We recently raised up the deck by an extra 3½” with the thought of placing new bridges high enough off the ground so the dogs can safely navigate beneath them, without hurting them or themselves. The next step is to raise up the track to match the new deck and bridge height. The problem is the wooden stringers. You guessed it, they’re rotted.
The wooden stringers were a means to an end, and that end was to run trains. Needless to say it’s impossible to run trains when there’s no track to run on. The track is, well, was secured to the stringers every 8″. But not anymore… With two dogs we could barely keep ahead of them and the maintenance. With three it’s a losing battle.
It’s hard to describe how much destruction the pups have caused to the Barkyard Railroad. Imagine most of the track on the ground from the bridges to downtown missing, parts and stretches ripped loose and removed a little at a time, until there’s no track left. Nothing left but the stringers, if they hadn’t already rotted away that is.
A New Approach
It’s obvious that the previous track arrangement no longer fits our needs. We need a new approach. While the roadbed brick production is ramping up to remedy the situation, we still can’t keep up with the destruction of the track on the ground.
Even so, the roadbed bricks can’t address the elevated stretches of track and stringers that need attention. The plan has always been to replace certain elevated stretches with trestles. The time has come. The old infrastructure allowed us to run trains until it didn’t. No regrets.
We’re no strangers to scratch building trestles. At least not the straight kind. We even scratch built a four foot long Howe truss bridge. It used to stretch across a pond with a waterfall. But like everything else, it didn’t hold up to the dogs or time.
The new part to all of this is the curved aspect. All the jigs and such we made way back when only deal with straight (tangent) sections. If you’d like to know more about the history of the railroad, you can refer to the Bit of History section below.
This trestle from the deck will have both 14′ and 20′ diameter curved sections, along with the tangent sections. So far there’s a design for the 14′ diameter sections, already 3D printed and ready for testing.
There’s added complexity in addition to the curvature. The trestle has to rise at more than a 2% grade, a full 6″ over 23′ of track. From 18″ at the deck to 24″ at the bridge. Previously the ruling grade was kept to 1%, but that’s out the window with the triple decker (re)arrangement of the upper loop.
The Particulars
To put things in perspective, each bent must rise above the previous by 3⁄16″, and each bent is about 8″ from the last, for roughly a 2.34% grade. For the 14′ diameter curves there are four bents per track section, 16 per circle, roughly 32″ long and 22.5° per section.
By the time the fifth bent is reached, the track has climbed ¾” from the start. The plan is to make the bents for each section all the same height, and account for the rise by increasing the height of the footings, to be cast of concrete. Each footing will be roughly 2″ x 2″ x 16″ long, adding 3⁄16″ with each successive footing.
The footings are reset to 2″ when the next track section is reached, again grown by 3⁄16″, and height of the bents for that next section increased by ¾”. The photo doesn’t do a very good job of showing it, but there are six 14′ diameter track sections, then a 20′ diameter track section, followed by a 3′ tangent section.
But that’s a lot of talk with no pictures to show what we’re talking about.
The previous bridge approach trestles sat loosely on concrete block “caps”, roughly 2″x8″x16″, placed together and leveled. Nothing attached the approach trestles to the block caps. They were free to be repositioned every time the dogs smacked into them.
And smack into them they did! I don’t know how many times I would find the approach trestles upended or cocked at an angle, with the track and the bridges on the ground. We’re hoping to anchor the bents to the footings this time around to at least slow down the “remodeling”, pun intended.
New Needs Means New Jigs
So far only the jigs for the 14′ diameter curves are reality. The design for the 20′ diameter section along with the tangent section will begin soon. The design is made up of a number of identical parts with a handful of unique parts to address the areas where specific size and shape is necessary.
The trestle jig parts either press fit or snap together. The only gluing required is to attach the progressively taller adapters to the standard bases. This is only necessary to allow 3D printing of all the parts without needing supports.
The standard base cradles the bottom of the trestle bent in a 3⁄16″ deep notch. Because the base is only 8″ wide and a standard 20″ tall bent is nearly 16″ wide at its base, the jig has index marks to align with the bent’s center leg.
