some good waterproof dressing. It is quite likely that to buy a belt
that has been filled means to buy one that perhaps has some bad leather
in it that would be seen in a dry oak tan belt, and also that the
adhesive power of the filled belt is not quite equal to the dry one;
but the points that the filled one possesses over the one not filled
are, first and mainly, "it is filled when you buy it with a preparation
that does not injure the leather in the least," and the preparation
you will fill it with, for it will be filled, will be engine oil and
water, a combination that will ruin any belt made and also get it in
six months into a condition that will make a permanent repair with glue
impossible, for machine oil and moisture are strangers to glue and will
ever be. More good belts are ruined by being soaked with engine oil
until the points come loose and then pulled out of shape than from any
other cause. Of course you may be able to keep a main engine belt that
runs through a damp wheel pit and basement, and through a long damp
tunnel to a main driven pulley that has two big boxes that are just as
close to the pulley as a first-class machine designer could put them,
and never get a drop of oil or water on it. But this is not likely.
One very common cause of trouble with engine belts is the fact that
such belts usually run under the floor, where there is considerable
moisture. Then the oil table under the average large Corliss engine
will leak around dash-pots and rocker-arm shafts, and some oil will
fly from the eccentric oil cups, get into the wheel, run around the
rim and get to the belt; if the belt is not filled a very few drops of
oil will make a large spot on it. Then, if an engine does not run the
whole twenty-four hours, while it is off, watch. A few drops of water
from a leaky valve stem whose bonnet drain is stopped up, as it will
sometimes be, has a way of getting through the floor and falling on to
the belt and running down the inclined inside of it until it finally
comes to the flywheel, which, with the assistance of its crowning face,
very kindly makes a nice pocket for said water and proceeds to drink
it up. Result: the glue is loosened and the belt may come apart in
consequence. Should there chance to be a point just at the bottom of
this pocket, it will get the glue soft enough to slip but may not open
up, which is much worse than if it did open up; for it may slip away
from the shoulder of the splice for half an inch, and when the engine
is put to work it may close down by running under the wheel and stick.
If it does, the result is that at no very distant day you will find a
break at that particular place, right across the face of the belt. The
reason is that the load was all taken off the inside half of the belt
by point slipping, thereby making the inside of the belt too long and
putting all the load on the outside. The outside will continue to do
all the work until it stretches enough to bring the inside back into
service again. During this week or month you have been pulling your
load with a single belt, not a double one, and after a short time
you will find the break referred to above in the shape of a clean,
well-defined crack extending across the belt parallel with the points
of the laps. Now of course you are going to send for the man who sold
you the belt and ask him to fix it. If he is a wise man and understands
his business, he won't do a thing but show you right under that crack
a point that does not come up to where it should come. Then the thing
for you to do is to say to him that the belt is examined every time it
is put into service and that you have noticed that the points he refers
to all come loose during a "run," that any one knows that a few drops
of water would not take any belt to pieces while it was running, and if
it was water, why did it not take it apart everywhere, etc? And finally
crush him completely by telling him that your men have no time to put a
pair of clamps on a belt in order to pull back into its proper position
every point that comes loose; that if they did do it they would have no
time for anything else, especially in the present case, and that if his
people had made the belt right the glue would have held, anyway.
After he has given you a new belt or repaired your old one, just take
my advice and box that flywheel up above the top of the eccentric oil
cup, at least 12 inches, and get some good, heavy tin or zinc and put a
tight roof over the belt, under the floor.
First put in a ridge pole out of 1-1/2-inch pipe, starting at the face
of the wheel and running in the direction of the main driven pulley,
holding it firmly in place at each end with a strong iron clamp. Then
solder into each edge of the strip of tin, which should be long enough
to reach beyond any possible leak through the floor or oil table, a
piece of 1/2-inch pipe, and put the tin over the ridge pole with a piece
of small pipe on either side. Ordinarily the belt goes out past the
cylinder; if it runs through a bricked-up runway on its route to the
main driven pulley, just fasten the two pieces of 1/2-inch pipe to
either wall and have the ridge about 6 inches higher than the outside
ones. Then every drop of oil or water that comes through the floor will
fall on to the roof and run down to the walls and be carried down to
the floor of the pit and have no chance to touch the belt.
