Planning and Designing a Model Railroad

Designing a model railroad can be a very daunting task for some people. But, it doesn’t have to be.

With the software available today, complex model railroads can be designed and checked, before you take the next step.

This eliminates the frustration and most design mistakes.

Markus Robinson is a 57 year old engineer living in northeastern Pennsylvania. When he was a kid, his Dad had a huge Lionel set (which he sold without discussion with any of the kids).

Markus has grandchildren who love trains and is at the design stage of his model train layout. He has kindly shared his progress with us and would love some feedback…

Tinplate Central Lines O-Gauge Layout Design Process

By Markus F Robinson

Layout History

The layout is my modification of John Long’s Classic Toy Trains Magazine November 1999 Tinplate Central Lines layout which in turn was based on April 1948 Model Builder article about the Pittsburgh Springer family layout.


The original April 1948 Springer-family version of the layout.



John Long’s 1999 O-27 implementation which was the starting point for my version.


John Long’s version (despite his claim that it is an O-31 layout) “sized to fit on four standard 4×8 sheets of plywood” (i.e. 16’ x 8’) is an O-27 implementation. The 0-31 equivalent requires at the minimum 17’ 4 1/2″ x 8′ 9″. Long tweaked the Springer family design (as discussed later) to create a layout “… ideal for those who like plenty of action … [that] would take a while for an operator to become bored with…” Long’s version “make[s] good use of 1950s-era Lionel components and accessories.”

I’m trying for a hybrid prototypical dealer-display-like layout with lots of track.

The current state of my Tinplate Central layout, version 6.8, is getting pretty close to implementation-ready.


Version 6.8 of Markus Robinson’s Tinplate Central O-Gauge Layout


The Tinplate Central Route-Map

It is difficult to appreciate the implications of a convoluted layout. So to be able to study what the layout actually allows one to do, I created a ‘flattened / straightened’ route-map version which does the same thing with fewer curves and no overpasses. Making use of several trestles the Tinplate Central layout above is a folded, twisted version of the route-map below.

The route-map captures the layout’s five main route options. Four trains can operate concurrently and independently on Routes 1 and 2, or on routes 3 and 4.

The route-map has the following characteristics:

  • John Long’s modification of the original Springer family layout adding switches 2 & 14 creating Route 2 and two independent routes [1 and 2] each supporting concurrent operation of two trains.



Markus Robinson’s Tinplate Central version Route-map (v2.0)


  • Markus’ addition of switches 16 and 17 results in two additional alternate routes [3 and 4] that together also support the automatic concurrent operation of four trains; one on route 3, and three on route 4.
  • Markus’ addition of a fourth spur line. The spur lines [collectively Route 5] offer switching, engine and rolling stock storage and interactive accessory opportunities.
  • Route 1 [Green] is a point to point route with a small loop at one end and a large loop at the other. As a train A enters the large loop it releases train B from the other end of the large loop, while train A continues to the other end of the large loop and stops.
  • Route 2 [Blue] is an oval route with a passing siding which allows two trains to operate concurrently by coming to a stop between switches 3 and 15.
  • Route 3 [Red] is an oval supporting a single train.
  • Route 4 [Magenta] is a hybrid comprised of Route 2 and the small loop from Route 1. It can support three trains by taking advantage of the Small Loop and the Passing Siding.


Overview of the Design Process

A brief overview of the design process for Tinplate Central follows.

