NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/rec/models/railroad/2439

Text File | 1992-11-23 | 10.2 KB | 245 lines

Newsgroups: rec.models.railroad
Path: sparky!uunet!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!cwatters
From: [e-mail] (Coyt D Watters)
Subject: Re: Past posts on building a helix?
Message-ID: <[e-mail]>
Sender: [e-mail]
Nntp-Posting-Host: top.magnus.acs.ohio-state.edu
Organization: The Ohio State University
References: <[e-mail]> <[e-mail]>
Date: Mon, 23 Nov 1992 19:33:37 GMT
Lines: 232


This is a summary of some of the posts (and email) I rec'd re: the helix.
I've tried to keep attribution correct.

Date: Tue, 6 Oct 92 20:14:04 EDT
From: [e-mail] (Van Kelly)
To: [e-mail]

Dear Coyt,

I've pondered a problem similar to yours on a few sleepless nights.  I
think I can make a few modestly innovative suggestions to ease your
construction headaches.

The bad news is that 20-24" height is outside of the range one can
tackle with the simplest "spiral" helix technique.  This is where you
just jig-saw a spiral cut in a 4'x 4' piece of thin plywood and then pull
the "center" up to form a helix with radius decreasing from about 22"
on the bottom to about 13" at the top.  This kind of N scale helix is
the most trivial to support and conceal, and makes train rescue
especially easy from above.  But it's obviously restricted to about
12-14" of rise with a single track, or 6-7" rise with a double track.

For a wider range of helix sizes with only modestly more work, I have
prototyped some new construction techniques that avoid some
traditional helix pitfalls: wasted plywood, weak butt joints, and
clumsy final grade adjustment.  Although I haven't yet used these in
an actual railroad, I did test them out on a reduced-size helix model.
They worked great there, but maybe you can point out some non-obvious
scaling-up flaws. I can assure you that my techniques are simpler than
those required for any  elevator or "vertical switch".

1. To eliminate *all* butt joints in the helix, which weaken the structure
   and can produce abrupt changes in grade, laminate the entire roadbed
   as one continuous curved piece using half-thickness material
   (e.g., 1/8" or 1/4" ply for N scale) and ordinary (yellow)
   woodworkers glue.  Start by cutting a bunch of identical 60 or 90 degree
   curved roadbed sections from the plywood,  say 4" wide for a
   double-track N helix.  The shorter the sections (60 instead of 90),
   the less plywood wasted, for just a *little* extra cutting effort.

   Layout semicircles or 3/4-circles of roadbed from these shorter sections
   on some flat surface, say, a sheet of 3/4" ply.  Laminate the second roadbed
   layer on top of the first, but staggered by half the length of a roadbed
   section. Use either clamps, screws, or weights (cinder blocks) to hold the
   layers together while the glue dries.  To eliminate the waiting,
   build from foam-core and *water-based* contact cement (Elmers).

   Now you have some long smooth pieces of pre-fab roadbed to work with.
   Sand them smooth while it's still really easy to do so.  Next you can
   continue laminating these longer sections into one continuous piece.
   You can even do the whole job in one fell swoop at the workbench, providing
   you drive some temporary screws/nails to keep everything lined up.  Or you
   can do the final lamination operations in-place on the layout, but you
   will need both clamps and temporary screws/bolts.

2. You want a support system that lets you fine-tune the grade
   at each level without affecting other levels.  The traditional spacer
   blocks between levels that some HO guys use don't allow this.  One
   solution that has worked in the past is to support the roadbed on
   vertically oriented lengths of continuously threaded rod stock, with
   nuts and washers holding the roadbed in place.  You can adjust vertical
   spacing at any time with just a wrench.  However, setting up a
   pre-lam helix this way is awkward, even if you do the final lamination
   operations in-place.

   A cheaper, faster alternative is to use simple vertical wooden studs
   (say, 1x2 poplar) inside and outside the helix, supporting the roadbed
   on eccentric cams screwed into the studs.
   These cams are just 1" diameter wooden circles (say, 3/4" thick by
   1" diameter dowel sections) with "center" holes drilled 1/8" to 1/4"
   off-center.   You screw these into the studs at intervals equal to the
   *nominal* vertical track spacing, but you can diddle the roadbed
   elevation up or down a full 1/4" - 1/2" just by twisting a cam.  When
   you've got it all set right, just screw'em tight.  A spot of glue and
   a brad or screw holds the roadbed in place wherever it show signs of
   not wanting to lie flat on a cam, or you can just mount another cam
   above the roadbed and clamp down.  Especially with this latter
   double-clamped construction, you have more than enough built-in
   "slip slop" to compensate for any expansion and contraction in your
   wood underframe.

Hope these ideas help.

