My helix

To keep the thickness down I constructed mine using 1/8 inch door skin and 1/2 inch homeasote. I used liquid nails between the two layers. The joints of the door skin and homeasote are stagered 6 inches. My helix is a 24 inch dia. and asembled with threaded rods. The track is Peco code 100 spiked down and the joints soldered. On the lower part were the joints were not far apart I also glued the track at those points. I am in Florida and the layout is in the garage, I have not had any issues over the years.

I'm the one who utilized a

I'm the one who utilized a double thickness of masonite, and I believe it is going to work out. BUT I would not recommend it,...toooo much work cutting duplicate pieces, gluing together, then sealing and painting numerous coatings to try and 'stabilize' it.

From what I've seen and read since starting mine, good grade 1/2" plywood seems to work out fine,...just be sure to seal off the plywood good.

Making it an oval shape, particularly that small of an oval seems to be adding unnecessary complications to the process,...both for the roadbed and laying the track....... just my observation.

double layers

Haven't done it myself yet, but from my reading I think the idea with doubling up is to avoid having to do biscuits and ensure smoother and more solid joints between sections. You just overlap the sections by half and glue everything together. Seems easier to me than biscuits. If you use trapezoids the way I've seen most people doing it, the sections are not that long, under 12" on the short side, and if you use those with biscuits you'd have to make a heck of a lot of biscuit joints. If you're cutting quarter-circle sections out of thicker plywood without doubling up then biscuits make more sense to me.

I'd certainly not bother with both biscuits and doubling up. I agree doubling up 3/8" plywood might be more sturdy than needed, depending on width and how far apart your supports are. You could use 1/4" plywood and cut that into trapezoids and double those up and avoid bothering with biscuits.

Not only is this a case of no-single-answer, it seems to me there are no two helixes made exactly alike once you factor in all the variables (like how to suspend the layers, material, joinery, number of tracks). Are you planning on threaded rods, blocks between layers, slotted strips on the outside or some other method?

Good input all around....

Bob - interesting to note that you are using a layer of homasote! I like that idea because the homasote provides sound-deadening qualities. With a 24" helix diameter, is it safe to assume that you are modeling in N scale? Otherwise, for HO scale, you would have a grade something north of 3%. I failed to mention in my original post that I'm modeling in HO scale, "modern era" as defined by the NMRA Recommended Practice 7.1 - so have to account (height-wise) for double-stack operations.

Brian - sounds like you've thought out the use of masonite for your application. I see no added complexity from the oval shape I'm planning to use. Using the trapezoidal sections means all you have to do is flip one piece over so that the angled faces countermand the urn for one section.

eastwind -  I'm thinking threaded rod at this point in time. If I make the trapezoidal sections wide enough then after I've built the whole thing flat (with track installed), I can simply drill a hole(s) through all of the layers and then insert a nut/washer combo as the threaded rod is inserted through the layers. Also, if the sections are wide enough, it may even allow for slots which would allow the threaded rod to be simply slid in from the side!

Paul,

I think you're missing a detail , from what I've read; the problem with growing a helix like you propose is, as you separate the turns, either the helix shrinks in diameter, or the length per turn increases.  You can see this in real life, if you have a spring-driven garage door.  As the spring stretches, it reduces in diameter; as it returns, it grows (it's hard to measure on a 2" spring, granted, but the change is there).  In the ultimate case, your spring would become a wire, if it didn't fracture first.  Your helix must do the same.

It's a 3D problem.  Consider that, "building it flat", your multi-turn helix has a separation of, for example, 1".  To be able to run trains, you need a separation of, for example, 4".  To stretch that coiled "spring", with a fixed length of track already installed (and for that matter, a fixed diameter of plywood), you insert your rods and begin separating the layers.  As you do that, either the helix diameter must shrink, or the length must grow.  But your track and plywood are fixed length.  Something has to give, and I expect it will.  I'm no helix expert, so I can't tell you where the effective length change will come from, or if you'll just see some dimension that becomes wonky, like the roadbed will buckle.   

Now, to be sure, the length change for the oval you describe isn't much, from 0" separation to 4" separation - I calculate it to be about 0.034", give or take.  But it does exist.

Anecdotally, I remember reading (LDSIG Journal, maybe?) that someone had encountered this when building a helix - I think the result was that the threaded rods developed a lean as the structure was stretched, which in turn imparted a slope to the roadbed (don't remember the details, don't know if it leaned in or out, and that may have been dependent on whether you raised the inside or the outside first).  I also don't know if, ultimately, the structure was usable.

