Pecan Street is very much the result of "trial and error": every now and then, something went slightly wrong or didn't work at all - or I simply changed my mind and approached things differently. It's what I call "accidental modelling".

Here's an account of how it all came together in the end, step by step (which is why there is also a certain amount of back and forth).


Any commercial products mentioned here are purely bona fide indications of what I have been using myself.
I have no connection to any manufacturing companies nor do I profit from listing any products or brands.




The prime question that needed to be answered was what two tracks within a depth of only 6 inches actually looked like in the real world - something that no track planning software is truly going to be able to show.

A piece of paper cut to the required width served as a simple set-up for this "proof of concept". It was always clear that it was going to be tight, but the one thing I didn't want was for it to feel overly cramped (which is why a boxcar was set up on one of the tracks).

This is one of those "instant moments": either it clicks, or it doesn't. In this case, I was happy to move on and actually start building the layout.



  For the two baseboard segments I returned to the construction method used on Little Bazeley Mk2 which provided a solid yet lightweight layout framework. Using a sheet of 10mm poplar plywood as the base, pieces of 20mm styrofoam insulation are glued to that board, which in turn is covered by another sheet of 10mm ply in "sandwich style".

Once the side- and backdrop are added, this results in a very sturdy yet lightweight layout frame and basebaord (which also results in comparatively silent running of trains).

There is, however, one obvious problem with this approach - there is no real cavity underneath the trackbed-baseboard. In order to be able to run cables underneath the baseboard top, the styrofoam blocks are set back slightly, providing some space at the front.



Even though the restricted size of the layout clearly limited the possibilities for structures, I wanted the two tracks and the shuffling of rolling stock to convey a certain meaningfulness. Having a modern-style warehouse as a backdrop and a "destination" wasn't very orginial, but since I already had the Walthers "thin profile background building" (something I would call "low-relief") it was both an easy and convenient choice to make.

At 1-1/8 inches (2.25 cm) the structure has a very shallow footprint, yet its length of 19 inches (48.3 cm) and its heigth of 4 inches (10 cm) result in a fairly sizeable building, making it all the more ideal for Pecan Street.


Since clearances are all important (and especially so if you only have 6 inches to work with), assembling the warehouse was the first step needed to provide a point of reference for where exactly the tracks could go. At the same time, the baseboard sides were added, and then a liberal coat of acrylic paint sealed the baseboard top - track ballasting and scenicking involves no small amount of a mix of water and glue to fix it all in place, and baseboards that haven't been sealed this way may well start to warp.
The low-relief warehouse kit is a very straightforward build, as was adding two Walthers SceneMaster modern-style wall lamps along with some initial weathering using Vallejo grey and black wash.

Having a narrow gap between rolling stock and the warehouse front not only results in a more prototypical appearance, it also provides more room to space out the two tracks a bit, making operation easier.

It is important to be very diligent in this step, using the widest and tallest item of rolling stock along with all locomotives intended for use on the layout, to ensure none of them fouls the warehouse.




With all the clearances worked out and marked down, it was time to lay the track - once I decided which track to use.

Peco is my track of choice (due to decades of trouble-free running) but that still left me with two options for an HO scale layout: regular code 100 or code 83 US style track.

While the code 100 track has a slightly chunkier rail profile, the most noticeable difference is the smaller size and closer spacing of the ties, giving the code 83 version a distinctly more authentic "American" look.

So why even hesitate?

The sturdiness of Peco code 100 track is simply unsurpassed, and since Pecan Street would require joining two segments every time it would be set up (and then disconnecting those segments again for storage), this seemed to be an aspect worth taking into consideration.
  Not that Peco code 83 is overly delicate, but the finer rail profile does make it feel somewhat less bombproof in comparison. As an added bonus, using code 100 track would allow me to fall back on a short piece of Peco setrack - the ultimate means of connecting the two segments in a sturdy and reliable way. Code 100 track would also allow me to run some older models with oversize wheel flanges.

As for the visuals, I had looked into that a few years prior (for a US layout that ultimately never materialized) by setting down a 50' boxcar on the code 100 track of my Little Bazeley Mk1 layout. Not surprisingly, with no code 83 track right next to it for direct comparison, the weathered and ballasted code 100 track didn't look too bad, in spite of the "wrong" size and spacing of the ties.

