Why are locomotives arranged like this?

Factual Questions
1

I drive a bus route that runs parallel to a set of train tracks. Twice a day a freight train goes by. While stopped at stop lights, I’ll sometimes glance over to watch the train go by.

I noticed that the locomotives on the train are often arranged in a forwards-backwards-forwards-backwards arrangement. Sometimes if there’s only two locomotives they’ll both be pointing forwards, but in many cases there are almost always 2+ locomotives arranged in alternating directions. Why is this? Since I myself have never seen a train go in reverse, wouldn’t it make more sense to have all the locomotives point fowards?

On a more mundane note, how many people does it take to operate one freight train? Is it one guy, like in a tractor-trailer?

2

If the pair of locomotives are working on a route where there’s no easy way to turn them at on or both of the ends, then it wold be easier to have them pointing in different directions, so the crew can just swap locos at the end. I think diesel-electric locos work equally well in both directions, and it would just be more convenient for the crew to work in a loco pointing forwards.

3

I’m assuming you’re talking about people actually on the train, rather than including the ones operating points and switches and so on (which, IIRC, are mostly automated now anyway)…

You only need one person to operate a modern diesel-electric locomotive “engine”; in fact, one person can operate multiple engines. Usually, though, when you see a train with more than one engine there will be two engineers on board. Unless there’s going to be a lot of adding/shedding wagons, that will be the sum total of the crew.

4

Side note/hijack: It’s not freight trains, but the Chicago area’s Metra commuter rail system (full size diesel trains carrying passengers on the same rail lines that carry freight trains) do have trains that run with the locomotive on the back end for some of their runs. The train I take to work will run “backwards” on the way into Chicago, park in the station and let off passengers, then just pull out going forwards along the same tracks and head out into the suburbs. Repeat process for each run of the day.

A number of the passenger cars have a little compartment up top inside that looks like an engineer compartment, though I’m not sure if they do any driving from there or just relay the track/signal conditions back to someone in the locomotive.

5

Giles pretty much has it.

Diesel-electric, diesel-hydraulic, and electric locomotives all do run equally well in either direction (and some even have a cab at either end). Even steam locomotives ran pretty well arse-backwards, but the trouble there was visibility, so in the steam days, most major stations and yards, and pretty much all terminals had a turntable. These days, most turntables are gone, and instead there is the cheaper and more space-efficient “passing loop” or a “runaround loop”. This isn’t a loop as you might imagine it though - more like a segment of line where there is a parallel track joined to the main line at either end. Basically, you stop your train on the main line between the switches, uncouple the lcomotives, drive forwards, throw the switch, drive backwards down the other line back over the trailing switch, throw that switch back to straight, then back up and couple to the other end of the train. The last loco is now the lead unit. There is also such a thing as a “balloon loop” which actually is a loop, and in that you simply drive your train around it and wind up facing the other way. Those take up loads of space though and are relatively uncommon.

So, in a “lash-up” of, say, four locomotives, you need the engineer’s cab facing outwards - but only on units 1 and 4. The middle ones can be any which way, and indeed some locomotives are specially designed to operate in the middle, because they have no driving cab at all! These are cheaper to construct, and are called “drones”.

Of course, that is the traditional way of doing it. On longer freight trains, it is increasigly common to use “distributed power”, which simply means putting locomotives in the middle or at the end of the train as well (still controlled by the guy at the front). This is for purposes of increased adhesion and braking power, and if you have other locomotives at the end of the train, you don’t need to do any shunting at all when you want to go home - just walk up to the other end and get in.

Some countries have experimented with a control cab in the rear car of passenger trains that are hauled by locos at one end, and this might be what you are seeing. The British even experimented with doing this in the steam days, with the fireman staying in the loco, and the driver controlling the train remotely.

6

Just to add, if a freight train is returning empty, it has more motive power than it needs, and the crew will often switch off one or more locomotives. It is even possible to switch off the lead unit (the controls will still work), and some folks like to do this for a quieter ride. Others say it’s a bit spooky driving a train in silence.

7

This is precisely how the regional trains in Toronto work. On one end you have the locomotive, and on the other end, you have one of these.

8

Freight trains NEVER operate with just one person. All trains (freight or passenger) are required to have at least a two man crew. The engineer is the title of the person who “drives” the train, and is found at the front of the train, at the controls. The brakeman (which is also sometimes the role that the conductor–the manager of the train–plays) is also required to protect the rear of the train, in the event that it needs to reverse, and also to jump off the train in order to flip manual switches, if the train needs to change tracks.

