commute. He notices
that he is averaging 2 MPH less than on his own bike, despite similar weather conditions.
This is a low-pivot URT Trek Y-bike
set up with static sag of about 1".
Item: A strong rider breaks his hardtail bike, borrows a FS mountain bike from
me. He normally outclimbs me by a good margin on long, steady 2000 ft. climbs.
A group of 5 riders go out on such a climb, and my buddy falls off the back,
complaining that the bike is a dog. The bike is 3.5 lbs. heavier than his hardtail.
The FS bike is strongly coupled.
Item: I am testing a Fisher low-pivot URT, a strongly-coupled design. The rear
suspension deflects suprisingly far with each pedal stroke. The bike feels slow,
especially up hills when I am pushing hard rather than spinning. The bike is equiped
with a rear-suspension lock-out. Everyone who rides the bike brings it back with the
suspension locked out. The reason each gave, independently, was to the effect that
the bike "ate my energy" with the suspension active.
Weak evidence? Sure. Sure is a lot of it, though. FS bikes are always heavier
than hardtails of comparable quality, and
certainly, price. This makes them slower up hills by itself. However, when you
see a strongly coupled design's shock being pumped through 30% of its travel on a smooth
road climb by a not-unusually-bouncy rider, it is not hard to believe that a regrettable
amount of energy is going away, energy that came from a rider's legs.
####Is a heavier bike noticeably slower up a hill? I don't buy it.
Moderator, aka Jim Papadopoulos###
###You do us all a disservice to suggest that lots of energy is dissipated this way. The
actual amount should be relatively
tiny, not sufficient to cause the feelings claimed for it. Have you given any
thought to my proposition that a suspension
bike WITHOUT a shock will also feel unpleasant and slow?###
That is, it isn't hard to believe if you've actually tried it, rather than theorized about
it. Jim, you've got to actually ride some of these things! We wouldn't be wasting
quite so much time arguing about silly stuff that a number of the prime arguers have never
tried.
####Sorry, it's far from silly. I'm well aware from riding 20 years ago that scores of
riders claimed exquisite sensitivity to
spoking pattern, steel alloy, seatstay tubing gauge, etc. It's doubtful that comparable
claims have any greater validity today.
####Secondly, I'm well aware of my tendency to IMAGINE that a bike feels a certain way, if
I know
the bearings are dirty or the steel is cheap. I daresay I could be quite prepared to
believe that a FS bike
feels a certain way, or even 'wastes energy'. Without good blind testing, all these
'feelings' are suspect.
####Thirdly, I have a faith in Newton's laws that means I don't always have to experience
something, in order to make an informed guess about how it has to go. When a sufficiently
small percentage of power is dissipated, it's very likely that
no-one can actually tell. ####
For instance, if you tell me that you have ridden steel, aluminum, and carbon frames
extensively, and that you don't perceive that they filter road vibrations differently, I
will respect your observations. If you haven't done the experiment, then
your theories are worth less. ###That experiment was done with radially spoked
wheels in the 70's. Calculation makes the same result plausible for frames, though of
course an experimental confirmation would be nice. ####
An experienced, honest rider, one with no axe to grind, can do a pretty good job in
back-to-back, A/B comparisons, in many situations.
####That's actually a supposition I'd like proof for. When I went looking, I couldn't find
it.####
John Olsen
Tue, 03 Mar 1998 16:26:07 PST
John Olsen writes:
> Item: 2 riders, one on a hardtail mountain bike, the other on a
> comparably-equipped Full Suspension mountain bike, are riding on the
> road. Mr. Suspension is working hard to keep up with Mr. Hardtail.
> They switch bikes, and the new rider on the FS bike is working hard
> to keep up. The FS bike is strongly coupled, that is, pedaling
> forces deflect the rear suspension with no other input.
...and cites other examples of poorer suspension performance than with an unsuspended
bike. I don't find this hard to believe but I don't attribute it to energy
absorption by suspension dampers or friction but rather to interference with pedaling.