Ultimately, my decision was helped by looking at prototype railroad locations around warehouses - and seeing a clear and distinct potential for hiding any perceived visual shortcomings of code 100 rail, since the tracks in those places often convey an entirely different look than what can be seen on a through traffic line.
Looking north from a road crossing on Titan Row in Orlando FL in 2021, we see two modern warehouse structures served by two sidings, along with a line that runs on to serve more customers.

The track here is far from being dilapidated, in fact the through track seems to have been re-ballasted recently and ties replaced. But ballast on industrial trackage often doesn't get tamped and regulated the way it would on busier stretches of railroad track.

As a result, ties often get covered by ballast both to the sides of the rails as well as in their center, and over time, grass and weeds will grow up in places.

A dry run using a piece of Peco code 100 track showed that emulating this appearance on Pecan Street would minimize the visual impression of the larger and more spaced out ties.

As a result, I opted for sturdiness and proven trouble-free running qualities and chose Peco code 100 track for the layout.




The simple track configuration involved in a "tuning fork" layout was thus laid using Peco code 100 track; Streamline flextrack for the siding in front of the warehouse (since this curves away from the single switch on this segment), and pieces of Setrack "snap track" for the siding running along the front of the layout (simply because I had quite a stash of this available, and it does make putting down a segment of straight trackage a breeze).


The switch is a Peco Streamline medium radius "Insulfrog" point, meaning it has an electrically insulated frog and - unlike "Electrofrog" switches (that have an electrically live frog) - requires no extra wiring to operate, being DCC-friendly straight out of the packaging.

The only snag that "Insulfrog" points can throw up with DCC operation is a potential shorting problem with some metal wheels as they travel through the frog. Even though this only seems to be the case with a very limited number of wheelsets, I decided to play it safe and add insulator joiners to the inner rails past the frog (as per the instructions printed on the Peco packaging of the switch).

The illustration below is taken from my Little Bazeley layout, but the principle remains the same for any layout.

Once everything panned out, the simple track configuration was laid down in virtually no time at all.

I like to make sure that the track is firmly and securely attached to the baseboard (no roadbed needed for industrial tracks) before weathering and ballasting, which is why I use Marklin Z Scale track pins. Due to their very small size these are very inconspicuous yet still hold the track down perfectly well.

A test run then makes sure that all electrical connections work properly.


There is also really no difference in this case between an old-fashioned DC wiring approach and a DCC friendly one. The only difference are the additional feeder tracks required due to the aforementioned insulator rail joiners on the inner rails past the frog of the switch; I went ahead and added additional feeders to both tracks at the end of the two sidings to give this layout segment an added boost in terms of feeding power to the track.

All the wiring follows the common colour coding of "black is at the back" (and therefore red wires to the front rails).



As a first step in weathering the tracks, I took the baseboard outside and sprayed them a dark matt brown colour using an aerosol spray can from the DIY store, masking parts of the switch in order to avoid conductivity problems.
Like most modellers, I have accumulated lots of "spare" items over the years, and I felt that Pecan Street would be an excellent opportunity to use items from this stash.
However, the rattle can that I still had left over from my Little Bazeley Mk2 layout build a few months back, proved to be a bad choice.

The spray paint went on as usual and without any problems, but when I tried to wipe the top of the rails immediately after, I found that the paint was too sticky to come off neatly - or even refused to do so at all.

I suspect that the aerosol/paint/solvent mix had deteriorated inside the half-empty rattle can, but whatever the exact reason, I had to apply a different approach than I usually would. Deciding to let the paint dry completely, I left it alone for a good 48 hours. After this, I used swabs saturated with isopropyl alcohol to wipe the top of the rails clean.


  This worked well (and without the damage abrasive tools inflict on the rail surfaces), although the paint was a bit more stubborn in some places, requiring two or even three swipes.

In the end, however, this unexpected problem was solved and all the rails had a shiny surface (necessary to provide good electrical pickup). The masked-off section at the switch was then touched up with a fine brush and rust-coloured acrylic paint.

In order to brake up the unrealistic uniformity of the colour, I then added a few touches of RailMatch's "sleeper grime" (an acrylic paint) as well as some Vallejo grey and black washes to the sleepers.

Even though this is a very straightforward and simple way to achieve a more realistic look, some (if not most of it) would ultimately be lost - since I was planning to "sink" the track and ties well into the ballast. However, weathering the track really is an essential part of recreating the atmosphere of the real railroads, so I just went ahead with it regardless.

