The Nepenthes, Gestalt, and Delirium Railway

1. Here is the summit of the railway at a busy time. A battery-powered NW2 hauling a train of beer-cars on the mainline; a couple of live steam engines sitting in the steam up bay, which has a railer at the end of the track; the lower-level branch line visible underneath; and Joe hanging around on the deck, enjoying the sunshine and the flowers; all help to make up the scene, and the Gestalt.

I start off by making these points to make clear that my emphasis is on the railway itself: the NG&DR is my garden railway, not a model of something else. So, I like the "free" in freelance, because then I am able to do things such as run trains in different scales and gauges; and run trains that are made up of carnival floats (for kids, and other non-railway people), in addition to models of past steam-age legends.

However, the description of Nepenthes that I like is that it is A drug that dispels the cares of the world. And I think that sounds pretty good. If my Railway becomes someone's Nepenthes then I shall be well pleased.

Gestalt is A whole that is greater than the sum of the parts and comes from the Uber Gestaltqualitaten of Christian von Ehrenfels. Briefly, my Gestalt is the spirit-lifting feeling that can come from a miniature train threading its way through the flowers.

Delirium is the effect of the Nepenthes and the Gestalt, and is admittedly heavy-duty: temporary and reversible severe confusion and rapid changes in brain function. But, if one ignores the technical gibberish and segues into delirious, the concept becomes rather inviting: thrilled, excited, ecstatic.

If my railway gets me thrilled, excited, or ecstatic, simply by arranging a miniature train to wander through a garden to the extent that I forget about real life for a while, then it truly is a drug that dispels the cares of the world. And so, if you understand any of this at all, then you will understand my railway's slogan:

The reason that this history is important to garden railway people is that, for the first five years or so after the new installation, the dirt settled substantially. And so the railway track settled, a lot, even though the dirt was tamped initially; and this settling caused repeated re-work. The settling has more or less ceased by now, eight years later; but there is a cautionary note for the wise in this history.

The gauge 1 trackplan is a folded double reverse loop with a cut-off from one loop to the other loop (the track alongside the station platform), and an additional dual gauge siding (top left, the siding joins to the proposed Gauge 3 line at the duck-under). The gauge 3 trackplan is not completely implemented, but, in the schematic, connects the dual gauge siding and another gauge 3 siding that then joins one of the gauge 1 reversing loops. The gauge 3 line does not follow the reversing loops because the track through the tunnel was deemed to have curves that would be too tight. There is a steam-up track at the high point of the line; this is connected to the continuous oval with a manually operated junction. The cut-off provides a continuous oval, used by both gauges, and this oval track is the default when all the junction points are in their un-activated positions. The reversing loops join together at the top left of the pond; thus the middle straight connecting the loops is just two or three inches long.

2. This is a view taken from the house roof in the early summer. The train on the right is a work train hauled by a Bachman Spectrum 45 tonner. The train is run around the track, and any plants that have grown out and interfere with the white and blue loading-gauge-sized "boxcar" are trimmed with the scissors kept on the back of the boxcar. The boxcar also is the dump for the horticultural debris, and it fills quickly at this time of the year. There is an NW2 with a white tank car train on the siding behind the tree.

Track was made in the house. First, by drawing what was required on craft paper. Then, ties were fixed to the craft paper with water-soluble glue (Elmers); followed by rail attachment and all other building tasks. On installation the craft paper was torn off; and any paper remnants and the glue disappeared with the first rainfall. The drawing step, which allows for erasing and re-drawing, allowed experiments and ultimate accuracy.

3. Junctions, on the basement floor, and installed (after a few years, the ties and the rails have weathered). The, short, piece of track between the points at the top of the image on the right connects the two reversing loops.

The detail images below show two significant differences in two junctions on the railway. The left image shows the use of spikes for rail attachment, the right shows the use of screws. My experience is that screws are considerably better than spikes because screws do not work out of the ties as do spikes. I note that screws are used on 7-1/2 inch gauge track; I suspect for a similar reason.

Since these images were photographed I have concluded that staggering the screws (or spikes) along the rail is generally a bad idea. I have had no problem with splitting the wood of the ties. But I have had problems with ties mis-aligning by twisting due to staggered fasteners. This has been particularly true with dual gauge track, which has three rails and consequent twisting torque inbalance applied to the ties.