Just when you want to push hard on the pedal, Lucy pulls the football away, and
Charlie Brown falls on his ass, so to speak. This same action can be simulated on a
rigid bike with either an elastic band for a chain or a wind-up spring in the rear hub.
You can put stroke into it, but you can't put much force into it, so the work
transmitted is reduced.
This has always been the complaint about suspension, not that energy gets lost. It
never gets out of the riders leg. Meanwhile he is bouncing up and down doing
springboard work.
Jobst Brandt <jbrandt@hpl.hp.com>
Wed, 4 Mar 98 8:35:31 PST
Jobst Brandt writes:
>...cites other examples of poorer suspension performance than
>with an unsuspended bike. I don't find this hard to believe but I
>don't attribute it to energy absorption by suspension dampers or
>friction but rather to interference with pedaling.
Jobst, you could be completely correct. I am assuming that the damper is a major
contributor, but I don't know this for a fact. The Lucy analogy is a good one, and
that could be a, or the, big player.
I think that we could do some order-of-magnitude calcs to see what kind of power a damper
is consuming, but I haven't had
time to do it yet. One could assume a 210-lb rider and bike climbing a 5
degree hill at some power level, thus, some speed.
Say there is 180 lbs. on the rear wheel, the damper has a 3:1 lever ratio, critical
damping, and it is being forced to stroke
at 50 RPM x 2 through a .75" stroke at the damper. Rough numbers, but
reasonable. One could calculate the power lost to the damper, subtract that power
from the power driving the bike up the hill, and get a delta-V. If this is
negligible, then the observed slowness could be from some power-delivery issue, or???
In support of the latter theory, the FS designs that seem to climb most nearly like
hardtails don't allow suspension motion
to create bizarre pedal velocity profiles. The bad ones feel funny when you pedal
them up even a smooth hill.
John Olsen
Wed, 04 Mar 1998 19:12:36 PST
John Olsen writes: > You could be completely correct. I am assuming that the damper is a > major contributor, but I don't know this for a fact. The Lucy > analogy is a good one, and that could be a, or the, big player. Actually I have a better example. Running across a trampoline is a good parallel to pedaling a sprung bicycle. My experience with this is long ago when riding a balloon tired clunk with low tires. I am sure the major part of this is what is often called eccentric work, in which muscles re-absorb work as in descending a stairway. I am not moved by the damper argument although I know all the energy it dissipates is rider effort. Interfering with the stroke of a pedaling rider is much more disruptive than some systematic loss. > In support of the latter theory, the FS designs that seem to > climb most nearly like hardtails don't allow suspension motion > to create bizarre pedal velocity profiles. The bad ones feel > funny when you pedal them up even a smooth hill. Yes and I believe you can see it in elapsed times in comparison runs. Jobst Brandt <jbrandt@hpl.hp.com>
Wed, 4 Mar 1998 22:46:10 -0500
Jobst Brandt wrote on 3/3/98 19:26: >I don't find this hard to believe but I >don't attribute it to energy absorption by suspension dampers or >friction but rather to interference with pedaling. Just when you >want to push hard on the pedal, Lucy pulls the football away, and >Charlie Brown falls on his ass, so to speak. This same action can be >simulated on a rigid bike with either an elastic band for a chain or a >wind-up spring in the rear hub. You can put stroke into it, but you >can't put much force into it, so the work transmitted is reduced. The effect of suspensions I know nothing about but the analogy I am forced to question. Since it is interesting in its own right, and since it involves another question that came up recently, I am hiving it off into a new thread. It happens that I just fitted my Trice with a new rear wheel whose freehub, I was surprised to discover, transmits its power through a spring. Or what amounts to a spring: a roller clutch that compresses a substantial amount as it is engaged. The result is remarkable resilience. It feels as though the chain is an elastic band. Very strange indeed. And, it turns out, very wonderful. Immediately I changed to it, the cadence I could maintain while climbing increased dramatically. For the first time in the month that I have owned the Trice, riding it ceased to feel constantly tiring. Some weeks