Those who have built helixes in the way you are proposing (build it flat, then stretch it) can comment upon whether the forces induced by "overcoming" this length change matter at all, or can be tamed in some manner, but I wouldn't do it without knowing that answer beforehand.

Blair

Good point....

...Blair. I've noticed in some of the "build it flat" discussions that it is built in sections (1/4 arc, 1/2 arc) and then each section is connected as it is "elevated". Perhaps the small increase in length (as you calculated) gets unknowingly accommodated / adjusted for along the way....

Yes,

I would expect that, building it piecewise, you'd probably not notice that - it would be accommodated in the joints.  

Blair

What need for Oval Shape?

Blair made a pretty interesting analysis.

I would ask why you specifically need that oval shape? If you have the room, the simple circular shape is easier to build.

I thought I did a pretty thorough analysis, and reanalysis, of where I needed to drill all of my holes for the 'L bracket supports of my roadbed for the helix,...YET I still found little discrepancies of as little as a 1/32" that produced little 'bumps' in my roadbed. I had to go back an enlarge a few of those screw holds to fine tune the situation. Having two short (1 foot I think you said) segments in each loop of your helix makes thing the much more problematic,...I think.

stretching a spring

Brian the 'stretching' problem Blair refers to has nothing to do with oval vs circular, the stretching aspect was warned about due to Paul Mac's plan to build the entire helix flat, all layers right on top of each other, and then hoist it vertically using the threaded rods. Hoisting a flat curve into an upward sloping spiral makes the curve tend to twist toward the center of the curve. Flat roadbed will tend to cant inwards (super-elevate on the outside of the curve), and the free end will tend to angle inwards.

If you build a helix with the layers hoisted as you build it up layer by layer you compensate for this twisting as you go. If you make all your measurements extra-carefully but with everything flat, your measurements will be off a little when you assemble the pieces into a sloping spiral.

The math is beyond me, but I can easily see it throwing measurements 1/32" off or more per layer, with that compounding as you go up.

One guy on the internet who posted a video of him doing a helix ended up with his threaded rods bent at wild angles because he pre-drilled the holes for the rods with everything flat and had no compensation mechanism built in for the stretching so the rods had to be bent. Honestly, it looked more like a how-not-to-do-it video.

I suspect that people who use double overlapping trapezoid sections usually build sections (say a quarter of a circle) flat, then attach them in the already-hoisted position, and when they do they end up with cracks where the larger pieces attach to one another, a wider crack on the inside, and that compensates for the effect.

There ought to be a way to do the math to exactly calculate the perfect dimensions of the trapezoids so that when constructed flat the curve gets slightly larger with each revolution, and then when hoisted the narrowing makes everything line back up again. Maybe some of the ex-engineers can do the math, or calculus, whichever it is. Sins and cosines, I haven't used all that stuff since I graduated and don't remember it. In practice people just force things to fit, like you drilling out your holes 1/32" to the side.

Space....

..the final frontier. Its a simple matter of acquiescing to the available space.

To not exceed the 1.85% grade without straight sections would require that I increase the radius of the helix. I don't have the space to do that but do have the space to extend it one direction.

My helix

Paul: My layout is HO, the percentage of grade is a little under 3% and I built it to acomodate double stacks. I have about an 1/8 inch clearance over the double stack.

My bad

Bob - when I read that you had a 24 inch diameter helix I inferred you were in N scale. You probably meant to say "radius"...   It's all good!

My mistake

I ment to say radius.

I used 1/2" plywood for my

I used 1/2" plywood for my subroadbed:

Mark,

Please tell us how you built it, what reinforces the joints, are the segments 1/4 turn, 1/8 turn, or less, did you build it flat and elevate, or elevate as you built it?

As for my earlier comments, I really expect that if you build it in sections, rising as you go, the difference in measurements between flat 45 degree segments and, say, the same segments at a 2% grade will be so small that you will happily never encounter them, as the tolerances on cutting the segments are of the same order.

The challenge lies purely in trying to build them flat, then elevate afterwards.  Glad to hear that there's a youtube demonstration of the problem, although I didn't find it with a cursory search.

