Question regarding grade transitions

[Lemosteam]

  On the Layout engineering thread I am contemplating a 5-6% (switching only) grade on my LIRR Industrial Service Inc. layout.  Like a tangent track handles lateral transition, how should I go about handling vertical transition; i.e. from horizontal to grade down, to horizontal grade at a lower level.  The angle may be 5% over 40" of length or 6% over 33".

Greatly exaggerated:
_______
              \
                \
                  \_______

   Are there any rules or worries?  One comes to mind and that's vertical coupler face alignment/engagement as one car is level and another is at 5% right at the apex of the slope.  Any others or am I simply fearing the worst? 

   Should I gradually radius the top and gradually fillet the bottom?  I fear this will add to the length forcing me to increase the grade...

[Cajonpassfan]

John, I don't think there is any formula, other than common sense, that will solve your theoretical challenges; too many variables such as rolling stock, power to be used etc. I would simply build a mockup and try it with whatever cars and locos you intend to use..... Shorter equipment will allow more dramatic changes, it's that simple.
Good luck, sounds like a fun project!
Otto

[nkalanaga]

I can run 2-bay MT hoppers up an 8%+ grade without trouble at the transitions, but can't push them with an Arnold Alco.  The loco trucks won't take the vertical curves.

At least as important as equipment length is coupler mountings.  Conventional truck mounted couplers will work better on grade changes than body mounts, but those on extended arms may not. 

Another factor is the length of the vertical curve.  At one extreme, a sharp bend will almost guarantee derailments, no matter what you run.  A gentle grade change will accept any train, as on the prototype, but may not fit on a layout.

As Otto said, experimentation is probably the only way to see what works for you.

[mmagliaro]

I don't have a "scientific" method for this either.  Here's my $.02.
I generally start with nothing more radical than 0.5% grade off a level.  I try to go 8-10 inches for each 0.5% increase,
until I get up to the grade I want.  So for a 2% grade, it sill take me4x8 = 32" to get up to 2%.   But this is just a rule of thumb.

Having said that, one way to look at it is based on the longest rigid wheelbase you plan to run over it, and whether your transition jump will cause any wheel flanges to lift up over the rails.   Imagine a 4-8-4 with 80" drivers.  That's about 1.5" across the 4 wheel centers.  Let's say the flanges are .025" deep.    Well, draw your proposed vertical grade
on a piece of paper, with a 1.5" rigid "wheelbase" (just a line) trying to go up on it, and see if any of those
middle wheels will end up hanging in mid air with the flanges up above the rails. 

.. (!  I hope you can visualize all this !)

Now, if you are just running small short-wheelbase things, this can be a lot more forgiving, and you probably only have to worry about coupler trip pins bottoming out on the rising track in front of you, or cars uncoupling.

5% in one shot strikes me as asking for trouble.
Since I think your wheelbases will be short, you probably only need 3" or so at each grade.   So 1.5%, 3%, 4.5% would
gobble up 9-10".

[Lemosteam]

Cajonpassfan, thanks I agree I might have to mock it up.

nkalanaga, thanks for the pointers.

Max, thanks I understand completely.  So rather than transitioning directly into the desired grade, you do this by gradually raising the grade over distance, almost like chording an arc in multiple sections until you get to a tangent at the desired grade.  This will add significant length to my transition.    

I understand your example at the bottom of a grade but do you do the same on the way down?  I'm imagining a set of drivers extending off the rails until the loco tips back onto the rails, LOL!

I think I may be down around 3.5% now, see the Layout Engineering Thread for my proposed shelf layout:
https://www.therailwire.net/forum/index.php?topic=28266.0  David K Smith, has helped tremendously express my concept in AnyRail...

[Rossford Yard]

The biggest problem is with a turnout at the base.  If a car is still coming down a grade, the frog will catch the coupler trip pins, so be sure to be very gentle at least a car length + from the frog.

