What is a Model Railroad? The Design of Sumida Crossing.

Temp Expwy 2204

What is a model railroad layout, as opposed to a toy train set? Is there a distinction at all? Are they clearly separate or just two ends of a range? I think most people who consider themselves “model railroaders”, even if they will admit to “playing with trains” (and most of us will), would say that there’s a very big difference between a child’s “toy train” and a model railroad layout. Where I suspect we differ is in just what that distinction is. And further, I would also say that model railroaders in different parts of the world have differing views.

Is “serious” model railroading about strict replication of appearance, operating behavior, or exact scale? The answer isn’t as simple as “yes” or “no”, as it’s about all of these things, and yet each of them is optional. And even the most serious model railroaders aren’t above violating the ones they do practice with a bit of whimsey at times: John Allen, an early model railroader well-known for realistic scenery and even weathering of his rolling stock, famously had a dinosaur in one yard pulling a boxcar.

And the reason I’m bringing this up here is that Japanese practice calls some of the assumptions that Western model railroaders have about serious model railroading into question. Can a temporary layout made from sectional track really be a serious one? How important are matters of gauge and scale? How detailed does scenery need to be? In designing Sumida Crossing, I had to re-think what I believed from years of immersion in U.S. model railroading practice, as reflected in books, magazines and shows. I consider the result a model railroad, and I’m certainly serious about it, but it violates many of what I used to see as the “rules” of model railroading.

Here I’m going to examine how the Japanese hobby, or what I know of it from an outsider’s perspective, differs from the North American and British practice I’m more familiar with (the latter mainly through magazines). And what that led me to conclude about what I wanted in a model railroad, and how to go about achieving it.


North American and British Model Railroading


If you’ve never read a model railroad magazine, you’ll no doubt be amazed at the high level of detail, craft, and research that went into the layouts featured there. Many of the larger ones have taken decades of continuous work by an individual to reach their published state. Others are club projects that are more modest in scale, yet still took several years for a team to create. Often they represent specific locations and time periods, carefully researched down to the level of how many boxcars could be stored on each siding. Other times they are more interpretive, but still soundly based on prototype (that’s model-railroad speak for “real railroad”) practice. Clearly these are not toys. And over time, as the technology of model-making has improved, expectations for what “the best” should be have risen. It can all be a bit intimidating to a beginner.

And yet, nobody starts out as “the best” at anything, and almost all modelers tend to have one aspect they’re better at, or more interested in, which they emphasize over the others. Individual circumstances also come into play: what a student or retiree living on a modest income in a small apartment can do is very different from what a middle-aged professional who owns a house with a full basement can. That disparity doesn’t make one more of a model railroader than the other, although it may make their layouts more or less impressive at a casual view.

In North America, layouts featured in magazines tend to be large, filling a basement, garage, or spare room. Often these represent a point-to-point original, but may be connected at the ends (out of sight) to allow continuous running. While British layouts can be that large, more commonly they’re sectional layouts intended for use at shows, which can’t be fully assembled at the owner’s home due to lack of space. And typically these are point-to-point layouts with the scenic portion perhaps a dozen feet long and two or three deep, although ones with continuous loops are not uncommon. Both tend to emphasize prototypical operation, running multiple trains to schedule, and following a simplified set of operating rules similar to those on a real railroad. Both also strive for highly realistic scenery and rolling stock (trains) appropriate to a specific prototype (real or representative).

At the other end of the scale, a “toy train” is typically set up on the floor (often under a Christmas tree, at least in the west) and consists of a simple loop of track, maybe with one siding, and one train with a small number of cars, not necessarily all in use at the same time in the real world. Trains tend to run in circles, often at unrealistically high scale speeds (a steam engine doing 200 mph isn’t at all prototypical, but that’s about how fast most toy trains will run at full throttle).

Let me summarize what I thought of as the “rules” of Model Railroading, before I started thinking about Japanese models and modeling:
- Scale: A model railroad should be built to a specific scale (ratio of model lengths to equivalent prototype lengths).
- Gauge: Track should be built to a specific gauge (distance between the rails) that is the scale equivalent of the prototype being modeled and be built in a way similar to the prototype (i.e., with ballast, ditches for drainage, and even appropriate weight rail for the era and modeled prototype).
- Place: A model railroad should represent a part of a specific prototype railroad or a fictitious one that is based on prototypes of a specific kind or region.
- Time: A model railroad should be set in a specific time period, defined by rolling stock used and scenic details (which can be fairly broad, and the same layout can be used to represent different periods by swapping out rolling stock and anachronistic scenery).
- Practice: Model railroads should not blatantly contradict operating practice of the modeled prototype (e.g., putting a caboose on a string of c. 1990 container well cars is not something most railroads were doing).

Exceptions


Now you can find exceptions to all of the rules. For example, many modelers use N-scale buildings at the back of HO layouts to create “forced perspective” (making the scene seem deeper than it really is), and HO-gauge track, which is correct for 1:87 scale, is also used with 1:76 scale OO layouts. And not everyone worries too much about rail weight. But these exceptions largely apply in ways that aren’t readily visible. You’ll generate a lot more heat from purists if you use On30 models (which represent 30” gauge trains) to replicate a 36” gauge prototype, as this is a visible difference.

