Shopping Cart
Your Cart is Empty
There was an error with PayPalClick here to try again
CelebrateThank you for your business!You should be receiving an order confirmation from Paypal shortly.Exit Shopping Cart

KCS 3rd Subdivision

A joint project of Steve Davis, Tom Vanden Bosch, and the Tuesday Night Gandy Dancers.

Click here to view a presentation given at the Indian Nations Division of the NMRA on the design considerations and goals for this layout.

KCS 3rd Sub Quick Facts

Railroad: Kansas City Southern

Section modeled: 3rd Subdivision

Location Modeled: Between Heavener, OK and Watts, OK

Scale: HO

Physical Location: Coweta, OK

Era: late spring 1982

Planning started: November 2008 (when current property was purchased)

Construction started: March, 2009

Current status: framing, flooring & drywall complete, track laid Poteau – Sallisaw.

Building size: 40’ x 60’

Layout space: 30’ x 52'

Design: walkaround, single level w/ “mushroom” second level along rear.

Layout height: 45” – 55” from floor (floor is raised in portions)

Minimum mainline curve radius: 33”

Absolute minimum curve radius (branch line, siding): 27”

Maximum grade: 3% (max grade on prototype: 1.5%)

Minimum main line turnout frog: #8

Minimum turnout on layout: #5

Train Control system: Digitrax LocoNet DCC, walkaround duplex radio throttles.

Track: Code 83.

Turnout control: Tortoise slow motion switch machines for all switches connected to the main line, dual control w/ dispatcher lockout, Heavener and Sallisaw Yards also Tortoise to facilitate single-button yard ladder/track control while switching. Other industry and branchline turnouts use Caboose Hobbies ground throw switches except where difficult to reach.

Typical mainline train length: 25 – 30 cars.

Maximum train length: 33 cars (or longer but will not fit in all sidings)

Typical “Dodger” (local) train length: 15-22 cars.

Crew/dispatcher communications: radios (furnished by layout host)

Why the KCS?

I grew up in the San Francisco Bay area and have been an avid SP fan for years. Moving to the heartland, I really loved and admired the Santa Fe. So why KCS? A few reasons:

My dad, a prototype modeler who introduced me to the hobby and is building a lot of the structures and rolling stock that will appear on this layout including the Arkansas River bridge, said “model something you are familiar with, that you can physically scout.”

That left me with many great options (MOP, MKT, ATSF, BN, Rock Island, Frisco, etc.) but many of those are widely modeled. Great KCS layouts do exist, but they seem to be more rare.

I have a certain admiration for the KCS, the smallest independent Class 1 RR in the US, interesting motive power and operating procedures, still growing and going strong!

The KCS runs through a scenically interesting part of eastern Oklahoma, with hills and even mountains, grades which still require DPU to conquer, rivers and bridges, including a landmark bridge across the Arkansas River south of Sallisaw that will be featured prominently on this model.

Interesting and still used branch lines, interchanges and connections.

Why the 3rd Sub?

The 3rd sub runs from Heavener to Watts, the 2nd sub runs from Watts to Pittsburgh, KS. Today these are both consolidated into the “Heavener Sub.”

I liked Heavener yard: the layout, fuel racks, etc.

I liked the vicinity of Heavener and the town and industries there.

The St. Clair Lime Co. operation (at Marble City) is interesting and is still in operation.

The Ft. Smith branch is still in operation and great to model, and it was operated by the BN at that time w/ KCS trackage rights. The BN is another RR I’m interested in.

Sallisaw had a great yard (removed just a few years ago) and still has an interchange with the MOP (now UP) as well as the MOP diamond.

I had a “Spot Location Inventory Chart” (“SLIC”) from Frank W. Bryan showing all the spurs, sidings and car spot locations for the 3rd sub. We didn’t have that for any other KCS sub.

I couldn’t model the whole Heavener Sub w/o more room or going double deck.

I wanted to model a full subdivision between crew change points if at all possible and include as many of the prototype features, towns and sidings as possible. This was achievable with the 3rd Sub.

Why 1982?

Cabooseless trains on the KCS started shortly after that and I preferred to run cabooses.

In 1982 there was still BN traffic on the Ft. Smith branch.

The MOP still existed as a railroad to interchange with (wasn’t yet absorbed into the UP).

I liked 2nd generation diesels (GP30, SD40) etc and 1st generation diesels. In 1982 KCS still had GP9’s and the occasional F unit! Power was interesting on the KCS during that time w/ some first generation diesels still operating.

The white paint scheme was unique and interesting (but regrettably hard to find – so we’ve been doing a lot of painting!)

Wanted to have more variety of rolling stock, i.e. some more modern cars.

Design Goals

In an earlier presentation I grouped this into “Givens and Druthers”. But thinking about it now, to me it’s all “druthers” except the one given: that being, I have a 40’ x 60’ building which is already divided so as to afford a maximum layout space of 30’ x 60’. So here are my “druthers”:

Linear/walkaround design.

Virtual point-to-point with automatic restaging (i.e. the layout does make a circle but that isn’t obvious – NO TRAINS run in a circle during an operating session). By automatic restaging, I mean for example loaded coal trains, which always ran South on the KCS, are ready to run south again without physical restaging or moving/swapping of trains or loads/empties between operating sessions. This also facilitated running some trains during “open house” sessions for visitors.

HO scale (good compromise for me between large enough to handle easily and small enough to pack in enough operation – plus I had an existing stock of buildings and rolling stock from a previous layout, the BN Avard Sub.)

