coaster

Sure, but these are just prototypes. I hope to actually mold them.

Hi Jim, Cool! The flanges on the cross are supposed to bottom out. This way, you don't get a bump when the rail disappears. For the ones that stumble or crashed, I can't quite tell from the video, but it looks like the receiving point is in the straight position, correct? That is definitely an issue with the printed switches, and likely where the CoNO bit it. Unfortunately, the parts are too big for Shapeways to polish, and even then, the polish is more of a sandblasting than polishing. I've been looking to find a high end printer that can do it and give us a much better finish, but so far no luck. Found an amazing one where the print bed was just 1-1/4" too short. Naturally. Anyway, I'll keep working on it, we'll get it right. I should have a big announcement in a few months, too! Thanks for beating on this, and for the video of all those trains running through it!

Load up on some Train/Boat weights: http://www.bricklink.com/v2/catalog/catalogitem.page?P=73090a#T=S&O={} All of my engines have 2-4 of these in them. Never have slipping any more.

An 8-stud gap would be really weird, as that would represent 14-stud centers. Most seemed to agree that a 6-stud gap (i.e. 12-stud centers) looked the best. I'd suggest we should measure centerline-to-centerline. So where the standard gauge spacing is 16-stud centers, I'd propose the narrow gauge should be on 12-stud centers (6-stud sleeper-to-sleeper).

Those are all excellent questions, none of which I can answer at the moment. The easy solution to have everything simply 3/4 of the standard (i.e., R30 base curve, 12 stud spacing, etc). This does, as you mentioned, render the R24 useless. It also means we'll never have a dual gauge layout, unless we make more specialty pieces. For the record, the "magic" radii in the small gauge are R30, R78, and R150. I think a poll is absolutely in order. We should come to a consensus as a community what the narrow gauge system should look like.

Hi @jrathfon, just realized I never answered you on this, but you have brought up an interesting point. The crossovers have always assumed connecting the switches directly together. If, however, we concede a straight section between them, then yes, not only is it possible to adhere to the stud grid, but you can also do it with integer length straights (you still need bizarre-o angles for the switches, but they're at least consistent).

My buddy and I likely have enough trains and track to handle the 9V section, and we're local to the area, so it won't be a big deal for us to get everything there. We're missing a couple sets here and there, but we didn't pirate the trains like other sets, so they're still built as they were from 25 years ago. We have a few buildings from then, but overall are pretty light in that regard. One point of note, there was a 90° cross for 9V, and I think it should be included somewhere in the layout as well. Maybe the loop at the top of the "G"?

Thanks everyone! Glad you like it! Right now, I'm not planning on it. It'd cost about the same to 3D print a duplicate version as buying it used off ebay, and the demand would probably be too low to justify the tooling to mold it. I can add studs wherever there is room. I just skipped some areas to keep it mostly consistent with the LEGO ones. However, I've deviated quite a bit already from LEGO's design (note the switch mechanism, guards on the frogs, and crosstie orientation), so it really wouldn't make much difference to me. Does everyone have a preference?

At first I thought this was getting a bit out of hand, but nah:

High-five for Inventor! That's what I use to design the tracks. Just save a copy of your model as an .stl file (on the save dialog, click options and change from cm to mm and set to high-resolution), and you'll be good to go.

Looks awesome, Andy! Jeff, I haven't forgotten about you, I've just been spending my time on the computer rather than in the shop.

We're more or less on the same page. The only issue I'll point out is the crossovers are strictly impossible at the uneven size switches, since there's no way to put an adapter between them. In other words, you couldn't make an R56 crossover work, unless you also want to make 2 different length straights to adapt on each end. Realistically, I don't even think it'd be necessary anyway, so I wouldn't bother. Here's my R104 solution, pretty much exactly what you posted: As for the narrow gauge stuff though, the same principles could be applied, and in fact the special angles will still work out to be the same, just at different radii. Dual gauge would be a different project altogether. @AussieJimbo hits on it exactly; it should be done 3 rail, but you need two versions of everything, one with a common inside rail and one with the common outside. It's doable, especially with 3D printing, but it's probably not viable as a molded solution. Honestly, these projects could be done in tandem, but I think they're different enough that they should be kept separate projects.

@Lowa Sure they do. If you put a single R40 curve on the end of the LEGO switch (going parallel), you're exactly 16 studs x 48 studs from where you started. If you tried to make an R56 switch, however, and still be 16 studs over, you'd end up at ...crunching numbers...57.69 studs; not gonna work. This is why the standard LEGO switch has that weird whip at the end. If they kept it circular but still wanted to allow for sidings, it would have to terminate at ...more crunching... 36.87°. They would have had to also make an 8.13° and a 14.37° curve (although, I think this would have been a preferable design to what they gave us, but that's a discussion for another time) As such, while you might need some weird angle curves, it's still only physically possible to have switches exist at specific radii without also requiring weird length straights.

