_THE SHIFTING LINK._
EARLY REVERSING MOTIONS.
In the engineering practice of the world, before the locomotive
and marine engines came into use, there was no need for devices to
make engines rotate in more than one direction. When the need for a
reversible engine first arose, it was met by very crude appliances.
Locomotives were kept at work, earning money for their owners, which
were reversed by the man in charge stopping the engine, and by means
of a wrench changing the position of the eccentric by hand. A decided
improvement on the wrench was the movable eccentric, which was held in
forward or back gear by stops; the operation of reversing being done by
a treadle or other attachment located near the engineer’s position. A
serious objection to this form of reversing gear was, that the abrasion
of work enlarged the slot ends, and wore out the stops, leading to
inaccuracy and frequent breakage. A somewhat better form of reversing
motion was a fixed eccentric, with the means at the end of the
eccentric-rod for engaging with the top or bottom of a rocker-shaft,
which operated the valve-stem. This was the form of reversing motion
used on the early Baldwin engines. Numerous other appliances, more
or less defective, were experimented with before the double fixed
eccentrics were introduced. Till the link was applied to valve-motion,
the double eccentrics--an American invention--were the most important
improvement that had been made on the locomotive valve-motion since
the incipiency of the engine. The V hook, in connection with the
double eccentrics, made a fair reversing motion in comparison to any
thing that had preceded it. The objection to the hook was, that, when
the necessity arose for reversing the engine while in motion, much
difficulty was experienced in getting the hook to catch the pin. As
a simple, prompt, and certain reversing motion, the link was readily
acknowledged to be far superior to any thing that had previously been
tried.
INVENTION OF THE LINK.
There is no doubt but the link was first applied to a steam engine
by William T. James of New York, a most ingenious mechanic, who also
invented the double eccentrics. James experimented a great deal
about the period from 1830 to 1840, with steam carriages for common
roads; and it was in this connection that he invented the link. His
work having proved a commercial failure, the improvements on the
valve-motion were not recognized at the time; although the probability
is, that Long, who started the Norris Locomotive Works of Philadelphia,
and brought out the double eccentrics upon the locomotives built there,
was indebted to James for the idea of a separate eccentric for each
direction of engine movement.
The credit of inventing the ordinary shifting link is due to William
Howe of Newcastle, England. This inventor was a pattern-maker in the
works of Robert Stephenson & Co., and he invented the link in 1842
in practically its present form. The idea of Howe was to get out an
improved reversing motion; and he made a pencil-sketch of the link, to
explain his views to his employers. The superintendent of the works was
favorably disposed to the invention, and ordered Howe to make a pattern
of the motion, which was done; and this was submitted to Stephenson,
who approved of the link, and directed that one should be tried on a
locomotive. Although Stephenson gave Howe the means of applying his
invention, he does not seem to have perceived its actual value, for the
link was not patented; and Stephenson never failed to patent any device
which he thought worth protecting.
The link-motion was applied to a locomotive constructed for the Midland
Railway Company, and proved a success from the day it was put on.
Seeing how satisfactorily the invention worked, Robert Stephenson paid
Howe twenty guineas (one hundred and five dollars) for the device, and
adopted the link as the valve-gear for his locomotives. This is how the
shifting link comes to be called the “Stephenson link,” and the credit
for this invention was not extravagantly paid for.
The capability which the link possesses of varying the steam admission
and release, did not appear to be understood by the inventor; nor was
the mechanical world aware, for some time after the link was brought
into use, that it could be employed to adjust the inequality of steam
distribution, due to the angularity of the connecting rod.
CONSTRUCTION OF THE SHIFTING LINK.
[Illustration: FIG. 12.]
As usually constructed for American locomotives, the link is a slotted
block curved to the arc of a circle, with a radius about equal to the
distance between the center of the driving-axle and the center of the
rocker-pin. The general plan of the link-motion is shown in Fig. 12.
