Midway to Another Obsession, II

In my previous post I described the lead-up to my shipbuilding project, the USS Midway in 1/800 scale. This time we’ll get into the nitty-gritty of the project.

So, as mentioned, my starting point was to use an Arii 1/800-scale model kit.  And also as mentioned, I soon discovered a few issues and challenges.  Not surprising when your target is in constant motion.  My original goal was to build the model in the post-1986 configuration, but the kit is much closer to 1971.  Once I figured that out, it became a matter of ferreting out which features were added, changed, or removed– and when.  Oh, and there’s that other minor matter: I had to decide when enough was enough.  Just how picky did I want to get with the accuracy?

For the first, the ferreting of features, I figured that photos were far and away the finest and fastest way to find diFFerences (couldn’t think of any synonyms that started with F, sorry).  But it turns out that there are official US Navy drawings on the interwebs that shows the whole 1986 project in fine detail.  Just a simple matter of comparing photos and drawings to the out-of-the-box model, right?  Well, sorta.  I’ll explain in a bit.  But  that’s the basic idea.

For the second, deciding on the level of accuracy and detail, that’s driven by the following factors:

  • Difficulty, i.e. just how many lifetimes would it take to execute the change?  Did I want to widen the hull?  Did I want to get every single bump on every single catwalk correct?
  • Cost.  Do I want to buy frets of aftermarket brass railings, additional aircraft and such?
  • Size.  How fine can you model with styrene in 1/800?

I have lots of styrene structural shapes and in varying thickness of sheet material, so I felt confident I could fabricate most things I needed to do.

Midway’s island as it comes in the kit, starboard side. Here’s where most of the major modifications take place.

Basically, I ended up deciding to focus on the island modifications.  There are major things going on there, as compared with the as-built version– and the 1971 version is not all that different from the 1945 appearance, really.  I also decided to do a few things with the weapons on the sponsons, as mentioned in the previous post.

So: a simple matter of determining the conversion factor between the drawings and the model so I could build the add-ons, right?  WRONG.  I carefully calculated the ratio and measured out parts, painstakingly assembled them to the island, and… and… they just looked wrong.  Something wrong with my arithmetic?  I started measuring stuff again.  Then I’d measure in a different place.  Then I’d reference the empirical measurements on the blueprint.  Nothing added up.  So I started measuring different dimensions on the model and comparing to the blueprint, calculating the scale.  And, guess what?  Depending on the measurement, I got scales ranging between about 1/710 through 1/790.  Um, that’s a 10% variance in some cases!  Guess what, people?  This model is far from being a 1/800 model, or really any scale at all!  Basically it’s too tall for its length.  And that’s being generous in a generalized way. If one were to draw a line down the middle of the points on the graph, best guess is somewhere around 1/755 in average, give or take.

Well, heck.

What this means is: basically build stuff so it looks more-or-less right, and adjust it so that it fits in the available space.

If that’s not enough to make you throw your hard-won eBay prize in the ashcan, we’ll keep going.  Actually knowing that took some of the pressure off, because one of the things I do best is carving and filing.  Okay, that’s two things.

Here’s what you need to add to the island structure, taking the large view:

  • The radar room and platform (as originally built), on the starboard side of the funnel. (There’s one included in the kit, but it’s not right.)
  • The comms room just ahead of this.
  • The upper comms room that sits on the island just forward of the funnel.
  • The radar room / deckhouse structure aft of the island, including its legs and cross-bracing.
  • The radar tower itself (bwahahaha!)

And one other thing: if you think I’m giving you all the dimensions, think again.  Download the drawing and go through the agony yourself!  You’re no better than me.



Modified island, starboard side. In this image, the forward leg of the aft radar structure and the adjoining boxy addition are not yet installed. Note the hole in the deck for installing the crane; this will be plugged when the crane is moved outboard.

So, refer to the photos above and below.  The white stuff is the stuff I added.