ago Jim Papadopoulous asked why it's difficult to maintain a high cadence when going slowly. Assuming compensatory gearing, the only difference at the pedals ought to be the greater change in resistance around their arc as the slower-moving cycle loses more momentum between power strokes. Well, it looks to me now as though that greater change in resistance may be the key: I suspect that the proprioceptive system senses the unwonted accelerations in pressure and reduces internal stresses by telling the limb to slow down. Adding a spring to the hub smooths out the accelerations in pressure so that the legs can maintain a higher cadence with all the usual advantages this entails. I ought to add that I would be unusually sensitive to this effect. My knees are dodgy, pedalling the recumbent Trice involves a different mix of muscles than pedalling my uprights, and I climb at a laughably slow pace. At the other extreme, I can imagine a racer whose knees have encountered so much stress that his proprioceptive system has learned to ignore even damaging amounts of force: he would feel no advantage to the spring and would feel its loss of instantaneous response a disadvantage. Aside from this slight loss of responsiveness, however, I do not see a disadvantage. Although it will not transmit all the rider's force the instant he applies it, it seems to me as though it accept all that he can apply, store all that it cannot transmit at the moment, and transmit the excess later. It may not feel equally efficient but it ought to be equally efficient. This, I suggest, would be an excellent bit of research for an undergraduate thesis: test for maximum cadence and maximum 30-minutes (or so) output using four bicycle dynamometers: conventional hub w. heavy flywheel, conventional hub w. light flywheel, resilient hub w. heavy flywheel, and resilient hub w. light flywheel. Charles Maurer 5 Grandview Court Dundas, Ontario Canada L9H 5C8 Telephone & fax: 905.627.7035
Wed, 04 Mar 1998 21:31:57 PST
jolesn@paccar.com writes: >[How about some order-of-magnitude calculations, say ...] You read my mind. Even the numbers. Our moderator writes: >[When a sufficiently small percentage of power is dissipated, it's > very likely that no-one can actually tell.] Is it small? Take John's example of the suspension moving substantially through it's stroke on each pedal stroke. Suppose a rider weighing X kgf and the damping force is P %age of the rider's weight. E.g., if the rider is 100kg and damping is 1kgf, then P=.01 in the following. I do not know what is a typical damping force. The work to compress the suspension is mostly returned except for what is lost in damping. Suppose the damping is purely compression or purely rebound. Then: Suppose PX kgf damping force, .03m travel, 60 rpm. That's PX * .03 * 60 kgf*m/min = PX * .03 kgf*m/sec For comparison Rider P kgf, 3 m/sec (10kph), 1% grade P * 3 * .01 = P * .03 kgf*m/sec So for our hypothetical rider, going from a rigid bike to a suspended bike of the same weight but using a squishy coupled suspension adds (PX * .03) / (P * .03) = X If the damping is .01, it absorbs about 1% of the amount of energy required to climb a 1% grade at 10kph. Put another way, it takes a 1% grade to a 1.01% grade. Suppose the rider is climbing a 5% grade and 3/4 of their energy goes to lifting, then the percentage lost to the suspension is .01 percent / 5 percent * 3/4 = .0015 or under 2/10 of a percent energy ``stolen'' by the suspension. In comparison if a bike weighs B and adding suspension increases it to B+S, the extra lifting in the above climb is ((P + B + S) / (P + B) - 1) * 0.75 if P=100kgf, B=10kgf, and S=1kgf, the loss is about 6/10 of a percent. Riders certainly _claim_ they can feel the difference in weight; if that's true it's reasonable they would also feel the difference in energy lost to suspension, which is about 1/4 of that. Furthermore, they would feel the suspension loss always, while the above energy loss from weight would only be felt on a 5% climb but not on flats. Questions: did I screw up my calculations? Do riders of coupled suspension really have 3cm excursions on each pedal stroke? Is damping really 1kgf for a 100kgf rider? Does 1kgf of weight make a difference in a double-blind hill climb? ;-D on ( Pencil and paperless ) Pardo ###Don, I induced Eric Soroos (then at Cornell, more recently in your neck of the woods) to perform blind testing with weights on bikes to be ridden uphill. I think his weight increment was closer to 2 kgf, and only one of the subjects could detect that 'sometimes'. I keep forgetting the details though..... maybe they appear in an article in the archive? Moderator###