Each standard base has two legs designed to press fit into the next. Each base is at an angle to the next. Each successive base is 3⁄16″ taller than the last, hence the glued on adapters. The first base has no legs and needs no adapter.
The design provides alignment for five trestle bents and stretches over 32″. Next are the cross arms that hold the bents vertical. There are two sets of cross arms, one set for the inner curve, and one set for the outer.
Each set has its own inside and outside components that snap together in the middle. These cross arm assemblies then snap onto the legs that connect the standard bases together. The arms have a block, a peg of sorts, that fits in the ½” gap in the horizontal members of the bent section.
Of course, now that I’ve 3D printed an entire set, it occurs to me that this will only work for those bents that have a complete section at the bottom! DUH! But this is why we mock up the models and test fit. Back to the drawing board!
This further reinforces the shortcomings of trying to consider everything from just the drawing board. My need to have hands on pieces to manipulate and consider other options that never would have come to mind is the reason why we’re testing the designs before committing to them.
Model vs. Prototype Considerations
There are many more considerations, like limitations of models as compared with the real thing. For the sake of this discussion, it’s not limited to just modelling, but modelling a prototypical railroad.
Curves
As with any model railroad, some “allowances” are made for the model compared to prototypical practices. The best example is that of curvature. Most model curves are far tighter than anything in the real world. Prototypical curves are far straighter than anything we have the space to model.
Track also doesn’t just change dramatically from tangent (straight) to curved. The use of easement curves on the prototype helps ease the train into the curve without abruptly slamming everything into a corner. Think of a spiral that goes from nearly straight to tighter and tighter curves.
Our tightest curves are ten feet in diameter, huge compared to the toy like four foot diameter curves. Even so, 10′ diameter curves are really beyond anything seen on the prototype, including the tightest industrial sidings and spurs.
Those 10′ diameter curves are only used where space is at a premium on our pike, like the station spur and the tight fit of the mainline behind the shed. Everywhere else, we simulate an easement by starting with a 20′ diameter curve leading into the 14′ diameter sections.
That doesn’t help much when they’re all 20′ diameter sections though. Other parts of the layout are “flex track” which allows us to bend it to any curvature. Here we’re able to ease in to the curve by slowly increasing the force when using the rail bender, creating a more gradual curve.
Trestles
So why all this talk about curves? Beyond the modelling consequences related to track alone, it also influences modelling the structures that support and convey the track off the ground. In this case, trestles. And because the curves are tighter than on the prototype, allowances must be made for those structures too.
The way the prototype did things, each bent grew from the top down, growing the length of the legs as necessary. This leads to sections of standard height, with only the lowest portion varying from one bent to the next. We chose scale twenty foot section heights, or 10″ each at 1:24 (half dollhouse) scale.
The key take away from this is everything is designed from the track level down when it comes to a trestle, each section depending on the one below it, in a standardized fashion. The reason for this top down design is to ensure the stringers are directly supported by each bent without the need for shimming.
The rail stringers are perched atop the bents directly because the bents are built exactly as tall as required. Shims would diminish the strength and stability of the structure. These stringers are made up of staggered members, whose lengths span three bents, landing on the two outside bents in a joint.
This is to ensure there is at least one solid member at every joint atop a bent, and bolted together, again to increase strength and rigidity. There’s a stringer beneath each rail, generally with three individual members to provide redundancy for failure of any one of them, and each spaced apart to provide an air space so water is not trapped between them.
Making Allowances
So why does that matter to us? It all goes back to the discussion about curves and the allowances we need to make in our models. That staggering works great for tangent track, but not so much for curved track on our model. The prototype spacing between stringer members is around 3″. That’s 1⁄8″ in scale.
Unfortunately, the spacing between the members on the curved sections far exceeds 1⁄8″! At scale, each 8″x18″ member is roughly 5⁄16″ x ¾”. Over the 32′ length, a scale 16″, the overlap would be more than ¾”. That’s more than double the scale 5⁄16″ thickness of each member!