  1. Because I am a neophyte layout designer, I looked through books and on the web for attractive layouts, subscribed to Classic Toy Trains, and joined Dan Morgan’s modeltrainsforbeginners forum. 
  2. I determined the maximum size layout my house would support.
  3. I determined the curve constraints posed by my rolling stock and engines.
  4. I began to tweak a design I liked using a non-model-train-specific computer CAD tool. In the process I developed two sets of schematics; the first a set of iterations of the convoluted actual layout with overpasses and interesting curves (see diagram above – Version 6.8 of Markus Robinson’s Tinplate Central O-Gauge Layout), and the second a flattened/straightened ‘signal control’ route-map version (see diagram above – Markus Robinson’s Tinplate Central version Route-map (v2.0)) which allowed me to truly appreciate the options related to interesting train operation and multiple, concurrent, independent running. (For anything but the simplest designs, such a route-map version of the design is essential to making intelligent tweaks to the initial design, to optimize the interest factor, and to maximize the long-term interest in the eventual layout.)
  5. I began to select accessories I found attractive from among suppliers.
  6. I endlessly mined the Internet for answers to specific questions related to layout design. In the process I began building a custom document capturing all of the stuff I considered important for the good design. A key early decision was which track supplier to use. For instance Atlas-O took itself out of the running because it doesn’t support an O-31 or O-32 curve. (Their smallest is O-36 “for reasons of compatibility” according to their catalog.) I ended up going with GarGraves. Another example of my search revolved around how prototypical I wanted my layout to look. I’m moving towards what might be considered a hybrid prototypical dealer-display layout with lots of track.
  7. I spent $115 purchasing the WinRail model-train-specific CAD tool. (Once I decided to do a serious layout, it became the most important $115 I’ve spent on the hobby.) Personally I think anyone that designs a layout without using a train-layout-specific CAD tool needs his head examined. Features like clearance calculation, Flex track optimization, short circuit detection, gradient determination etc., etc., etc. offered by a good tool like WinRail are ESSENTIAL to good design.)

I’ve finalized basic layout (i.e. route options, number of turnouts, number of signal blocks, exact size of benchwork and location of Access Traps). But I still have a ways to go to get to a really good layout. I still need to

  • Tweak track curvature.
  • Finalize locations of things like decouplers, signal blocks, etc.
  • Understand the layout’s true power requirements.
  • Assure optimum layout on the fixed benchwork.
  • Really understand the features required for a DCC (command control) implementation, or a more classic “block-based” implementation, and deciding how far to go in either direction.


Design Constraints

The starting point for my design process was to identify an existing design that I liked and then establish my design constraints.

  • Maximum Size: 18′ 4″ ‘ x 8’ 11″(with a notch in upper left corner to accommodate an existing chimney). This maximum size was based upon the constraints of the room where the layout will be built.
  • Curve Requirement: O-31 based upon my existing engines and rolling stock.
  • Overpass Clearance & Grade: Assuming a 6 1/2″ overpass clearance, this has lead to a maximum layout grade of approximately 3.3%. This ‘Mountain Grade’ is a function of the exterior elevation change run of 222.3″ (2.9% grade) while the interior elevation change run is 214.2″, (3.03% grade) without considering curve impact which increases the grade.
  • Track Clearance: My starting point was initial ‘safe’ clearances of 4” for straight track and 6” for curves. However I was able to substantially optimize these requirements using the WinRail CAD tool to get much tighter clearances. Clearance optimization is important for any layout that is space constrained.
  • Train Lengths: Having an idea of what is going to run on the layout is important for getting a feel for how trains will look once the layout is built. So I “built” three trains; a 6′ 2 3/4″ long passenger train driven by my 4-6-6T Tank Engine, a 10′ 11 3/8″ freight train pulled by my 4-6-4 Hudson, and a 4′ 7/8″ maintenance train powered by my 4-6-0 Camelback. By creating scale images of these trains one can get a feel for how they will look running in the layout.

Evolution of the Design

Here are some “snapshots” of key stages in the design process that have gotten me to the current version of the design.

