Interfaces

There are some conceptual things to grasp when considering interfaces. These things are not difficult to understand, but they, or their implications, are not necessarily obvious. It is important to understand the characteristics of interfaces that make it possible, or impossible, to connect modules usefully, or at all, in order not to be disappointed when it is too late.

As an example, initially I had assumed that the module baseboard to baseboard interface design would not depend on what was laid on the baseboards. It is possible to approach the design this way; but this approach is just too abstract and general in practice. Soon it became clear that there are trade-offs which are worth consideration. Thus, for example, limiting design to what seems sensible for the configuration of the target 0, 1, & 3, gauges, and ignoring general possibilities, has the immediate benefit of bounding interface widths in a practical way. Modules for N or H0 gauges probably would be designed differently, with less emphasis on interface track configuration, and more on the human-scale maneuverability of a complete module.

What follows is the result of considerable work and re-work. The outcome being that the railway project and its interfaces have been successful: the railway is in use, and the design objectives have been achieved. So, I am recording the technical details of the railway interfaces both for me and anyone else who finds them interesting or useful.

The top drawing is not reduced to its most basic form; this is because it reflects the railway components as made, supporting three different gauges on two different dual gauge interfaces. However, the drawing does show that adding secondary track gauges to an interface, which is extra capability, can, and usually does, cause restriction elsewhere.

If the secondary gauge rails were not present, then the drawing would be basic. Which means that any two interface instances could be overlaid and the [primary] rails would line up. But, if secondary gauge rails are added to the drawing, as shown, then the rail configuration, being asymmetric about the interface vertical centreline, would result in some rails not lining up. This is because, with secondary tracks shown, the drawing view is of just one end of the module. The view of the other end would have the secondary rails mirrored about the interface centreline. As indicated in the drawing label, the module end shown is referred to as the initial end; the other is the final end. And, when modules A and B are connected, in that order, the final end of A is joined to the initial end of B. The terms initial and final are inherited from the vectors of the abstract design of the modules.

The portable railway does not have an interface as shown with four rails. As indicated earlier, the drawing is an amalgam of the two single track interfaces that the railway does have; and each of these has three rails. The two interfaces correspond to two dual gauge tracks: gauges 0 and 1, and gauges 1 and 3. In each case the gauge 1 track is primary, which means that the two interfaces can be joined together, provided that only the gauge 1 track is used. More on the topic of multiple track interfaces follows, after commentary on latching.

In contrast to track configuration, latching is basic: all modules have exactly the latching configuration shown at every interface, including junctions or other configurations with more than two interfaces. Thus any two interfaces can be latched together, provided that the module baseboard, or something else, does not prevent it. All interfaces have provision for a pin and latch on the left of the centreline when viewed from underneath, a hole and a keeper on the right. Basic latching reflects the initial assumption of interface unfettered generality referred to in the introductory remarks of this description.

The bottom drawing shows a bolt hole interface to a latching mechanism. This latching mechanism could be drawn, in abstract and generalized form, so that any conforming latch and keeper could be accomodated. This is too hard, I have no intention of getting into designing and producing over-centre latches. In practice, a commercial latch is selected and purchased, with consequent pros and cons. The portable railway selections are Southco stainless steel latch 97-50-110-22, and keeper (the part to which the latch attaches) 97-57-105-24. These are a little expensive at $10-15, according to quantity purchased, for one latch and one keeper; however, they are of good quality and work well.

Here is an image of the initial end of a single track G0/1 module from the portable railway. There is no keeper fitted to this module; otherwise it stands comparison to the drawings above.

The extra hole to the left of the painted 6, not shown as part of the interface design, originally was intended to aid separating modules if they became locked together. This was a valid concern that was solved by changing the positioning pin design. Originally, it was intended to use commercial pins used in industry to position machine tool parts. These steel pins are parallel, 1/4 inch long, and are pressed into place at the non-protruding end. This did not work in 1/8 inch thick aluminium. The pins both jammed in their mating modules, and came loose in the soft aluminium of their host modules. The new pin design is conical to prevent jamming, and is screwed to its host module. The pins are made of brass solely for ease of machining. The extra hole has not become obsolete, but has become useful for jig location purposes, described below.

In this view the module that the train is on is a single track G0/1 module. To the right of that is a single track G1/3 junction module. Toward the bottom of the image and also just at the rear of the locomotive, these two modules are joined to double track modules; the joins are marked by the green spirit levels.

The battery powered G1 locomotive is pushing a G0 flat waggon.

The junction has moving point frogs; this obviates the need for check rails.

The specification of the double track separation has a mildly curious feedback effect on single track widths, including the module baseboard width. At the interface, (which is emphasized, anything can happen away from the interface), and assuming baseboard symmetry about the interface centreline, the track separation of ten inches means that a single track baseboard cannot be wider than ten inches. Otherwise, it would not be possible to join more than one single track module to a double track module. The portable railway convention is to make the single track module baseboards 9 inches wide, symmetrically across the interface centreline; which explains the one inch gap between the single track modules in the image above. Another convention is to make the single track interface 6-1/2 inches wide, again symmetrically across the interface centreline. It follows that, at an interface, a double track interface is 16-1/2 inches wide, and a double track baseboard is 19 inches wide.

The view of the underside of the module shows a latch and a keeper. This view also indicates that a full complement of latches and keepers is not always necessary.

The hole to the right of the painted 2 is in the centre of the double interface; currently, this has no function, but is referred to as a service hole; it was drilled in the first modules for the portable railway in case it became useful. In this context it is noted that the holes in each single track interface, intended as module separation aids, also have become useful service holes. They are used in a module production jig to ensure the ten inch interface separation. These holes, due to their original intended use, are offset by 1/4 inch from the centres of the interfaces.

The remaining convention concerns which rails have rail joiners, and which are left as plain rail ends. Not to specifiy a convention results in a lot of busy work when a railway is erected. The convention is that the rail(s) on the pin side of the single track interface centreline are fitted with joiners, those on the hole side are left plain. The effect of this convention can be seen in the images above. The joiners are lightly soldered to the rails.

Portable railway modules do not have legs; instead they have an interface for legs. This means that any leg design suitable for railway erection, and which matchs the module leg interface, can be used. Telescoping aluminium legs have been made, and are in use, because of a desire to enable considerable railway height adjustment; these legs are time and money expensive. However, under the right floor conditions, literally, broom handles can be used for support. And, both styles of leg can be used concurrently.

The module leg interface is an aluminium tube 1-1/8 inches internal diameter with a steel washer embedded as shown in the drawing. The washer spreads any point loads within the module. The depth of this socket is not critical.

Most modules have six leg sockets; in practice, two, three, or four of these are used to support the railway.

As a practical matter it has been found to be highly desirable to prevent legs falling out of the module sockets. To do this, the existing legs have been fitted with magnets that attach to the steel washer at the bottom of a socket.

This image shows a junction module interface and use of the leg interface. The track had not been fitted at this point.