1. Pulleys are not generally "positive action". In other words...pulleys slip.
The exception to this rule is sheaves for use with timing belts and chain &
sprocket, but these are special types and I'll assume that that's not what
we're talking about in general. If you don't care about synchronization, slip
can be good. It is a built in torque limiter and overload protection.
2. Gears drive other gears. As a result the two coupled together rotate in
opposite directions. Pulleys drive other pulleys via a belt or rope. As a
result they generally rotate in the same direction.
3. Gears can usually transmit much more power and torque (steel drives steel)
than pulleys, which involve friction, and are much more compact.
4. Gears and pulleys both have a rotational ratio equal to the ratio of their
diameters, but gears also follow the ratio of their numbers of teeth. For this
reason there is no net lost motion and the ratio remains constant and exact
forever since it is the result of the ratio of integers, not real numbers. For
instance, if you drive a 10-tooth gear with a 20-tooth gear and turn it
100,000,000,000 revolutions the driven gear will turn EXACTLY 200,000,000,000
turns, no more no less.
5. Pulleys can exibit the phenomenon of mechanical advantage using the
familiar "moving pulley" arrangement. For example, if you connect a rope to
the ceiling and loop it through a pulley connected to a weighton the floor and
lift the weight with the free end of rope, it will feel as if it weighs half as
much since the pulley is moving upward.
6. Gears must be located at precise center distances from each other so that
teeth mesh properly. Pulleys may be located at varying and long distances from
each other and are used extensively for lifting (like an elevator) and
long-distance mechanical coupling.
7. Pulleys are cheaper and easier to manufacture. You also don't have to worry
about matching pitch, pressure angles, etc.
and to "Don's" list I might add efficiencies - both mechanical and system
common gear boxes - worm, epicyclical, spur, and the like - are usually in the
40-85% efficiency range, depending on load arrangement, load level, and gear
a pully with a sleeve bearing is rated at 98% efficient in a small size wire
rope at greater than the minimum proper "D/d", and belts approach 90% when
and most gears require lubricants and scheduled replacement of the lubricant -
especially after the first 50 hours of run-in.
Is there any elegant way to step up the
speed of a bicycle sprocket to use a belt
to transmit power to the rear wheel?
Or would any additional machinery always
email@example.com (Hobdbcgv) in
The belt *would* decrease efficiency if it were a v-belt. There is
friction into and out of the groove...
That crappy sproket/chain drive is the epitome of efficient pants leg
David A. Smith
The spur gear efficiency is usually rated at 98-99% including bearings
and it is one of the highest for the common machine elements. In gear
boxes, several gears stages are sequentially connected and the
efficiency decreases with the number of stages. A spur gear reducer with
an 85% efficiency should have at least 8 stages and its ratio most
likely will exceed 10000. It doesn't sound like a common case.
That is for one stage only whereas gearboxes with an efficiency of
40-85% have several stages. Usually, the belt drives are at the end of
the efficiency charts.
Maybe you should think about what you hear.
A chain is inherently more efficient because it doesn't have to
_bend_ and hence suffer losses due to hysteresis. A chain link
pivots on a (variably!) lubrucated hinge which requires much less
energy than to flex a belt... cetrainly a tradition V belt.
Also, belts rub on the pulley as they enter and leave the groove;
which is similar to the side-plates of the chain rubbing, but can't
be ameliorated by lubrication with a belt!
A chain is also less-sensitive to variations in tension.
Chains do suffer from some of the same problems as belts such as
sprocket(pulley)-wear, stretch, misalignment axially and radially
which can have significant impact of overall efficiency.
The only improvement in chains is with involute meshing. Such chains
and sprockets are available for industrial applications, but don't
expect to be able to find them in a bicycle shop.
The drive can already be 95% or more efficient.
An Engineered chain drive can be 99% efficient.
To gain consistent efficiency greater than that by any means is an
It is probably more beneficial to work on making the rider's power
input more effective.
Adding more components seldom makes a system more efficient.
Especially when you consider that the 20W to 100W that you gain (at
a peak cyclist output of 2kW) by a more-efficient transmission
system are negated by carrying the extra mass (and hence weight) of
And let's not forget one important aspect; reliability.
/"\ Bernd Felsche - Innovative Reckoning, Perth, Western Australia
\ / ASCII ribbon campaign | I'm a .signature virus!
I've never seen cable belt drives used outside of an engineering laboratory. I
tried on occasion to put Berg components in low- volume products where the high
price was not an issue, but I couldn't tolerate their abysmally long and
uncertain delivery times.
I've often cited Berg's catalog as an example of how to do it right, with
complete specs, dimensions, tolerances, and contact information on every page,
printed in bold black so a tenth- generation copy was legible, printed on
newsprint so they were cheap enough to distribute everywhere, and small enough
to tuck in your back pocket for reading in the head.
They just couldn't deliver the goods when I needed them. I haven't heard from
them in years. I see they're on the Web and eBay now; maybe they've changed.
The words "engage" and "disable" imply the existence of clutches, which would
likely be heavier and bulkier than a derailleur.
The cycling industry, like any other consumer product industry, optimizes
products for their efficiency and velocity in leaving the showroom. Clutches
would have to be hidden from view; derailleurs are crude, but can be made to
look shiny and 'high-tech', whatever that is this week. ;-)
A pulley could act like a clutch by rotating.
In what other ways does a pulley mimic a clutch's behaviors?
Did you mean the slack v-belt drive used in lawnmowers and such?
Fine for v-belts.
I was thinking about cable chains, which have teeth overmolded on a steel
cable. They can run with zero static tension.
The overmolded material is polyurethane. No lubrication is required.
The tension member is stainless steel cable. Doesn't stretch much.
I'm curious how they'd survive here in SoFla. Many polyurethane products
disintegrate in a year or two, even if protected from the harsh sunlight.
Oddly enough, there's a bicycle track not far away. Maybe the color would sell
them for a while. Anybody who would buy a bicycle just for the track is
probably price- insensitive. Might be a niche market there.
I'd expect the cable chains to use just a bit more power than an oiled roller
chain, because the cable must flex as it passes over the sprockets.
Most of the friction in a conventional chain
drive probably comes from the links
rubbing against the pins and the rollers
rubbing against the pins and teeth.
The friction would increase with both
tension and speed so the % power loss
should be constant under all conditions.
For cable bending, the % loss should
decrease at higher speeds.
The issue is, how much energy is wasted
due to polyurethane rubbing against the
firstname.lastname@example.org (Mike Halloran) in
What is the most common failure?
How do they make the loop?
Carbon fiber would stretch even less. Has
anyone tried graphite?
Another question: Stainless steel costs
less than 2 dollars a pound. Why aren't
they making conventional bike chains and
clusters out of SS?
At the current prices of clusters, it would
be cost effective to make them out of a
Ni Cr super alloy, at least the smaller 12,
13 and 14 tooth sprockets which wear the
After flat tires, bad drives are the reason
most bicycles sit unused in garages or
A year or two isn't bad.
They have stop watches. They are going
to know if there's a significant advantage.
Officials have to OK anything new.
Negligible for a 12" front sprocket -- a few
ounces of bending force vs the 200 pounds
tension in track bike chains. The 2" dia
rear sprocket may be less than 1% as
All conservatism is based on censorship of
-- Bret Cahill
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.