One of the most difficult things in the operation of large stations
where a large number of belts are used is to keep them thoroughly
clean and free from moisture and machine oil, the latter especially.
One very hard problem that confronts all designers of machinery is the
prevention of oil leakage from boxes. In several plants with as many as
six dynamos of the same kind and the same design, at least four of the
six have leaked oil every time they were run. The others did not leak
as a usual thing, and all were equipped with the most modern methods of
holding oil.
[Illustration: FIG. 80.]
Now we come to the building of the belt, and we will notice only such
points as interest the engineer or buyer. The first thing is to see
that the laps are of uniform thickness, so that the belt will run
quietly; and it should be absolutely straight when unrolled on the
floor. If it has a long, graceful curve in it, look out; for it will
not run straight on the pulleys until it has stretched straight, and
by that time one of its edges may be ruined by coming in contact with
the floor or some other obstacle. Next notice how long the leather is
from which it is made. It should not show more than 52 inches, and
then there will be 4 inches hidden by the point that is out of sight.
Then see that the joints are broken properly. For instance, find the
center of any piece of leather on one side of the belt, and then look
on the opposite side and see if the joint is right under your center
mark. It should be by all means, for right here lies the most important
thing about the construction of leather belts. A belt whose laps are
all the same length, and which has all its joints broken correctly,
will put the same load on the glue throughout, and that is what must
be done in order to get the best results. See Fig. 80. Here we have a
belt that is 36 inches in width and a double ply. Now suppose there
is a draft of 9360 pounds on this belt, that from point _A_ to point
_B_ is 26 inches, and that the points are 4 inches long. Now we have
26 inches plus 4 inches plus 4 inches times 36 inches for the number
of square inches in the glued joint. This equals 1224 square inches;
the total pull on the belt divided by 1224 will equal the load on each
square inch of glued joint, and will equal in this case 7.65 pounds.
Now instead of assuming distance _A--B_ in Fig. 80 to be 26 inches, let
the lower joint get out of step with the upper ones, and conditions get
vastly different. We will suppose that the dimensions are as given in
Fig. 81, as was the case with a new belt that was measured less than
one month before the observation was made and we have the following:
Joint _A B_ is now only 10 inches, and we have 10 inches plus 4 inches
plus 4 inches times 36 inches which equals 648 square inches, and the
lead on the joint is now 14.44 pounds. You will readily perceive what
an important part in the life of the belt, and the life of everything
around the belt as far as that goes, the proper breaking of the upper
and lower joints is. Of course the belt maker will tell you that his
glue is just as strong as the leather itself, and he is about right
as long as you keep the belt free from oil and water; but when the
belt becomes filled with oil the glue rots and loses its strength much
faster than does the leather.
[Illustration: FIG. 81.]
No good belt needs any posts along the sides to make it run straight
and stay on the pulleys. If the pulleys are in line and the belt
straight, it will run straight. All belts should be made to run
perfectly straight on pulleys, first on account of the local
advertisement that it gives to the man who has charge of them; second,
if they do not run true, they will be on the floor or wrapped around
the shaft in a very few minutes, should they ever slip. Another very
important thing in the care of belts that carry heavy loads is that if
any of the points do come loose so far back that they will not return
to place without putting on the clamps, put them on by all means; as
the restoring of this point to place means that you will still retain
in service all of your belt, as you will not do if you glue it down
where it is and thereby cut one side completely out of service.
HOW TO CLEAN BELTING
We submit the following as the best and proper way of cleaning a
leather belt. It may seem simple, but it is safe and effective, as has
been proved by many people who have thus restored old and dirty belting
which had become almost or quite unfit for use.
Coil the belt loosely and place it on edge in a tank in which it may
be covered with naphtha; a half barrel makes a good receptacle, but
something with a tight cover would save the loss by evaporation. Put
in enough naphtha to cover the belt completely and allow it to remain
for ten or twelve hours; then turn the belt over, standing it upon the
other edge. The vertical position of the belt surfaces allows the dirt
to settle to the bottom of the receptacle as it is washed out, and
permits naphtha to get at all the parts.
After the belt has remained in the naphtha another ten or twelve hours,
or until sufficiently clean, raise it and allow the naphtha to drip
back into the tank. Then lay the belt flat, stretching or shaking it
until almost dry. You will find that the naphtha will not affect the
leather nor the cement in the center of the belt, but may open the
joints at the edges; in which case the old cement should be scraped off
and the edges recemented. Your belt man will know how to do this. The
belt will now be somewhat hard, and should be treated with a reliable
belt dressing before being replaced on the pulleys.