The second difference occurs where the wide gauge rail crosses the narrow gauge rail. In the left image, an active point, here called a secondary point, which moves at the same time as the junction primary points, guides and supports the wheel. In contrast, in the right image a static check rail is used which only guides the wheel. There is a potential de-railing problem in the latter case, which affects the narrow gauge wheels. The wide gauge wheels are not affected, their path is identical to that going through a normal frog. Wheels, particularly small wheels, do not do well as they cross the large gap in the right image; this is the gap that is filled by the secondary point shown in the left image. Wheels can drop into the gap as they cross it, and this can lead to either or both of wheel flanges snagging on a point and de-railing, or to significant rocking of the vehicle that the wheel is supporting, and any problem that may result from that.

This problem occurs because of a desire for the junction to handle a wide range of wheels. There are at least three standards for gauge 1 track and wheels (G1MRA, NMRA, and NEM), which are not properly compatible. And, somewhat amazingly, manufacturers seem quite happy to invent their own variants. Add in the significant differences caused by the different scales, ranging from 1:32 to 1:13, and the resultant jumble of sizes means that normal railway track-design procedures, e.g., controlling wheel tread widths and track flange clearances, do not work.

My dual gauge junctions with secondary points have been successful in dealing with these issues, to the point where, normally, I never think of them. However, the junction shown in the right image is the manually operated junction from the steam up track, and I have become somewhat wary of it. The potential problem occurs in traffic coming from and going to the steamup bay, which is slow speed traffic on the narrow gauge track. Sure enough, I have had gauge 1 equipment de-rail as it is moved onto and offof the mainline. I am inclined to stick with an active point in the future.

Remote point operation requires the introduction of some sort of system. My inclination is to guess that the pneumatic system, which has had a very low component failure and maintenance rate, is better than an electric system would have been. I have been very happy with pneumatic control; it works in all weathers here in Colorado. I have heard that frozen condensate can be a problem in damper climes.

Most of the railway's locomotives are controlled by one controller, shown in the lower image. This is a re-packaged Spektrum 2.4GHz surface racing car controller with a 300 degree linear potentiometer replacing the standard 85 degree, short-travel, trigger throttle unit. I have made a couple of other controllers, dedicated to particular engines, which have used logrithmic potentiometers instead of linear, and I recommend this to anyone thinking of using a similar system. The advantage is more controller travel at low speeds for any given regulator change; thus there is more sensitivity and better control accuracy, see Using a Logarithmic Taper Potentiometer (in Radio Control Installation).

The controller shown has three channels. My diesel outline engines use all three: regulator, reverser, horn; my live steam engines use one for the regulator if they have slip eccentric valve gear, two otherwise.

But I find the continuous oval to be too much like a track around a Christmas tree with a train chasing its tail. I like to see trains running down different tracks and demonstrating the effect of the reversing loops. I have to admit to getting a bit of a charge from a visitor suddenly getting animated and saying "Oh, its going the other way!" (it's that Gestalt again). And visitors, particularly kids, certainly like to see trains switched onto the track going across the stream and through the tunnel. Also, using this simple mode means that reversing travel direction has to be effected by a human "crane-shunt". Thus, using just the continuous oval is the convenient and simple, but boring, mode.

A second mode of operation is to route a train by switching points according to one's fancy. Now the problem with that mode of operation is that it is surprisingly easy to derail a train with badly set trailing points. Almost always, when this happens, it is because someone spoke to the operator, distracting him from his job. The other times occur when the operator distracts himself. Ideally, trailing points are forgiving and allow themselves to be pushed out of the way; but this does not always happen if the rolling stock is lightweight, and cannot happen if the junction is dual-gauge and has a secondary point.

The third operation mode is a variation of the second: disciplined instead of casual point switching. It sounds terrible, but it is not in practice. Once a train has entered a reversing loop, in order for the train to exit in the normal way, the junction points must be switched. This is very obvious when actually thinking about a canonical reversing loop; but, as indicated above, experience on the NG&DR is that, with casual point switching, humans forget and get it wrong - leading to a derailment. However, a rigorously applied single rule that the points always are switched once the train is through, works very nicely, leading to a train using all the available reversing loop track, in both directions, without much thought and without derailment. Some emphasis on the usefulness of this discipline can occur when, for some reason, normal operation is changed and the train is reversed to exit the loop. In this case, the just-switched points are wrongly set, which is easily forgotten, and derailment-in-reverse occurs.

On the NG&DR, a benefit of using the reversing loop track exclusively, as occurs using the third operation mode, is that a second train can be parked on the cut-off track, i.e., in the station, at the same time as the first train is running. The points on the junctions protecting access to the station track never are switched in the third operation mode.

4. The Carnival Train. Hauled by the NG&DR workhorse 45 tonner at the lowest point of the railway.

Image 5 is of an NW2, (purchased by the NG&DR from the Atlantic Coast Line, hence the colour scheme), hauling some USA Trains tank waggons with a paint scheme loosely modelled on the Coors beer tankers that I drive by when I go to the Colorado Railroad Museum. The locomotive is a USA Trains NW2, converted from track to battery power (twelve AA re-chargeables), controlled by the 2.4GHz radio already mentioned. The locomotive has a MyLocoSound diesel card, and a SpeedMax Electronic Speed Control.

