APPENDIX 4: DESIGNING SHANDIN LOOP
Gary Green
A partial view of Shandin Loop. The wye turnout is at the top of the photo. The #8 turnout is at the
left and the switch points of the #6 turnout are just visible at the far left.
This is a drawing of a portion of the Shandin Loop with the different elements of track highlighted in different colors. While not always true, the colors designate the following elements:
As I thought about how to go about designing a loop module set, I decided early on that the track plan would determine the physical structure of the loop, not the other way around. I wanted to design a multi-track loop so that there would be room for placing trains on the track, staging of trains, and passing tracks for trains passing through the loop. The three tracks of the loop have been more useful at setups than I ever imagined. I spaced the loop tracks 3" apart to allow plenty of room for fingers. I added a continuous loop to the inner siding. I’ve forgotten whose great suggestion this was, but it wasn’t originally mine.
As I began to draw the loop track plan, it seemed best to make it symmetrical from side to side. I knew that fitting the track elements together would be a real headache and wanted to do it only once. My approach was to design one half of the loop and then to copy a mirror image of it onto the other side to end up with a drawing of the complete loop track plan. My drawing tool is AutoCAD LT which provides some capabilities not found in other drawing tools I’ve experimented with.
I plotted actual easements using the application contained in Dale Muir’s web site rather than using the bent stick or other method. The reason for this was simply that it was easier for me to draw and manipulate the easement elements this way. I plotted the easements full size and then reduced them to HO scale for use in the drawing. I plotted turnouts using the NMRA standards so that I’d be sure to know where both switch points and frogs really were. Fortunately, the Walthers turnouts I used fit the NMRA standards reasonably well. This isn't the case with all turnout manufacturers.
When the track plan was complete, I added the outline of the module sections. My rules were that the nose of the loop would comply with the Free-mo standard requiring a 24" wide face with 6" of straight level track before the switch points of the wye. The module set was to be five sections, four in addition to the nose section. The track was never to be closer than 3" to the inner or outer edge of the sections. The sections had to be made of with straight, not curved, sides. These rules left me very little room for variation and made planning the shape of each section easy.
My first significant mistake was to build one section of the loop to see if it would work and could be used successfully with other modules. While the experiment was successful, the ends of the section didn’t form exactly a 45 degree angle. The result of this is still visible as a mating point between sections that doesn’t close completely. A related mistake which I tried unsuccessfully to avoid was assembly of the rest of the loop sections on a flat floor. The floor turned not to be flat after all. The result is visible as another location where the mating point between sections doesn’t close completely. If I were building the loop again, I would assemble all the sections at the same time, clamped together as a full module set, on as flat a surface as I could possibly find.
All turnouts were converted to make them DCC friendly, and all electrical gaps filled with gray ABS plastic. All frog flangeways were filled with plumbers epoxy putty and then opened up to match the Mark IV NMRA track gauge flangeway requirement. Code 83 turnouts and flex track are Walthers and Code 70 turnouts and flex track are Shinohara.
To insure that the track was placed according to the plan, 11’ x 17" printouts of the plan were glued like tiles onto the plywood substrate. Cork roadbed was then glued down according to the plotted centerlines, and the track glued down using the "centerline" created by butt-joint of the cork roadbed.
The following comments describe my approach to each track element called out on the drawing above:
1. The #4 wye turnout has the same diverging track geometry as a #8 turnout. If made to NMRA standards, the diverging switch point has a radius of 117" and the curved closure rail a radius of 67". Knowing this information, I decided to let the turnout act as a substitute for a spiral easement for the adjoining 48" curve.
2. The 48" curve begins just as close to the frog of the wye as I could accomplish. The gaps for the polarity reversing unit are just beyond the frog of the wye.
3. Since the main track is modeled as the middle track of the loop, I could have let the #8 turnout provide an easement for the end of the 48# curve. I decided, however, to insert a short easement of 45 feet to provide a smoother flow from the 48" curve into the straight track of the turnout (which leads to the inner siding).
4. Combined with the short 45 foot easement, the broad radius curves of the diverging track of the turnout results in a very smooth route into the middle track (the main line) of the loop.
5. The #6 turnout leads to the outer siding of the loop. Note that the radius of the diverging switch point and the curved closure rail a #6 turnout is 43", less than the 48" recommended minimum radius for Free-mo. I expect that the point angle and straight section through the frog make the effective radius somewhat larger.
6. After a short straight section, the outer siding curves at a 48" radius into the outer track with a radius of 54". No easements were provided.
7. After leaving the #8 turnout, the main track has an easement 80 feet long before a constant radius curve of 51".
8. The inner siding has a very large radius spline of unknown radius that joins it to the inner track of the loop.
9. The curvable turnout was adjusted so that the radius of the diverging track was 48". The outer track then had a larger, unknown radius.
10, 11, 12: The loop tracks continue at a constant radius through the loop. Bridge rails are 1" long so that they interrupt the flow of the curve as little as possible. My experience is that they tend to bend slightly when installed just from the pressure of the track joiner. In any case, the visual impression is of an uninterrupted curve through the bridge rails. The inner and outer tracks (the sidings) of the loop are Code 70 on N-scale cork roadbed. The middle track is Code 83 on HO-scale cork roadbed. Transitions between the two heights are extremely gradual. The overall effect of the smaller rail and lower track is less than I imagined it would be, and I’m sure that most never notice the difference. To insure reliable operation, I did not elevate the outer rail of any of the loop track. Gaps separate the loop tracks electrically from each other so that track detection and signaling can be installed at a later time.