Island, port side (i.e. facing the flight deck). Note the bracing on the bottom side of the radar room, and the struts on the back side of the large radar antenna.


Island, view from forward. Note the overhang of the comms rooms to starboard (to the left in this view).

Here’s where I will stop and point out the one single thing I’m most unhappy about: those two projecting comms rooms on the starboard side of the island.  They’re definitely oversize, especially the aft room.  I couldn’t decide whether it was enough of an overage to cut them off and start over, and eventually decided just to keep them.  Oh well.  Once all the other details are added to the superstructure, they’re lost in the clutter somewhat anyway, or at least that’s what I tell myself.

I originally started to rough-in some railings, using trimmings from .010 styrene, but I jumped the gun.  Should have done all of that last.  A lot of them got knocked off with handling.  If you want to spend a couple dozen dollars you can get brass railing material in 1/800; I didn’t bother.  It’s that choice we all have to make about what is good enough for us.

So here’s a few comments about the construction.  I used sheet .020 styrene for all the flat surfaces, which is a great material because you can cut it with scissors.  The legs under the radar house are H-column trimmed to fit.  The radar tower (a challenging sub-assembly) is four legs of thin styrene rod, I believe 1/16″.  Crossmembers are HO-scale 2X2, which is .022 X .022″. The three decks on the tower are .010″ styrene.  The front-right deckhouse leg consists of two laminated strips of .020; the inboard piece extends flush to the deck whereas the outboard layer overlaps the deck edge.  The radar antenna was scored with my Exacto in a crosshatch pattern to simulate the gridwork.

Island add-ons, from port quarter.

I added most of the stock details from the kit; a few are excluded due to later mods.  For instance, after the guns were removed there was no need for a gun director and so it was also uninstalled. There is a section of starboard catwalk that must be trimmed off to make room for the passage at front right of the radar room/deckhouse.  I replaced the radar on the mainmast with the one that would have gone on the lateral radar platform (trimmed slightly on each end to fit).

Completed island, starboard.

There are a couple of relocations that become necessary.  First, the secondary (aft) mast does not get installed in its original location but is shifted to the left and is attached to the port wall of the Radio & ECM Room just forward of the funnel.  Second, the main crane must be shifted outboard to clear the new superstructure elements.  When I did this, I carefully trimmed off the mounting post and used it to fill the hole in the deck, then just cemented the crane to the new location.  I strongly suspect that this was moved on the real ship in the 1966 rebuild as well. A truly great modeler would have taken the time to ream out the spaces in the crane girder, but I lean slightly to the impatient, lazy end of the spectrum (my family would disagree) and I did not go to that trouble.

Side note: my styrene cement of choice is methyl-ethyl ketone (MEK).  Plastruct sells it in a bottle with a brush applicator, but I refill my bottle from a can that I bought at Ace Hardware at a fantastically-reduced price per ounce.  Just be careful that you don’t overdo it or you can ruin delicate parts– and especially paint.

But First!  (Don’t you hate it when you hear that?)  Don’t attach the deck just yet.  The kit comes with an option of installing the No. 2 (starboard side aft) aircraft elevator in the up or the down position.  I decided to go with down, which meant having the hangar doors open (i.e. omit the kit part), which meant, having to install the hangar deck.  I installed some supporting members and cut a piece of .040 styrene for the hangar deck.  Since not much of it’s visible I only did a portion of it.  This also means that you should prepare a couple of aircraft now so you can install them inside the elevator opening.  At this point you want to skip ahead to the discussion of eras and the appropriate air groups for your chosen era. That discussion occurs in a future blog entry which hasn’t been written yet, so I suggest you either wait a bit, or become clairvoyant.