9

That looks quite similar to the Tri-Rail that runs in South Florida.
Tri Rail Picture

It never turns around, the driver just transfers from one end to the other.

Further reading of the Wiki article indicates that “The service began with MKO F40PHL-2 locomotives and Bombardier BiLevel Coaches originally destined for GO Transit, but diverted from the original GO order.” So it’s no wonder they look the same, except for the palm trees of course.

10

Those GO units look wonderfully retro to me. I’m sure they’re high tech and all, but there’s a sort of “optimistic futurist 1930s” look about them. I think it’s the multiple sloping windshields.

There is actually a bit of a safety argument against having trains pushed from the rear. The things are perfectly safe on the track, and are not necessarily more prone to derailment than any other train, but if they do come off the metals, there is a tendency for the cars to scatter acros the countryside being pushed by the loco before it can stop, moreso than if the heavy locomotive is leading and if it derails, the cars tend to follow with less chance of broken couplers and more chance of staying relatively upright. It’s a bit like pulling a trackless toy train across carpet, then trying to push it backwards - or a n00b trying to reverse a trailer in his car.

11

Aren’t the front/back symmetrical diesel electric locomotives called “pushme-pullyous”’?

Is there a different limit on how many cars in a row you can have before you need a locomotive, based on direction? That is, if you made a simple train with many cars and one locomotive at one end, could you make it as long to go one direction as you could the other? For one thing, on a curve, pushing means there is additional force pushing outward on the curve, whereas pulling means that additional force pulls inward. If centrifugal force mattered much on curves, I’d think you’d do better by cancelling some of it by having the train in tension rather than compression.

12

Theoretically, I suppose, you could push more cars than you could pull. This is because you can only pull cars up until you reach the design limits of the drawgear (couplers and associated bits and pieces). The actual number of cars would depend on weght and other factors. Probably before you reached this limit, if you only had locomotives at one end or the other, you would start having problems with the air brakes, so I’d guess that would be the determining limit, in either direction.

Once you start using “distributed power”, or locomotives in the midldle somewheres, then there is basially no limit. Let’s say you have two freight trains, one in front of the other on the same track. You can just couple them together, join up the air and the electrical jumpers, and bingo, you’ve got one train, controlled by one locomotive. The limit of the longest trains then comes down to the maximum length of freight yard sidings and passing loops along the way.

Curves are not so much of a concern (in this area) as you might think. Trains don’t tend to want to pull themselves into the inside of a curve rubber band-style - if you consider the long rows of rubber-tyred wagons pulled by airport tractors, they can hold tight curves even without the benefit of rails. It’s not uncommon, in this country at least, to see photographs of a freight train passing itself in the distance. Curves are more of a problem in terms of increased demands on locomotive power, especially on grades.
The issue of tension and slack in the couplings is an extremely complicated one. Generally, the longer the train, the more difficult it is to drive properly without risking a break-away, derailment, etc. Distributing the locomotives makes it a bit easier. There are times when you want the train taut, when you want it slack, or when you want bits of it in different tension. With the locos all at one end (either one) you can get problems with the air brakes, because a brake application will take time to propagate down the cars, and this can be a Bad Thing - probably worse with the power at the rear because a breakaway is more likely as the brakes start to bite at the back of the train first, and you get loads of tension on the couplers. Tension in braking can be roughly controlled by applying the train brake (all brakes apply), applying the locomotive brakes alone, or by “bailing off” (all brakes except the locomotives).

In very general terms, a taut train is good, except when moving off.

13

Yes. I’m not as familiar with freight consists, but for the passenger trains in the NYC area:

Metro-North/LIRR electric trains have a minimum of 4 cars (6 for LIRR because of some of the wider street grade crossings) and a maximum of 12 - more than 12 cars in a set draws more third rail power than the grid can handle. Technically if the third rail power allowed, they could run in an infinite number of cars. Although these cars are called MU (Multiple Units), each car is also considered a locomotive, since it has all of the equipment on board to control and pull the train. It’s a similar situation with subway cars. Most of the IRT (numbered line trains) are in 5 car sets, with a cab car at either end and three trailer cars in the middle- and have two 5 car sets attached together to make a 10 car train. The BMT/IND (lettered lines) have so much different equipment that practically every rule applies depending on what it is, but ALL equipment will have a car cab at both ends.

