NEB&W Guide to Steam Locomotive Components

Last Update: 2009-02-07

Locos Table of Contents
Rolling Stock Table of Contents
The following discussion is not intended for a student of steam technology but to how these features might relate in terms of modeling steam.

Valve Gear

When the Newcomen engine was first used, they had to hire a boy whose job it was to open and close the valves on each stroke. He stood at ground level and manipulated the valves by means of strings. I believe it was one Humphrey Davies who got bored with this, so he tied the various strings to appropriate parts of the machinery so that the valves were worked automatically. This also speeded up the engine. His employer rewarded him by firing him as he had worked himself out of job (no unions then to maintain a featherbedding job of valve puller).

Now think of a loco. Steam is admitted to one side of the piston by the opening of that valve first, the piston moves down the cylinder, the first valve then has to be opened to the exhaust while the other valve opens to live steam, and the piston moves back. In other words, the valves move when the piston is at either end of its stroke, and thus the piston is connected to the wheel 90 degrees out of phase with the connection for the valves. If the piston rod is connected to the driver at say "6:00", the connection for the valves might be either "3:00" or "9:00". The choice will determine if the engine moves forward or backward.

There were two basic types of valve gear on steam locos, inside and out. The 19th century loco used the inside type, known as Stephenson Link Motion. For each cylinder, there were a pair of disks off-center mounted on one of the axles, or in other words, two pairs of disks. Each pair had one disk 90 degrees ahead of the piston connection and the other, 90 degrees behind. The disks acted as cams. Each disk was surrounded by a ring and a rod that connected to either the top or bottom of a curved slotted link. As the engine moved, this link rocked back and forth. (I've never understood why they didn't use just one disk and pivot the link, as they did when they moved the valve gear to the outside until just recently. The connecting rod was pivoted to transmit the reciprocal motion outside to the steam chest. This fixed the rod, so the link had to move up and down to change the position, so it couldn't also be fixed in position with a pivot.)

The connection to valve chest was a rod that slid in the link. If the engineer moved the connection to the top of the link, that motion was conveyed to the valves. If the rod was moved to the bottom, the opposite motion was conveyed, thus reversing the engine. Now consider if the connection was set in the middle of the link. There would be no motion, so the valves would not admit any steam to either side. If the connection was moved up or down just a little, the valves would only operate with partial movement, so only a little steam would get in for each power stroke. When the engine was started, the connection was moved all the way to the end of the link to feed the maximum amount of steam per stroke, thus getting the most power. Once the engine was up to speed, the engineer would move the connection toward the center to conserve steam.

At about the same time they moved the linkage on the outside, they improved it with another connection. A set of levers was added between the piston rod and the valve rod. The idea was that this added and subtracted some of the motion of the piston rod to the valve gear. On part of the stroke, while the valve gear was say pulling on the valve rod, the motion from the linkage was pushing, and the net effect was no or little movement. At another part of the stroke, the two added together. The net result was to make the valves open and close in a jerky but snappy motion, rather than the slow and steady motion obtained from a single rotating source.

The actual first use in America of Walschauerts, the first outside type of valve gear, was on the "William Mason", built in 1974, but it didn't catch on at the time. In 1904, it was used again, on the B&O's "Old Maude" and this time everyone took note.

The other two major types used in America was the Baker and the Southern. The second most common was the type invented by A.D. Baker and I believe the first use of this type was in 1911. It differed, I believe, in that there was no sliding joints.

The Southern valve gear was invented by William S. Brown, a loco engineer on the Southern R.R. It was first used in 1913, and by the time of the USRA designs in 1918, was popular enough to be used on their light and heavy 2-10-2's. Some other USRA engines were also built with it, such as New Haven's USRA 0-8-0's and 4-8-2's. Even so, not that popular in the long run. A key spotting feature would seem to be that the crescent-shaped slotted link is horizontal.

The Steam Chest

The original valve most commonly used was called a "D" valve, due to its shape. This had the advantage of steam pressure keeping it tight.

Imagine a card table with three closely spaced one-inch holes drilled through the top, in a line. The middle hole is connected to a vacuum cleaner under the table. Your job, Mr. Phelps, is to cup your hand so as to connect the vacuum cleaner to first the left-most hole and then the hole on the right. As you move your hand back and forth to cover each hole, the other is exposed to the atmosphere. The point is that all the time, the vacuum cleaner is sucking down to keep your hand on the table.