More test running and checking of all electrical connections after this stage is essential to discovering (and if needed ironing out) any bugs. Once the functional aspects of the track are all settled, it's time to add ballast to the weathered track.



Having ballasted model track for a good thirty years in HO/OO, N and Z scale, I felt confident that my array of proven methods accumulated over that period of time would provide me with a trouble-free part of putting together Pecan Street - especially since for the first time I would not be attempting to produce a well-maintained and neat looking section of ballasted tracks.

I was in for something of a surprise - and an extended episode of accidental modelling. Initially though, everything went according to plan.


Track maintenance work at Abingdon Va, 15 October 2018 (Video)

  The colour of ballast varies greatly on the prototype, depending on the type of stone used and how much soil and rust is deposited on the trackbed by traffic on the line. A little bit of research into the area and era modelled (on location if possible, otherwise online and through books and other media) really helps to get it right.

Unless you are basing your layout on a specific location with its own unique characteristics, railroad sidings serving warehouses in the US commonly seem to use grey, light grey or buff coloured ballast.

I have been using Woodland Scenics ballast (actually made from crushed walnut shells) for decades and saw no reason to change that.

For this layout, I opted for my favourite, the "fine" grade light grey ballast (B74), which is usually marketed for N scale and looks really good on HO track. With Pecan Street, however, one fundamental aspect of ballasting was different.
Usually (i.e. on all my previous layouts) I would try to achieve a fairly neat appearance where the ballast was evenly spread, with no loose particles on top of the ties or clinging to the rails. I would achieve this by sprinkling on only small quantities from a teaspoon and then straightening the ballast out using my index finger, a fine paint brush, and a toothpick.

The basic idea is that you can always go back for seconds and don't want to use too much at once in order to avoid a mess. It's also at this point that your Peco code 100 track (dismissed as "toylike" by some modellers) really starts to shine.

This time, though, that "mess" which I would always try to avoid was exactly what I now wanted for Pecan Street.

  What I didn't bank on was that I would find myself in a little bit of a mess of a different kind.

In spite of not having to worry about not getting ballast everywhere, I still tried to apply some amount of restraint.

Once I was happy with the general look, I proceeded to permanently glue the ballast down.

For decades I have done this using what most modellers use: a mix of water and white glue (approximately at a 1:1 ratio) with a drop of washing up liquid (to break surface tension).


The mix is then carefully applied from a dropper, which proved a better choice here than a syringe (both came from the vet). The idea is to really soak the ballast, as this ensures that the water/glue mix doesn't just cover the surface. Mixing white glue and water in this way is a time-tested method for ballasting track, and once applied it's best to leave everything to dry thoroughly for at least 24 hours.
  And this is when things started to behave quite unlike what I was expecting from my past experience.

To kick things off, I found that the white glue I was using dried perfectly clear, but with a distinct and very glossy shine. While this would be great if you modelled a scene during or after a rainfall, it was not what I wanted.

Maybe it was just that make of white glue. Common modeller's wisdom has it that the cheapest white glue you can get is the best for this, but after testing out an additional two or three brands I found that none of them really dried with a matte appearance.

I noticed that all the labels only specified that the glue would dry clear but made no reference to a glossy or matte appearance.
On top of this, I then noted a second, completely unexpected problem with the stretch of track that I had already ballasted.

Even though I didn't just throw down the ballast with wild abandon, I found that after I had soaked the mix of Woodland Scenics ballast and scenic scatter material (still using the "trusted" mix of white glue, water and a drop of dishwasher detergent), it had a light but still distinct tendency of expanding compared to what it looked like in its dry condition.

The result was a bit more "ballast and weed cover" than I had aimed for. While the visuals were fine, I found that running a "test car" (in the form of a Walthers boxcar) over the newly ballasted track did reveal two or three spots where the ballast interfered with the wheel flanges.

It wasn't that big a deal and easily rectified by scraping away the few grains of ballast causing the problem. Any small screwdriver or similar tool will do the job.


  In this case, however, I used a cheap calculus scaler (I got mine on eBay) that makes access easy and allows for a pin-pointed removal of stray ballast.
I was starting to realize that, contrary to my initial assumption, covering ties in ballast was actually more challenging than creating a clean and neat spread of ballast around the rails and ties.
So I made a point from this moment on of putting down less ballast than originally intended, especially between the rails.