I figure you could probably lay out flex track from the middle of the 5% grade, and fix the other end on the flat part , and wouldn't the track sort of assume a parabolic grade that would gradually rise?  If it didn't flex naturally vertically, you are probably too steep.

[RWCJr]

John Armstrong's book Track Planning for Realistic Operation addressed "vertical curves", or grade changes. He recommends transistion grades at the top and bottom. As you noted transistions take up a lot of linear track space. It could be very helpful to make a test track to see just how gradual your transistion needs to be. My favorite locomotives are Northern type steamers, and they require fairly long and gradual starts and ends to grades. But todays N 4 axle and many 6 axle diesels have very flexible drives allowing shorter transistions as far as engine tracking and traction. But as others have noted, rolling stock and coupler matching will need to be checked, both uphill and down, since there will be different forces on the wheels.

[pnolan48]

In my experience the top of the grade (convex) is more troublesome than the bottom (concave). I think it's easier to ease the concave part than the convex part, just by the way we often build our roadbed.

[kelticsylk]

I don't have years of experience with creating vertical transitions, but I have had success with them. I would think a major obstacle would be the flexibility of the roadbed you are using. On my layout, I use 1/4" (6mm) plywood laminated with 1/2" extruded styrofoam. It's strong enough to do the job and flexible enough to form "natural" vertical curves. I didn't have to work at creating transitions, they form as you raise or lower the roadbed to create the grade.

Most people use stiffer material like 3/8", 1/2" or even 3/4" plywood. I can see where the vertical curves would be an issue. I would think the curve could still be formed "naturally" if you fastened the roadbed to the uprights so that you take advantage of the natural tendency of the wood to bend down from its own weight. This does require that you use one long piece of roadbed through the entire transition curve...

The thickness of your roadbed will dictate how fast the transition will occur. The two uprights shown are where the roadbed sections meet and need to be fastened to make the thing work.  There can be as many intermediate supports as required. Just make sure they are mounted to your benchwork in a way that allows you to adjust them. I didn't and now I have to do it over.

A similar method is sometimes used to make motorcycle tanks. The metal top is bent around the curve of the side and tack welded to hold it in place. You don't force the metal to bend, you "encourage" it.

FWIW

[nkalanaga]

Frank:  A good point, but, NEVER have a roadbed joint at a change of grade!  The curve will tend to straighten out at the joint, resulting in the kink one is trying to avoid.

In your drawing, if the roadbed is a stiffer material, you might also want to add a lengthwise brace under the "level" part, making a T girder out of it.  I've found that if one doesn't, the transition tends to continue through the rise, making a hump (or dip, at the bottom) in the "level" portion, with the riser acting as a bearing rather than a pier.  Here "level" means "constant grade", not "horizontal".

[mmagliaro]

Cajonpassfan, thanks I agree I might have to mock it up.

nkalanaga, thanks for the pointers.

Max, thanks I understand completely.  So rather than transitioning directly into the desired grade, you do this by gradually raising the grade over distance, almost like chording an arc in multiple sections until you get to a tangent at the desired grade.  This will add significant length to my transition.    

I understand your example at the bottom of a grade but do you do the same on the way down?  I'm imagining a set of drivers extending off the rails until the loco tips back onto the rails, LOL!

I think I may be down around 3.5% now, see the Layout Engineering Thread for my proposed shelf layout:
https://www.therailwire.net/forum/index.php?topic=28266.0  David K Smith, has helped tremendously express my concept in AnyRail...

YES  definitiely I do the same thing at the top of the grade (I think that's what you meant by "on the way down").
If the engine crests a 5% grade right back onto level track, yes, the front wheels will be hanging in mid-air until the rest of the engine gets over the hump. It's not that extreme, but it does happen a little.
Yes, I make the grade drop in steps at the top as well as the bottom.

[spr1955]

I have a logging line switchback that is about 5.5% and run an Atlas shay on it without any problems (grade and curves). My only advice to add to the transition approach, which is what I used, is to make sure your subroadbed joints are not in the "transition zone". Keep them along the grade once you have it and way before and after your bottom and top .

David P.