Fitted to slide in the link-slot is the block which encircles the
rocker-pin. The eccentric-rods are pinned to the back of the link;
the forward eccentric-rod connecting with the top, and the back-up
eccentric-rod with the bottom, of the link. Bolted to the side and near
the middle of the link is the saddle, which holds the stud to which the
hanger is attached; this, in its turn, connecting with the lifting arm,
which is operated by the reversing rod that enables the engineer to
place the link in any desired position.
ACTION OF THE LINK.
Regarded in its simplest form, the action of the link in full gear
is the same upon the valve movement as a single eccentric. When the
motion is working, as in the figure, with the eccentric-rod pin in
line with the rocker-pin, it will be perceived that the movement can
not differ much from what it would be were the eccentric-rod attached
to the rocker. Here the forward eccentric appears as controlling the
movement of the valve. Putting the link in back motion brings the end
of the backing eccentric-rod opposite the rocker-pin, the effect being
that the back-up eccentric then operates the valve. When the link-block
is shifted toward the center of the link, the horizontal travel of
the rocker-pin is decreased; consequently, the travel of the valve is
reduced; for, with ordinary engines, the travel of the valve in full
gear equals the throw of the eccentrics, the top and bottom rocker-arm
being of the same length. The motion transmitted from the eccentrics,
and their means of connection with the link, make the latter swing as
if it were pivoted on a center which had a horizontal movement equal to
the lap and lead of the valve. The extremities of the link, or rather
the points opposite the eccentric-rods, swing a distance equal to the
full throw of the eccentric. The variation of valve-travel that can be
effected by the link, is from that of the eccentric throw in full gear
down to a distance in mid gear which agrees with the extent of lap and
lead. The method of obtaining these various degrees of travel is by
moving the link so that the block which encircles the rocker-pin shall
approach the middle of the link.
When an engine is run with the lever in the center notch, the supply
of steam is admitted by the lead-opening alone. In full gear the
eccentric, whose rod-end is in line with the rocker-pin, exerts almost
exclusive control over the valve movement; but, as the link-block gets
hooked towards the center, it comes to some extent under the influence
of both eccentrics.
A thoughtful examination of Fig. 12 will throw light on the reason
why the proper position of a slipped eccentric can be determined by
the other eccentric when the engine is on the center. In the cut, the
crank-pin is represented on the forward center; and in that position
the eccentric centers are both an equal distance in advance of the
main shaft center. It will be evident now that the valve must occupy
practically the same position for forward or back gear, as each of the
eccentric-rods reaches the same distance forward. Putting the motion in
back gear would bring the backup eccentric-rod pin to the position now
occupied by the pin belonging to the forward eccentric-rod.
VALVE-MOTION OF A FAST PASSENGER LOCOMOTIVE.
Reducing the travel of the valve by drawing the reverse-lever towards
the center of the quadrant, and consequently the link-block towards the
middle of the link-slot, not only hastens the steam cut-off, but it
accelerates in a like degree every other event of steam distribution
throughout the stroke. To explain this point, let us take the motion
of a well-designed engine in actual service, which has done good
economical work on fast train running. The valve-travel is five inches,
lap one inch, no inside lap, lead in full gear 1/16 inch, point of
suspension 9/16 inch back of center of link.
EFFECT OF CHANGING VALVE-TRAVEL.
When this engine is working in full gear, the steam will be freely
admitted behind the piston till about eighteen inches of the stroke,
when cut-off takes place; and the release or exhaust opening will
begin at about twenty-two inches of the stroke, giving four inches for
expansion of steam. Now, if the links of this engine are hooked up so
that the cut-off takes place at six inches of the stroke, the steam
will be released at sixteen inches of the stroke; and at that point
compression will begin at the other end of the cylinder.
WEAK POINTS OF THE LINK-MOTION.