Thu, 05 Mar 1998 08:26:09 -0800
I think Jobst is right. The "walking on a trampoline effect" is very noticable when the an experienced rider is sprinting or making an hard effort out of the saddle. The clearest sensations of the differences from the rider's point of view is apparent in these circumstances. It is also in these cases that the forces on the bike are large enough to cause large displacements of the suspension elements and proportionally larger losses. These would be difficult to decouple. There is a learning curve associated with out of the saddle pedaling on FS bikes though. A rider who puts in the time can adapt to the "walking on the trampoline" phenomena, just as it is possible to learn to walk on a trampoline, and learn to transmit more power. The differences are subtle changes in pedaling technique (for lack of a better way to say it for now). I doubt that one ever becomes much better in pedaling a softly sprung FS bike in an all out sprint. There is no room for adaptation of technique in that case. That's not as important in MTB riding as it is in track and road, so it's not the end of the story. It would be if the discussion was limited to road racing bikes. There is certainly a small cost in total power transmission in the process of adaptation (beyond damping losses) and the peak will suffer. You can't get it all back even if you practice a lot. There may be physiological costs as well, a reduced ability to optimize the system. Dunno. There is also a psychological aspect of sprinting or climbing hard on a suspended bike. Until a skilled rider spends some time on it, they do not like the feeling of the trampoline. This is not a measurement of power, but it is an element of the discussion in many cases when a skilled rider (but one who is new to fully suspended bikes) makes comments on this subject. However, the rider is not out of the saddle in some of the circumstances where there seem to be differences in power transmission efficiencies between suspended and rigid bikes. In fact, part of the technique of riding a FS bike is learning to stay in the saddle whenever one can. In these cases the effect Jobst cites is less significant since the rider is supported by the saddle. In my experience, the differences in speed are still there in this case, and these could be discernible while climbing. They are probably undetectable while riding on the flats. While the suspension components still are moving while the rider is seated and they are responding to both pedal forces and to inertial forces, the displacements and velocities are very small. So the losses are also very small. There are a few ways to verify this and it may be the simplest experiment one can do in this investigation (though it isn't that simple). There are losses while climbing (seated) due to the inevitable (unless you have a lot of money!) weight differences between the two types of bikes that are unrelated to the losses in the dampers themselves. Jim asked about the magnitude of the losses Mr. Olsen reported, and I haven't seen the reply. But things like "off the back" and "dropped" may have some reasonable interpretation that indicates the sort of small losses that can be attributed to small differences in vehicle weight are working here. I did a very simple article on vehicle weight and climbing speed for MTB Pro years ago. I was arguing against the purchase of very light parts on a performance basis for recreational riders. However, having been off the back the small amount (recreationally?) I accounted for in the analysis, I can say that these small differences are in line with the some of the differences in performance Mr. Olsen reported. Go here for the story: (keep in mind that it is a simple story written for a popular audience, not scientists) http://www.bontrager.com/TheProf/896/lwparts1.htm If one could build two bikes of equal weight and equal performance in other respects (bearing losses, wheel mass, tire pressure, etc, etc), one could eliminate the losses in seated climbing attributed to the differences in weight. This could easily be done by adding weight to the lighter of two bicycles (one suspended, one not) built with identical components. My guess is that, in seated climbing, the losses due ot the small losses to the damper would be difficult to detect. Summary: The "walking on the