Four years ago, right here in

Four years ago, right here in this blog, I posted this sketch of my plan. The four sections along the west wall (the town of Park) were operational and I was propelled by the enthusiasm of my students for model railroad operations to envision a more complex railroad which could keep as many as 20 students busy in a variety of roles and operating scenarios.

The scheme is a small branch that crosses a Class I mainline, sharing trackage through a junction yard in a port town with carfloat connections to other coastal destinations. The mainline is represented by staging on either end of the yard, the branch has a terminal town at one end (the town of Day) and at the other end track continues past the modeled town of Park into staging that represents a link to another railroad further along the branch.

Last year we started operating the yard and installed a temporary section to represent points east. This year we're building that section in the southwest corner, left of "to lower level staging" in the plan.

Jeff Allen

Back when I was cooking up

Back when I was cooking up that original plan I could tell that the deck-to-deck separation on the Day branch extension was going to be tight. I don’t have much running room from the junction where the branch takes off from the main and can’t afford the space for a helix: everything has to hug the walls to leave some room for “normal” classroom activities.

Here’s the grade profile:

The total railhead-to-railhead separation is just 5 inches, give or take an inch if I take grades to extremes! I’m going to fabricate the upper deck from ½” homosote laminated to ¼” plywood sitting directly on risers located around the staging tracks to minimize unsupported spans. So the actual clearance isn’t going to be much more than the minimum tunnel height!

Here's the plan. Sorry about the sketches, that's how I work and I'm too lazy to redraw everything for better communication:

I’ve operated on several layouts with hidden, difficult to access staging. Some with cameras, others with occupancy detectors. Workable, but I can see why experienced folks swear off of hidden staging.

Jeff Allen

Am I out of my mind?

What can I do to mitigate the worst of the problems? I’m considering hinging the upper deck panels to be able to get at hairballs in the staging area below. That would be a drag, but would it be worthwhile? Any other ideas?

Jeff Allen

Clearance

5 in railhead to railhead is, in my opinion, way too tight.  By the time you get the upper benchwork, wires, etc, you are down to less than 4 inches.  There is NO way to work on track, you can't see anything, to have to remove ALL the equipment in front of things to reach anything in back.  Anything protruding down or sagging from the upper level (wires, switch linkages, tree trunks, uncoupling linkages, etc.) is liable to foul the lower tracks.

Yes, you can make it work.  It will be good right up until something goes wrong, then it will be horrible.

Thanks for cheering me up.

You don't like the hinged upper deck idea?

Jeff Allen 

Or...

#1-I generally don't like having hinged areas that are scenicked. Too much potential for something to go wrong and fatally injure some time-consuming work.

#2-Take that idea and flip it: make the staging area hinged to drop down for track work.

#3-What about continuing the staging on a downward grade of about 1%? Keeps all the trains collected on the lower end and adds additional clearance as you go.

#4-Drawer staging. I've seen several instances where the layout builder just made a giant traverser that could hold all or most of a train on each track. Consider putting a short length of track on a stationary segment opposite the entrance end to provide a motive power escape spot. Use slide barrel bolts on the topside to align each track with the entrance lead and possibly as power conductors from the main to each track. Eliminates expensive and space-consuming ladder switches.

Hope I've given you some good food for thought. 

Mike in CO

Good ideas

Thanks Mike, I'm a glutton for food for thought!

I completely agree about tilt-up scenes, especially in this case where I'd like to have an icing platform close to the wall (that would get crushed when tilting up). I might be able to put the icing platform outboard on that spur that I was treating as a team track - somewhat different operationally, but it would allow the tilt-up solution.

I've thought a lot about various drawers and tilt-downs for the lower deck, but it seems to me that when there's a problem on that level it's better to untangle it without shearing the derailed rolling stock. 

Good thought about continuing the downgrade, but I've calculated that even with the steepest possible grades the length of track would still not provide enough clearance to get eyes and hands to the inner staging tracks in the even of a derailment. I'm pretty sure (not certain) that some sort of access hatch is my only option. What I'm really wondering is if making that entire section an access hatch is worth the cost in scenic restrictions, operational interruption and section joint alignment headache when compared to the effort to fish derailed trains out of the hole when, as David points out, something goes wrong. 

Jeff Allen

Contributing to my delusional frame of mind...