Metro-North’s main diesel trains are the GE P32 locomotive and Shoreliner single-level trailer cars - each consist also has a Shoreliner cab car on the southern end, for when trains operate in push mode with the engine in the rear. This equipment can handle up to 10 Shoreliners for one cab, although MNRR never runs more than 7 at a time. Note that the less # of cars, the higher the acceleration level, which is why the 2 and 3 car sets used on some of the branches have faster rides than the 6 or 7 car trains that run in and out of Grand Central.

Long Island Railroad uses DE-30 engines and C3 multi-level trailers. Although there are C3 cabs, these aren’t widely as used since one DE-30 can only push or pull up to 6 cars. Any more than that (including all of the diesels that run into Penn Station) and a second locomotive is needed, which is put on the opposite side of the train. The Cannonball, the Friday night summer getaway train to the Hamptons uses a locomotive at each end and 12 C3 trailer cars. I’ve heard that thing is a bitch to stop, once it’s running at full speed.

New Jersey Transit is the only railroad in the area (besides Amtrak) that uses Electric Locomotives, as opposed to electric MUs. One of these can pull up to 12 single level trailer cars (always with a cab car at the other end) or 8 multilevel cars (again with a cab car at the other end). More than 8 cars and they’ll put an electric locomotive on both ends. Note that these trailer and cab cars work the same with diesel engines, although as of yet the multilevel cars haven’t been used much on the diesel lines.

I’m not sure what the limits are for Amtrak, but I’ve noticed that if they add a second locomotive (either diesel or electric), they’re usually both at the front end, in the configuration described by the OP. Because of this, most Amtrak trains can NOT operate in the opposite direction, and must use a loop to turn around. The exception is the Acela, which has an electric engine at both ends and 6 passenger cars, but rather than the rear car just providing extra power, both engines work together as a push-AND-pull operation, which helps maintain the high speeds and smooth ride which the Acela is known for.

14

Nothing much to add here other than that LA’s Metrolink commuter rail system also uses Bombardier equipment and switches off between using the push or pull configuration depending on which direction the train is headed. The push/pull system has been the subject of a lot of discussion and controversy out here, mainly because of this incident, which was just back in the news this morning.

15

I think you’re talking about a siding here–at least, that’s the term we’d use in North America. Handy things; good for allowing trains to pass in opposite directions too.

There are still also wyes in places here–these allow trains to basically make a three-point-turn, just like you might in your car. The train can’t be too long, of course (unless you have a massive wye), but I recall being on one train of about a quarter-mile in length that was wyed. Backwards and forwards, waiting for switches to be thrown–a siding would have made more sense

You’ll still see this in the Canadian west, most notably near the Spiral Tunnels. Great long heavy freight trains negotiate a 2.2% grade through the spirals, often crossing themselves in the process. Sometimes, in the middle of the consist, you’ll find a locomotive or two, which will be removed once the train gains the summit of Kicking Horse Pass.

16

The OP’s questions have mostly been answered, but I’ll throw in a bit more detail on US freight railroad practice in case anyone’s interested.

It’s interesting to note the differences in typical European and American practice. In both cases, turning a locomotive at each end of a run is considered inefficient, but the solution differs by region. In Europe, most road locomotives, whether diesel or electric, have cabs at both ends, since trains tend to be rather short affairs pulled by a single locomotive. In North American practice, locomotives have single cabs, usually closer to one end than the other, while the trains are relatively long and pulled by multiple locomotives controlled from the leading cab.

The single-cab arrangement traditionally has been preferred in North American practice because the added expense of a second cab (with duplicate control hardware and provisions for crew) on each locomotive was considered excessive for the utility gained. With two locomotives or more used for most trains, it was simply more expedient to run them in pairs tail-to-tail, with the cabs at the outer ends. This practice was begun during the early diesel years, when locomotives had full width and full height carbodies, but carried over when railroads went for the so-called ‘road-switcher’ arrangement, with narrow hoods covering the prime mover and other operating gear. Early road switchers had high short hoods, with a less-than ideal view ahead for the engineer, and a few railroads ran them interchangeably long- or short-hood forward; by the 1960’s however, the manufacturers were building road locomotives with low short hoods for better visibility and running them short-hood forward became the dominant practice.

One side note: early diesels were often sold in set of two ‘A’ (cabbed) units back-to-back, with one or more ‘B’ (uncabbed) booster units coupled in between the 'A’s. Once the road-switcher configuration became dominant, the use of uncabbed boosters pretty much died out, although, I believe, a few may still be found in use today.