In the last issue, I described a method for constructing plain track for my layout representing the Pembroke Southern in 1906 [Gourley 97]. As you will recall, one of the goals of this layout is to demonstrate the use and construction of highly detailed and accurate track. In this issue, we will investigate the construction of a turnout with the same level of fidelity.
The ARE&MOWA was actually relatively new at the turn of the century - minutes of the proceedings of their annual meeting only go back as far as 1898. Consequently, the best practices that would later form the basis of recommendations were still being determined. These practices did not receive the committee's approval until 1910 [ARE&MOWA, 1910]. A subset of the approved dimensions are shown in Table 1. Interestingly, critical dimensions such as the total length of the turnout or the switch length have not changed very much over the years.
Even so, regardless of whether the standard turnout dimensions had been determined by 1906, it is unlikely that a little pike like the Pembroke Southern was laid with any respect for the wisdom of a committee in Chicago. However, no better information has come to light, and so I am laying track according to the practices that seemed to be gaining favor at the time.
From the various discussions and submissions among the ARE&MOWA papers, I was also able to piece together a picture (Figures 5 and 6) of a turnout at the turn of the century. The accuracy is for the most part confirmed by photographs.
Questions such as this are important because tie plates, joint bars, rail braces, and the bolts that hold frogs and points together are the types of details that I want to represent on my track.
Before these details can be added, some of the artifacts of typical model track need improvement. For example, typical "rail joiners" and "throw bars" do not have much in common with the prototype's joint bars and switch rods. The effort of applying details is wasted if we cannot devise a robust but non-intrusive method to join lengths of rail and operate switches.
Before we begin, I should mention that this technique was not the only one considered. I also contemplated using simple PC board ties, which certainly yield the required accuracy, but it is difficult to make them match the adjacent stained wooden ties. On the other hand, the contact cement method that I described in the plain track article does provide the desired appearance, but the accuracy and robustness are suspect. Finally, I considered using the Scalefour Society's riveted construction with the ties turned upside down. This method does satisfy the requirements for accuracy and appearance. However, I don't own one of the fancy riveting machines, and narrow North American ties are not available in plywood as required for this type of construction.
I ultimately settled upon a combination of a couple of ideas. In areas where gauge is not especially critical such as at rail joints and turnout leads, brass sequin pins were driven into the ties and sanded level before the rail was soldered to them. Because it relies on the wood of the roadbed and tie to control the gauge of the track, I do not think that this method is appropriate for locations where the gauge must be tightly controlled.
For frogs and points, I wanted a means that will not vary with the changing seasons. This meant
that the two rails must be rigidly joined together. I prefer solder in a case like this, and so, PC
board needed to figure in the equation somewhere. As shown in Figure 1, the PC board was
hidden beneath the ties, and the rail was supported on small studs (actually pieces of code 70 rail
which are roughly the same height as the ties) that were in turn soldered to the copper cladding.
The board and studs are covered by ballast, making
the construction virtually invisible except in places
where the ballast is well below the tops of the ties,
such as around points.
In my initial test of this construction method, I planted narrow PC-board strips between every third or fourth tie. While this certainly worked as I intended, ensuring that all the studs were at the same level was quite difficult. In the future, I shall use a single long piece of PC board for all the studs around the frog, and another for those at the points. The rest of the turnout will be laid using the brass pins as described above.
The only way that I could expect to construct track with this kind of accuracy was to lay it out on a template. Using a CAD package on my computer, I laid out such a template using the dimensions that I had found in my research. This was then printed and glued to the roadbed beneath the ties. Because the ties get sanded anyway, the added thickness of the paper does not cause abump as it might if I were using, say, PC board ties.
This in turn meant that I could build the frog to a precise angle by pinning the two frog rails to a template. The frog rails themselves were filed to the correct angle by clamping them in a jig described by [Gittins 96] and shown in Figure 2.
A .010 x .030 inch brass strip representing the reinforcing bar on the points, snuggles nicely along the web of the Railcraft code 55 rail. The reinforcing bar plays a functional role in this model as well as a cosmetic one. In many cases, track modelers notch the foot of the stock rail to provide a recess where the point can fit. The prototype did not do this and so, I wanted to avoid it with my switch. To do this, I needed to remove a lot of material from the foot and web of the point rails. The resulting points would have been very flimsy were it not for the reinforcing bars.