Now, back to the ship.  There are a few other comments I need to make about the construction.  The kit’s instructions have you install the deck-edge whip antennae at a rather early stage of the construction.  If you want the ship rigged as if it is in port, then go ahead.  If you want to show it in operation, read on.  While at sea these antennae swivel out and are basically parallel to the ground… um, water.  This is for obvious reasons.  But the kit doesn’t mold them that way.  So you must cut them and reattach the whips at a 90-degree orientation. This makes them more delicate (ask me how I know).  If I had it to do over again, I’d just wait until nearly everything else is done before tackling that part of the assembly.  I also had to replace three with pieces of .018 wire since they mysteriously vanished. The wire antennae looks so much better than the chunky plastic ones that I wish I had just replaced all of them!  Oh well, maybe some other time.

And speaking of some other time, this blog post has become rather lengthy, so I will address the painting and finishing, and the aircraft complement, in a third installment yet to come.  I must go attend to a family matter (funeral of my mother) next week, so there may be some delay, but fear not.  actionroadlogobang


Backdrop Painting 101

How many of us just nail up a piece of hardboard and start slinging paint at it?

Most model railroads do not exist on an infinite plane.  Rather, and especially with modern layout design philosophies, they represent something closer to a series of dioramas placed end-to-end.  Personally, I fell in love with the diorama concept as an impressionable lad on visits to places such as the Denver Museum of Natural History, or the visitor centers at Mesa Verde and Dinosaur National Monument.  The Denver museum’s dioramas tend(ed) to be rather larger in scale, but the concept is the same: a foreground consisting of a three-dimensional recreation of the subject, and a two-dimensional backdrop painting to convey the illusion of depth.

Depth is what we’re after, here.  Whether you’re building a display for a naval vessel, a B-17G, an Astin Martin, or a model railroad, the backdrop is key to making the eye believe that the exhibit exists in a larger world than it actually does.  Accordingly, it behooveth the one who modeleth to acquire certain skills that pertaineth to creation of said backdrops.  And that is our purpose in this series: to learn how to create convincing backdrops behind our model scenes that enhance (not detract from) them.

In this installment we’ll start at the beginning of a backdrop project.  Next time we’ll have a look at the application of these concepts to a specific situation.  You non-train people, bear with me and just substitute language applicable to your discipline in place of all the railroad-related items.  The principles are the same.

As always, I start with certain assumptions.  First, I assume that you are not a professional artist of some kind.  I assume this because, if you are such, you will probably tear me to shreds, and frankly I’m not sure I can bear that.  So, for our purposes you agree to at least pretend that you are not already an expert artist.  Good enough?  Good.  Second, I assume that you are capable of holding a paint brush or at least a pencil.  Third assumption is that you have some idea where your railroad exists in the world– even if your world is imaginary.

Now, for the steps.  Planning is key, not to state the obvious.  But how many of us just nail up a piece of hardboard and start slinging paint at it?  Before you go any further, the first step is to put down that there paintbrush, pardner!  Set back a spell and listen to a tale from a man called Jim…

First thing you need to do is to stop, think, and visualize your scene.  You’ve built some benchwork, laid and electrified your track, maybe even crafted some landforms.  This tells me that you’ve decided on a location for the scene.  It’s on the plains.  Or it’s in the desert.  Or it’s at milepost 23.6 on the Moffat Line.  Wherever it is, you have defined where the scene exists.

Now that the scene has been located in space, you need to locate it in TIME.  This may or may not be obvious to you, so let me elaborate.  Take a moment and look out your window.  Now, what time is it?  Where’s the sun?  It’s causing objects to cast shadows.  Where do they fall?  How long are they?  What’s the sky look like today, and how’s the weather?  What time of year is it?  Now, step away from the window, picture your scene, and write down its time-based attributes.

Sadly, or perhaps fortunately, the weather doesn’t change much on most model railroads.  On mine it’s always winter, so I designed my scenes to reflect typical December-to-February conditions.  Your choice will have a lot to do with how you proceed with your scenicking, not just how the backdrop will appear.