  • Tinplate Central Lines v1.0: My O-31 capture of John Long’s O-27 Tinplate Central layout (see diagram above – John Long’s 1999 O-27 implementation which was the starting point for my version.) was the starting point for my design work. Carefully studying his implementation, his measurements, and actually laying it out using a non-train-layout specific CAD tool allowed me to discover that while he claimed the layout was O-31, actually it used O-27 curves which would derail my running stock. This first version also allowed me to get a basic feel for how large the benchwork would need to be to implement an O-31 version of Tinplate Central.
  • Tinplate Central Lines v2.0: This is a O-31 implementation of John Long’s O-27 Tinplate Central layout but also showing my planned addition to create two new independent routes. John Long’s concurrent Routes 1 and 2 are shown. I laid this version out limiting myself to O-31 curves and straight track to create a “Manhattan Geometry” version of the layout (only horizontal, vertical, and 45-degree diagonal straight sections.) It was way too early to start messing with optimized curves, diagonals and the opportunities offered by Flex Track. Having studied John Long’s implementation and created version 1.0 of the Route-map shown in the diagram above – Markus Robinson’s Tinplate Central version Route-map (v2.0), I realized that with the addition of just a couple more turnouts and a short length of track I could double the interest factor of the overall design.


Tinplate Central version 2.0 showing routes 1 & 2 and my planned addition to create routes 3 and 4


  • Tinplate Central Lines v3.0: This version shows the layout with the new independent routes 3 and 4 as well as raised terrain, tunnel placement, elevated track shown in YELLOW and gradient change track shown in RED, rolling stock and placement of some Lionel accessories and water features. Concurrent Routes 3 and 4 are shown. The idea at this stage of the design process was to understand how the terrain would have to look to accommodate the required changes in gradient. I also needed to give some thought to how to disguise the access traps in the middle of the layout so that any portion of the layout would be accessible by a reach of no more than three feet. Disguising an access trap as a water feature is the simplest solution. At this point preliminary calculation indicated that the layout required mountain grades that with their curves probably exceeded 4%.

    Since I’m interested in creating a prototypical layout, by placing some accessories on the layout, I could get a feel for how cluttered and realistic the layout might look. Basically this layout lends itself to six different model areas; a rural farm scene on the plateau in the upper left, a forest scene on the center plateau, a manufacturing hub in the lower left, a main station lower center, a freight/repair yard on the right hand side of the layout, and clustered around the left-side lake is a hobo encampment. Fine tuning modeling for those areas can happen later.


Tinplate Central version 3.0


  • Tinplate Central Lines v4.0A: This image shows the layout without elevated terrain, but with switch numbers, decoupler locations, locations of tunnels and cut, roads, and additional accessories. This version still uses John Long’s placement of access traps.


Tinplate Central version 4.0A identifying turnouts, decouplers, and accessory options.


  • Tinplate Central Lines v4.0C: The image below shows the locations of control blocks (for non-DCC implementation and/or accessory signal control) as well as a study of 3′ reach limit from the periphery and from the edges of John Long’s placement of the two 2’x2′ access traps. I saw that the right-side trap needed to be shifted right to improve accessibility to this crowded part of the layout. (See diagram below). With the addition of another spur line, I will need to add a seventeenth control block.



Tinplate Central version 4.0C showing the sixteen control blocks and illustrating that there are reach challenges accessing the right-hand side of the layout.


  • Tinplate Central Lines v5.0: Based upon a suggestion from of one of Dan’s modeltrainsforbeginners forum reader’s in response to a forum post of mine about prototypical track options, the image below shows a partial attempt to implement the layout using Atlas-O O-36 curves which is Atlas-O’s smallest curved track (other than their O-27 curves which will not accommodate my rolling stock.) Unfortunately, O-36 curves throughout were too large for my benchwork constraints.

    What was great about this attempt though was that the Atlas-O website had a link to the freeware version of the WinRail software with the Atlas-O track library loaded, so I got exposure to this CAD software. In this version I’m beginning to tinker with optimized Access Trap locations. Note the use of the WinRail CAD tool ‘clearance’ feature to optimize placement of track. My largest piece of rolling stock is a 15″ passenger car with truck pivots 9.8″ apart. I assumed the standard 3.1″ maximum width of an O-gauge piece of equipment. This data is used by WinRail to calculate required clearance. The result is that instead of using conservative estimates of 4″ clearance on straight track and 6″ clearance (center to center) on curves, I can precisely calculate the clearance required.