5. The scale of this locomotive can be deduced from the locomotive number, which is 2906; the first two digits indicate that the scale is 1:29. Much of the NG&DR rolling stock is marked in this way. For example, the 45 tonner in image 4 has the number 2001, indicating 1:20.3 scale. Some items do not follow this convention, when the number has some significance with respect to the modelled prototype.

6. This image shows the beer tanker train running in the opposite direction, emerging from the South end of the tunnel.

The blue coal train shown in images 7 and 8 is a favourite; and, of the live steamers on the NG&DR, has the size and running characteristics that are the best suited to the railway. The scale of the train is considered to be 20.3, although the locomotive is freelance and scaleless as purchased. The locomotive is a Roundhouse, butane-fired, live steam, Lady Anne with cosmetic alterations reflecting an Englishman's idea of a nineteeth century Colorado narrow gauge locomotive. The locomotive is fitted with radio control for regulator and reverser. The leading boxcar and the coal gondolas were made from Hartford kits, and these are very nice 20.3 models of equipment used on the D&RG in the 1880s. The caboose is scratchbuilt and is based on a design by NorthEast Model Products. There is a description of this train in Trains Special to Me.

7. Non-railway visitor: "Is that real coal?" NG&DR: "Yes, that's real coal."

8. This train is equipped with link and pin couplers as were the prototype waggons. On the NG&DR, in general, rolling stock is equipped with Kadee couplers. So is this coal train at the front and back, the links and pins are used only within the train.

The green train in images 9 and 10 is a model of trains that I grew up with. The locomotive is a Southern Railway Schools class, and the coaches are non-corridor suburban equipment used for local and commuter service. There is a description of this train in Trains Special to Me.

This train is very free-running, and, unlike the Roundhouse locomotive in image 7, this Aster locomotive has no braking effect when the regulator is closed. Driving the Schools, particularly with a three coach train, requires 100% concentration. Precision-open regulator uphill, full-shut regulator downhill, and timing and precision of the changes, all are important. A speed-sensitive drag-car is partly made; and it is hoped that this will steady the performance. I like this, 1:32 scale, train; but really it is too fast for the NG&DR.

9. Looking North, this view shows the railway environment in the summer. Visible are the stream as it enters the pond, Red-Feather Celosia and other annuals. Some Japanese spurge is poking up over the rocks; and there are a few hens & chicks and a variegated sedum in the foreground.

10. Looking East plainly shows Hosta, variegated iris, impatiens, and coral bells. There are many other plants that are hard to make out.

I have just one gauge 3 train; for the most part, I have been too busy with gauge 1 and the NG&DR to spend time on gauge 3. The train I have is all from Garden Railway Specialists (GRS). The locomotive is a live steam model of a Great Western Railway 0-6-0 tank locomotive. This locomotive is a GRS-modified electric motor powered kit; unfortunately, it shows it, this is not a well-balanced steam locomotive. I have carried out a significant amount of work on this locomotive, which is described in ReBuilding a GRS Live Steam Pannier Tank. One problem remains, which is that the butane burner howls unpleasantly. Another task remains, which is to fit radio control.

The train is a typical British goods (freight) train from years ago when the standard configuration for a goods waggon was two-axles, four wheels, and no braking system when running. Train braking was left to the brakes on the locomotive, brakes on the guards-van, and skilled operators. I like the size of gauge 3 equipment, which, interestingly, at 22.6 scale, is smaller than Colorado narrow gauge equipment on gauge 1 at 20.3. This is because of both the scale difference, and the size difference of the prototypes.

11. A gauge 3, 1:22.6 scale, British goods train. The green carpet with the tiny white flowers is Irish moss. In the foreground are a bunch of healthy hens and chicks. Behind the second track, which curves toward the bridge over the stream, are Artemesia, several kinds of hens and chicks, and miscellaneous other plants, including daffodil leaves that have not yet died back. In the top left corner are flower boxes on the patio; it is too early in the year for the flowers to show.

12. A goods train running onto a dual-gauge siding. The background has Japanese Spurge, Thyme, Juniper, and other early summer plants.

The train in image 13 is a model of another train that was part of my growing up. The Golden Arrow, between 1920 and 1960, was the way to travel between Paris and London. In its glory days it was all Pullman, all first-class, and all important. Due to trivial, and lowly, circumstance I saw this train hundreds of times during my childhood and adolescent years. There is a description of this train in Trains Special to Me.

