Designing my train layout - 6 - Why it works for me

In my last posting, I explained the process that led to my final design. In this posting I'll try to explain why the design works for me.

The first aspect of this layout that I feel happy about is that of staging. Let me digress a bit and show you an old Marklin layout that I built back in 2003:

This layout was also essentially an oval but the key difference compared to the current design is that I had placed some bare staging tracks at the back of the layout separated by a view blocker (1/4" plywood). The intent was that viewers would stand at the front of the layout and watch trains stop at or pass through the station. While operating this layout, I realized that I could have made a second station instead of the bare staging tracks.

In the current layout, one possible style of operation would be to simulate time-table driven movements through a station. In this mode of operation, each station can act as the staging area for the other station.

In the diagram shown above, each green arrow represents a train. As you can see, it is possible to have as many as eight trains staged around station 1 in order to run a timetable sequence of trains passing through station 2. Of course, in a similar manner, I can use the station 2 as staging for running traffic through station 1. A word about the term "timetable": Unlike the US where most trains on freight railroads are extras, in India, Most trains are actually listed on the timetable.

In addition to timetable driven operations, I also plan to use the layout as two bidirectionally signaled single lines. This style of operation is often used in places where traffic density is higher than can be supported by dedicated double lines. This happens near big cities and on hill sections.

Finally, I have made provisions for four storage sidings for trains not used in the main sequence of operation. Since I tend to be very careful in my rolling stock purchases, I don't thing I will exceed this relatively modest amount of storage space for a while.

Since there is a considerable amount of available space in the center of the layout, I could add storage sidings there in the future if I need the space.

So this concludes the postings on the design of the layout. Next, I will cover the construction as well as some tips and tricks that I have learned in the process.

Designing my train layout - 5 - double track

While leafing through the Armstrong track planning book, I came across this diagram that changed my thinking drastically. The thing that caught my eye was this text from one of the captions: "In model practice, where mainline length is severely limited, two and three track railroading may actually surpass single-track operation in variety"."

Most of the trunk lines in India are double track (with the Chennai-Mumbai section being the only important route with a lot of single track). Upon reflection, I realized that what I wanted was a lot of parallel action. Furthermore, the station configurations could be made more realistic. I then decided to give double track a try. I started by taking a fresh look at the "water wings" configuration from a double track perspective:

The results were surprisingly good. I was able to fit two stations that each had a single passing loop shared between the two tracks. I have, in fact, seen this layout in some small stations on double line sections in India:

Although promising, there were two issues:

• Lack of visual separation between the stations. The two stations are physically close together even though there is a nice long run between them. The only obvious solution was an oval. Now, I had initially rejected ovals because I though they looked boring. However, it was clear that a good old oval is what would work best for me.


• I came to the conclusion that I wanted the stations to have different configurations. Specifically, I wanted a conventional double station as shown in the next diagram. Although configuration B would be more prototypical and offer greater operational flexibility, I was willing to settle for configuration A. The point is - it looks like a regular double track station.
  

At this point, I showed the last design to my friend Dale Schultz who has built a large Marklin layout. Like myself, Dale is interested in using computers to automate train operations. He is a very sharp thinker whose judgement I greatly respect. He told me I was on the right track but suggested that I reduce the number of sharp curves. The next design I came up with was this:

This was now getting interesting. Notice that the lower station remains curved. This was intentional - I had come to feel that at least one of the stations ought to be curved for aesthetic reasons. The two problems were that the passing sidings in the upper station were too short and there was only one cross over and it was too far removed from the nearest station. The next step was to integrate the crossover with the upper station, thus:

This was definitely an improvement. There was a pleasing curve to the upper station and both stations had passing sidings of acceptable length. Dale suggested investigating curved turnouts. I looked through the Walthers catalog and found that the Shinohara #7 curved turnout would be perfect for my needs. This turnout has radii of 24 and 2.8.5 for the diverging and normal routes. This is the next design I did incorporating curved turnouts for the crossovers.

I felt that I was closing on the final design. The last tweak to the design was to confine all the curved turnouts to one side of the layout - mainly for aesthetic reasons.

And that did the trick. I was now fully satisfied with this design - The full process had taken about seven weeks but it was time well spent. A 3D view is shown below that includes the benchwork. I am deliberately not using the space in the center of the layout for the time being. I may eventually come up with some use for the space but it will be left vacant for the time being.

In my next posting, I will explain the operational possibilities of this layout.

Designing my train layout - 4 - more single track designs

With the "folded dogbone" eliminated from consideration, I took a look at some alternatives, starting with the "water wing" configuration. The following is an initial sketch to explore this configuration and it seemed promising:

After some thought, I was able to come up with the following design. As you can see, I was actually able to fit three stations into this design. I should note that the third station (on the straight section at the top of the layout) was really more in the nature of a what-if exploration. I was still thinking in terms of two stations, the main ones being on the ends of the two lobes. The two main stations are well separated and the passing loops are surprisingly long.

