Hi all, This is my first post. I built an HO layout starting in 2017 and I am starting my first N scale layout now. I have laid some N scale track (and converted a Minitrains loco to DCC+capacitor and light) for an HOn30 mining spur on my HO layout, but that's about it for my N experience. In any event, the reason I am switching to N scale is to be able to model more prototypically in a small space, which is probably why a lot of you are here. A multilevel layout provides me the best opportunity to keep curve radii large and realistic, which means that I will need one or two helices. I looked at other multi-level solutions, and tried quite hard to come up with a good single-level design, but eventually settled on a helix as the best solution. I've watched and read the same videos and articles that you probably have, and, yeah, it's a project. One of the things that bothered me most about helix construction was the inefficient use of plywood; it seemed as though my choices were to have a bunch of plywood end up in the trash, or cut the circles into halves and spend a lot more time fiddling and jigsawing...and still throw away a lot of wood. This seemed like a really good reason to explore 3D printing, as these printers just print the important part and don't bother with the waste. Oh, yes, I do know there is a 3D printing forum but I wanted to post it here as I feel there could be a lot of community benefit even for non-3D printing enthusiasts who don't bother reading that forum. Even if building a helix is the *only* reason someone might buy a 3D printer, it might end up being the same cost as building a helix out of plywood, and you'd end up with a free printer. I am fortunate enough to own a printer already and have become decently proficient at CAD. I use Fusion360, which is free for consumer use. I'd like to share the build here and of course welcome any feedback as I go. It might work and I might discover the reason this technique has not been widely embraced. I will say that I'm already on the 32nd design version and have made a ton of mistakes already, with more surely to come. I am using the threaded-rod technique with 3D printed roadbed spans between the rods. My printer is not large enough to print semi-circles or even quarter circles, so each span is 1/8th of a circle (it turns out that 8 sets of rods is probably the right number anyway, so this is not a big deal). I am modeling the San Jose / Los Gatos California area in the 1950s and trains will not be very long. As a result I am hoping that 12" radius corners at 3% slope won't be too steep, but I know that's more aggressive than most people are comfortable with. I am using Atlas Code 55 flex track. I enjoy designing things and spent a lot of time building features that would help speed up assembly (and disassembly), and allow easy adjustments as needed. I will share some of the progress up to this point below - I think it's promising enough to be worth a look, so here we go!
One challenge with 3D printing is that it's slow. The best way to speed things up is to print less material! Shown below is the first version I printed which took about 90 minutes to print, versus 50 minutes for the black sample shown in the previous post. So the name of the game is to remove as much material as possible while maintaining strength in the the right directions. The more time I spent looking at and working with these spans, the less material I realized I needed and revised the design. It reminds me of a quote I like: "I would have written less if I only had more time". The current design shown in the first post is quite strong radially and strong enough (I think) vertically. Overcoming the propensity to sag between spans was a challenge without making the print taller. As you can see below, there is a buttress offset to the inside of the span (red arrow) to help with vertical stiffness. This is the tallest part of the print, but since it is offset, it doesn't encroach on the tracks below or above. I wanted it on the inside so that it could also catch a train if it falls off, and I didn't put one on the outside as well (which would have made it stronger) to allow my hands easier access. I also ended up slotting the holes for the rods (blue arrow), which makes assembly a lot faster!
The last design feature that was important to me was some sort of guide to make laying the track easier. I've laid flex track before and don't mind it, but with the helix my only goal was to make it easy and reliable. I ended up with a design that encapsulates the ties and locks them in...without being so tall that it interferes with the wheels or trucks. I will also use some drops of glue to hold the ties down permanently, but this little lip does allow me to snap the flex track in place, and join them with rail joiners, without any glue. You can also see those little gold pills. These too are 3D printed and each one is angled at 3% so that the track spans are not torqued flat when tightening down the bolts. My first versions did not use these pills and it took me awhile to figure out why my grade was so inconsistent! I could have incorporated the angle into the track spans themselves, but I liked the idea of one universal track span design that could be graded at any angle (by using different pills), and be used for a clockwise or counter-clockwise helix, so I settled on separate pills. There is one on top and one on the bottom, each facing a different direction. The square "tail" on the pill helps me see which direction the pill is facing, uphill or downhill, since it's hard to tell with the naked eye. Here's a video showing the assembly process: https://photos.app.goo.gl/pftCgxeBJ1iRptUE9
Nice work! I see some folks in the hobby diminish the usefulness of FDM printers in our miniature empires, but this is a prime example of how useful they can be. If you're ever ready to upgrade the Neo, the current generation of FDM printers have finally crossed that speed gap and would probably knock that 50 minute print time down to 15-20. I upgraded from a standard Ender 3 to a Kobra 2 and I'm having fun printing FDM again due to the faster print speeds. Keep us posted on how this goes and feel free to check out our 3D printing forum to share projects, ideas and talk shop! Cheers -Mike
Cool approach. Minimize the rise per revolution as much at you can. Can the roadbed be thinner? Here's why. NMRA RP height clearance is 1.73 inches from top of rail. If the road bed plus track height is .27 inch you get a 2 inch rise per rev. Your 12 inch radius circle's circumference is 75 inches, giving you a 2.7% grade which is steep. Short trains. Check if long cars' overhangs hit the chamfered cross pieces on the inside of the circle. Putting them on the outside would fix it.
Increasing the radius to 18 inches lowers the grade to 1.76% while a 24 inch radius will yield a grade of 1.33%. A 30 inch radius will yield a grade of 1.06% and a 36 inch radius will result in a grade of .88%. Notice that doubling the radius reduces the grade by one half. An added benefit of the larger radius is that it enables easier interior access to the helix which may come in handy when things don't work and play well together. The overriding consideration is, of course, how much room one has for the helix. That 36 inch helix will take up a minimum area of 6+ feet in width and length.
In depth discussion, not sure if you need to join nScale.net to see it and the server has been slow lately: https://www.nscale.net/forums/content.php?254-An-optimized-helix
Something else that I forgot to mention, the discussion in the above posts referenced a 2" elevation change for one revolution. Obviously more than one revolution would be necessary for a double deck layout. Lots of track consumed with no scenery or operational sidings. Just one of the trade-offs involved in model railroading.
Super cool idea for using an FDM printer! I would add that the trackwork in this helix, with code 55 track, must be flawless. The slightest kink, undulation or rough joint will have rolling stock trying to check gravity, ESPECIALLY on a 3% grade AND with such a sharp curve radius. My model railroad (flat/no grades) uses Atlas c55, and I found my most picky units (Kato PA-1s) require smooth, flawless trackwork in c55. It is far less tolerant of less than stellar trackwork than Code 80 which is far more forgiving. The shorter rail height of c55 requires lower-profile wheels (older locomotives and rolling stock with pizza-cutter flanges will not operate smoothly on c55), which makes trackwork all the more critical. Code 80 allows taller flanges, like the pizza cutters, and is this more tolerant of imperfect trackwork and more reliable operation on tight curves or steep grades. Since the helix is not likely to include modeled scenery, why not design the tie height to accommodate c80 track? The reliability gained is worth the extra redesign time, unless you have a significant amount of sections printed already. Just food for thought.