Abortive partial implementation of Tinplate Central version 5.0 using O-36 curves



Tinplate Central v6.0 Benchwork laid out using the WinRail CAD tool


  • Tinplate Central Lines v6.0: Having purchased the deluxe version of the WinRail CAD software with the library compiler feature, I begin by defining the benchwork. This isn’t a really interesting image, because it just shows the benchwork. However, it illustrates an important concept when doing design. I always capture a ‘snapshot’ of the current version of the work so that if I screw up as I’m tweaking the design, I can revert to an earlier version and restart from there. In this case I wanted an ’empty’ design just showing the benchwork. Note the position of the right access trap, relative to two of the nine pieces of benchwork I will require. In a later version I further shift this access trap to the right.
  • Tinplate Central Lines v6.1: One great feature of the Winrail CAD tool is that it has a HUGE library of supported track and accessories. It has a complete library for the GarGraves Phantom O track which was the other suggestion for good prototypical track Dan’s forum reader made to me. The layout is not complete because the ‘overpass’ and gradient sections haven’t been added yet, but I’ve determined that I have enough space to add the 4th spur line, and I’m placing yellow blocks to represent some of the accessories to make sure my optimized track layout will have room for them.
  • Tinplate Central Lines v6.2: With the track all laid out, I’ve optimized the placement of the right access trap to improve the reach off to the right of the image at the top of the next page. I’ve also turned on the WinRail short-circuit feature. Once I use WinRail to implement the layout’s blocks specified in version 4.0C on page seven these short-circuit problems should be addressed.


Tinplate Central version 6.1


  • Tinplate Central Lines v6.4: Here I’m working on two important tasks; 1) maximizing the length of the gradient change portions of the track so that my overall gradients are as mild as possible. Initially my worst grade was a curved 3.9% (probably over 4% once the impact of the gradient curves is taken into account.) Now I’m down to a curved 3.03% on the inner loop which is the steeper of the two. And 2) optimizing curvature so that I fit the largest radius curves possible into the design. I started out with the GarGraves predefined curve set (the smallest I’m using being their O-32), though I would always select the largest curve that would fit in each area. But gradually where appropriate, I’m swapping out these curves for curves that the WinRail CAD tool calculates using Flex Track. This leads to a much more prototypical layout.


Tinplate Central version 6.2 showing potential short-circuits


(Note that it would be much harder during design to just start with all Flex Track because the Flex track optimization tools require so many variables to come up with a solution. By constraining the problem with pieces of straight and curved track that together get the layout “mostly where I want it” later optimization with Flex Track is much, much easier.)

I’ve got a ways to go with this optimization strategy, and I also want to play around with different sized turnouts. Version 6.4 of the layout only uses the GarGraves O-42 remote right and left turnouts. There are places where a Wye turnout or one with a less sharp curve would be better.


Tinplate Central v6.4 optimized to maximally reduce grades.


  • Tinplate Central Lines v6.4 (Visio CAD): In the image below I’m examining the size of the benchwork with respect to the size of the room. It’s a tight fit although the existence of large doors to other parts of the building and to the outside makes it less cramped than it seems. If you look close, you can also see tentative positions of various decouplers. (I could have done this in WinRail, but I already had the layout in the other CAD tool, so I saved myself some time by just importing the image from WinRail.)


Tinplate Central v6.4 verified in the space where it will be built. Big doors on the left, right and bottom make it less cramped than it seems.



Tinplate Central v6.8 with the left-side loops optimized


  • Tinplate Central Lines v6.8: Examining the image above, one sees that swapping an O-72 turnout for an O-42 turnout, tweaking the location of several turnouts, adding some straight track (to allow for a decoupler) and iterating with Flex Track has made the two loops on the left side of the layout cleaner and addressed some clearance issues. (I tried multiple different options before I got something that worked. This is why the design iterations bumped up from 6.4 to 6.8.)

Note to readers: If you’ve gotten this far, thanks for your interest. I am most interested in any observations, comments, suggestions, criticism (or just gossip) you may have about this project. You can reach me at or post your comments to Dan’s modeltrainsforbeginners forum. Thank you!

This entry was posted in Model Train News. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *