A SHUNTING PUZZLE LAYOUT
Scale speed - Couplers - Setting up the game
Once you have built a shunting puzzle layout with sound baseboard foundations and track laid down in a trackplan arrangement which allows for the rules of your chosen shunting puzzle and provides smooth and reliable running of your stock, you will want to operate it - after all, you have set up the playing board, now you want to have fun actually playing the game. In this case, you might find a few useful or interesting hints here. The actual rules of various shunting puzzles, however, are covered elsewhere (e.g. here for Inglenook Sidings and here for the Timesaver) - this page is all about making things move on a shunting puzzle layout.
Freshly ballasted industrial track at the cardboard factory in Deisswil, Switzerland (note the welded joint between heavier and lighter type of rail) [click for a larger picture]
Track which allows for trouble-free running of locomotives and stock is essential on a shunting puzzle layout
Many good looking model locomotives are made to run way too fast and are sent banging into freight stock or made to whizz them around points (turnouts) in a way which makes derailments almost inevitable.
Class 09 Shunter (U.K.), max. speed: 27mph
ALCO S1 Switcher (USA), max. speed: 60mph
(Pics © Mark Shipman / © All American Trains)
The first and very basic step to reliable (and realistic) operation on a shunting puzzle therefore is to slow down and run trains at scale speeds.
Obviously speed ranges of shunting locomotives vary according to location and era, but shunting moves in yards and on industrial tracks are usually very slow. Sometimes this is reflected in the maximum speed range of the prototype (at 27mph, the British Cl 09 shunter can't do anything else than trundle along even at top speed), but even an Alco S1 switcher would never do anything close to its potential 60mph when actually performing switching duties. "On the job", locomotives sorting out and dropping off freight stock (or even passenger stock) will usually be working at speeds of 5-10mph.
The best way to track the speed of your models is to measure a straight and level stretch of track and then note the time the locomotive takes to cover it. In 00/H0, a model travelling at a scale speed of 5mph (7,5 km/h) will take 12 seconds to cover a distance of 1 foot (30 cm). If you're travelling along at 15mph (23 km/h), it's still 4 seconds. In N scale, a model travels 0.55 inches (1,3 cm) per second at a scale speed of 5mph (7,5 km/h).
Scale speeds usually don't come for free, they require good equipment (both locomotives and track) which is well maintained. As far as track is concerned, good electrical conductivity is a must - if it can be achieved by regular cleaning, that's fine, if not, all electrical contacts need to be improved and maybe additional track feeds will be necessary. Maybe upgrading to a better control unit will be inevitable, perhaps opting for a model which features inertia and maximum (i.e. low) speed settings. If a locomotive still refuses to run slowly even under these perfect conditions, re-motoring and/or re-gearing may be the only alternative to dumping it. However, more and more model shunters are geared to prevent non-scale speeds and come with flywheels to ensure smooth running even when the locomotive is crawling.
Maximum speed of the Swiss Federal Railway's Ee 3/3 shunters is 40km/h (25mph). The gearing of Roco's H0 scale model automatically ensures scale speeds. (Picture is (c) Roco)
In most cases, the fun to be gained from operation is in direct proportion to the amount of coupling and un-coupling involved. Unfortunately, the vast majority of model coupling systems supplied with ready-to-run models seem to assume that once you've coupled two items of rolling stock together, you will never ever want them to part again. No matter whether you're looking at the tension lock coupler (UK), the hook and horn coupler (US) or the drop loop coupler (Europe), none of them is suited if you're thinking of eliminating the famous "hand of god" (more humbly referred to as "Godzilla's paw") and have some "look no hands" coupling and un-coupling. All of these "standard couplers" require most unsightly uncoupling devices which take the form of oversize ramps in the best of cases.
There are many answers to this problem. On a DCC (digital command control) layout, there's the option of having rolling stock equipped with digitally controllable couplers - no uncoupling devices to be installed, and yet stock can be un-coupled anywhere on the layout. On a conventional layout, couplers which separate by means of (electro-)magnetic uncoupling devices have proven their worth in terms of appearance and function over a long period of time. There are many different brands available, but probably the most famous (certainly in the USA) is the Kadee (R) coupler, of which various different models, fitting US, UK and European rolling stock, are available. Once again leading the way in terms of quality of ready-to-run stock, more and more US outline H0 scale models are now being factory-equipped with the maker's own brand of electro-magnetic couplings which are, in general, compatible with Kadee couplers and Kadee uncoupling magnets. Some more recent UK outline models feature NEM-362 coupler pockets which make conversion to Kadee-type couplers a matter of seconds by simply plugging in the corresponding coupler type (Bachmann leads the way in this field). On the European continent, Kadee type couplings ystems are mostyl restricted to US outline modelling, as almost every European manufacturer has develeoped and refinded their own "close-coupling" systems. Most of them also allow for delayed uncoupling (a piece of rolling stock can be moved to any point and dropped off after it is uncoupled without re-coupling as long as the loco is pushing it), which really is quintessential for shunting puzzle operation.