Once you select a time of year, I’d strongly urge you to gather a selection of photographs taken around that same time.  Even if you’re not reproducing a real scene, find or take some photos of something similar, for the rules of light and shadow are the same regardless.  If you are building a real scene, it is even more imperative that you do this.  Obvious?  Probably, but there are practical reasons for this.  More on this later.

Next, you must build the physical structure of the backdrop.  There are a few choices you can make, so let me discuss my efforts for a moment.  I have built basically three scenes on my layout requiring backdrops.  Two are corner scenes and have curved backdrops; one of these was later extended so I had to add on about 16 feet of additional scenery.  Options are driven by requirements; the curved sections were built using Masonite (hardboard); the extension of the Moffat Tunnel “diorama” was built using foam-core board, and after my sad experience with that, I went back to Masonite for the Front Range section.  For that part I also chose to try jigsawing the horizon and placing skyboards behind it.

So, what are the requirements that drive your options?  I’ll enumerate a few; you can probably think of even more that apply to your situation.

  • Distance from the average viewer, and height of that viewer’s eyes.  This influences the height of the backdrop (i.e. if it’s set back 36″ from the edge of the layout, you may need to make it fairly tall).
  • Support options.  Will it attach to a wall, or must it be free-standing?
  • Shape: flat, or curved?
  • Budget: how much do you want to spend?
  • Domestic: are you able to hammer, saw, screw to walls, and otherwise disturb your household and/or neighbors?  If you live in an apartment, your landlord may have something to say here.
  • Width of scene.  If it’s less than the width of available materials, you might choose a pre-cut material such as foam core rather than a more elegant lumberyard solution.  (Joining backdrop sections can be challenging.)

Back to my experiences.  First of all, using Masonite for the curved backdrops was entirely satisfactory.  Just be sure you have adequate framing to secure it.  It also proved to be good for the long Front Range backdrop– I cut the top edge, you may recall, and that gives my mountains a hint of three-dimensionality that is sometimes hard to get from a flat painted surface.  Masonite is durable and flexible.  Just be aware that it can get rather heavy in large pieces, and can chip or separate with handling if you’re too rough.  Foam core, on the other hand, seemed a perfect solution for a huge section that was wall-mounted.  That is, it was perfect until I airbrushed it with diluted Acrylic paint for the base colors.  The next day when I inspected the dried paint, I found that the moisture had caused the boards to curl over!  It’s like un-ringing a bell; it can’t be undone.  Well, almost can’t be done.  It turned out I had some pieces of 1/2″ trim in my spares pile, and nailed these along the upper edge of the warped foam-core boards to bring them back into a semblance of flatness.  Misting the boards with water helped them relax a bit.  Moral of that story: don’t get foam-core very wet, and I strongly recommend that you secure all edges of it until your paint is good and dry.

Okay, now that you’ve decided on a material, cut it and attach it to the backside of your benchwork.  As mentioned, curved pieces must be securely attached.  However, remember what I said about requirements and options?  My Front Range section actually goes down the middle of my garage, and I thought it wise to make it removable.  So, I built some slots on the back of the scene and slid the backdrop down into them.  This is easy since I can walk right up to the back, but there’s an unexpected implication of this: when building your foreground scenery, you cannot attach to the backdrop.  Plastered hills cannot “lean” on the board, although if you’re good you can shape them to touch it when everything’s installed.  Did I mention that I’m good?   🙂

I did say that we’d do a project in the next installment, so I will try not to get ahead of myself too far here, but there are a few general things you’ll do anyway.  The first of these is to prime the surface.  I’d use a white latex wall primer.  This way your colors will work when you get to that step, and you’ll also be able to see your sketch lines.  Next, referring to your photos or sketch, draw the outlines of major landforms or skyline features such as tall buildings.  Anything that’s a major terrain feature should be sketched, but the most important one is the horizon.