@M_slug357 I'm just referring to the narrow gauge tracks alone. I agree, I think dual gauge will be too complex to implement. The question is simply: In a strictly narrow gauge layout, if you have a double track bed, what is the center-to-center distance between those parallel tracks? Standard gauge would be 16 studs, but I think that looks like too much and would vote for 12 studs.

More or less, yes. When you compare real world examples of narrow gauge track to how we would space out the straight sections in LEGO, the 12 stud centers look more realistic. Compare: This is probably a discussion point for the community at large. Once we make that determination though, we have to do some math. The 12 stud spacing fails when we look at switches. Switches only fit the stud grid at certain intervals (I have a formula for it), and you have to have the right starting point for it to work. The R24 isn't it; it should be R30. Moving up then, we don't hit the next viable switch radii till R78 and R150, so those, naturally, should be wide-radius curve sizes. However, if we want to get into dual gauge, then we obviously have to stick with the 16 stud centers. Does this make sense?

On the narrow gauge, no. I've only framed it out; it was something to do as a break from the standard tracks I've been working on (If you're not familiar with it, see the BrickTracks thread here). My intention long term was to develop the narrow gauge line as well, assuming I can get the standard stuff off the ground. EDIT: I should clarify, it depends entirely on how you want to view the track system. You could follow the standard R40 model exactly with centerline distance between parallel tracks at 16 studs, and then your geometry is exactly the same. However, that looks really weird with the narrow gauge tracks, as they're simply too far apart. 12 stud centers look a lot better.

The radius actually matters; it can't be chosen arbitrarily, which sadly, the existing narrow gauge track was. R24 is useless for everything when it comes to switches; it should have actually been R30. Nominal wide radius is R78 (or R150 if you want to go big). Lowa, shoot me an email or pm if you want to discuss. I have the skeletal model for the narrow gauge tracks worked out already. No sense in either of us reinventing the wheel.

I don't fault anyone for being ambitious. However, there also need to be realistic expectations about what you'll be able to accomplish, and they clearly bit off more than they could chew. Also, the project wasn't sold as a "we want to try something," rather they gave the impression they were set and ready to go, when in reality they were clearly far from it. Whatever issues they had should have been realized long before they took money from anyone, or at a minimum should have been resolved in months, not years. While I wasn't a backer, I think those that were are within their rights to be disappointed.

True, but it is certainly more than just Codefox that's unhappy. The next part though is an issue. There was nothing stopping them from making a reliable metal track, other than perhaps their own capabilities. I think this is really why people are upset: if you weren't actually capable of making the track, why undertake it in the first place?

Based on the messages I've suddenly been getting, yes.

Age is certainly a factor, but natural latex just simply will wear faster than a Buna-N or EPDM ring.

I think the square rings should be better for traction, but I'm not sure how well the elastic bands will hold up. They're meant to provide spring tension, not a continuous rubbing contact. Even the stock LEGO ones crap out after so many hours.

You could put one of these on the inside to hold the fence: http://brickset.com/parts/614026

For that scale, you might almost be better off skipping the second hole altogether and use an element that gives you a small offset, perhaps like the Technic crank:http://brickset.com/parts/4119475 Not to be pedantic, but o-rings do come in a square profile as well. It'd be interesting to see which would actually provide better traction.

The parts are printed in Nylon. I haven't found anyone with an ABS printer suitable for printing the switches (they're too big for most beds). Honestly though, I haven't looked all that hard because the 3D printed parts are just a stepping stone in the development processes. I intend to mold these in ABS, so finding a printer wasn't all that critical. I just needed functional parts as a proof of concept. The Nylon is still pretty tough, and from what I was able to observe at the show, the rails themselves looked to be holding up. For paint, I just used a glossy dark grey spray paint from the hardware store. I picked one that said it was suitable for plastic (specifically, this one). The paint did seem to give it a bit of a hard-coat, but also added thickness to the studs, even with a light coat, making it more difficult (though not impossible) to assemble. I'm sure there are better paints to use; I wanted to do a couple pieces just to see how it looked. Color match isn't terrible, but it does bring out the roughness of the printed part. I tried sanding and polishing one area first; that was a waste of time. They almost need a coating with a lacquer or something first to smooth out the surface. Overall though, I'm happy with how they performed. Seeing those monster passenger cars go through it was pretty satisfying!