This attribute which the link-motion possesses, of accelerating the
release and compression along with the cut-off, is very detrimental
to the economical operating of locomotives that run slow. High-speed
engines need the pre-release to give time for the escape of the steam
before the beginning of the return stroke; and the compression is
economically utilized in receiving the heavy blow from the fast-moving,
reciprocating parts, whose direction of motion has to be suddenly
changed at the end of each stroke, and in helping to raise the pressure
promptly in the cylinder at the beginning of the stroke. A locomotive,
on the other hand, that does most of its work with a low-piston speed,
would not suffer from back pressure if the steam were permitted to
follow the piston close to the end of the stroke; and a very short
period of compression would suffice. If the engine, whose motion we
have been considering, instead of releasing at sixteen inches, could
allow the steam to follow the piston to twenty-two inches of the
stroke, after cutting off at six inches, a very substantial gain of
power would ensue. And this would be well supplemented by avoiding
loss of power, did compression not begin till within two inches of the
return stroke.
WHY DECREASING THE VALVE-TRAVEL INCREASES THE PERIOD OF EXPANSION.
Increase of expansion follows reduced valve-travel, from a similar
cause to that which produces expansion when lap is added to the edge of
a slide-valve. When the valve is made with the face merely long enough
to cover the steam-ports, there can be no expansion of the steam; for,
so soon as the valve ceases to admit steam, it opens the steam-port to
the exhaust. When lap is added, however, the steam is inclosed in the
cylinder, without egress for the time that it takes the lap to travel
over the steam-port. An arrangement of motion which will make the valve
travel quickly over the port, has a tendency to shorten the period
for expansion; while making the valve travel slowly over the port,
has the opposite effect, and protracts expansion. A valve with, say,
five inches travel, has a comparatively long journey to make during
the stroke of the piston; and the lap-edges will pass quickly over
the steam-ports,--much more quickly than they will when the travel is
reduced to three inches. In a case of this kind, there is more than the
mere reduction of travel to be considered. Suppose the valve has one
inch lap at each end. When it stands on the middle of the seat, it has
a reciprocating motion of two and one-half inches at each side of that
point to make. At the beginning of the stroke, it has been drawn aside
one inch (we will ignore the lead), but still has one and one-half
inch to travel before it begins to return. On the other hand, when the
travel is reduced to three inches, the valve has only one and one-half
inch to travel away from the center; and, one inch being moved to draw
the lap over the port, there only remains one-half inch for the valve
to move before it must begin returning. This entails an early cut-off;
for the valve must pass over the ports with its slow motion, and be
ready to open the port on the other end, before the return stroke. Thus
a travel of five inches draws the outside edge of the valve one and
one-half inch away from the outside of the steam-ports, three inches
travel only draws it one-half inch away, and a greater reduction of
travel decreases the opening in like proportion.
INFLUENCE OF ECCENTRIC THROW ON THE VALVE.
As reducing the travel of the valve diminishes the port opening, a
point is reached in cutting off early in the stroke where the port
opening is hardly any more than the port opening due to the lead. This
is what makes long steam-ports essential for a successful high-speed
locomotive. The best-designed engines give an exceedingly limited port
opening at short cut-offs, and badly planned motion sometimes seriously
detracts from the efficiency of the engine, by curtailing the opening
at the point where a very brief time is given for the admission of
steam. The magnitude of the eccentric throw exerts a direct influence
on the port opening when cutting off early. A long throw tends to
increase the opening, while a short throw reduces it. The long-throw
eccentric will draw the valve farther away from the edge of the
steam-port, when admitting steam for the same point of cut-off, than
a short-throw eccentric will move its valve. For an ordinary 17 × 24
inch locomotive, the throw of eccentric should not be less than five
inches, unless the engine is intended entirely for slow running. There
are many engines running with eccentric throw less than five inches,
but they are invariably slow unless the valve-lap is very short. With
an ordinary lap, an engine having an eccentric throw of 4½ inches needs
so much angular advance to overcome the lap, and provide lead, that the
rectilineal motion of the eccentric is very meager at the beginning of
the stroke. That is, the center of the eccentric is traveling downward
in its circular path, which gives little motion to the valve, just
as the crank gives decreased motion to the cross-head when near the
centers.
HARMONY OF WORKING-PARTS.