In fact, if you turned up the machine, it might suck so hard you couldn't move your hand. If you greased the surface, then you would be able to slide it back and forth.

Well, this is analogous to the steam chest. The vacuum cleaner represents the exhaust connection to the stack. The top of the card table represents the live steam coming in from the boiler.

When they went to superheating, the temperature of the steam was raised several hundred degrees than it had been. This was too much for the lubrication they had available under the conditions of the D valve. Instead they had to go to a piston valve, which didn't have the pressure of steam on the sliding surfaces. A piston is round, the D valve used a square steam chest, so that all engines built with superheating had this double round look. It was possible to add to retrofit a square steam chest with a piston valve, but often this required a new outside steam line.

Superheating came in just after 1900 and the outside valve gear was successfully introduced in 1904, on the B&O's "Old Maude".

In other words, 19th century locos had square steam chests and inside valve gear. Twentieth century ones had round chests and outside valve gear. C. 1902-'10, you might see either a square chest with outside gear, or a round chest tilted inboard for the inside gear, both combinations somewhat rare and definitely limited to locos built during this short time span.

The Firebox

The power of the loco was developed in the firebox - the bigger the firebox, the more powerful. (Not fast - the size of the drivers determined speed.) The firebox was located at the end of the loco nearest the source of fuel, the tender. During the 19th century, it was between the last two drivers. C. 1900, it began to be placed above the drivers, allowing it to be wider, but this only worked if the drivers themselves were small enough. Otherwise, the firebox would not be deep enough. Within a few years (early 1900's), the firebox became so big, it had to sit behind the drivers, supported on its own set of wheels. At first, only a pair, but then later, four wheels. (The C&O built a class of locos, called "Alleghenies", that had a 6 wheel trailing truck because the firebox was so enormous, but dieselization came along before this next step became common.)

The firebox is squarish while normally the boiler section around it is round. This means each set of staybolts running between the two were of different lengths. If the boiler section here was also made rectangular and thus paralleled the firebox, it gave a sort of shoulder-pad look to the boiler. This was called a Belpaire firebox, although I think technically the firebox itself was standard and it was the boiler design so modified. (And this reflects the modelers' tendency to label the last whole segment of the boiler the "firebox".)

The Drivers

Nineteenth century locos had counterweights that were placed between the spokes and bolted on. Twentieth century ones were cast on, in a sort of crescent moon shape. (Don't have a date when they switched.) The main driver, the one that had the rod from the cylinders attached, had a bigger counterweight.

In the Dec. 17, 1932 Railway Age, there was an announcement about the new "double-disc" driving wheel invented by the Scullin Steel Co. (which modelers would come to call a Scullin Disk Driver). The article said they had been in use and being tested for the last 13 months (in other words, late 1931) on the Frisco, the MP, the D&SL, and the NYC. The claimed advantage was that it allowed for more counterweighing. In some cases, they used just a single Scullin on the main driver, on others, for all the drivers.

Boxpok drivers become the norm toward the end of steam. In rare, rare cases, the center driver could be a Boxpox while the others would be spoked. According to an article in the Nov. '34 Railway Mechanical Engineer, the design (and the castings themselves) came from General Steel Castings. The name was a take-off of "box spoke" because the spokes had a box-section.

I think there was also a Baldwin disk driver. Don't have the particulars on this.

Some modelers have taken a spoked driver and added an overlay of thin styrene to simulate these cast drivers. Don't have a practical way to convert a disk driver to a spoked one.

The Cab

The term probably comes from the same source that gives us "cabin", "cabana", "cabinet" and maybe even "caboose". In popular use, as shortened form of "taxicab". Diesels have cabs, cranes have cabs, trucks have cabs, and of course steam locos have cabs.

Cabs were no big deal (in terms of design, etc.) to the prototypes, but in terms of modeling (and overall look), I think the cab is one of the most distinctive aspect. Many roads had their own unique style, although the style changed over years. For instance, the NYC had a two-window type with a horizontal axis up until WWI (which I call the Alco standard cab) for lack of a better term. In the '20's, they were made much shorter, so the axis changed to vertical, and they slowly gained length until by the '40's, the horizontal distance was greater. ANY NYC loco built within the years of a certain style cab style tended to get that cab and not the same cab as might of been found on the same type of loco built earlier. (Confusing? A NYC Pacific, Ten-wheeler, Consolidation, Mikado, etc. all built in the 1910's got the standard Alco cab. A Pacific built in the 1920's would have the same type of cab as a Mikado built then, and NOT the same as the Pacifics built a decade earlier.)