The problem with this design was that the stations could simply not accommodate more than one passing loop due to the curves involved. also, there were no good locations for storage sidings of reasonable length. My other concern was the S-curves at the entrance to the passing loops. In hindsight, I realize that I ought to have looked into the possibilities of using curved turnouts. In any case, this configuration just did not look right to me.

At this point, I decided to relax some of my requirements and explore more exotic configurations. I present a few of these below

None of them seemed to be very appealing. At this point, I decided to take a few days off from 3rd Planit and re-read John Armstrong's classic treatise on layout design.

As I shall explain in my next posting, this break turned out to be a good idea. and I was able to break my mental log jam.

Designing my train layout - 3 - Folded dog bone

Building on my experience with my previous layout, I decided to start by investigating single track main line designs. Basically, what I wanted was not much more than this schematic:

In other words, an oval with two stations. I wanted a reasonable amount of separation between the two stations. I wanted the train to have at least a short run between the stations. This schematic may seem very simple but it should be noted that complex behavior can arise from deceptively simple systems. Consider for example the difference between a simple and double pendulum. While a simple pendulum can be fully described by equations for simple harmonic motion, the tip of a double pendulum can move chaotically. In an analogous manner, even with just two stations, it is possible to create complex patterns of train movements. Having decided on a two station configuration, the next step was to see how long a main line I could fit within the space available available.

I prefer to do my design work directly on the computer using the 3rd PlanIt layout CAD software from trackplanning.com. I have been using this package for many years, initially using to design sectional Märklin layouts. I think that 3rd PlanIt is a very powerful package and there is indeed quite a bit of a learning curve. However, once you get the hang of it, 3rd PlanIt just seems to get out of the way and lets you explore designs very quickly.

The first general shape I investigated was a folded dog-bone. I have seen this shape used in many track plans and tried to get it to fit in the space available.

The nice thing about this shape is that there are good locations for locating the stations allowing decent separation between them. However, the main problem is that the curves at the end of the "bulbs" necessarily have to be very tight. Nothing more than 20 inch radius curves can fit in the space available. another problem is that access to the interior of the layout would be problematic.

In order to use curves of minimum 24 inches radius, I discovered that there were two options, neither of which were attractive.

1. Overlap the bulbs: This allows two large radius bulbs to be accommodated. However, as previously mentioned, I was not keen on a multi-layer layout. A second problem is that by having the bulbs on different levels, the main line and the stations must necessarily be on an incline.

2. Widen the layout: Widening the layout allowed the use of 24 inch radius curves but the increased width unacceptably cut into walkway around the layout. One more problem that I discovered was that the station passing loops could not be made very long. I don't really want very long trains but this seemed a bit ridiculous - barely 41 inches for the station on the top.

Clearly a folded dogbone was not going to work for me. In my next posting, I will describe some exotic variants of the folded dogbone that I tried.

Designing my train layout - 2

One more digression before I begin describing the design process for my current layout. Every layout that I have built so far (five in all) has incorporated lessons learned from previous layouts and the current one is no exception. A schematic diagram of my old layout is included below.

When thinking about layouts from an operational viewpoint, I find it useful to think in terms of schematic diagrams (which are similar to electronic circuit diagrams) rather than actual track diagrams.A video of this layout in operation under computer control is included below:

The focus of this layout was a station on a single line (somewhere in Germany near the Swiss border since all the rolling stock was either Swiss or German!). There were two storage yards and trains would proceed back and forth between the yards through the station. From an operational view point, this allowed me to model meets and overtakes.

Some of the problems (and lessons learned are describe below):

• Since the layout featured only a single station, there was really only one viewing position. After a while, this became a bit boring since watching the storage loops was not much fun. This led to the requirement that the next layout would have at least two stations.
• It became very tedious to sort out problems in the lower level storage yard since access was so limited. This in turn was due to the fact that I did not have the room for a helix to connect the levels. I therefore decided to avoid multiple levels in the next layout.
• Unused freight yard: I created a freight yard for the station because I wanted to combine some switching action along with mainline train movement. The intention was that freight trains would pick up and drop off cars at the station. It was hard enough to get reliable automatic operations on the main line so this little freight yard was never utilized. Automating a layout is enough of a challenge and in any case, I did not really enjoy switching very much anyway.
• Inadequate storage for unused trains: This is a perennial problem with railway modelers. There was simply no place to add any extra storage sidings on the old layout. I therefore decided to incorporate additional storage sidings in the next layout.
• Automating a layout requires a large number of sensors since the computer needs a good understanding of the location of a train. There were not enough sensors on the old layout and the sensors were not well located. These problems arose since I had not really designed my layout with automation in mind. It became clear that sensor locations have to be considered before designing the layout.