Blair

helical stretching

the same is true for any grade, the true length of the diagonal, the hypotenuse, is longer than the horizontal distance.

the following table shows the circumference of circles of various radii and their additional length due to grades from 1-5%.   For a 32" radius, the additional length is 0.04" for a 2% actual grade.

it doesn't appear to be significant for small grades.

 rad   circ     1%     2%     3%     4%     5%

  24  150.8   0.01   0.03   0.07   0.12   0.19

  28  175.9   0.01   0.04   0.08   0.14   0.22

  32  201.1   0.01   0.04   0.09   0.16   0.25

  36  226.2   0.01   0.05   0.10   0.18   0.28

  40  251.3   0.01   0.05   0.11   0.20   0.31

  44  276.5   0.01   0.06   0.12   0.22   0.35

  48  301.6   0.02   0.06   0.14   0.24   0.38

@ Greg,

actually, reading the two-decimal-place table, for a 32" radius it states 0.04"; given rounding, that could be anything from 0.035" to 0.044999999" (sorry, no repeating bar over the nines), so I'll take my back-of-the-envelope calculation of 0.034".

Blair

.

would you mind sharing how you calculated 0.034?

doubling 3/16" ply with overlapping

https://forum.mrhmag.com/post/pokey-progress-onward-upward-jan-15-12199338

Hi Blair, here are some particulars of my helix

 I used a 4 1/2" railhead-to-railhead separation between tiers.

Here's a plan view of the helix:

It was a four track helix, with the inner radius being 36", and the outer radius being 42". The curves were split in half, with a 24" straight section inserted to increase the track length on each tier and reduce the grade. Grade on the inner track (steepest grade) was 1.65%. giving a compensated grade of 2.43%.

The helix was hung from a 3/4" cabinet-grade birch plywood header that was attached to the ceiling. You can see the header and the top tier here:

This made sure the threaded rods were all in tension, which meant the helix was stable in it's normal state (if the rods were in compression the helix would be unstable in it's normal state, as it would want to collapse sideways. Only the top deck subroadbed connections to the helix would stabilize it).

Here's how I installed the tiers initially:

Each tier semicircle was cut from one 4X8 sheet of plywood. The straight sections were cut from the removed center piece of each sheet of plywood. 

Scrap sections of the plywood were used as splice plates to tie each semicircle to the straight sections as pieces were lowered into position as construction progressed. The splice plates were wood-glued on, with screws going down from the subroadbed into the splice plates for added strength. You can see some of the splice plates here:

The splice plates stretched about 3' long each, from the ends of the circular roadbed sections right across the 24" straight sections. This minimized any waviness that might have occurred across shorter splices (I had LOTS of extra plywood, and I wanted to minimize any issues with warping plywood).

Just for laughs, here's some additional stats on my helix:

And here's my parts list (minus wire to power the thing):

This construction approach resulted in a very solid foundation for the trains. It was pretty inefficient use of the plywood, but the trade-off there is that I only had four splices in each 360-degree turn of the helix. And I had lots of plywood scraps for splice plates and such on the rest of the layout!

I didn't mean to hijack this thread. Hopefully it will give the OP some thoughts for his own helix, and maybe trigger some ideas more suitable to his situation.

WOW

Mark P, what an elaborate and beautiful construction. How long did it take you?
I'm in awe.

Thanks for sharing!

Mark - this is exactly what I was hoping for. It's great that you took the time to share with us in some amount of detail how you built yours.

A great example of whats called "paying it forward".....

I hope to be able to do the same when I get mine built.

Cheers!

...what an elaborate and

 railandsail,

Thanks for the compliment!

The helix took 364 days to build. I started the day after Thanksgiving 2005 and finished on Thanksgiving 2006.

I did other things on the layout as well, but nearly all of it was necessary to get the helix to completion. I was thoroughly sick of running nuts up and down those threaded rods, and of laying track, for some time afterwards!

sure, Greg.

I use 3PI a lot for sketching stuff.  You'll probably call it cheating, but then back when I used a calculator instead of a slipstick, I was accused of cheating too...

Since 3PI doesn't give true path lengths for even simple multi-segment paths*, I just drew a 1' straight and a 32" semicircle, and added their lengths, then doubled it; all numbers are rounded to 3 DP.

result, 2*(100.531+12.000)

=225.062;

If you use these line segments and build an S*, then select  the path of the S and give it a 4" elevation change, the total path length becomes 

2*(100.547+ 12.002)

=225.098; difference is 0.036, so somewhere I dropped two thou.  My bad.  

By the way, I wasn't picking a fight, if you thought I was; I just noted you had quoted the wrong elevation change from the table.

Blair

* This little test has pointed out a couple of challenges in 3PI which i wasn't aware of.