X
BELTING, ITS USE AND ABUSE[7]
THERE is no class of appliances so little understood by the ordinary
steam engineer and steam user as belts, which may be seen by the
quantity of belting sold annually. Where one can point to a belt that
has been in continuous use for twenty years, you can find hundreds that
do not last one-fourth as long. Why? Not always because the buyer has
tried to get something for nothing, but as a rule, when they do, they
get nothing for something.
[7] Contributed to Power by Wm. H. McBarnes.
The average belt is a poor one, and the average buyer will not find it
out till he has used it for some time. If you weigh the belt dealer up
as a man who is trying to rob you, beat him down in price, then get him
to give from 5 to 40 per cent. off, he will enter a protest, and, after
some explanation, will come to some terms with you. Have you gained
anything by your cleverness? Well, hardly. Belt dealers and makers,
like almost all other dealers in supplies, aim to get nothing but
first-class goods; but second and third, and even fourth-class goods,
are made, and you get the quality you pay for. In the second place,
belts wear out quickly because they do not get proper care. To let a
belt run one moment after it gets too slack is bad practice, for it is
apt to slip and burn all the staying qualities out of it. Another good
reason why it should not be run slack is that the engineer or belt man,
to save work, would be tempted to put on a dressing or, worse yet, put
on resin to make it pull, and, in the language of Rex, "the man who
will put resin on his belts is either a fool or a knave," for it is
sure to spoil his belt if continued for any length of time.
In an emergency, as when some unforeseen substance has found its way
to the belt, it may be necessary, to keep from shutting down between
hours, to use some of the so-called dressing. We know from experience
that engineers will go to almost any extreme to get out of a tight
place--circumstances sometimes make it necessary to keep a belt running
when it should not--but this should not be allowed to any extent. To
allow a belt to run too tight is just as bad, for it will make short
life for the belt, hot boxes and scored shafting. There is not one
in twenty who takes the time or can splice a belt properly; it is
generally done in a hurry, any way to make it hold together, with the
understanding that it cannot talk; but it does. How often we see boards
nailed up or rims tacked on to keep belts from getting off the pulleys.
All of this is good for the belt dealers.
This is not all the fault of the engineer or the belt manufacturer.
Often belts are made uneven, and soon get out of shape, even with the
best of care. We sometimes find a belt that ordinarily runs easy on
the pulleys and does its work with ease suddenly inclined to run to
either one side or the other of the driven pulley. This is caused by
one of two things--either the belt has been too slack, or the load
increased for want of lubrication, or other causes. In either case it
will run off if you insist on applying the power. The remedy would be
to take up the belt, thoroughly oil the journals, or take off the extra
load--maybe a combination of all. Still a little extra work making the
belt tighter will enable it to run well and do the extra work just
as long as the extra tension can be maintained. Then it may appear
perplexing and run to one side of the driven pulley when the driven
shaft gets out of line with the driving shaft. In a case of this kind
the belt does not run to what is called the high side of the pulley,
but to the low side. Another peculiar indication: If two shafts are
parallel and there is a high place on the pulley, then a belt will run
to the high place; but if the shafts are out of line, or, in other
words, are not parallel, and the face of the pulley straight, then the
belt will run to the low side or that closest to the driving shaft. The
remedy would be to line up your shafting.
The object of this chapter is not to say how belts are made, but to
impress upon the minds of belt users that to get the best results,
belts, like all good servants, must be well cared for, and all
responsibility should rest with one man, just as with your engine or
any high-priced machine.
XI
A COMPARATIVE TEST OF FOUR BELT DRESSINGS[8]
DURING January, 1905, a comparative test of the working efficiency
of four belt dressings and preservatives was made by T. Farmer, Jr.,
and the writer. The test was made on the regular belt-testing machine
of Sibley College, Cornell University, a full description of which
appeared on pages 705-707 of Vol. 12, _Trans. A. S. M. E._ This machine
tests the belt under actual running conditions, though our belts were
in somewhat better than average condition. The four belts were new
4-inch Alexander No. 1 oak-tanned single-ply, and were 30 feet long.