All of these lessons have been incorporated in the current layout to various degrees.

Designing my train layout - 1

As a child I grew up with model trains. I was given my first train set when I was just two years old. Later, at the age of eight, my parents bought me a great Triang-Hornby train set when passing through England. Modeling activities stopped in school and college. While attending graduate school at Syracuse University, I bought a small loop of N-gauge track and used to run a train around my computer, much to the amusement of fellow students.

When I resumed active railway modeling in 1992, I started with Märklin, a well known German model train company. One of my best friends had Märklin train set and I always wanted those trains. Photos of my old Märklin layouts can be seen here and here. For a variety of reasons (chronicled in more detail in this article on my web site), I decided on a new direction.

I finally dismantled my Märklin layout towards the end of 2007 and cleared the train area in my basement for a new layout. As such, I had a clean sheet of paper to work with. My layout design was constrained by a few factors:

1. Space: My old layout was just 4 feet by 10 feet. I felt that there was some scope for expansion. After careful measurement and negotiations with my wife, I determined that the actual space available for the layout was was 14 feet by 7 feet. However, there is a furnace close to the layout and I need to ensure that there is enough room around the furnace. Hence, I lost a triangular a nibble from this space.
   Now, 98 square feet is not much compared to the large spaces available to the typical American modeler. However, I felt that even this amount of space would be a bit of a challenge. Larger layouts are more costly both to build and to maintain. Its all too easy to bite off more than you can chew in this hobby! The shape of the available space is shown below. The available space is the area in white with thick borders. The diagonal section on the left side of the diagram is caused by the need to leave enough room around the furnace. Another issue is that this has to be a free-standing layout since I need access to things stored around the room.
2. Operating style: This played a very important role in shaping the design of the layout. The one clear requirement in my mind for this layout is that it must support fully automatic computerized operation. Consequently two sub-requirements became clear:
   a. A continuous run is absolutely essential. A point-to-point layout is simply not suitable for automated operation given the need to switch locomotives around their trains.
   b. There need not be any provision for local switching in industries and freight yards as is common in most American layouts.
3. Carpentry skills: I am not much of a carpenter - I just don't have a very steady hand and am a bit impatient. As a consequence, I wanted the benchwork to be as simple as possible. A major consequence of this constraint is that I decided to go in for a single level layout. Apart from being much easier to build, its much easier to debug computer control programs if the trains are visible at all times.
4. Scale is HO (1:87.1): This is because I have a lot of European outline rolling stock in HO that I would like to run on this layout. Also, a large variety of HO scale equipment is available from manufacturers around the world.
5. Minimum radius for the mainline must be at least 24 inches: This would allow the operation of most modern equipment although even this radius is too small for longer cars like the Amtrak Superliner car which is 85 feet long for example (nearly a foot in HO scale). There will be a fair bit of overhang for the longer cars on 24 inch curves but at least, they don't derail.

After spending a week or so considering these constraints, I updated my copy of 3rd PlanIt layout design software and got to work. In the next posting in this series, I'll describe the actual design process.

Productivity

A few years ago, I read an interesting little article by Kenneth Rogoff, the well known professor of economics at Harvard, called "Artificial Intelligence and Globalization" in which he says:

But I wonder whether, even within the next few decades, another factor will influence our work lives even more: the exponential rise of applications of artificial intelligence.

At the time I read this, I found it funny. I have worked for many years in AI and related fields. We have an inside joke: "if it works, its not AI". Now I am less sure.

I have recently started watching an interesting show called How Its Made on the Science channel. As it's name suggests, this show is all about manufacturing technologies. I was struck by the high level of automation in many of the manufacturing processes that they present. Clearly the level of productivity must be high in these automated factories.

This then led to consider other examples of highly productive organizations:

1. Craigslist: Serving millions of daily requests with just 30 employees.

2. The UPS Worldport sorting system in Louisville, Kentucky: check out this this amazing video about how they handle vast numbers of packages with just a few employees to keep watch.

3. Google's data centers: Consider the vast number of machines in one of their data centers supervised by a handful of employees.

4. Plentyoffish: This is a very popular match making web site that was created and run (for the most part) by just one person. Apparently, the founder just hired the first employee recently (only after the company had grown amazingly large).

I don't think I am alone in beginning to wonder where this will lead us. I just came across an article in the Economist web site in which he author makes some interesting points. In a nutshell, the author says that jobs are ever more specialized requiring a lot more training. And even when trained, the skills become obsolete ever more quickly.

I would add a couple of points to the Economist blog posting:

• As one gets older, retraining gets harder. This is partly because distractions (house children, finances etc) increase as one gets older. Another factor is that many people are resistant to learning things from people younger than themselves.
• Another source of concern is that many of the new job opportunities are in the discretionary side of the economy. I think that jobs based on discretionary expenditure are much more vulnerable to downturns.