And this brings me to a discussion on point-of-view and how that influences your backdrop.  Any good painting essentially displays the subject from a particular spot, or point of view, and all the picture’s perspective supports that.  (Unless you’re trying to screw with the viewer like in some Escher drawings!)  But a backdrop on a railroad can be seen from any of a number of vantages, and this is a real disadvantage, so to speak.  Here’s the problem.  Let’s say you have a structure like a ski lodge or a row of storefronts.  The facades do not present much of a problem, as they are probably parallel to the surface of your backdrop.  But what about the side walls?  The roof?  Draw it so that it looks right from one angle, and it will look wrong from another!  Go back and look at the header image of this article.  The tops of the ski lodge buildings are painted to look right if one is a certain height– like mine– and if I’m standing so that the physical structure to the right is presented at the same evident angle.  But if a small child views it from below, then he cannot see the top of the blue building whereas the tops of the brown building stick up awkwardly.  It detracts from the realism, and there is no perfect solution for this.

If you’re interested in my opinions, here they are.  First, do your best to find what you think is the most common viewing angle, and design the perspective from there.  Second, if there’s a way to make the painted structure tall enough that the roof can’t be seen from any angle, so much the better.  Third, if at all possible, place such structures as far into the distance as possible; this reduces distortion of the perspective.  If that’s not possible, I would suggest that you cut the facades from foam core and fix that to the backdrop, giving a hint of three dimensions, and not show any side or top surfaces.  I’ve done this with the pillars on my model of Union Station and it is surprisingly convincing.  But just realize that no backdrop can perfectly represent three dimensions in two, as viewed from every angle.  This is my strongest argument against using enlarged photo montages for backdrops.

Okay, so now we have the primed backdrop installed on the layout.  Now it’s time to have another discussion on what you will paint, before you paint it.  This discussion can be summarized as:  Light is Everything!  Everything.  Absolutely… everything.  Remember talking about the time of day and year and stuff like that, up above?  Now it’s time to get really really specific.  You must decide where the sun is, at this moment.  That will dictate which sides of objects are highlighted and shadowed.  It dictates where shadows fall on the ground (yes, you will paint those!)  It decides how bright, and how sharp, your scene is.  It also influences how you place your layout’s floodlighting, which is a subject for possibly another post.  The only way to avoid this decision process is to decide that it’s night.  So, go get a bucket of black paint and use it on the backdrop.  You’re done.  Wasn’t that easy?

Oh wait.  It’s daylight.  Dang it.

So, did you ever wonder why sundials aren’t perfect tellers of time all year long?  Or why that one scene in National Treasure, where they find Benjamin Franklin’s glasses behind a brick at Independence Hall, just doesn’t work in the real world?  Because of that whole the-axis-of-the-Earth-is-tilted phenomenon.  In other words, the Sun is not at the same angle at 9:00 AM in January as it is at the same time on June 20th.  Not even close.  And this is why you must decide on a season, preferably a month, and a time of day, for your scene.  A funny characteristic of light beams: at our distance from the sun they are effectively parallel.  This means that all shadows hit the ground at the same angle. This is actually a good thing, because you can make a template to help you judge this.  Easiest thing in the world, so long as you know how high in the sky the Sun is, and what angle relative to your backdrop it is.  Do a little three-dimensional geometry (see previous post on the usefulness of geometry) and cut out a right triangle with hypotenuse cut at the sun’s angle).  Use that when you start painting shadows and such.

Oh, and one other thing.  Which direction is north in your scene?  All else hangs on that.  (South, if you’re in the southern hemisphere…)

I made a few references to perspective above.  In case you never had Art Class in 7th grade, I’d suggest that you google up the concept of perspective and get more familiar.  As it turns out, there are two very different types of perspective that we will utilize in our scenes.  One is what I will call the vanishing-point concept, and the other is that of distance perspective.  So, the vanishing point refers to a place in the far distance where all lines parallel to our line of sight converge.  It’s Infinity, kind of a visual black hole.  Turns out that drawing objects is a little more complicated than that– there may be two or more such points arranged along a horizontal line, for example– but it will help you get things closer to real appearance.  Look at any photo and you will observe this.  See the header photo above and study the brown lodge for a moment.  I purposely shot the photo standing in a location where the perspective was most convincing.