... is the dramatic increase in operational reliability we've experienced since one of my students took on the task of screening rolling stock for coupler height compliance. Last year all rolling stock was tested for "free rollingness" and non-complying cars set aside. A couple of weeks ago one student screened out cars that didn't meet the Kadee coupler height gauge test, and our overall reliability has been MUCH better. We still haven't screened for wheel gauge and I suspect that we'll weed out some problems there (though I'm pretty sure some of the turnouts are non-conforming too). 

I have a bulk pack of metal wheel sets, ready to swap out with any existing plastic ones. 

Those steps, along with rerailers at key points on the lower deck tracks, should reduce the incidence of things "going wrong, ... horrible."  The question is, will it be reduced to tolerable levels? Fish out the wreckage, or clear the top deck and lift it for skyhook resolution?

Jeff Allen

Have you considered??

Hi Jeff, 

Even that track separation may be unobtainable because of the length of the vertical transition curves needed to get you on and off the grade when using bogie US stock and standard knuckle couplings. 

I've posted this diagram before that illustrates the "trap" -

just because the grade is 2% on the main grade as drawn to get the same vertical separation with vertical curves included then the actual grade may have to be far steeper than your planned 2%.

Others may be able to provide some guidance as to how much length is needed for a vertical curve from flat to 2% without uncoupling on US bogie rolling stock. As a general rule of thumb to prevent uncoupling and potential runaways, for safety you should always keep at least 1/2 the knuckle height engaged while transitioning on or off the grade. Don't be too surprised to find that the vertical transition curves could eat 1/3 of the total length of the grade. 

You WILL NOT be able to pull the types of vertical curves I can get away with because I run short rolling stock one at a time over this type of vertical curve one at a time, so avoiding the uncoupling problem. Even this vertical curve is longer than 6" long. There are also some tricks for making a smooth vertical curve this tight by modifying the rails. 

I hope that this isn't too discouraging, but it is better to be forewarned. The rules are different when you stop "playing on the flat". 

...

I'm not clear on how it would be acceptable to the Fire Marshal, the school district or your own sense of liability risk tolerance to have any form of obstruction across a doorway of a school classroom, as presently shown in the plan. The clearance curve near the car float still would be viewed as an obstruction. I doubt there's any legal way you can extend track beyond the turntable and car float, across the door in any manner.

Wow, great feedback.

This forum is a treasure, thanks Joe. 

John, your work has certainly taken the problems of vertical curves to the extremes and I sincerely appreciate your insights. I spend some time with my Algebra students studying the way these curves work and why they are important, since they are parabolic and hence relevant to the quadratics they are learning about:

My calculations run from grade to grade and allow me to install whatever length of vertical curve I need according to coupler displacement allowances. I usually aim for 48" minimum horizontal length, but that's one of the variables I can manipulate for my clearance problem. In this case it affects the critical clearance at the point where the staging ladder dives under the upper deck, but it doesn't impact the final total clearance I'm struggling with.

Chris, excellent observation. I'm planning to treat the design of that section in a separate post, but suffice it to say you are quite correct. The fire marshall came through a few weeks ago in advance of the Maker Faire event that took place in my classroom and I had the opportunity to review the plan with her then. I was pleased to hear that she's receptive to the facts that my design will be a) normally open, no obstruction of the fire exit andb) when in operation outside of normal class hours will include an approved panic bar (just like the one on the exit door) to throw the gate open. The actual installation will still require her approval but she's not opposed in principle. 

Great feedback, keep it coming! 

Jeff Allen

At the top of the hill

Hi Jeff,

The real problem at the top of the hill is not so much bottoming out in the middle of the wagon but the way the trailing coupling lifts as the leading uphill wagon comes off the grade relative to the coupling on the next trailing wagon that is still on the grade or transition curve.If this height difference is more than 1/2 a coupling height between the 2 wagons then you are heading into breakaway territory with a BIG WHOOSH. A minor track jiggle could be enough to trigger a WHOOSH if you are better than 1/2 a coupling height out. That's why transition curves can eat up so much horizontal distance.

Working the geometry out on that one is a whole lot more complex when you try to track where the coupling heads are in space relative to the track and each other. Things like car length, bogie centres, distance from bogie pivot to coupler head all come into it. Any vertical play in either of the couplings further complicates things. That's why I've suggested the 1/2 coupling height difference as an indicative safe limit. 

At the bottom of the hill it doesn't matter as much - you just leave stuff behind "on the flat". That stuff won't roll too far or too fast 

Plan for failure or plan for success ?

My two cents worth .....