It’s also widely known that many “N-scale” structures are not very close to 1:160, as there is nobody enforcing design standards (unlike model trains, where NMRA standards for interoperability brought consistency to many aspects long ago). As long as these aren’t critical structures (like engine sheds) or placed adjacent to structures of significantly different scale, this is an exception that can pass unnoticed in most cases.

Modular layouts at shows in North America tend to be operated as simple round-and-round demonstrations (unlike “exhibition layouts” in other parts of the world, where prototypical operation is often a key aspect of the show). And, at least most of the time, they don’t represent a single railroad, place or time, as each modeler builds a module to their own design and a steam-era freight may operate through a module depicting a contemporary container terminal (there are modular layouts that are more consistently designed). But these show layouts are mainly an excuse to show off the individual modules (which represent some modeler’s skill) in a way that is interesting to observers. It’s not the layout that’s the model, but the individual parts.

While “the rules” aren’t absolutes, they do seem to be very widely honored, and when they’re violated it is typically in ways that are less obvious (and likely because some other modeling goal outweighed that aspect in the designers mind).


Japan


Japan turns most of the “rules” on their heads.

First, Japanese manufacturers mix 1:160 scale Shinkansen and 1:150 scale narrow-gauge trains on “N-Scale” track, which is prototypical for the Shinkansen but 26% oversize for typical Japanese narrow-gauge (worse than the 20% difference of On30 vs On3). And layouts often mix both kinds of trains. Although this could be seen as a kind of forced perspective, it’s not usually done in a manner that provides that result. While you could choose to model one or the other, most contemporary cities of any size have both, and in particular a Shinkansen in isolation would be unprototypical unless all you modeled was a rural stretch between major stations. So as a modeler of Japanese trains, unless you want to scratchbuild, you’re hostage to the products available from the manufacturers.

In Japan, model railroaders often don’t have space for a permanent layout, and will use sectional track (Kato Unitrack or Tomix FineTrack, for example) assembled on the floor, with minimal or no scenery (perhaps just a few ready-made buildings to set the scene). There’s no attempt at “realism” in the scenery or track. Imagination plays a greater role here, but they’re still operating trains in and out of stations, or yards, or both, and often more than one train at a time. That’s not to say all Japanese layouts are like this; there are permanent layouts in Japan, but they appear to be the exception rather than the norm.

And then there are the people who have no layout at all, and take their trains to hobby shops, clubs and even restaurants where a model railroad is available to rent. These often consist of multiple loops of track, which can be rented separately, so multiple operators can run trains without interfering with each other, and the whole gives the appearance of a large city with multiple rail lines. An individual may make station stops, and even obey signals (if the layout has them), but isn’t doing much more than running in a closed loop. There’s no “prototype” per se, unless you count Japan as a whole as the prototype. And there’s nothing that coordinates the entire layout to be representative of any specific time or place.


Modeling vs. Play


So where, if anywhere, do you draw a line between modeling and play? I’ll claim it’s intent. If all you’re doing is running a train fresh from the box in circles at unprototypical speeds, maybe stopping occasionally as the whim strikes, or changing direction, then you’re playing with trains. And we all do it sometimes. But if you’re trying to recreate some aspect of a real-world railroad, then you’re operating a model railroad layout, even if you gloss over the aspects that aren’t pertinent to you.

Any model is a simplification of the real world, that’s the essential meaning of the world “model”. Each modeler is going to pick and choose which aspect(s) to replicate most closely, and those choices will be different ones. Some people may use written train orders and schedules, others may just pick up cars or stop when they arrive at the station. Maybe you don’t care about scale speed, but you’d never use the same track for both a Yamanote commuter train and a suburban commuter train. If you try to replicate the real world somehow, through detailed scenery or trains, or adherence to some subset of real-world train operations (e.g., not suddenly reversing direction because you missed a station stop, which is generally prohibited on most railroads), then I’ll claim that you’re doing more than simply playing.

Scenery is one of those aspects. A space-constrained modeler working on a bare wooden floor with “from the box” trains, but strictly adhering to a schedule, or a simplified operating rulebook, is just as much a modeler as someone who’s running highly detailed trains in meaningless circles through beautiful lovingly-crafted scenery. They just have different focuses.

We can, and should, honor those rare modelers who do it all (accurately detailed models operating to strict rules through exacting scenery). But you’re certainly capable of being a model railroader if your aspirations, or your skills, don’t run that high. And even those “do it all” modelers probably started with a subset at one time. Fully-formed model railroads do not spring from the forehead of Zeus, or even Bill Walthers. They’re created through an evolutionary process, with earlier ones either discarded or added-to over time as the modeler’s skills and knowledge improve, and interests change.