The Harriman-controlled roads had a unique variation on the cab roof, a sort of scalloped cutout. This might have been just any old variation, but these locos have served as the basis for some of the Varney engines, MDC engines, and the Bachmann 2-8-0, so this design is over represented in the hobby.

Basically 19th century cabs were made of wood and 20th century ones were made of steel. But there is another facet of this. When loco boilers were small, the cab could be placed up high enough to see over the boiler and still be within clearances. (Think of a diamond-stack loco where the stack is much higher than the top of the cab.) But the size kept growing and the cabs were raised ever higher. By 1900 or so, they were high enough that clearance became an issue. Now if you look at the typical clearance diagram, the top corners are rounded, and so the cab roofs were also rounded to fit - pretty much the same profile as the PRR roundroof box cars, also made to just fit within the clearance diagram.

A wood cab could be made with clapboard or other types of typical wood construction. However, most American wood cabs were given a panel look and you don't tend to see the individual boards. Steel cabs had rivets, but these were pretty tiny and more like pepper grains in HO than the "half-grapefruit" sized rivets of early cast metal models. (You would be better off filing off all the rivet detail than leaving the D cup size rivets of the worse offenders.)

The front of the cab extends forward of the back of the boiler and the cab is made with a vertical front. Occasionally, the front of the cab follows the line of the backhead of the boiler, which slopes back. This cab shape is sometimes called "sports model".

Sometimes the back of the firebox was so wide, the cab was cut out at the bottom.

The extreme northern roads (CP and I think some CV) had a distinctive "all-weather" cab, where the back was enclosed.

The Domes

Technically, there is only one dome on a loco and the other similar looking affair is a sandbox, but it is okay in modeling parlance to refer to both jointly as domes. The steam dome is to collect steam from the highest point of the boiler so it doesn't carry over water. The sand dome, oops, I mean sand box, is forward of this, generally centered over the drivers or near the front of the drivers, and generally the pipes leading down for the sand are exposed.

In general, the builders would build anything and everything to suit the purchaser, but left to their own designs, Baldwin favored a ringed dome and Alco, a gumdrop shape.

Headlights

Earliest engines weren't run at night so it took awhile for headlights to become standard (around 1850).

In general, oil headlights were housed in a rectangular box, with a nubbin on top which I think acted as a chimney. At first they were so big and so was the balloon-type stack, while the smokebox was short, so the headlight was placed on a bracket. Later engines had the headlight on the top of the smokebox.

In 1897, George Pyle devised a practical steam-powered generator which made the arc headlight possible. The arc light was typically housed in a cylindrical housing. It, too, had a nubbin on top, to hold the top electrode. I can't think of any reason while the shape of the housing would change from a box to a cylinder.

At first the electric or incandescent light used a carbon filament which was way too fragile to be bounced around on a loco (or an auto, or a bicycle), but in 1911, ductile tungsten made a much studier and longer lasting light. In 1915, federal regulations required electric lamps although I don't know if the arc headlight was included under that, or not. Also saw a reference that in 1920, all locos had to have electric lights so I bet this was one of those requirements that kept getting extended.

The first electric headlights were rather large, about the same size as the arc lamp. The main thing (from our modeling viewpoint) was that the nubbin at the top was eliminated.

With early locos, the boiler was small and sat low to keep the center of gravity down. The headlight almost universally was set at the top of the smokebox. As engines grew, they had a choice of locations, including centered on the smokebox or sometimes higher than center. (The Central Vermont seemed to favor the headlight bracket itself was centered which put the headlight off-center.) On articulated locos, in later years the headlight was put on the pilot deck, which turned around a curve (because it was attached to the front set of drivers) before the boiler did.

The Mars light was an auxiliary one to the main light. The Mars light oscillated in a figure 8, to draw attention to a moving train approaching a grade crossing to warn motorists. There was a short reference in the June '37 Railway Mechanical Engineer that it was being used on the CNW's "400" trains between Chicago and the Twin Cities. This must have been the first use or among the first, as they had to explain it to their readers.