I have restricted access to staging on the Cornhill and Atherton RR, but I spent a lot of time on bullet proof build of trackwork and testing. I decided that the amount of work needed to make the staging hingable, slideable or whatever or the restrictions on the build of the deck and scenery above were not worth the effort, when compared to the low likelihood (and unknown nature) of failure.
Yes, it will probably "fail" in someway at some time in the future and when it does I'll take it on the chin and deal with it.
It's a mindset for me in that you can plan for failure, or plan for success and I prefer the latter because all of the work is constructive rather than defensive.

Rob Clark

That's the spirit!

Thanks Rob, after considering all of the options and constraints I've decided to proceed on exactly that basis. I think Dave Husman's original assessment still stands, vivid and undoubtedly accurate:

So I'll just do my darnedest to make it work as well as possible! For example, I don't think the small amount of extra headroom I can gain by increasing the grades will materially alter the "horribility" of dealing with a problem, so I've opted to keep the grades minimal (thereby also allowing less severe vertical curves at the transitions), and to adjust turnout locations so that none of them fall on the vertical curves, and are all fully exposed. Then, as you say, lots of care and testing in the track work. 

Having established the lower deck elevation, I had to cut a few inches off of the legs of the old IKEA shelf unit I'm using to support the panel frame at the end of the branch:

That frame is now in place:

That slight taper on the far end was a design change that will allow the fascia opening to the lower deck to be closer to the outermost staging track. Cars will still have to be pulled to get at inner tracks. The staging tracks stop at the end of the shelving so I may be able to get at the ends by popping up through that opening, though the turntable mechanism above it could be limiting.  

Once the students return from Thanksgiving break this week, it'll be on to the rest of the panel frames on this branch! 

Jeff Allen

Two things

Hi Jeff,

2 things you might like to try to make sure that this will work for you. 

1. Make a test rig for coming off the grade and onto the flat using the calculations you have made before you attempt to build the grades into the layout for real. As I've said previously, the top of the grade is far more critical. If the calculations are OK or not, then building a quick a test rig will verify them. Its called an "experiment". Remember the top of the hill is more critical than the bottom of the grade, but a quick test rig there may also be a worthwhile exercise. Make sure you can get at least 2 full carlengths clear of the vertical transition curve. This ensures everything is either fully on the grade or flat after the test. 
2. Now when/if you make the test rig, By design, rail doesn't bend well vertically and our vertical curves are probably far tighter than the real world would ever attempt in 1:1 scale. 
   Here's the best construction tip I can give you for making the test rig. If you cut through both the foot and and web of the rail from below with a thin Dremel disc (or even a thin hacksaw blade) every 1/2" or so, leaving only the rail head itself uncut, then you will have a very flexible rail that will bend both ways into a smooth vertical curve without kinking. How this works - going from flat to grade at the bottom of the grade, the gaps caused by the cuts open up to give the rail flexibility. Going from the grade to the flat at the top of the hill, the gaps caused by the cuts close up to give the rail flexibility.

As an exercise - cut 3 sections of rail say 6" long, 2 lengths are the standard rail, the other length has been "doctored" as in #2 above. For one of the standard lengths of rail, have the students try to bend the standard rail vertically. Don't worry if they kink it because it will twist and kink sideways before it will bend smoothly vertically. Why did it kink that way??  Because of the shape of the rail and because the "elastic limit" of the rail has been exceeded. When you try to bend a rail vertically (either way) you are trying compress or tension the heaviest part of the rail's cross-section, but the weakest (thinner part the web) will yield first and kink. A kink is a "permanent deformation" This leads to stress/strain analysis used for testing materials to prove that a material is "fit for purpose" and won't deform permanently under load. 

Pass the other 2 lengths around an have the students feel how much easier the doctored rail will take a vertical curve. Emphasise that you want the to feel how hard it is to "just start" to get a vertical curve, so that the rail is just starting to bend and not to kink like the other rail did.

Will the rail be safe on a vertical curve?? Will it break? How do you know?? Have the students measure the rail cross section. For the rail head itself take the measured dimensions and consider the rail head cross-section to be rectangular rather than rounded. This is the narrowest section of the rail where any structural failure will occur under load. Discuss the concept of "point loading" where the weight of the wagons can be considered to be a load applied at a single point where the wheel meets the rail. For a selection of wagons/locomotives, weigh each and divide by the number of axles. This will give you a "maximum point loading" for your rolling stock in pounds. Stress is simply Force/Area - calculate the stress value This is your Design Stress.  Use the 1/2 hard specification for the Yield Stress to work out a Safety Factor. (simply Yield Stress/Design Stress). Given the low point loads, I'd expect this Safety Factor to be> 100, probably even heading towards 1000. So the weight of the trains shouldn't break the rails, even though that have been significantly weakened.   