The reinforcing bars were simply soldered to the web of the points after the appropriate piece of the rail head had been removed. Because of this sandwich-like construction, the points are very stiff, which is something that we may want to keep in mind if we are ever going to build a 33' switch.
The end of the reinforcing bar extends beyond the heel of the point to serve as the hinge upon which the point pivots. When I arrived at the detailing stage, a few NBW castings turned this end of the reinforcing bar into a plausible representation of the point hinge.
Finally we come to an area where model track
typically betrays itself - the switch rods. Indeed,
calling them switch rods flatters them with more
prototypical correctness than they deserve. In a
typical model installation, the switch rod is a piece
of PC board that is soldered to the two points. It is
usually too clumsy and wide to masquerade as a real switch rod, much less one with clips and
insulation. Typical switch rod models better deserve the name that modelers have given it -- the
throw bar.
Some British modelers have been using a throw bar beneath the roadbed. Two posts are soldered to the points and pass through holes in the roadbed to engage the throw bar. Thus the model switch rod does not need to serve any mechanical function and can be a true representation of its prototype. I have decided to adopt this approach in my own modeling, although it is too early yet to report the results. I will report further on the subject in a future article.
Having never actually seen such a system in the flesh, I was not sure how the posts were connected to the points. Since I don't trust a soldered butt joint, especially with the stresses applied by a tortoise switch machine, I opted for the system shown in Figure 4. The end of the post was filed to a thin shape that would accept the foot of the point rail. This was then soldered to the face of the point reinforcing bar. The resulting bump can be disguised as a part of the switch rod clip. Incidentally, I would not recommend this means of attachment to a standard HO modeler, as the wheel flanges will catch on anything that is affixed to the gauge side of code 55 rail.
Finally prior to painting, I detailed the turnout. The locations of the various parts are shown by
Figures 5 and 6. This proved to be a fairly time-consuming process. There are many bolts and
rivets that require representation, even leaving aside the spikes. The following details were
installed on theturnout:
Switch rods: Made from .010 x .020 inch styrene strip glued between the points. The clips are evoked by nut-bolt-washer castings glued to each end, and NBWs epoxied to the point rail. The Pembroke Southern had no signals, and so, there would have been no need for insulated joints in the middle of the switch rods.
Rail braces: I initially intended to make these from polyurethane castings, and spent a day making several discarded masters. Ultimately, the mold had a big bubble in it, and the castings were unusable. As there wasn't time (why do I always wind up modeling to a deadline?) to redo the mold, I opted instead to make a punch and punch them from brass shim stock. The prototype was forged or pressed steel, and so, this seemed like a reasonable approach. While the resulting parts are not as crisp or uniform as the castings would have been, they really contribute more than any other detail to the prototype appearance of the turnout.
Bolts and rivets: The frogs and points on these early turnouts were bolted and riveted together. Using the prototype diagrams as a guide, I glued NBW castings on with epoxy where appropriate, and placed dabs of epoxy to represent bolt and rivet heads elsewhere.
Joint bars: The joint bars were mastered with styrene and NBW castings, and cast in
polyurethane. This, it turns out, is a time-consuming process, and a better solution to the joint bar
problem needs to be found.
Tie plates: The only tie plates represented on this turnout were those beneath the points. At first, the naturalchoice for making these seems to be some kind of metal shims. However, we need to remember that the points will move back and forth across these shims, which will leave the metal shiny. I thought about using paper or coloured plastic, but in the end, elected to paint them on using epoxy. For obvious reasons these details must be installed before the switch is functional.
References
ARE&MOWA. Proceedings of the Eleventh Annual Convention of the American Railway and Maintenance of Way Association. Vol 11. ARE&MOWA, Chicago, 1909.
Gittins, Paul. "Easy - 'V'sy". Scalefour News. Number 93, p23.
Gourley, Rene. "Canadian Track for the Turn of the Century", The Proto:87 Journal. Number 5, January 97, pp6-7.
Gourley, Rene. "Proto:87 Switch the Old Fashioned Way", The Proto:87 Journal. Number 3, January 96, pp 3-5.
