Power Transmission by Belt: Part 1

By Staff
1 / 5
2 / 5
Figure 1, basics of power transmission by belt.
3 / 5
Formulas for making belt-power calculations.
4 / 5
5 / 5

Editor’s note: This is the first of a two-part series on belt
power transmission.

At one time, perhaps up through the 1940s, it
was very common for an industrial complex to have one large power
unit to power the whole operation by means of one or more long
lineshafts. A flat belt drive was then used for each individual
machine. Steam engines were commonly used as the main power unit,
but gas engines were also well represented. Diesel engines started
to come on strong in the 1930s and 1940s, but the whole concept
started to lose favor as individual electric motors became smaller
and cheaper and electricity became more readily available.

By the 1950s, it was common for each lathe, milling machine,
sewing machine, etc., to have its own electric motor. Ironically,
each machine still had its own internal belt drive from its

Farm uses for belt power continue to be strong, and today every
engine show and festival features some kind of apparatus that uses
a long, flat belt.

The subject is not particularly technical, but is complicated by
many factors, some of which are the numerous rules, formulas and
superstitions involved in the belt-power world. Our goal is to boil
it all down to a set of simplified rules that always work.

Getting Started

First, the basics: The best way, in my opinion, to understand
the mechanics of power transmission by belt is to study Figure

A weight hangs on the axle of a pulley that is suspended from
the ceiling by a rope or belt, as shown. Each side of the rope
holds one half of the weight. In this case, the initial tension on
each side of the rope is 100 pounds. Now, grasp the crank handles
on each side of the pulley and apply clockwise torque to the pulley
as shown by the arrows. Don’t pull down or lift … just turn! The
effect is to increase the tension T1 and decrease T2. The amount
added to T1 is subtracted from T2. When T2 gets small enough, the
belt slips and the pulley spins. This is a most important

Next, we must discuss how much T2 can be allowed to decrease
before we would expect slipping to begin. For the expected range of
the coefficient of friction between most belt/pulley combinations,
experiments show that T2 should not be allowed to get much below
about one-half of T1 for drives with no automatic tensioner, or
about one-third of T1 for those with such a device. This is also a
key concept.

See the formulas on the chart at right for making all belt-power

Belt Creep

The concept of creep is not generally discussed, and therefore
is largely not understood. The physics is simple: You have a slack
belt running onto a driven pulley and a tight (stretched) belt
running off – or the opposite for a drive pulley. Since all belts
are somewhat elastic, the belt has to change length while it is in
contact with the pulley!

Normally, this stretching or unstretching occurs near the exit
from the pulley, but as load is increased, this point begins to
move around toward the entry side. At the point where the creep is
starting to occur very near the entry side, slip is about to start.
Creep results in about a 1 percent loss of overall speed. This is
not slip!

How Crown Works

Refer to Figure 2, which illustrates why a belt will climb to
the largest diameter on a pulley. This is often incorrectly
attributed to centrifugal force. The main force at work here
results from the fact that the initial point of contact of the belt
on the pulley is on a larger diameter than where it settles as it
continues around, thus the belt is continually being led

Figure 3 shows a belt running on a pulley with an exaggerated
crown shown only for illustration. Each side of the belt tries to
climb the slope with the result that the belt runs in the center.
This much crown would quickly cause internal separation in the

An example of a crowned surface that causes, rather than
corrects, problems is that of ridged roadways where car tires try
to climb the slight slopes. The road is like the belt and the tires
are pulleys!

18 Considerations for a Successful Flat Belt Drive

Here are the key things we need to know before we can design a
successful flat belt drive for power transmission. Note that the
following is simplified and compromises have been made. However,
the values are conservative:

1. Belt speed should be in the 3,000-to-4,000-foot-per-minute
(FPM) range. Speeds up to 6,000 FPM are sometimes used, but are not
recommended. Stay below 5,000 FPM if any pulleys are cast iron.
(Belt speed in FPM = pulley RPM x pulley diameter in feet x

2. Plan on about 3 HP-per-inch of belt width as a convenient
horsepower for a single-ply leather belt at speeds in the
3,000-to-4,000 FPM neighborhood. Reduce this proportionately if
belt speed is lower than this. For example, for one-half the belt
speed, figure on one-half the HP, and further reduce the figure by
half if either pulley is 4 inches or less in diameter. But, on the
other hand, if both pulleys are large (8 inches or more) and belt
speed is up in the 4,000 FPM range, we can figure on 5

3. Minimum pulley diameter is 3-1/2 inches (absolutely never
less than 3 inches). For very low power, there is no strict limit,
e.g., sewing machines, governors, etc.