Hitherto we have regarded the link as merely performing the functions
of transmitting the motion of the eccentrics to the valves, with
the additional capability of reducing the travel at the will of the
engineer. Otherwise, the motion of the link is intensely complex; and
its movements are susceptible to a multitude of influences, which
improve or disturb its action on the valve. A good valve-motion is
planned according to certain dimensions of all the working-parts;
and any change in their arrangement will almost invariably entail
irregularities upon the link’s movement, which will radically affect
the distribution of steam. A link-motion schemed for an eccentric
throw of 4½ inches will not work properly if the throw be increased
to five inches: a link with a radius of 57 inches can not be changed
with impunity for one of 60 inches. Any change in the position of
the tumbling-shaft or rocker-arms distorts the whole motion, and any
alteration in the length of the rods or hangers has a similar effect.
That the link may perform its functions properly, all its connections
must remain in harmony.
ADJUSTMENT OF LINK.
A very important feature of the link is its property of adjustability,
which serves to neutralize the distorting effect of the connecting
rod’s angularity. As has already been explained, the angularity of
the main rod tends to delay the cut-off during the backward stroke,
while it is accelerated in the forward stroke. With the ordinary
length of connections, this irregularity would seriously affect the
working of the engine. But it is almost entirely overcome by the
link, which can be suspended in a way that will produce equality for
the period of admission and point of cut-off for both strokes in one
gear. Perfect equalization of admission and cut-off for both gears
has been found impossible with the link-motion; and designers are
generally satisfied to adjust the forward motion, and permit the back
motion to remain untrue. The point about the link which exercises the
most potent influence on adjusting the cut-off, is the position of the
saddle, or of its stud for connecting the hanger. This stud is called
the point of suspension. Raising the saddle away from the center of
the link will effect adjustment of steam admission; but in locomotive
practice the saddle is nearly always located in the middle of the link,
there being practical objections against raising it. Equalization of
steam distribution is produced by placing the hanger-stud or point of
suspension some distance back of the center line of the link-slot, the
distance varying from ⅛ inch to ⅞ inch.
Moving the hanger-stud affects the link’s movement in a way that is
equivalent to temporarily lengthening the eccentric-rod during a
portion of the piston-stroke. The length of the tumbling-shaft arms,
the length of hanger, the location of the rockers and tumbling-shaft,
the radius of link, and length of rods, all exercise influence on the
accurate adjustment of the valve-motion.
SLIP OF THE LINK.
In equalizing the valve-motion, and overcoming the discrepancy of steam
admission, due to the angularity of the connecting rod by moving the
link-hanger stud away from the center of the slot, a new distortion is
introduced. The link-block being securely fastened to the bottom of
the rocker-pin, moves in the fixed arc traversed by that pin, which is
nearly horizontal. The action of the eccentric-rods on the link, on the
other hand, forces the latter to move with a sort of vertical motion
at certain parts of the stroke, making it slip on the block. Moving
the hanger-stud back tends to increase this slip, which will become
excessive enough to seriously impair the efficiency of the motion if
not kept within bounds by the designer. Where the slip is very great,
the motion will not be serviceable, a consideration which can never be
overlooked; for the block will wear rapidly, producing lost motion, a
very undesirable defect about any part of a link-gear. With the long
rods which prevail in locomotive practice, designers have no difficulty
in keeping the slip within practical bounds; but with marine engines it
is sometimes necessary to sacrifice equality of steam admission to the
reduction of the slip. The greatest amount of slip is in full gear, and
it diminishes as the link-block is moved towards the center.
Placing the eccentric-rod pins back of the link-arc, as is almost
universally done in this country, has a tendency to make the link
slip on the block; and care has to be taken not to locate these pins
farther back than is actually necessary for other requirements of the
link-motion’s adjustment. Auchincloss, who is a recognized authority
for designing of link-motion, gives four varieties of alterations
capable of reducing the slip when it is found too great for a
practicable motion. His resorts are, either to increase the angular
advance, reduce the travel, increase the length of link, or shorten the
eccentric-rods. One, or a combination, of these methods may be adopted,
as the designer finds most convenient.