This would make sure there would be no problems with wheel flanges, and potentially "underdone" spots could always be touched up later.


This did not, however, fix the "white glue glossy finish" problem. Was it possible that the formula for white glue had changed across the board? Whatever the reason, I quite unexpectedly found myself having to look for an alternative solution. There were several options, but I wanted something that wouldn't be too far removed from what I had been using previously. And so I found myself purchasing a container of Woodland Scenics Scenic Cement™.
It solved my problem instantly, drying to a clear matte finish and not "pushing up" the ballast. For whatever reason (and it can't be too sophisticated), this scenic cement worked while my trusted home-made mix didn't. Of course there can't be too much of a difference in terms of components (the scenic cement incorporates a whetting agent, too) but since it worked and I wasn't going to be needing gallons of it, that was all fine by me. Sometimes there's nothing else to do but embrace accidental modelling when it happens. From here on out, I will be using Scenic Cement.

Something I never do, though, inspite of many positive reports from other modellers, is "mist" the ballast with isopropyl alcohol (IPA) before applying the water/glue mix. Even when using a perfume atomizer, I have found that no matter the distance from which the IPA is sprayed on it tends to disturb the dry ballast - which is the exact opposite of what this procedure is supposed to achieve.

  So instead of any "pre-treatment", I simply apply the scenic cement (or, previously, water/glue mix) by taking advantage of its capillary force.

Because the scenic glue has a lower surface tension than the ballast and scenic material (thanks to the "wetting" agent), applying the mix close to the edge of the ballast (area #1 in the picture) results in the watered down glue being drawn in and seeping into the dry ballast by itself - or rather by its capillary force (2).

Ultimately the effect is similar to blotting paper or household tissues sucking up liquids, and it leaves the ballast almost completely undisturbed (3). The illustration used here shows a piece of Z scale track, but the principle obviously applies regardless of scale.

Letting the capillary force do its magic does however result in a slightly slower procedure. Starting at the edges of the area to be treated, patience is required to then let the scenic cement work its way properly into the ballast and scenic material (which might take a minute or two). Additional scenic cement is then only applied to the areas which have already soaked up the mix, saturating them to the point where the mix then seeps into the adjoining area of dry ballast and scatter material. It is definitely a patience game, but then ballasting shouldn't be rushed anyway.


Since ballasting is best done in stages over several days, I felt this was a good point in time to assemble the second baseboard in-between ballasting sessions. It's an almost identical twin, constructed using the same methods a baseboard #1, except that I opted for one less piece of styrofoam. Having six of those instead of seven would give me more wiggle-room to run the wiring.


Pecan Street obviously doesn't allow for a lot of scenery, but what little could be done on segment #1 would be situated towards the backscene and fulfill two purposes: convey an atmosphere of a more rural than urban setting, and hide the fact that the warehouse is nothing more than a flat frontage essentially just stuck right onto the backdrop.

Taken together, those two points would ideally result in a number of trees, but the layout simply doesn't allow for that, so it would essentially have to be shrubbery and possibly something akin to low-relief trees. It is this kind of challenge that fuels my appreciation of theatrical layout design, since art directors in the entertainment business often have to work with similarly restricted spaces, be it on a theatrical or a movie production stage.

Since the shrubbery will be "growing" up from the baseboard top to the backdrop, providing a base which provides an additional vertical structural support will make it easier to bring the shrubbery up higher so the required effect will actually work.

This base is made up from four pieces of narrow styrofoam. They are shaped to provide a slope and then covered in a very thick layer of green acrylic paint - normally I would cover the styrofoam with a layer of plaster of Paris but this seemed like overkill in this situation. When all will be done, nothing of that slope should remain visible. The next steps on this will, however, only be tackled once all the ballasting in that area is done.



Following the "heavy ballasting" around the sidings, the area around the switch was ballasted with a decisively lighter touch (and therefore more in line with what I was used to from previous layouts). The idea here is that the track is in the process of being somewhat rehabilitated, as the tamped ballast along with a small pile of fresh ballast besides the switch is supposed to indicate. At a later point, the two slabs in the background will support a few lengths of switched-out worn and rusty rails.


The narrow strip at the back of the left-hand section of Pecan Street doesn't allow for much in terms of scenery. The up side of this is that a lot less scenic material is needed compared to what would usually be the case. The possible down side: the scenery needs to be carefully compressed in order to look convincing.
I have always built up my layout scenery in layers. This allows for a gradual build-up and blending, which avoids an overly uniform look and also makes tweaking things a lot easier.