Pilots/Cowcatchers

The public knows these as "cowcatchers" but the industry and modelers call them pilots. Neither term makes too much sense. The device doesn't catch a cow, it throws it off the track. A pilot is a guide. Perhaps the front wheels were known as the pilot wheels because that is what they do, and the device got its name from that. The device was not to provide humane treatment to a wandering bovine but to sweep any obstacle from being hit by the wheels, because that would easily derail the loco. (Unlike a highway vehicle would could ride over something in the road, you only have to lift the wheels of an engine an mere inch so the flanges would clear the railhead.)

The first "cowcatcher" was used on the John Bill, a wedge-shaped device so massive it required its own set of wheels. Nineteen century locos had a wood pilot that extended way far out in front. I think this might have been because the wood wasn't strong enough except in compression. Twentieth century locos favored a stubby pilot made of scrapped boiler tubes set in a vertical fashion. Some roads had a hallmark design. Switching locos didn't have a real pilot but instead a pair of footboards, which I guess for the slow speed of a switcher was sufficient to prevent obstructions from reaching the wheels. The New York Central used footboard pilots on all their freight engines, including all their road engines.

Feedwater Heaters

A feedwater heater used steam from the exhaust to preheat the water going into the boiler. Thus it would make sense to find these devices up near the stack. Sometimes these were exposed and make a big visible difference and sometimes they are pretty much buried in the smokebox and barely visible.

I have to go look this up, but I think there were three main types. One was the Coffin type, which was shaped like an upside "U" and very very odd when hung outside on the front of the smokebox. The Elesco was a crosswise cylinder, which looked like a St. Bernard's keg when mounted outside. ("Elesco" probably is a contraction of something with "company". Maybe even "L.S.Co.") The Worthington was shaped like a box and according to Model Railroader, gradually became the favorite.

Air Pumps

The steam from the loco was tapped to power the device that pumped up the air for the air brake system. (Thus you would not expect to see any air pumps on locos pre-air brakes and any older engine would be expected to have been given air pumps if used into the 20th century.)

At first air pumps were mounted on the left or fireman's side. The engineer typically sat on the right side and controlled the valve gear by a rod that went out to the valve gear (with or without a power reverse to assist him moving this heavy machinery). If the air pumps were on his side, it would get in the way of the rodding.

In general, the more powerful the loco, the greater the need for air. Thus a small loco might have one pump, a big engine, two cross-compound ones.

At one point, the ICC decreed that above a given weight on the drivers, the loco had to be equipped with a coal stocker, as the power requirements were too great for manual firing. The Rutland moved the air pumps on a few of their 2-8-0's to the front of the smokebox in order to lighten the weight on the drivers and thus get under the wire. Other roads sometimes placed the air pumps on the front of the smokebox, which creates a very distinctive look.

As boilers grew in diameter, there wasn't clearance for air pumps on the side, probably the major reason why they might be placed on the smokebox door. Modern steam placed a pair on the pilot deck and obscured them with a shield. (When any fluid like air is compressed, it heats up. The air was run through a series of pipes to allow it to cool but along with the front mounted air pumps, they made this into a radiator - probably from seeing this done on automobiles.)

Running Boards

With 19th century locos, the trainmen could walk onto of the square steam chest, so the boards could stop short of the cylinders. Not so with the round chest when piston valves were introduced. Around 1930, when the whole concept of streamlining started to take hold, the running boards were made straight down the length of the loco.

At first the running boards were made of wood. Probably around 1900, along with so much else, running boards were switched to metal. In later years, the running boards were etched with a diamond or other pattern to make them less slippery, but early 20th century locos had smooth metal. (At one point, all HO models had smooth boards, particularly on cast-metal boilers with the boards cast on. Then some manufacturers started making the boards a separate part and added a diamond tread, but that isn't appropriate for all locos.)

On the prototype, the handrail stanchions were made different lengths, so the handrail itself would stay level, even as the boiler changed in diameter along its length. On many models, the stanchions are all the same, so the handrails run crooked. Even worse, model handrails are often oversized, so the effect is even more pronounced.

Smoke Deflectors

As steam locos grew out to the clearance diagram, their resulting sort of streamline effect (even on locos not given streamlining shrouding) caused the exhaust smoke to hug the profile of the boiler and get in the engineer's eyes. Various attempts were made to deflect the smoke upwards by means of sheet metal around the smokebox. Whether this actually worked wasn't too clear as I understand proponents of each side were equally adamant in their effectiveness. But they sure do alter the look of the loco.