That means the stringer members can only span between bents without a noticeable deformation that would look totally out of scale. Of course, the shorter than prototypical member length will also be noticeable, but not as much so once the tie strips are in place.
Generally the prototype used longer ties to provide a walkway of sorts on at least one side of the track. These longer ties, bridge ties, are also more closely spaced on a bridge or a trestle. That along with a railing made it much safer to walk along the track that high off the ground.
Back in the days of steam, red hot cinders could fall from the ash pan onto the trestle timbers and set them on fire. A fire barrel filled with water or sand was placed every so often along the walkway, on even longer ties so as not to block the walkway, to provide a ready means to extinguish a fire.
Model track comes with one option, standard length ties. For that reason, a slide for the table saw was created that allows notching wooden “guard rails” to fit over the ends of the wooden bridge ties to be assembled into 8″ tie strips, long enough to span between two bents.
Unfortunately, that only works for tangent track sections. It would need reworked to accommodate curved sections if not totally remade just for curves. In this case, I was overly obsessed with true to prototype realism in my modelling.
That’s a holdover from my HO scale days where anything that’s out of scale stands out like a sore thumb, making everything look toy like. It definitely kills the illusion of realism. In the case of the notched tie strip guard rails, not all railroads notch them.
For the sake of simplicity and rapid production, the Barkyard Railroad will no longer notch its guard rails.
A Bit of History
Back when we first moved here to Mount Dora in 2014, we had to tear up all the track we had laid at the old house in Wekiva. There were plenty of projects that took priority over getting the railroad out of mothballs and back up and running. Slowly but surely we renovated pretty much the entire house.
One of the earlier renovations was the garage. With just two stripes of concrete and a dirt floor, it was a carriage house in every sense of the word, complete with carriage doors. You guessed it, they were rotted and needed totally replaced.
But first we had to do something about that dirt floor. It was like silt, a very fine mix of dirt and sand, stirred up into a cloud at the least provocation, sticking to our legs, ankles, and feet. Better plan on taking a shower if working in the garage.
Nick helped us install a plywood floor over the dirt, using the foundation blocks and concrete stripes to support a 2×4 framework covered with ¾” tongue and groove plywood. Benches soon followed along with a new table saw.
About that time we had to replace the dilapidated fence between us and our neighbor to the west. Most of the fence panels were rotted away to nothing, but some of the wood that still had some life left in it was saved as raw material for building trestle bents.
Early Trials
It was a brave new world learning to use the new table saw and fashioning a crude template to hold the pieces of a trestle bent together while assembling it. In case you’re wondering, a trestle is made up of individual trestle bents, lashed together with horizontal girts and diagonal cross braces.
The bents themselves have their own cross braces and other means of securing the legs together which divide the bent into multiple sections. So the crude template used small chunks of wood, strategically placed, and screwed to a chunk of plywood.
Repeatability was questionable at best. That is to say no two bents were interchangeable. Originally each of the legs was cut individually, and required new setups for the three different angles in involved. Getting a repeatable length was nearly impossible.
It soon became apparent that a better quality template was necessary. The new template was custom cut on the table saw to the correct angles, making dado cuts to hold the entire length of the legs in position, along with the horizontal joining members.
Other changes were made to increase productivity as well. The new design accommodates using the template as a “sled” for the table saw to trim all legs to the proper angle and length in one operation. Runners to fit in the T-slots were attached to the back, holding it in perfect alignment with the saw table.
That Was Then, This Is Now
All that seems so far away now, around Halloween of 2015, sitting in the living room, assembling trestle bents into an approach trestle for the scratch build Howe truss bridge. Fast forward to the task at hand today. The simplistic jigs I fashioned back then for holding the bents together did little to align or secure them in place.
And they were only meant for the straight sections to boot. Back then I would have been happy to have even the tangent trestle I built survive, but too many other things had to happen first, and many false starts, before we could think about a permanent garden railroad.
Again, permanent is a relative term. The dogs and the elements would beg to differ with the “permanent” moniker.
There is much more to come. Stayed tuned.