The first stage in building up the background scenery to the side of the warehouse was to cover the styrofoam base with white glue; this serves as a base for a generously applied mix of Woodland Scenics "fine turf", mixing "green grass" (T45, light to medium green) and "weeds" (T46, medium to dark green).

This produces a slope covered in low grass of various shades of colour. It provides a pleasing enough look, but more vegetation is clearly needed. Adding different colour shades of Woodland Scenics "bushes" (FC146 medium green, FC147 dark green, and FC149 forest blend) results in another layer, and a more interestingly varied backdrop scenery.


Having more dark colours towards the very back helps to accentuate the perception of visual depth. For the most part, this second layer is sufficient, since the background scenery is supposed to be just that: its role is not to attract attention but rather quite the opposite - to just blend in with the backdrop (which, being of a deliberately toned down "not so sunny sky" tone, also doesn't detract from what is going on in the foreground).
  Using low-relief buildings imitates a stage trick and is a prime example of theatrical layout design.

But in order for this visual deception to work, it is important to disguise the true dimensions of such "stage props" - the unhindered view of the shallow sidewalls of the warehouse meeting up with the backdrop is a dead giveaway at first sight.

The deception works as soon as that view is obstructed, preventing a clear visual perception of where exactly the warehouse ends and the backdrop begins.

This is what stage designers do both in theatres and on film production sets, and in this case the effect is easily accomplished by adding some sprawling vegetation in the form of trees and tall bushes.

This sprawling vegetation is essentially a first layer; refinements can be made anytime at a later stage. It is formed of Woodland Scenics "bushes" (again mixing FC146 medium green, FC147 dark green, and FC149 forest blend) and "foliage" made up of various colours (F51 light green, F52 medium green, and F53 dark green).

1 + 1 = 1

With the left hand segment of the layout having come along quite a bit, work started shifting to the right hand segment - and with it the question of how the two segments would ultimately be joined up for operating sessions. Although actually, that question had been answered long before, since the possibilities of joining up the two segments with a piece of setrack was an important point in going for Peco code 100 track.
Previous experience with modules has lead me to trade visuals for truly secure physical and electrical track connections across segments, and nothing beats having a short piece of track as a "bridging" element.
The two segments of Pecan Street are joined up by a very short (41mm) piece of track from the Peco setrack range - a sturdy and reliable connection. The tracks on both segments have been set back accordingly in order to allow for a snug fit. Once both segments are fully scenicked, the "connector track" can be weathered in order to tone down the visuals.


A true "tuning fork" layout only has a single lead track running up to the point, but early on in my planning I decided to add some additional visual interest and operational potential by adding another point on the lead track with a very short spur that would hold one single piece of rolling stock only.

In order to give that single car spot some sense of being another place that would have the odd boxcar or mechanical reefer dropped off and picked up, I decided to go for the Dutch company Artitec's low-relief kit of a cold storage building that I had bought years ago (and is no longer produced).

The kit is interesting in that it is made from resin rather than plastic; this allows for larger pieces with details moulded on.

As a result, the entire model is made up of only a handful of parts. Somewhat inspired by the rendition on the box, I gave the few parts of the kit I would ultimately be needing two coats of paint from a rattle can - the first black (for contrast and texture), the second a light spray of white.
  The transformation from the original light yellow colour is quite striking, and the rather finely moulded details are nicely highlighted by the darker shades from the first coat of spray paint.

As straightforward as this model is, it would need some adapting to fit the reduced footprint available on the narrow layout.

First off, the already low-relief model was reduced further in depth from the original 5cm to just a tad over 2cm, resulting in a setup almost identical to the warehouse scene on the left hand segment.

To give the model a bit more stability I cut a 2cm thick piece of styrofoam to shape and glued it to the back of the building's front

Since the roof moulding didn't really work with the reduced depth of the model, I replaced it with a simple piece of styrene and then added the moulded loading platform and canopy as per the instructions.

As with the warehouse on the left hand segment, getting the clearances right with this building is the first step as these determine where exactly the rails will go.

And again the object was to get cars fairly close up to the loading dock edge, as this will eventually make things look a lot more realistic.

Once the measurements were all in, checked, and then double and triple checked, laying the track was next.