The second concept, distance perspective, is simply the recognition that the further things are away from you, the more air there is between you and it.  Air is not clear.  Big surprise, huh.  Being about 78% nitrogen, it’s actually quite blue, and sometimes contains other particulates or vapor that make it even more murky.  That’s why distant ridges covered with green trees actually look blue, or gray, or black.  Your distant palette will be in faded blues and grays.  Nearby forests and objects will be much closer to true color, so for those you would break out the ocher and reds and greens.  Again, see the header photo and you can see how I utilized this concept.

Now, go back to the scene you’re building.  Study your photos if you have them, and decide on the depth of the scene.  How distant is the horizon?  How close are objects in the picture?  A deep but distant scene will have relatively flat perspective. A close scene will have noticeably different object sizes from foreground to background.  You will use the vanishing-point technique to get the sizes and positions right.

So that was a rather lengthy introduction to the subject.  Next time we’ll bust out the paint and brushes, and make a mess.  Stay tuned.  actionroadlogobang

Track Geometry, or Two Parallel Lines

What they don’t teach you in geometry class is that low-grade 5/16″ plywood has a mind of its own, especially if it has ever gotten wet.

Navajo Mine Railway Track
Jointed rail on the Navajo Mine Railway

Of all the math courses I took in junior high and high school, the one in which I actually got decent grades was geometry.  I suspect it’s because the subject made sense.  It wasn’t dealing in the abstract but instead it had immediate real-world applications.  (Apologies to all the math teachers out there; I get it that you have to be able to crunch the numbers before you can apply them.)  In fact, that geometry class may have been my favorite one at Fairview High, despite the teach’s terrible 1970’s comb-over.  I mean, really.  When you can see shiny scalp between the bryl-creamed strands of hair, who exactly do you think you’re fooling?

But I digress.  Geometry, and especially trigonometry, have proven more useful in my world than almost any other subject I learned in school.  Yeah, I got mediocre grades in English!  A person can use trigonometry to build stuff.  That and an understanding of the concept of pi are sufficient tools to design things like, say, a model railroad.  Well, that and a few carpentry skills, which I picked up from my dad and my other favorite class, wood shop.  (Why not two favorites?  Why choose between pizza and a hot fudge sundae?  Gimme both, say I!)

What they don’t teach you in geometry class is that low-grade 5/16″ plywood has a mind of its own, especially if it has ever gotten wet.

So let’s fast-forward about 21 years after my graduation from high school. One week the family was out of town, so I used the evenings to throw together some benchwork for a layout expansion in the spare room. My design called for using the conventional cookie-cutter method to create the subroadbed.  I visited Hacienda, came home with all the requisite materials **, and stockpiled them on the back porch for the following day when I would start cutting.  Unfortunately the next day it rained while I was at work. Not to worry, I said to myself; the wood will behave if I use sufficient supporting risers between the subroadbed and the supporting 1X4 framework. I should have worried.

(** Why use such thin plywood for decking?  Because, in turning a helix, it enabled me to reduce the grade just a smidge since the structure above the lower levels was a bit thinner.)

Turns out, my materials had acquired some new un-flatness, and I was never really able to completely train it out of the wood. Plus, thin plywood doesn’t warp consistently in every direction anyway; it actually does have a bit of a grain.  And here’s the thing: laying tight-radius track on a grade with reverse curves is a recipe for disaster anyway. Throw in a few unplanned humps and dips and you will never reliably keep a train together. It’s like installing multiple yard humps out on the mainline; those Kadees just slide apart. (Hm, a model hump yard with scale retarders?  Who wants to give that a shot in 1:87?  I can tell you, the hump part will work just fine!)