There also wasn't any standard shape to these, so even with a model that has them, you could change the shape and alter the look.

The Tender

Railroads were developed from the mine trams of Europe's coal mines and the first steam engines used to pump out water would have used the fuel at hand, coal. Even the pioneering locos in this country of the 1830's were designed to burn coal, but at first they couldn't get the coal to burn fast enough to keep up the steam. And American was covered with vast forestlands which were deemed a hindrance to the spread of civilization. So wood became the predominant fuel. The tenders were basically rectangular water tanks on wheels with the wood piled on top.

It took the development of an adequate combustion chamber inside the firebox to make coal more practical, even as the woodlands were being depleted. For instance, between 1853 and 1859, the PRR ran tests on the relative merits of bituminous coal vs. wood. Coal proved so superior that all the wood burning locos were retired only three years later (1862). Coal, however, had to be shoveled, not tossed by hand. The tender was modified by making the front half for coal, the back half for water, with a sloped partition between. In order to keep the weight balanced, the front coal section was made higher (coal plus the spaces between the pieces being less dense than water).

Oil was another alternative - it being burned in lamps back to prehistoric times. In 1887, the PRR tested an oil burning loco. While they found that on a weight basis, oil was almost twice as effective as coal (1:1-3/4), substituting the 8,000 tons of coal being used daily would have used over a third of the entire US's production of oil at the time. Oil was only used on railroads near oil fields where it was much cheaper. A tender for an oil burning loco has a tank inserted in the coal bunker section.

Pretty much any tender could be used behind any steam loco. In the final days of steam, they often took the tender of a larger loco that was being scrapped and using it behind a smaller loco that still had some service time left, so you might see a Niagara tender behind a Hudson, for example.

The builder's photos of the first "superpower" locos like Hudsons show ridiculously small tenders. Generally it wouldn't make much sense to have more supplies (i.e., a bigger tender) than a loco would use on its run. Earlier locos had to be serviced at frequent intervals and the small size of the tender reflected that. The steam of the c. WWI era could probably run much longer between servicing, but they probably designed the tender without realizing that. (This is a guess.)

As tenders grew in height (or sometimes the coal bunker was rebuilt taller), the sides had to curve in, again (like the cab), to fit the clearance diagram. Also, because both the cab and the tender have to fit within the same diagram, the curves should be in line. Locos stayed on-line (mostly) and thus different roads had different clearances. Generally the Western roads had more generous ones. So you can't also put one tender from one road behind another loco. (I speak from experience. We tried to replace the Broadway Limited Niagara tender with one of their C&O ones, but the tender was higher than the loco. The NYC had some of the worst clearances.)

Locos that ran in areas where there were track pans to pick up water on the fly didn't need to carry as much water. Therefore, the length of the coal bunker in relation to the water area was greatly increased. (Track pans were very high maintenance so they were rare - but the locos that ran in that territory tended to be the ones everyone likes to model.)

Vanderbilt tender was a very distinctive type where the water compartment is round like a tank car. A round container is the strongest way to contain a liquid, so why more tenders weren't built in this fashion, I don't know. I'm not sure of the source of the name itself, as the New York Central, the road controlled by the Vanderbilts, did not use this type of tender to any extent, if at all. However, there was a use of a round firebox in a so-called Vanderbilt firebox/boiler, most notably on the famous 999 - maybe the name just was carried over from that.

The centipede tender was designed solely to fit on to existing turntables. All these wheels were placed nearer the center of the tender, allowing the end to overhang the turntable. (The name was taken from its resemblance to the bug of the same name, which in Latin literally means "100 feet".)

A slopeback tender made it easier for the engineer to see when coupling up in reverse, so capacity was sacrificed for visibility. You would only see a slopeback tender on a switching engine. Another variation for sake of sight was to move the sides of the coal bunker in, the so-called "clear-vision" tender. (Typically, perhaps universally, a slopeback tender had a clear vision coal bunker, but a clear vision bunker typically was not slopebacked.) On the Rutland, the switcher tenders had a clear vision bunker but not slopebacked, at least ones used into the end of steam. On the D&H and NYC, the slopeback feature was also rare. On the other hand, the PRR loved the slopeback feature.