Following the same procedure for laying track and wiring it all up as on the first, left hand side segment, the simple track configuration (which is indeed identical to the one on the left hand segment) was again made up of Peco code 100 track Streamline flextrack for the siding in front of the cold storage building and pieces of Setrack "snap track" for the siding running along the front of the layout.

No changes to the switch arrangement either, using a Peco Streamline medium radius "Insulfrog" point and adding insulator joiners to the inner rails past the frog and extra wire feeds to all three segments of track, as per Peco's instructions printed on the packaging of the switch.

  And again, Marklin Z Scale track pins were used to attach the track firmly and securely to the baseboard.

Wiring such a simple track configuration throws up no difference between an old-fashioned DC wiring approach and a DCC friendly one - although one could argue that the way I brought together all three pairs of feeder wires has a distinctly "old school" look to it.

After the all important testing using a volt meter to make sure that all connections work properly and provide eletric current to all segments of the track, weathering is next.



Avoiding the mistake of using the contents of an old rattle can, I spraypainted the rails, masking off (again, as always) the part of the point which is made up of moveable parts and also ensures connectivity for the elctrical current.

Painting the rails outside on a hot August afternoon in the midst of the 2022 heatwave made the paint dry really quickly, so that it was impossible to wipe and clean all the rail surfaces before the paint started to set. But based on the experience with the left hand segment, I knew not to worry and just let the remaining paint dry properly. A day later I used swabs saturated with isopropyl alcohol to wipe the top of the rails clean.

In order to brake up the unrealistic uniformity of the colour, I also repeated the process of darkening the ties, except this time I only used a generous splash of Vallejo black wash (which, when dry, is quite transparent).

Once all is set for the moment, making sure all electrical connections still work as they are supposed to is certainly a good idea before moving on to the next steps.


I did this following my usual best practice (applying current from a standard 1980s DC controller and measuring the current in all relevant places with a volt meter) but then also got out my old Roco/Atlas GP40 from 1984 for some actual test running - just a simple moment of inspirational nostalgia.


Prior to ballasting a few scenic items needed to be put in place, and one of them was a grade crossing. Apart from providing an opportunity to add some scenic interest, the crossing was intended right from the start to mark a boundary for switching moves when operating Pecan Street in "tuning fork switching puzzle" mode.

In essence, it simply limits the length of a train (and therefore the number of cars that can be switched in one move) by having a rule in place which states that the grade crossing cannot be blocked during switching.

For this layout, I made a point of using as many items as possible that I already had (i.e. purchased at one time "for later"), and in my stash I found a number of concrete grade crossing panels from BLMA (since taken over by Atlas).

  The nice thing about these was that they not only fitted my very approximative era very well (they have been used since the 1980s), but also happened to be the same type I observed in April 2022, at a crossing over a single Norfolk Southern track at Memorial Hospital in Roanoke Va.

The BLMA model consists of two center panels and four side panels, but given the restricted space I opted to use only half of these components - which I would clearly have to raise a bit since the packaging stated that they would work from code 60 upwards.

The short stretch of road was built up around the track using styrene sheets cut to shape, and the BLMA panels brought closer to railhead level using styrene shims. This is one of those occasions where a delicate balance needs to be struck - setting the panel in between the tracks too low will look terrible, setting it too high will foul the couplers. As always with clearances, it is best to use actual stock to make sure things fit; in this case the central panel sits at a visually realistic level - close, but not too close, for couplers to clear it and run over it safely.

  The crossing will be guarded by a pair of signals with working lights, using a very nice brass item bought on eBay.

Holes to accomodate the signals are drilled, and the road is given a coat of medium-light grey; this will serve as the basis for some overall as well as detail weathering later on.



In essence, Pecan Street is a self-contained layout. And whilst there are no plans to add any extension (such as connecting it to more modules) for the foreseeable future, adding the possibility to do so at this stage makes sense.
  This also makes sense since there is very little additional work involved in simply cutting an opening into the sideframe (once again making sure that the tallest and widest item of rolling stock in use will pass through without problems).

On the layout itself, a scenic break is needed to disguise the actual end of the track regardless of whether it can be extended beyond the baseboard frame or not.

A chance discovery of a 2022 model of a low relief bridge from the Bachmann (UK) range of "Scenecraft" resin models provided just the kind of "scenic disguise" I was looking for. Although intended for British layouts, the design seemed a good fit for a layout based on an East Coast location.
  In order to work visually the depth of this very low relief model had to be increased quite a bit.