trampoline effect may be dominant during large out of the saddle pedaling efforts. Power transmission is likely to be reduced in these cases. It is possible to reduce the losses with adapted pedaling technique in some cases, just as it is possible to learn to walk on a trampoline. The trampoline effect is not present when the rider is seated. In this case the probable explanation for differences in climbing speed may be the additional weight of the fully suspended bike. These losses are also small, but small losses can be "big" enough to matter to a lot of riders. One more note: I believe rear suspension is what this discussion is about. Pedaling losses are not as apparent in bicycles with suspension in the front only. This is not to say that there is not a similar loss, only that it is much smaller and more difficult to detect. Keith Bontrager
Thu, 5 Mar 1998 08:53:26 -0800
Jobst Brandt wrote: >Actually I have a better example. Running across a trampoline is >a good parallel to pedaling a sprung bicycle. My experience with >this is long ago when riding a balloon tired clunk with low tires. >I am sure the major part of this is what is often called eccentric >work, in which muscles re-absorb work as in descending a stairway. That's the same mechanism I've thought could explain poorer performance on an (unsuspended bike) frame that flexes a lot in the bottom bracket when pedaling hard. In the past, you've expressed doubt that bottom bracket flex wasted any significant amount of energy; is it a question of degree (the effect is much larger with suspended bikes), or ...? Tom Ace crux@best.com ###Commenting more on Brandt's note than Tom's: Assigning blame to 'eccentric work' is possibly but not necessarily true. (I happen to think that this issue is one of the great ones, deserving lots of time and thought to reason it out.) My optimistic sense of how humans (and animals, birds etc.) use their muscles is that if there is ANY way to avoid substantial eccentric work, the organism will figure it out. So my bets are laid more on issues of 'muscle control' and on the feelings of a firmly constrained path which allows us to push hard without any surprising jitters or wobbles. This is not an energy issue. When the structure offers a flexibility comparable to our own bones and tendons (magnified by various lever arms) then I expect that power production will go to hell. How the suspension ties into this is a little mystifying. You can set up a bike in a certain gear so that the suspension doesn't 'pogo' due to pedal torque. But it still might from bobbing your upper body or accelerating your legs up and down. Would all of these cases necessarily perturb the pedalling.....? Moderator####
Thu, 5 Mar 1998 12:48:53 -0800
The moderator wrote: >I induced Eric Soroos (then at Cornell, more recently in your neck >of the woods) to perform blind testing with weights on bikes to be ridden >uphill. I think his weight increment was closer to 2 kgf, and only one of >the subjects could detect that 'sometimes'. I keep forgetting the details >though..... maybe they appear in an article in the archive? Anecdotal example only: When riding my commuting bike with and without the water bottle battery, which weighs somewhere over a kilogram, the difference is just plain not noticeable to any significant degree. However, that same kilogram hanging off a the seatpost (as in tools in a bag) or on the handlebars (like an aero bar) are immediately noticeably. This is mostly likely due to the height of the extra weight and its influence on how easily the bike leans and recovers. I would imagine that a small amount of weight fairly evenly distributed, centered fairly low, and not in rotating parts would be very difficult to distinguish. Wayne Lim ####The great thing about 'weight addition tests' is that they're so easy to perform. Both riding speed reduction, and sensitivity to extra weight, are easy to explore for a few kg at least [if anyone cares to do it]. Moderator###
Thu, 5 Mar 1998 16:45:50 +0100