4. Pulley width should be at least 1-inch wider than the belt, 2
inches is better, more if possible for large-diameter pulleys.

5. Crowned pulleys are desirable; in fact, virtually necessary.
The amount of crown varies so widely among manufacturers and is so
poorly covered in the literature that it is difficult to give a
good rule. For leather belts, about 1/16-inch for small pulleys and
1/4-inch for large will usually suffice. This is measured as
difference in diameter, not radius. More is necessary for low
speed, less for low-stretch belts.

6. Ideal belt length puts the drive and driven shafts about
20-to-25 feet apart.

7. The tension, or pull side, of the belt should always be on
the bottom. Vertical belt drives are less desirable, but are okay
if belt tension is sufficient.

8. Always run the grain side (smooth side) next to the pulley.
Useful power transmission can be as much as 30 percent greater this

9. The belt should be kept clean and free from water and oil.
Wipe off any oil immediately. Leather belts can be cleaned with
naptha or gasoline.

10. Only recommended belt dressings should be used sparingly
(once every few months). Neatsfoot oil compound is often
recommended. Resin-type dressings that harden cause severe internal
belt wear. Some oily dressings may cause loss of gripping ability
and/or stretching. Oil-type dressings are never used on
rubber-based belts.

11. Belt creep is responsible for about a 1 percent decrease in
load pulley RPM. Remember, creep is not slip.

12. Belt joining must be done so that belt sides are in line. If
an endless belt can be used in a particular situation the belt can
be cemented. This is the best method of joining. The belt must be
carefully tapered with a draw knife on each end over at least 4 or
5 inches of length. Liquid hide glue (or that recommended by the
belt supplier) is used. Glued joints must run in such a direction
that any external belt contact will not tend to open the feathered

13. Initial belt tension should be about 40 pounds per inch of
width. This means that the bearing force, the force separating the
pulleys, will be about 80 pounds per inch of width. Proper belt
tension requires good shaft bearings. For less strenuous use, a
good rule is to keep belt tensions high enough so that slip does
not occur. Slip causes heating and glazing of belts, thus
shortening belt and pulley life. Note: Centrifugal force reduces
belt-to-pulley contact force, especially on small pulleys.

14. Do not force belts onto pulleys. This can stretch a belt
unevenly and put a permanent weave in the belt. Such belts will be
unstable and run from side to side.

15. Center-cut belts, those cut along the spine area of the
hide, are best. They are stronger than side leather and run truer
because any differences in stretch characteristics on one side are
matched on the other side.

16. Belt flop is a potentially hazardous condition that is
usually caused by torque pulsation of driver or driven pulley or by
shifting of driver or load. Sometimes the length of the belt,
torque pulsations, vibrations, etc., all combine to produce a
particularly unstable condition. Sometimes an idler will help.
Frequently, the best fix is to cross the belt. This of course
produces a reversal of rotational direction, which can be dealt
with by “moving things around” or maybe reversing the engine’s
rotation (if possible). Crossing belts causes significant belt
wear, especially if the belt is not very long.

17. Good quality three- or four-ply cloth/rubber belts are
roughly equivalent to single-ply leather belts for a given

18. Often it is desirable to use an idler on the slack side of
the belt. This is always applied nearer the driven pulley, perhaps
3 to 5 feet away. This tends to improve the contact angle (i.e.
more wrap), and adds slack-side tension when heavy loading would
otherwise tend to reduce it to the point of slipping. Idlers should
not be crowned.

Special thanks to George B. Loughery and the Hay Creek Valley
Historical Assn. for permission to publish this article, which
originally appeared in the Association’s Summer 1994 issue of The

Contact George Loughery at: 3219 Westview Drive, Reading, PA
19605; (610) 929-3794; g19605@hotmail.com

Contact the Hay Creek Valley Historical Assn. at:?P.O. Box 36,
1250 Furnace Road, Geigertown, PA 19523; (610) 286-0388;

Gas Engine Magazine
Gas Engine Magazine
Preserving the History of Internal Combustion Engines