Van Kelly
[e-mail]

From: [e-mail] (Bill Brown)
Newsgroups: rec.models.railroad
Subject: Re: Helix Calculation
Date: 6 Oct 1992 20:26:23 GMT

Well - it shouldn't be too hard.  The circumference of a circle is

    2 * pi * radius

So one turn is equal to a run of 2 * 2.14 * 22" = 138"  One of these
babies builds mileade in a hurry - sounds like you would need about
7 turns, which is 80' or over 2 scale miles!  All normally hidden!

A grade of 2.5% (maybe a little steep, but maybe in the ball-park) gives
a gain in elevation of just under 3.5" per turn.  Probably the minimum
clearance one could live with, allowing for maintenance, derailments, etc.
and also taking into account the thickness of the roadbed.

A while back John Armstrong had a (HO) trackplan in MR where he put at
least one passing siding on a helix.  Sounds like a good place to hide
one!

Elevators are another problem - I've day-dreamed about both abd a helix
seems simpler IFF you have the room for it.

If I blew the arithmetic - well - that's the way it goes!

From: [e-mail] (Urban F)
Subject: Re: Helix Calculation


>Space restrictions dictate that the radius be MAX 22".

If the radius is 56 cm, a full 360 deg turn will be 351 cm.
Assuming a level separation of 7 cm, the grade will be 2%,
which should not be any problem.

From: [e-mail] (John Scott McCauley Jr.)
Subject: Re: Helix Calculation

In article <[e-mail]> [e-mail] (Coyt D Watters) writes:
>Does anyone have a method for designing a helix?  I am building a
>Ntrak compatible layout, but due to space restrictions, am going to
>have a second layer about 20-24" above the first layer.  I want to
>build a helix to bring the loco from the lower level to the upper and
>vice versa.
>
>Space restrictions dictate that the radius be MAX 22".
>
>Anyone have any suggestions, or should I just build an elevator long
>enough to handle my "longer" trains?
>
>-Coyt
>
>
>

I've also thought about building a helix. I think the usual method is to form
the roadbed from 90 to 360 degree cuts from thin plywood or foam core board.

However, using the above method you might wind up with lots of waste material
and nasty joints where the cuts meet.

How about this? (NB: I haven't tried this!) Take a 2'x2'x1/2" of foam core
board, cut it up to make lots of 2'x1/2"x1/2" strips (with a paper cutter
and a jig maybe?), and then glue several of them together to make the roadbed:

               l  l                Rail
             ========              Tie
        --- --- ---            Foam Core strips
           |   |   |   |
        --- --- ---

The 1/2"x1/2" strips of foam core board should be flexible enough to bend
to a 15-22" radius and 2-4% grade. You can stagger the long-way joints
to increase reliability.

The advantage of this method is that there is little wasted material.
The big disadvantage might be cost -- finding flexible, yet strong
material might cost more than plywood.

    Scott


From: [e-mail] (Ken Johnson)
Newsgroups: rec.models.railroad
Subject: Re: Helix Calculation


In article <[e-mail]> [e-mail] (Bill Brown) writes:

>Well - it shouldn't be too hard.  The circumference of a circle is
>
>    2 * pi * radius
>
>So one turn is equal to a run of 2 * 2.14 * 22" = 138"

Not quite.  The circumference is 138 in.  but the track rises (say) 5in.
on one turn of the helix.  So the distance of track per turn of the
helix is sqt(138^2 + 5^2), about 138.1

The gradient of the track is then 5/138.1 or 1 in 27, which is very
steep.  My back of an envelope calculation suggests you need about a
40-inch radius curve to bring the gradient down to 1 in 40, which is the
steepest incline you're likely to be able to climb.

If space is tighter than that, how about doing what the Victorians did:
keep two sets of locomotives (one set on each level) and link the levels
with a rope worked incline?

-- Ken Johnson
From: [e-mail] (Fred Mikkelsen)
Subject: Helix in a layout
Summary: I've only heard of a few successful helicies.

I've been reading on the discussion of the helix. I can only say one
thing:

     N O T !


A helix would seem to be a good way to gain elevation in tight space,
but to get 24" at 2% you need 100'. Usually, this is totally out of
view. This often is *more* track than the rest of the layout combined.

Additionally, a single helix becomes a total bottleneck. That means
at least two tracks (one for each direction) are needed in the helix.

At my club, we discussed building a helix as a way of achieving double
levels. We arived at some design constraints:

Entering a helix would be done only when it would require an operator
quite a bit of time to get to the next viewing angle. Typically, two
turns would be the maximum.

A Helix is best suited for moving things from a staging area. At the
end of an op session, waiting two minutes to put a train back is worse
than waiting two minutes to get a train out.

People from other clubs that I've heard about using a helix have all
expressed dissatisfaction with the delay factor. Big suggestion...use
a helix as a delay line first, then consider the topological advantages
of double stacking, etc. You'll get better results.