Okay, now move forward another decade. We’ve moved to a new house and I’m rebuilding the railroad in a larger space, using many of the components from the prior installation. This, of course, means that I’ve incorporated the same warped roadbed in the construction. Truth is, I never operated the former version enough to realize how bad the track geometry problems were. Now, however, I have a much longer, fully ballasted mainline and I’m operating longer trains with (this is significant) a lot of 86′ cars. Turns out, cars less than 50′ are Labrador retrievers, pretty forgiving of squirrelly operating conditions. An 86′ flatcar, on the other hand, is a rich sorority girl; it wants to be treated like a little princess or it uncouples, or goes off the rails. I had a whole yard full of prima donnas.  So, what to do? Go our separate ways, or fix the relationship?

This is where it’s useful to ponder the seemingly-obvious fact that we’re modeling actual railroads, only smaller. Real railroads face similar challenges, although they’re usually smart enough not to install a 40-degree curve on a mainline! No, they have track geometry problems of their own. Take a look at the header photo, which I snapped of the track at my former employer.  What solutions do real railroads bring to the problem?  Well, two immediately come to mind.  One is to contract a rail grinder company to visit and to smooth out the railheads.  Another is to get the section gangs or other contractors to level the roadbed, lifting track when necessary, and get everything more-or-less on the same plane.  It turns out that the first solution is tough to employ on a widespread basis in scale applications, whereas the second holds far more promise. And that’s where I am going to spend the rest of our time today: how to level out your track by lifting and re-ballasting.

Okay, first a couple of assumptions. First, I use flextrack, basic old Atlas Code 100 in my case, but any flextrack will do for this. You must be using some kind of flex track or scale “panel” track; hand-spiked track is beyond our scope.  Second, I’m ballasting my track and so must you. The type of ballast is up to you. On my railroad I go mostly for a darker shade to represent the D&RGW’s slag and scoria that they used for years. It’s a blend of 50% commercial black ballast and 50% coarse sand.  In other words, sand that’s roughly the same size as the commercial ballast. Third, I’m using mostly cork roadbed, because it’s easy, firm, and has a good profile. It’s also possible to use craft foam cut into strips; I actually use large sheets of this to underlay my yard trackage, but use multiple strips to elevate mainline bypass tracks around the yards.

Other tools and materials?

  • A short straight-edge, such as a small level, between 8″ and 12″ in length.
  • A tack hammer.
  • A nail set.
  • An inexpensive chisel.
  • A medium-soft paint brush
  • Track spikes, something a little longer than normal since we are lifting the track higher than normal.
  • A bottle of white glue such as Elmer’s (don’t go with the cheap dollar-store stuff!!!).
  • A squeeze bottle for diluting the glue, such as your previous Elmer’s bottle.
  • A spray bottle, such as you might use for plants or laundry or training the cats. Just be sure you tell your wife that you’ve borrowed hers. A better solution is to purchase your own dang bottle…

Now that you have all that stuff in a pile, let’s examine your track. First of all, you probably already have some idea where the problems live, but I’d suggest the following: assemble a short train of long cars, such as piggyback flats. These are best because there is little in the way above the deck line and you can most easily observe the action of the couplers. Now, run the train slowly around the railroad, walking with it and observing any bobbing of the ends of the cars. See if there are places where one car’s coupler lifts away from the adjoining car. Sometimes it’s obvious and other times it may take a few runs. Once you identify the spot, mark it somehow.  You’ve found a high spot, which means that on one side or the other is a sag.  The sags are what we must locate, and this is where the straight-edge (your short level) comes into play.  I use a level not because of its leveling features, but because it’s thicker than a ruler. This aids you in the process.

Place the level atop the railheads at the high point, and move it each way, eyeballing it from the side. You should be able to see where the track dips away from the straight-edge.  Go out some distance in each direction from the high point. This can be tricky if it’s on a curve and even more so on a curving grade, as there’s a helical shape introduced into the track now. More on this later.

So, now you’ve found a sag.  This is important since it’s far simpler to lift a sag than to undercut a high spot.  We will focus on dealing with the sag.