Particular care was taken to keep them free from oil and dirt. The
belts were first tested as received from the manufacturer, after which
each belt was treated with one of the dressings and again tested.
[8] Contributed to Power by William Evans.
The dressings were two semi-solids, designated No. 1 and No. 2; a bar,
No. 3, and neatsfoot oil, No. 4. As the first three are proprietary
articles, it was not thought best to give their names, though any one
familiar with the actions of belt dressings will readily recognize No.
1 from its peculiar curve. In applying the dressings, we followed
directions carefully, and in the case of Nos. 2 and 3 exceeded
them. The belt was given a five-hour run, during which two or three
applications of the dressing were given, and then it was set aside in
a warm place to allow it to absorb the applied dressing. After thus
"soaking" for at least forty-eight hours, the belt was again run, this
time for three hours, with one more application of the dressing. As
No. 3 was a bar of sticky dressing, it will readily be seen that this
precaution was not really necessary. No. 4, the neatsfoot oil, was not
applied during the last run, as we were afraid of getting too much
oil in the belt. As this oil is so extensively used by engineers for
dressing belts, special care was taken to get the best possible results
with it.
[Illustration: FIG. 82.]
In Fig. 82, the result of the test with the neatsfoot oil is shown
graphically. This curve is platted to show the relation between initial
tension per inch of width and horse-power per inch of width. One reason
for the drop in horse-power in the treated belt is that the slip was
materially increased; in the lowest tension at which any power at all
was transmitted, about 15 pounds per inch of width, the slip ran up as
high as 25 per cent.
[Illustration: FIG. 83.]
In Fig. 83, which shows the comparative value of the four dressings,
the highest horse-power delivered to the belt was taken as the
standard. The horse-power delivered by the belt was divided by this
standard, and the result, expressed in percentage, was used as the
percentage of available horse-power transmitted. This comparison shows
the great superiority of dressing No. 1 at all times, and especially at
low tensions. In looking at this chart, it is well to remember that No.
3 is a sticky dressing.
As the time of the test was so short, we were unable to determine the
ultimate effect of the dressings on the leather of the belts. We could
only approximate this by a chemical test and a close examination of the
belts at the end of each test. The chemical analysis showed no ammonia
or rosin in any of the dressings; No. 2 had a trace of mineral acid,
and all had oleic acid as follows: No. 1, 0.27 per cent; No. 2, 29.85
per cent; No. 3, 3.5 per cent; No. 4, 0.7 per cent.
The practical test showed no ill effects except from No. 3, the sticky
dressing, which ripped and tore the surface of the belt. The high
initial tensions caused overheating of the journals, even though we
kept them flooded with oil. On the low initial tensions there was
no tendency to heat, even when the maximum horse-power was being
transmitted by dressing No. 1. In the latter case we oiled the bearings
once in every two or three runs (a "run" comprised all the readings
for one initial tension), while in the former we oiled the bearings
after each reading and sometimes between them; even then we were afraid
that the babbitt would get hot enough to run. The readings for each
run varied in number from two to a dozen, but only the one giving the
maximum horse-power was used in drawing the curves. The belt speeds
during the tests varied between 2000 and 2500 feet per minute, most of
the tests being made at about 2200 feet per minute.
XII
BELT CREEP
THE question of the minimum amount of slip of a belt in transmitting
power from one pulley to another reduces itself to a question of creep,
for it is possible to have belts large enough so that with proper
tensions there will be no regular slip. With a difference in tension on
the two sides and of elasticity in the belt, creep, however, is bound
to take place. What does it amount to and what allowance should be made
for it? asks Prof. Wm. W. Bird of the Worcester Polytechnic Institute
in his paper under the above title.
[Illustration: FIG. 84.]
In Fig. 84 let _A_ be the driver and _B_ the driven, _T__{1} the
tension in the tight side of the belt and _T__{2} in the slack side,
the pulleys and belt running in the direction indicated. One inch of
slack belt goes on to the pulley _B_ at _o_; at or before the point
_p_ it feels the effect of increased tension and stretches to 1 + _s_
inches. It now travels from _p_ to _m_ and goes on to pulley _A_
while stretched. At or before reaching the point _n_, as the tension
decreases, it contracts to one inch and so completes the cycle.
With a light load the belt creeps ahead of the pulley _B_ at or near
the point _p_. If the load is heavy, the creep works towards the point
_o_ and the belt may slip; this also takes place when the belt tensions
are too light even with small loads.