> I think that we could do some order-of-magnitude calcs to see > what kind of power a damper is consuming, but I haven't had > time to do it yet. The losses in the damper can't be great. On my testrig (speed 8 m/s, 20mm bump every 4 meters, input power 2-3 Kw) the shockabsorbers barely get warm, and on an unsuspended bike some of this heat would be generated in the muscles. I'm told studies at Oldenburg University also gave a slight increase in power needed for suspended bikes on an uneven surface, but test subjects would not ride the unsuspended bike at comparable speeds (test were done to validate the model) An impedance mismatch between bike and rider is probably much more important. I seem to remember (but this is from a lecture 20 years ago, and not really my subject) a Harrisons muscle theorem, which stated that muscles need to be pre-stressed for efficient contraction. I think there might be an anology here between suspension interaction and ovalized chainwheels Marten Gerritsen, m.s.gerritsen@sms.utwente.nl
Thu, 5 Mar 98 15:52:44 PST
Keith B. wrote: >There are losses while climbing (seated) due to the inevitable >(unless you have a lot of money!) weight differences between >the two types of bikes that are unrelated to the losses in the >dampers themselves. Jim asked about the magnitude of the >losses Mr. Olsen reported, and I haven't seen the reply. But >things like "off the back" and "dropped" may have some >reasonable interpretation that indicates the sort of small losses >that can be attributed to small differences in vehicle weight >are working here. I didn't reply because what could I say? I am a bit taken aback by Jim's assertion that bike weight makes no difference on a hillclimb. I would buy that a rider would have a hard time noticing such a weight change in a double blind test, but I assert that, by simple Newtonian physics, his performance would change quite noticeably. Since the force that must be overcome to climb a hill is just W*sin(theta), if two riders of equal power, but different rider + bike weight, are climbing a hill at equal power, the difference in speeds is inversely proportional to the difference in weight. A 2-lb. difference in a 200 pound rider/bike combo is a 1% difference in speed. Integrated over the time necessary to climb a 4000' hill, this makes about an aircraft-carrier-length of gap between the two riders, with only a modest weight difference. But I'm sure that Jim would agree with this. I didn't respond with hard data because my examples were anecdotal, and I don't have hard data, as was, I think, clear from my text. John Olsen, who can really feel differences between two bicycles sometimes, and who may NOT be crazy. -------------
Fri, 06 Mar 1998 04:57:41 -0800
jolsen@paccar.com wrote: > > I am a bit taken aback by Jim's > assertion that bike weight makes no difference on a hillclimb. I would buy > that a rider would have a hard time noticing such a weight change in a > double blind test, but I assert that, by simple Newtonian physics, his > performance would change quite noticeably. I too was wondering what Mr. P was thinking on that one. There are a lot of things that are fuzzy and questionable in this debate, but Newton still rules, eh? ####Right, but not always in obvious ways.... (Did he race a FS bike?) If you added a couple of grams to your bike, ostensibly slowing would occur. But in fact power is hard to hold fixed, and rider sensitivity is limited. So nobody really cares. ####But let's say your weight was increased as much as 1%. Then speed on a steep hill would decrease 1%, for perfectly fixed power. I'm not sure that would show up on a speedometer at 8 mph, nor that you would know FOR SURE that your power had been held fixed. So to determine definitely that you were slower would take a variety of repeated runs, with the hope that your fitness change could be factored out! In other words, without simply hefting it or measuring on a scale, I think there's no way to know if your bike is even a couple of pounds overweight. ####I'm thinking that if you're racing at maximum effort, and end up less than 100 yards from the leaders after an hour of climbing, then saving two pounds could put you WITH them, so would be a meaningful investment. But if you're a recreational rider, interested in fitness or relaxation, saving 30 seconds on your way back to the chuck wagon, or saving 1% of your effort, doesn't seem particularly meaningful. Moderator#### I'm off to Portland for a day of schmoozing at the bike show up there so I'll chcak out of this for now. I'll buy the beer if you show up Saturday evening. Monday is reserved for experimenting with the frictional characteristics and dynamics of a small bent piece of fiberglass and metal on fluffy frozen water at Bachelor or thereabouts. Sunday and Tuesday too if El Nino cooperates. Weight seems to be an advantage in this pursuit... Keith Bontrager
Fri, 06 Mar 1998 8:44:06 PST