Assuming that your track’s been nailed down already, here’s what you do.  Take the chisel and whack chunks out of the plywood.

No.  No.  I’m only kidding!  That is not what you do.

(Take 2) Locate any track nails in the area, take the chisel, rest it on the railhead, and use its sharp edge to gently lift the nails up from the sub-roadbed. Yes– we are using the chisel as a crowbar. Once you have loosened all nails in your sag, gently loosen the track from the ballast.  I’m assuming here that you had glued the ballast when first installed. Sure, you could just demolish the first ballast installation, but that’s really not necessary.

— And here’s where I am going to take a giant step backwards and pretend that I was smarter back when I first installed the ballast. Here’s the secret: you can do all of this before you put the ballast in the first time! Saves much time. If you haven’t yet ballasted the track, simply adjust this process.  You’re smart and I’m confident that you can manage it.

OK, back on track, so to speak.  Now that you have your sag loosened, you want to try to get the railheads as level as possible, longitudinally and laterally. (Also, it’s kinda important that you don’t work on too much track at a time, lest it drift off alignment.)  Okay so far?  Now you have a loosened piece of track hovering above the roadbed. Take a Dixie cup full of your preferred ballast, and gently sprinkle it down the center of the track  Not much, mind you.  Just enough to work between the ties and get under the track. Work this in with the brush, tamp by gently tapping on the track, and re-check things with the straight edge. Wash, rinse, and repeat as necessary until the track has been leveled across the sag (without creating a new hump. That would be defeating the purpose, you see). Once it’s to the desired grade, ballast outside the rails, being sure to fill any voids caused by lifting the track.

Now, we fix it all in place.  Mix some glue about 50:50 with water in the small squeeze bottle.  Gently spray the track with the spray bottle to moisten things.  Using your squeeze bottle, flow the glue mixture into the ballast, thoroughly saturating it.  Be sure you saturate it all!  Now, go away and do something else for a day or so.

After the ballast has hardened, gently push the track nails back into place.  If you didn’t move anything, they should find their former holes and go right in.  You really don’t want to pound on anything at this point, but you can use the tack hammer and nail set to seat any stubborn nails if absolutely necessary.  But remember: be gentle! And accurate too, let’s not forget about that. The nails are not really holding the track down now; that’s being done by the ballast and the glue.

Polish the railheads to remove any extraneous glue or haze, inspect for any stray chunks of ballast that might interfere with a flange, and send the test train back over the spot.  You should notice a marked improvement.  Chances are, there will be other sags and humps, sometimes revealed by fixing one like this.  I had to lift an entire section of track, two or three feet long, to deal with a serious dip. Just fix whatever you must, until your trains stay on the rails better.

The result: You will be so glad you leveled your track. It makes such a difference. actionroadlogobang

A comment on superelevation. This, for those who may not be familiar, is simply the technique of banking railroad track on curves, much as in Nascar or at Indy, right? It seems like a really good idea, so I tried it on my railroad.  Um, it was a disaster. Here’s why: I’m dealing with pretty tight radii, down to 24″ r at times. I know, I know, curves that sharp look terribly unrealistic, but space constraints are a reality.  Well, if you think about superelevation, you are essentially laying track around a bowl. If you have long cars such as 85′ passenger cars, auto racks, or pig flats, what happens is that the trucks lean up so that the outboard axles are pushed higher. Combined with the fact that the trucks are already turned to the side, the outer ends of the trucks will probably hang up on the underside of your car and derail. If you superelevate a curve on a grade, it’s even worse–it’s like climbing a corkscrew.  I ended up abandoning the effort, removing all shims from my track and reverting to level track. In some ways it would almost be better to lean it out than in, at my curvatures! Sub-elevation?  Hm.  But no, I draw the line at some things.  I would not recommend superelevating any curves at less than about 48″ radius in HO scale, proportionally adjusted for other scales.