The point may be easily appreciated by imagining the belt to be of
elastic rubber. Professor Bird gives formulas for calculating the
creep, and tests made at the Polytechnic to determine the modulus of
elasticity. He concludes that the answer to his opening question is
that for the common leather belt running under ordinary conditions the
creep should not exceed one per cent. While this is sometimes called
legitimate slip, it is an actual loss of power and cannot be avoided by
belt tighteners or patent pulley coverings.
The smooth or finished side should go next to the pulley because
the actual area of contact is greater than when the rough side is
in contact; consequently, the adhesion due to friction is greater.
Moreover, the smooth side has less tensile strength than the rough
side, so that any wear on that side will weaken the belt less than wear
on the other side would.
XIII
ROPE DRIVES[9]
THERE seems to be considerable difference in opinion regarding the
various ways of applying rope to the sheaves in rope driving, viz.,
multiple- or separate-rope system, continuous-wrap or single-rope
system with the rope from one of the grooves running on a traveling
take-up device, continuous-wrap or single-rope system with the take-up
working directly on all the wraps.
[9] Contributed to Power by R. Hoyt.
[Illustration: FIG. 85.]
The multiple- or separate-rope system on a horizontal drive where the
distance between centers is great enough so that the weight of the
rope will give the required tension, having the tight or pulling part
on the lower side and the sheaves of the same diameter, as in Fig. 85,
should be very satisfactory, as old or worn ropes may be replaced by
new ones of larger diameter, or some of the ropes may be tighter than
others and still not alter the efficiency of the drive. It will be
noticed in this case that a larger rope does not alter the proportional
pitch diameters of the rope on the driving and driven sheaves; but if
one of the sheaves is larger than the other, as in Figs. 86 and 87, and
a new or larger rope is substituted for a worn or smaller one, or if
some of the ropes are a great deal tighter than others, a differential
action will be produced on the ropes owing to the fact that the larger
or slack rope will not go as deeply into its grooves as the smaller or
tight one. Consequently the proportionate pitch diameter on the rope
on the driver and driven sheave will be changed. The action will depend
upon whether the large or the small sheave is the driver. If the driver
is the larger, and of course assuming that the slack or large rope is
weaker than the combined tight or smaller ones, then it will have less
strain on the pulling side; but if the driver is smaller, then the new
or large rope will have greater strain on the pulling side. Whether the
driver is larger or smaller, a large or slack rope affects the action
oppositely to a small or tight rope. Fig. 87 shows how the action is
reversed from Fig. 86.
[Illustration: FIG. 86.]
[Illustration: FIG. 87.]
For clearness we will exaggerate the differences in diameter in the
sketches and figure the speeds that the different size ropes would
produce. We will take _A_ as normal, _B_ 1 inch farther out of the
groove, producing a difference in diameter of 2 inches; _C_ 1 inch
deeper in the groove, producing a difference in diameter of 2 inches.
In Fig. 85 assume for the normal diameter of driver and driven 40
inches, 42 inches for _B_ and 38 inches for _C_, with a speed of 200
revolutions per minute for the driver. Either _A_, _B_ or _C_ will give
200 revolutions per minute for the driven sheave, omitting slippage,
of course. In Fig. 86 say the normal diameter of the driver for rope
_A_ is 60 inches and of the driven 30 inches, a speed of the driver of
200 revolutions per minute will give the driven sheave a speed of 400
revolutions per minute; _B_, with the driver 62 inches and the driven
sheave 32 inches diameter, will give the latter a velocity of 387-1/2
revolutions per minute. With _C_ the driver is 58 inches, the driven 28
inches, and the speed given the latter 414-2/7 revolutions per minute.
In Fig. 87, the normal diameter of the driving sheave being 30 inches
and the driven 60 inches, a speed of the driver of 200 revolutions per
minute will give a speed of the driven member of 100 revolutions per
minute. With _B_, if the driver is 32 and the driven 62 inches, the
driven sheave will have a speed of 103-7/31 revolutions per minute;
_C_, with the driver 28 inches and the driven sheave 58 inches, will
give the latter a speed of 96-16/29 revolutions per minute. So it will
be readily seen what effect a large or a small rope would have.
[Illustration: FIG. 88.]