John Olsen writes: > I am a bit taken aback by Jim's assertion that bike weight makes no > difference on a hillclimb. I would buy that a rider would have a > hard time noticing such a weight change in a double blind test, but > I assert that, by simple Newtonian physics, his performance would > change quite noticeably. Since the force that must be overcome to > climb a hill is just W*sin(theta), if two riders of equal power, but > different rider + bike weight, are climbing a hill at equal power, > the difference in speeds is inversely proportional to the difference > in weight. A 2-lb. difference in a 200 pound rider/bike combo is a > 1% difference in speed. I think you are overlooking the high variability of human performance through psychological influences. Typically a rider given two largely identical bicycles, one painted to look like titanium, the other orange for instance, can easily be convinced the "titanium" bicycle is lighter and faster if he does not suspect he is the subject of an experiment. Under such conditions, better performances have been achieved on the heavier bicycle merely by expectations. This is part of the appeal of racing on a bicycle with "unobtanium" parts. It psyches out the competiton. Jobst Brandt <jbrandt@hpl.hp.com>
Fri, 6 Mar 98 12:00:55 PST
Jobst Brandt writes: >I think you are overlooking the high variability of human performance >through psychological influences. Typically a rider given two largely >identical bicycles, one painted to look like titanium, the other >orange for instance, can easily be convinced the "titanium" bicycle is >lighter and faster if he does not suspect he is the subject of an >experiment. Under such conditions, better performances have been >achieved on the heavier bicycle merely by expectations. Jobst, the effect you describe absolutely exists. However, we are talking about hard and fast Newtonian physics here, and the effect of weight on limited-power climbers is very clear. Do the experiment with identical electric motors, and it would work every time. Do enough trials, and the effect would be borne out statistically even with suggestible humans providing the limited power. I am merely contesting the assertion that weight makes no difference to uphill performance, if, indeed, that was the assertion. John Olsen ###I don't think that's what I said, but I think it's given a boost to the discussion! Moderator#####
Fri, 06 Mar 1998 18:11:23 PST
John Olsen writes:
> Jobst, the effect you describe absolutely exists. However, we are
> talking about hard and fast Newtonian physics here, and the effect
> of weight on limited-power climbers is very clear.
I don't believe Newtonian physics is entering into this matter, hard or soft.
Psychosomatic effects are a serious impediment to assessing the effects of equipment
changes, when human output is the goal.
> Do the experiment with identical electric motors, and it would work
> every time. Do enough trials, and the effect would be borne out
> statistically even with suggestible humans providing the limited
> power.
Therein lies the problem. You cannot get a statistical proof of smallchanges on a
bicycle, and these are small changes. The weight of a bicycle in the scope of
discussion are also in an area where the benefits of the heavier equipment give advantage
elsewhere. Obviously, a suspension bicycle should be worse on a long monotonic
hillclimb, both for its weight and its interference with pedaling.
> I am merely contesting the assertion that weight makes no difference
> to uphill performance, if, indeed, that was the assertion.
I didn't see that anyone claimed "weight makes no difference", only that it
could not be directly measured in most cases because the rider could cover the deficit,
especially if it is small.
Jobst Brandt <jbrandt@hpl.hp.com>
Fri, 06 Mar 1998 9:16:14 PST
Tom Ace writes:
> Jobst Brandt writes:
>> Actually I have a better example. Running across a trampoline is a
>> good parallel to pedaling a sprung bicycle. My experience with
>> this is long ago when riding a balloon tired clunk with low tires.
>> I am sure the major part of this is what is often called eccentric
>> work, in which muscles re-absorb work as in descending a stairway.
> That's the same mechanism I've thought could explain poorer
> performance on an (unsuspended bike) frame that flexes a lot in the
> bottom bracket when pedaling hard. In the past, you've expressed
> doubt that bottom bracket flex wasted any significant amount of
> energy; is it a question of degree (the effect is much larger with
> suspended bikes), or ...?