For your consideration,       

John, can I send you a ticket?

I like the way your mind works John, I'd love to see you in action in front of a group of my students! 

If I understand your #2 step above you're suggesting what I would call a "kerf" in the bottom of the rail to ease the bending stresses. I've used those when bending homosote roadbed sections in the horizontal plane and the kids get the idea pretty well, so this should make sense to them. But, my vertical curves are probably too subtle to need this step - a change of 2% in about 48 inches - which unmodified Micro Engineering code 83 flex track seems to conform to without any modifications. 

Your point loading topic plays right into an important element of the curriculum: surface area. How much surface area of a circular wheel is in contact with the linear rail? Mathematically, zero! But the "real world" considerations mean that the steel tire will "mash" slightly to create an ovoid area of contact, and the total area times the pounds-per-square-inch force will equal the total weight of the load. We perform this very calculation with a student on a bicycle, carefully measuring the pressure in the tires and the area of the tire prints on pieces of paper. Seeing that product equaling the weight of the student plus the bicycle always blows the thirteen-year-old minds. 

I can tell you enjoy blowing minds!

Jeff Allen

Vertical to Horizontal

Now consider the energy required to move an object laterally / horizontally vs vertically. The same amount of mass on a smaller footprint makes it quite easy. Think of those Kung Fu movies! A bicycle wheel will move easily under a two hundred pound rider when they are knobbies on pavement vs a road tire. Why? (Hint is surface area). Ask me how I know. Ok. Let’s not. 

Imagine the same railcar with steel wheels on steel rails. Maybe the tread taper isn’t only for curves?

More on vertical/horizontal...

Oh, don't get me started on conical tapered tread profiles Neil! 

Here's how I introduce the rationale for rails: to pull a buggy forward on a dirt road, a horse has to lift that buggy vertically the distance that the wheels have sunk into the dirt. Maybe only a half inch, but that's still quite a heavy lift.

That same horse can pull a much heavier car with hard steel wheels riding on hard steel rails because it barely has to lift it at all.

I get a double whammy out of that second photo because it was taken a block from my school. We can walk to the spot where it was taken and see that building still standing.

More blown minds as history comes alive!

Jeff Allen

Length of vertical curve

  VC's are calculated from PIVC(point of intersection of vertical curve) to PIVC so they're half on the previous grade and half on the ahead grade. By definition they do not steepen the calculated grade. We had a thread many years ago about the top and bottom change of grade affecting couplers and IIRC it came out that a rate of change of 1% per car length would keep everything coupled.  So a 2% grade uphill going to a level yard would need a VC of about 2 feet ( one foot each side of the PIVC) for cars up to 87 HO scale feet long. One percent of grade change per foot gives .12 inch of car pitch so the extended coupler pitch would give .03 inch of mismatch (using a 4:1 car to coupler overhang ratio),a kadee coupler head is about 5/32 in or 0.16 so only 1/5 of it is misaligned.   Thus  4% grade to level yard would require a 4 foot VC , half on the hill side and half in the yard side of the PIVC, etc.  Note a 2% up to 2% down VC would need to be 4 feet long, 2 feet up and 2 feet down,and a 4% up to 4% down would need 8 feet.. If 40 foot cars instead of 87 foot cars were used the length of curves would shorten accordingly. ( a 4% up to 4% down would be a 44" VC with 22" each side of the PIVC . These are approximate calcs based on what I recall from the old thread. .....DaveB

If it was me I would skip the

If it was me I would skip the dead end section above staging and just build the staging and keep it on the level. If one of the staging tracks needs to be used as a lead so be it. Right now it looks like you are trying to put 10 lbs of crap in a 5 lb sack and that never works well. You would be better off with less railroad that works than more that does not.

There is no way your layout surface will permit enough room for hands in between levels let alone a hand and a tool. If I had to put my hands in there I would loose skin going in and out not fun. You mention clearance is about what is available in an HO tunnel portal. After contorting my hand into one of those I can go about 3 inches in before my forearm wedges in the opening.

Now you may have much smaller hands than I do but that is still in my opinion not enough space to make the project worth while.