Trains never out of reach/out of sight while being operated on the subdivision.

Single level was a preference (I’m not that tall and didn’t want to have to stand on a stool a lot or sit on a chair to operate. It also seemed that double deck would be hard for me, not being a club, to maintain).

Sidings and spurs as per prototype (if a siding diverges to the right when traveling north on the prototype, it should diverge to the right when traveling north on the model). I was able to get the spurs as per prototype thanks to Frank’s SLIC charts.

Realistic, interesting and challenging car movements, switching, dispatching, signaling and interaction of trains and yard crews.

Staging yard for pre-made trains and through trains.

Working yard where cars are classified and trains are assembled.

Above all this RR was designed with operation in mind.

Preferred a CTC layout.

Wanted a “sincere” design (where trains don’t reverse direction or travel through the same scene twice).

Operators should be able to follow their train throughout the run between crew change points without having to stop, walk around the layout or get out of sight of their trains.

Train direction must be consistent (on the KCS 3rd, the operator’s left, when facing the track, is always North – operators looking at the layout are always looking East).

Prefer accessible (non-hidden) staging where space allows.

Limit reach to any tracks that might be switched to no more than 30” depth – 24” preferred.

Ability to traverse the subdivision in a prototypical amount of time using a fast clock.

Run the same trains & schedule as the prototype.

Wanted 4’ aisle widths for operators.


Most of the goals above were accomplished, but of course there were some compromises:

I would have preferred longer runs between towns, but just didn’t have the space w/o eliminating too many towns or going to double deck. Runs aren’t long enough to support intermediate signals except in a few cases, so most signals are absolute.

I had to eliminate some interesting towns such as Panama which I would have preferred to include.

With only 420 ft. of mainline run, that’s only 7 scale miles of track. Using a 6:1 fast clock and trains running at an average of 20 mph (including siding dwell times) it will only take 2 RR hours to traverse the sub, not 3 as per the prototype.

Had to compress some aisle widths to 3’ (most are still 4’). I kept 4’ at key passing areas and around the main yard where the most people will congregate and mark up for trains.

To fit the RR and the aisles into the space I couldn’t do a full 100% walkaround. This means for Northbound trains, crews can follow their train only so far and then the train disappears into benchwork and down a helix. To do this, the computer (using RR & Co. automation software) will take over and run the train down the helix and into staging. To make this appear as prototypical as possible and meet my other goals, NB trains stop at the northern crew change point (Watts, OK) and operators relinquish control there (simulating a crew change). Then the computer takes over. Likewise, Southbound trains will be run by computer into Watts and the crew will pick up their train at Watts to run southbound as per the prototype. 

Branch lines to Ft. Smith and the St. Clair Lime Co. at Marble City run separately from but parallel to the KCS mainline. Ideally they should run on a separate piece of benchwork or peninsula.

The MOP is provided for interchange purposes and crosses the KCS via a diamond at Sallisaw as per prototype but is mostly hidden trackage (breaking one of my goals!) and thus computer controlled except when near Sallisaw, when the Sallisaw operator will take over running those trains for interchange and switching purposes.

For operating interest I compressed the 24 hour day into 18 hours (3 real hours using the 6:1 fast clock). The day modeled is thus 5 am – 11 pm, and any trains running after 11 or before 5 on the prototype will run earlier or later so as to run between 5 am and 11 pm.

There was only room for 12 staging tracks, 16 were ideally needed. As a result, 1 of the 3 coal trains was not modeled (but all 3 did not run in each direction every day), no seasonal unit grain extras are modeled, and two trains will have to be staged on the layout (waiting for crews at Watts and Heavener).

There was no passenger service on the KCS in 1982. I wanted some passenger service and wanted to model the Southern Belle. I couldn’t bring myself to run a true KCS Southern Belle behind F units in 1982, so I compromised and postulated that the states of Missouri and Louisiana contributed funds for a state-sponsored Amtrak “Southern Belle Service” using Amtrak FP45 power (typical for the era) and equipment, like the Heartland Flyer (which came later) and other passenger trains that came and went. This is plausible but not prototypical. Also both Belle trains, being relatively short, stage from the same track, and thus must both operate at roughly the same time so that they can pass on the layout.

I kept Heavener Yard track-for-track to the prototype, except, left off one track, and added a couple of crossovers to help keep yard jobs off the main line and facilitate switching since the yard will likely be busier than the prototype.

Not all locals ran on the same day in a given week (though on certain days all but one did). Specifically, on M W F the Ft. Smith Dodger ran, and on Tu-Th the Waldron Dodger ran. I picked Friday for my operating day, when all but the Waldron Dodger ran, and ran the Waldron Dodger on Friday also. I assumed a busy week and extra power available at Heavener that Friday.

Most trains refueled at Heavener and I modeled the fueling pads for this. Due to space on the layout and locating the fueling pads at N. Heavener as per prototype, we may not be able to model refueling of Southbound trains, just Northbounds.

There was a wye at Heavener that I wasn’t able to model due to space constraints. I may yet figure out how to fit that in.

Some of the scene depths (to keep aisle widths above at least 3’ and not cut any peninsulas) are fairly shallow at only 18”. Would have preferred 30” minimum throughout the layout.

The Arkansas River bridge is not curved but the best place for it happened to be a curve so we built it as a series of straight bridges, keeping the individual sections as close to the prototype as possible. The bridge piers were individually cast.