RADIUS OF LINK.
Among the constructing engineers who plan link-motion, there is
considerable diversity of opinion about what radius of link helps to
produce the best valve-motion. The distance between the center of
axle and center of lower rocker-pin may be accepted as approximately
correct, although some designers slightly increase beyond these points.
On the other hand, the locomotives sent out from a leading building
establishment have the radius of link drawn ¾ inch per foot short of
the distance between the axle and rocker; and the claim has been made,
that the arrangement produces an excellent motion.
A committee of the American Master Mechanics’ Association have placed
themselves on record on this subject by asserting that the distance
between the centers of axle and rocker-pin is the proper radius for
the link. That same committee recommended that the link-motion should
be planned to give as long a link-radius as possible, subject to the
first-mentioned conditions.
It must be noted that the middle of the link-slot is the radius arc. I
knew of a case where the links for an altered locomotive were finished
out of the true radius through the edge of the slot being taken as the
radius-curve.
INCREASE OF LEAD.
Most of the men who are at all familiar with the valve-motion are aware
of the fact, that, with the shifting link, the lead increases as the
link is notched towards the center. Where the valve has 1/16 inch lead
in full gear, it is no unusual thing to find it increase to ⅜ inch lead
opening at mid gear. The phenomenon is better known than its cause is
understood.
[Illustration: FIG. 13.]
The relative positions of link and eccentric centers of an engine,
when the crank is on the forward center, are shown in Fig. 13; the
link being represented with the block in the center, which represents
mid gear. It will be observed that the centers of the eccentrics _f_
and _b_, from which the rods receive direct influence, are both some
distance ahead of the center of the axle, the one above, the other
below. The eccentric-straps to which the rods are connected sweep round
the eccentric circles, and are controlled thereby. When the link is
moved up or down, each eccentric-rod pin, where it attaches to the
link, describes the arc of a circle with a radius drawn from its own
eccentric. If both rods were worked with a radius from the axle-center,
the link could be raised and lowered when the engine stands on the dead
center, without moving the rocker-pin at all; but, under the existing
arrangement, the link is influenced directly by one or other of the
eccentrics, whatever position in the link the block may stand. When the
engine is standing on the forward center, with the link in mid gear, as
shown in Fig. 13, it will be readily perceived that the block stands
at its farthest point away from the axle; for the rods are so placed
to reach their greatest horizontal distance ahead, and consequently
in this position the lead opening is greatest. If the link be now
lowered, the backing eccentric-rod will immediately begin to pull the
link back: and, as the pin of the forward eccentric-rod approaches the
central line of motion, it will also keep drawing the link back; so
that, by the time the link is in full gear, the lead opening will be
considerably reduced.
[Illustration: FIG. 14.]
When the engine stands on the back dead center, as shown in Fig. 14,
the eccentric centers will be on the other side of the axle, and the
eccentric-rods will be crossed. While in mid gear, the link-block is
drawn closer to the axle than it would be in any other position of the
link; and consequently the lead opening is greatest. If the link be
now lowered, the forward eccentric-rod will approach its horizontal
position, and consequently reaches farther on the central line of
motion, so it will push the link-block away from the axle, thereby
decreasing the lead. Pulling the link into back gear has a similar
effect.
The tendency of a link-motion to increase the lead towards the center
is made greater by shortening the eccentric-rods. Increasing the throw
of eccentric inclines to accelerate the lead towards the center, since
it throws the eccentric centers farther apart. For slow running,
hard-pulling locomotives, where increase of lead is a disadvantage,
the tendency to increase the lead is sometimes restrained in forward
gear by reducing the angular advance of the backing eccentric. This
expedient is, however, not necessary where proper care and intelligence
have been bestowed in the original design of the motion.
In studying this part of the valve-motion, a young machinist or
engineer will obtain valuable assistance by cutting a link template out
of a piece of pasteboard, and using strips of wood as eccentric-rods.
With these he can test on a drawing-board or table the various
positions of the link, and note, in a way that is easily understood,
the effect of changing the link into different positions.