The Kadee (R) coupler - looks good, works well (Picture is (c) Kadee)
Using Kadee couplers with Continental or British outline stock is straightforward if the models have NEM-362 coupler pockets, in which case Kadee has a range of different couplings suitable for specific models (e.g. the Bachmann Cl 08 shunter takes the no.19 coupling [pictured above, middle]). In the case of stock without NEM pockets, some minimal amount of surgery is required (e.g. Hornby freight stock is best equipped with the no. 46 or the no. 36 coupling [pictured above, right]). More information on how these couplers work can be found here.
Just as important as the couplings are, of course, the uncoupling devices. They, like the couplings, should work reliably. Most layouts running stock which is equipped with Kadee-type "magnematic" couplers use uncoupling magnets which are installed between the track running rails.
The magnets don't look all too prototypical, of course, but there are a number of ways to make them less conspicuous (e.g. using them as part of a yard pathway crossing the tracks). Stronger magnets which are installed below the track are available, but they require a lot of advance planing and have a strong tendency to interfer with the running of traisn because their strong magnetic field can casue unwanted side-effects such as rolling stock being drawn to the magnet of its own accord or metal axles and loco metal underframes being "locked on".
A Micro-Trains N scale uncoupling magnet in place on an Inglenook Sidings layout
[click for a larger picture]
If necessary or desired, an Inglenook Sidings layout can be operated with one uncoupling device only. It would, however, be more advisable to use three uncoupling devices (one each at the throat of each siding) if possible
A Timesaver layout requires four uncoupling devices located in specific positions - it won't work with less, and it doesn't need more if couplings are used which allow for delayed uncoupling
The positioning of uncoupling devices is also very crucial for a succesful shunting puzzle layout. There should at least be enough of them to allow for all the necessary uncoupling moves. Some may want to reduce the number of uncoupling devices to a bare minimum and will accept having to sometimes make lengthy and not very protoytpcial shunting moves such as pulling the entire string of rolling stock from its siding in order to get at one single freight car, while others may want to spread a larger than needed number of uncoupling devices around the tracks. In any case, the trackplan and operating rules of a shunting puzzle need to be studied carefully in order to know what the minimum requirements for uncoupling devices are. The rest is, as so often, a matter of personal taste.
Setting up the game
Finally, when all is well on the layout and running smoothly, it's time to start playing the game. With other games, this would mean perhaps throwing a dice or picking a card - but how do you start a game on a shunting puzzle layout?
With both classics, Inglenook Sidings and Timesaver, you need to know which freight car (8 of them in a "standard" game of Inglenook, 5 in John Allen's original way of switching the Timesaver) is required to go where before you can start. Obviously, there are almost endless ways of doing this, ranging from stone age to high-tech, but it will always involve handling tokens which represent the individual cars and where they are to be moved. Alan Wright used his now famous "tiddlywink computer" (a mug which held a tiddlywink (i.e. a token) for every piece of freight stock on the layout) from which the required number of 5 would be drawn, and the order in which this happened at the same time determined the order in which the cars were required to end up. You really don't need more than that, but I have found that using a system heavily influenced by the classic American way of creating switching orders, the card and waybill system, will make things unmistakeably clear even to someone having their first go ever. The following illustrations apply for an Inglenook Sidings layout, but they're just as valid for a Timesaver layout (which due to its US roots will probably feature car cards anyway).
The system is quite simple. On a classic Inglenook Sidings you have 8 freight cars, each of which needs its own card as a means of referring to this specific car. This can be as simple or as elaborate as you care to make it - perhaps a slightly more elaborate form could look something like this:
For this US prototype Inglenook Sidings layout, every car not only has its own card but cars have also been selected so that the colour of the car (there's only one yellow, one green, etc.) allows for immediate spotting of the car. All in all, there's a total of 8 cards of identical size which are shuffled and then the five uppermost are put down. The order in which this takes place also serves to determine the order in which the cars should end up.
Here, another two cards need to be drawn, but so far you can tell that the car which is required to be the closest to the switcher is the the brown D&H boxcar, then the yellow ACY boxcar, followed by the green Vermont Railway boxcar.
It takes a bit of work to make all the cards you'll need (though it can be done in 5 minutes if you don't feel like having any fancy special effects), but it's worth it, and it adds quite a bit of atmosphere to operating a shunting puzzle layout.
Still keeping with an Inglenook formula shunting puzzle, you can also let your computer do the work and generate a few random shunting lists (maybe in advance of operating sessions) by using a nifty little piece of software, the Inglenook Random Wagon Selector (runs on Windows PCs).
This software is (c) William Pearson and made available here as a downloadable zip-file with kind permission and courtesy of Mark Kendrick. I am not aware of a similarly tailor-made software for a Timesaver layout, but you can easily set up and configure one of the numerous switch list generating programs available.
Shunting tracks have their own atmosphere - here, an autumn fall of leaves has added yet more of it
(Berne main freight yard, Switzerland, October 2002) [click for a larger image]
Text and illustrations not labelled otherwise are © Adrian Wymann
Page created: 11/OCT/2002
Last revised: 03/AUG/2005