I think the BB flex, which is out of plane with the motion, causes so little in-line
deflection as to be insignificant. A fairly large lateral swing of the BB is causes
by a tiny displacement in-line with the pedal stroke. I have often cited the use of
Alan aluminum bicycles for hill climbs by professional teams. These same bicycles,
that had standard steel frame tube diameters were never used on the flat or on descents
because they wobbled so badly from BB swing and head tube to rear wheel twist. From
this I think it can reasonably be deduced that BB flex does not affect power transmission
at cadences typical in hill climbs.
Jobst Brandt <jbrandt@hpl.hp.com>
Sun, 8 Mar 1998 02:10:42 -0600 (CST)
Tom Ace wrote:
>
>Jobst Brandt wrote:
>>Actually I have a better example. Running across a trampoline is
>>a good parallel to pedaling a sprung bicycle......
>>I am sure the major part of this is what is often called eccentric
>>work, in which muscles re-absorb work as in descending a stairway.
>
>That's the same mechanism I've thought could explain poorer performance
>on an (unsuspended bike) frame that flexes a lot in the bottom bracket
>when pedaling hard. In the past, you've expressed doubt that bottom
>bracket flex wasted any significant amount of energy; is it a question
>of degree (the effect is much larger with suspended bikes), or ...?
In normal bicycle pedalling (unlike a Soloflex workout station)there is no apparent reason
'eccentric work' MUST be performed. In seated pedalling, eccentric work is
unavoidable only if the pedal load is decreased so fast that elastic spring pushes the
foot UPWARDS on the downstroke. I can't believe this happens, so I place no stock in the
occurrence of negative work in that activity.
Speed is important in this subject. When you walk slowly across a foam pad, you do absorb
(dissipate) energy when the foam does work on your rising foot (ignoring the leg's weight
for now).
When you run, conclusions are harder to draw, since a thrust into the air is an aid to
lifting your leg rapidly. Negative work might be avoided, yet trampoline running
might still be difficult because of other 'force production' issues.
I might mention some highly simplified models of walking and running on flexible surfaces
by Prof. Tom McMahon at Harvard. (Published in Scientific American, and in greater detail
in his book MUSCLES, REFLEXES, AND LOCOMOTION, Princeton Univ. Press, 1984. See Ch. 8,
Mechanics of Locomotion.) They led to the development of a more compliant running
surface which may have helped in setting some world records. (I wasn't totally
convinced by McMahon's reasoning, but his data are very interesting.)
Jim Papadopoulos
Tue, 10 Mar 1998 14:38:31 -0700
This is my attempt to summarize what has been said on the issue surrounding "whether increased weight makes a difference". I'm not attributing comments to particular people, because I might be remembering wrong who said what, and I don't think it's that important anyway. ------------------------- It was suggested that some full-suspension bikes are inefficient, and that it's noticeable in riding (more pedalling effort required to get the same results). Suspension damping was pointed to as a possible culprit, energy being dissipated in the damper. A calculation was put forward that indicated that such losses would be on the order of 1%. The assertion was made that a 1% increase in required power is not noticeable, as has been shown in blind testing where people were unable to correctly identify when a few pounds had been added to their bikes, based on how the bikes felt. This seems to have been misunderstood to mean that a 1% increase in weight does not have an effect, whereas I believe it was intended to show that the very real 1% increase in power is not detectable to the rider. There is some dispute about whether a required 1% increase in power would make a rider 1% slower than some other arbitrary rider, thus causing a noticeable effect if the two were climbing a hill together. Other suggestions were made as to why a full-suspension bike would feel harder to ride. The basic principle is the "trampoline effect", which I don't think anybody has objected to, although there may be disagreements on the subtleties. ------------------------- Comments from me: Observation: two full tall water bottles weight 3.25 pounds. This is an amount of weight that people routinely add to/remove from their bicycles. A 1% increase in total weight corresponds to about a 6% - 9% increase in the weight of the bicycle alone. This might be more noticeable when doing something like bunny-hopping, where the weight being lifted by the feet and hands is increased a lot, rather than hill-climbing, where the weight being lifted is being increased a little. (Somebody else mentioned this before.) Jean-Joseph Cote