Sumida Crossing’s Design


And this brings me back to my own layout. When I bought my first Japanese train, the Yamanote E231, all I had was a simple oval of track with two stations. This evolved into the original Kitchen Table Layout, which had a Woodland Scenics grass mat and some pre-built Kato buildings for scenery (and the throttle plunked down in the middle of the oval, between buildings). This wasn’t model railroading, although it was a lot of fun.

But I am a model railroader, or at least I consider myself one. And I wanted something more. But I also have somewhat limited space, and much more limited time, and want to have something other than bare plywood and flex track for the next decade. I’m too impatient for that.

So I decided to tear down my old HO layout in the basement, and reuse the space (but none of the materials) for a new N-scale layout intended to let me run trains, in a somewhat prototypical manner, in scenery that called to mind Tōkyō, or at least the Tōkyō I know from books, online information and photos (I’ve never actually been there).

My intent was to create a “sceniced” model railroad (rather than the bare plywood foundation of one) quickly, then go back and add details to different parts over time, rather than trying to bring a small area to a highly-detailed “final” state before moving on. And to use Unitrack, rather than ballasted flex-track or something similar, both for speed of construction and so I could take the track apart when re-working the scenery. It also makes it easier to move the layout if necessary, since I don’t need to carefully engineer all the “edge of board” joints, yet I can still break the layout up into 2’x4’ “tables” that will fit in my car.

Following the typical Japanese minimalist approach to scenery, I decided to start with simple scenery. The “water” parts are just the underlying wood, painted with semi-gloss paint. Maybe I’ll add something more dimensional, like cast resin, someday, but not until after I finish the final versions of the bridges. And I’ve made use of “temporary” scenery: the roads are simply gray paint over the underlying foam, someday to be replaced with “real” roads with curbs, and painted lines, parking spaces and sidewalks. Two automobile bridges are simple mock-ups of cardboard and foam-core, and the expressway deck is another foam-core mock up. Someday both the bridges and the expressway will be replaced with scratch-built models, but that’s not a priority.

The landscape is mostly carved foam (and a bit of plaster cloth) painted with flat interior latex (someday I’ll add scatter to represent grass and underbrush, but it’s not a priority). It provides a three-dimensional shape to the city (Tōkyō is built on a plain, and much of it is created land around the edges of Tōkyō Bay, but it’s still not a flat city). Structures are mostly pre-built Kato and Tomix ones, somewhat accurate for 1:150 scale, and representative of Japanese buildings, but lacking a great deal of detail that needs to be applied (interiors, signs, repainting the exteriors in “flat” colors). And while the track is all sectional Unitrack, it’s laid on the same kind of roadbed that would be used for flex track, and perhaps someday I’ll replace the Unitrack with ballasted flex.

The track is just an oval, or rather three sets of double-track ovals. One reason for this is that I lack space for more than two stations, and a point-to-point layout with only two stations is a bit more minimalist that I want to be. This way, I can use the layout to represent multiple lines with a succession of stations that the trains proceed through, which to my way of thinking is more prototypical than a linear out-and-back two-station layout would be, at least for the urban 10-car trains I plan to model.

As for operations, right now the “commuter line” is still incomplete, and I mostly just run trains in circles, although I do attempt to do so at prototypical speeds, and I don’t mix Shinkansen with non-Shinkansen on the same loop. I even try to keep my Chūō-Sōbu trains (East-West lines) separate from my Tōhoku (North) or Tōkaidō (South-West) ones, although I’m less successful at that. Once the commuter line is finished, I plan to try operating to schedule with a Fast Clock, with trains interrunning from the commuter line to the subway line (prototypical for some lines in Tōkyō, although I really need some specialized train models I don’t have to do that correctly).

I have a model railroad. It’s not sophisticated. It’s not highly detailed. And it has the usual scale and gauge problems inherent in Japanese N-Scale. But after a year and a bit of construction I don’t have to look at plywood anywhere other than the deliberately unsceniced return loop at one end. And my Shinkansen can run on a separate oval from the narrow-gauge trains, even if this does mean I’m running 1:160 scale trains next to 1:150 scale. I have something that gives the sense of contemporary urban Tōkyō I wanted, and which allows me to have a number of trains operating in proximity without the “spaghetti-bowl” of track often seen on space-constrained Japanese layouts.

All things considered, I’m pretty happy with what I’ve achieved so far. I’m also looking forward to the day I do get to make real roads, to finish the bridges, or to add ground foam, trees and other little details (not to mention cars and pedestrians; the city is a bit empty at present). A model railroad isn’t a fixed result, it’s a dynamically evolving object. And we aren’t all John Allen, much as we might like to be. Put a bit tritely, it’s the journey that matters, not the destination. And I think we can learn a bit from the Japanese about what that journey can look like while still being a model railroad.

Other website changes:
- Added my new E259 to the Roster page.
- Updated my Kato DE10 page to document the DCC decoder install. There’s also a thread on the JNS Forum about this.
- Added photos from the DE10 DCC install to the Decoders Photo Album.
- I’ve separated a bunch of the information about model train motors, power and DCC decoders out of the “Electrical” section (which is now more narrowly focused on the layout electrical systems) into its own “Model Train” section.