Unfortunately, I have been unable to get the model to run successfully on the NG&DR. The final blow to my attempts came when I found that the locomotive, which needs a fairly large radius curve, just would not negotiate a junction of the NG&DR because of a seven-feet radius in conjunction with a gradient change. The train runs well enough on large tracks; at the Sacramento Summer SteamUp, for example, shown in the other image.

13.The Golden Arrow posing on the NG&DR. Below, it is running in fine style at the Summer SteamUp

With respect to the railway and its creator, this is not a bad state of affairs. The railway has run its course and has matured to the point where its shortcomings are becoming a nuisance, even mildly irritating. The railway and the process of creating it have, of course, been instructive. And, should there be a second garden railway in the life of its creator, the lessons learnt from the NG&DR would live on.

The secondary things about the railway, good, bad, and speculative, are what follows. The reader may feel that this all has been written before, elsewhere, by others; but that is important, too.

• Trackplan Overall, I like the trackplan. The dual capability of a simple hands-off mode of operation and a more interesting and absorbing mode where trains can be controlled to do other than chase their tails is well worth having. I should give such multiple capability serious consideration as a requirement for another garden railway.


• Gradients The main failing of the railway is excessive gradients. Really this matters only because I like running with other than electric motor motive power, live steam being the obvious example. Any form of motive power requires control. Live steam locomotives only offer a regulator, which must be controlled remotely in small gauges; gasoline engines are like steam engines, and need a remotely controlled throttle; clockwork locomotives are controlled with an on-board governor; and so on. However, electric motors are controlled by the physics on which they are based; they are inherently self-controlling, which is amazing and beautiful. When I run battery powered trains, I set the regulator once and work the junction points. When I run live steam, I get busy.

  The standard way to control live steam locomotives is to maintain a constant load by making the railway dead level; this is un-prototypical, but it works as a control mechanism. So, perhaps it makes sense to have a level continuous oval for tail chasing, and reserve the gradients for the absorbing modes.

  More generally, my experience with the NG&DR leads me to the conclusion that, even with remote control, gradients of up to 1% (1in100) are ok, but more than that starts to introduce lack of realism, and other disappointments. 2% (1in50) is about the sensible limit for gradient. The NG&DR has 3-4% (1in33 - 1in25) gradients. Abrupt gradient changes also are to be avoided; they are difficult to traverse smoothly, and, on a curve, can cause unsprung rolling stock briefly to become three-wheeled, which causes othewise agreeable stock to give up and jump the track.

• Size The reason that the line has excessive gradients is that I tried to do too much with the available space. I copied the thinking from, say HO and N scale, electric-powered modelling layout design into the steam-powered garden. Also I chose an enormous loading gauge (ten inches square) in order to accomodate, well, anything. So, if I did it again in the same space I would accept size restrictions on the trains, either 1:32, or even 0 gauge 1:44, for main line trains, or small tank engines for 1:20; and this would enable more reasonable gradients. The other way to do it, of course, is to consume more real estate, i.e., take over the whole garden.


• Curve Radii The bigger the better. I never have seen a model or miniature railway with curves that are too large in radius. Not being able to run the Golden Arrow on the NG&DR because of sharp curves and gradients remains a considerable disappointment.


• Track Height I really like ground level track. For me, in a Garden Railway, Garden is just as important as Railway. However, a steam up bay should be accessible when standing upright. I should make this a high priority item for another railway. A sloping plot of land has some real advantages here.


• Nice to Have
  • Not having to put out a train every time you need one must be very nice. Garden Railways Magazine has published articles about storage systems, including running trains into the house basement (better than very nice). My guess is that the feasibilty of train storage is well worth thinking about. Although, inadvertently leaving out a live steam locomotive in freezing conditions could be painful.
  • Failing actual storage, a railer is a good idea. I fitted one to the end of my steam up bay, which turned out to be good and bad at the same time. Quickly, I found myself using the railer to load trains, but then engines in the steam up bay were in the way. Railers are easy to make; two, one for engines on the steam up bay, one for trains on a siding, definitely is a nice-to-have.
  
  
• Detailed Item
  • The track making method shown I like a lot; being able to draw, full-size, is very useful. However, when this track is installed, over time, the ties move if they are not prevented from doing so; battens are one way of preventing movement. The problems with battens are the amount of work involved, and the rigidity of the final track - it becomes sectional track, not flextrack, and that can cause a distortion problem with temperature extremes. However, in any future track-making I should come up with a method of restraining tie movement, at least for bridges; ties buried in ballast are not so much of a problem.

14. A Drug that Dispels the Cares of the World