If you do build this thing I wish you the best of luck and hope you are successful.

My head hurts!

After reading this interesting but complex thread, my head hurts. Math is my least favorite subject so I've never tried to calculate vertical curves coming off or on grades, I just figure them out by visual means and lots of testing.

One question comes to mind about tight  clearances with staging... What about cleaning track?

Too much cart behind the horse?

Access could be easier if there were fewer tracks. Try a staging track instead of a yard? Put it at the front where there is access from the people-walking area. If this is too little for operation, maybe add just one more track, also near the front.

This may be less than half a loaf, but it should be a lot easier to chew.

DrJolS

Plan B

Now we’re getting somewhere! If I build all four tracks I can certainly abandon the back ones if necessary.

Track cleaning is certainly a key concern. I use a couple of different dry light friction track cleaning cars and one of those track cleaning pads on a stick, but as has been pointed out the latter is going to be a pain in the caboose.

Since I dismantle the whole layout every year I’ll be able to do a thorough polishing of the rails and light application of graphite when I set it up. Elsewhere on the layout I’ve found that to be a remarkably effective and durable treatment, especially the polishing. I use 1500 grit paper followed by rubbing with jeweler’s rouge, then buffing, then swiping with graphite. The layout is in operation for about 8 months out of the year, a bit too long for that treatment to hold up so that’s certainly a problem.

DaveB, thanks for that excellent summary of the practical rules of thumb for vertical curves - exactly the ones I use to establish my minimum VC lengths. Since I will be running some full length passenger equipment down a 2% grade into level staging (shorter equipment on the upper deck branch), two foot vertical curves would be the minimum. Once I decided clearance was going to be absurdly small no matter what I did, I gave myself the luxury and increased confidence of 48”.

Rob, you’ve hit the nail on the head for what I call “plan B.” The reason I’m even venturing to attempt the double deck approach is that I’m currently operating with plan B!

It’s a 10 foot section that includes two staging tracks for the main, two for the branch, and switch lead for Idora yard. We’ve been operating with it since the middle of last year. That yard lead is much longer than necessary (great lesson in yard switching), but the staging tracks are pretty short, long enough for the branch trains but not for the main, at least as we’d like to run it.

And I really want the variety of operations and car handling of the produce district terminal branch.

Plan B is also my fallback if I can’t get approval for the gate that crosses the exit. As most of you are aware, those of us in California have become pretty sensitive to fire hazards!

Incidentally, there’s also a Plan C. If I can’t get approval for the gate, I’ll modify the branch to be served by the car float.

Let’s see, so far the overwhelming majority of feedback reminds me that I’m a fool, but if I gerrymander carefully I can extract a mandate to proceed!

Jeff Allen

Increasing separation

Hi Jeff,

If you can go to the double decker, Is it possible to increase separation by reducing the thickness of one or both decks??  For example, if you leave the homasote off the top deck and go directly to a thinner cork layer instead then if the rail goes to where you initially planned height-wise, then you have a thinner top deck and have increased you deck separation by at least 1/4", if not more. If you do away with the homasote and go for something thinner on the lower deck, then you can either run that grade slightly steeper or further, but you have now gained better than 1/2" in deck separation. While 1/2" doesn't sound much, when every fraction of an inch counts, it's about a 10% increase in deck separation for a bit more effort when spiking tracks - and on straight yard type trackage, there needn't be that many spikes if there is some glue assistance? You can now make your support piers slightly longer to compensate for the increased deck separation.. 

Here's what 3 mm or 1/8" ply on piers looks like  https://forum.mrhmag.com/post/corrimal-colliery-and-its-incline-a-different-slant-on-rails-12206968  This requires more piers, which would be an accessibility problem which could probably outweigh any deck separation benefits, bit if it had an some framing for support to reduce the number of piers required (even on top of the upper deck) you are now approaching a 3/4" increase in deck separation. If the top deck is only screwed to the support piers, then the lower tracks can be got at relatively easily if you need to. As shown in the link, the piers are glued and screwed at their lower ends. A possible Plan D??  

The only problem with the 3mm ply option is that the trackpins protrude through the bottom of the ply.These "nasty surprises" can be either amptutated with a Dremel or simply pulled completely (safest option) once the track is ballasted on the top deck. Having the track ballasted also increases the rigidity of the thinner top deck.by effectively turning the tracks into support framing.that will help prevent both longitudinal and transverse sagging.