This was achieved by adding a structure made up of three pieces of styrofoam cut to size. The result in itself is a very crude contraption, but since the model covers it up, this is all it takes.

Black colour on the inside disguises the styrofoam and creates enough of a "dark passageway" feel to make it work as a scenic break - one of many techniques I like to use that originally stem from theatrical stage design. For the moment, that was it; the obvious rough bits and edges would be patched up and hidden in scenery after the adjacent scene of the track crossing had been worked on.


The narrow nature of Pecan Street poses a number of challenges, some of which have already been mentioned. Most of those boil down to the fact that there simply isn't a lot of room for scenery - and the grade crossing is certainly one of the areas on the layout where this is felt the most.

I have also mentioned how, in such cases, I like to fall back on Frank Ellison's modelling approach involving theatrical stage design - and this is perhaps also most apparent in how the grade crossing is set up on Pecan Street.

In its raw modelling state, the road crossing the tracks curves away slightly in order to provide at least some illusion that it doesn't just hit the backdrop in a head-on collision, but it still clearly ends right there, a mere 3.5 inches (9 cm) away from the track.

In order to conceal this, a number of steps would be necessary to create an illusion of depth, and one central prop was a rather nice model of a truck, manufactured by a company named Boley, that I had bought cheaply and put aside years ago for "some future project". Now, it would literally take center stage.

But before that could happen, the stage itself had to be set. The idea was to add a flat photographic backscene to the backdrop, and then add some semi-relief props to make the scene look threedimensional and create the impression that the road crossing the tracks actually went somewhere in the background.

After finding a suitable royalty-free image on the web, this was reduced to the required size, cut out and glued onto sturdy 2mm cardboard, and given a protective layer of matt varnish. This was then in turn glued to the backdrop, and a first batch of scenic material applied to cover both the edge of the card and fill in a gap between the overbridge and the backdrop. For the more three-dimensional props, two large scenic trees were cut in half and then glued into place.


The trees came from a range that can generally be termed "cheap model trees from China sold on eBay". They are somewhat infamous for their overall identical appearance (which of course explains the cheap price) but also rather popular, since it is fairly easy to improve their looks.

Usually the excuse for giving these a go is the sheer number of trees needed and the cost involved, but I still happened to have these lying around after a curiosity purchase and successfully (I felt) "pimping" them with Woodland Scenics foliage for use on Little Bazeley.


  Building up more layers of scenery (mixing Woodland Scenics fine and coarse turf, foliage and bushes of various colour shades) allows the foreground to blend with the flat backscene, further enhanced by some strategic "shadows". As on a theatrical stage, it is all aimed at deceiving the onlooker into believing that the scene has more depth and reaches further back than it actually does.

The scene could be left this way, but in order to fool even a lingering observer a bit longer, the aforementioned truck is placed as a view block which also deflects attention from the flat backdrop to an actual three-dimensional prop in the foreground.

In order to achieve this illusion, some further theatrical trickery is necessary - what seems to be a truck is actually only part of one.

Cutting down the rear of the vehicule at a 45 angle not only allows a longer vehicule to be squeezed into what little space there is, it can then also be placed in a skewed position (another theatrical trick for props that are supposed to fool our sense of perspective).


Taking apart the truck for the procedure was also a good opportunity to fit a driver behind its wheel - a suitably cut-up Preise figure from a set of track maintenance workers. It's another stage setting adage: things that are there may not always get noticed, but things that are missing most certainly will.

In a perfect world, I would have turned the front wheels somewhat, but simply thought of this too late - so I simply imagine the driver is just about to apply the steering wheel...



With the crossing scene set up, the rest of the remaining portion of the second, right-hand segment, was to be given a fairly straightforward treatment in terms of ballasting and background scenery.

  So out came my array of transparent household storage boxes, holding the contents of various bags of ballast and scenic material (mostly from the Woodland Scenics range).

The usual pieces of styorofoam were cut, glued down and given a heavy coat of dull green acrylic paint in order to serve as contours for the background scenicking in front of the backdrop.

As for the ballasting I decided to generally keep it fairly neat on this part of the layout compared to the warehouse segment. The basic plan was to see how things would look once completed that way, and then add more ballast or weeds on top of the already ballasted track if needed.


  The procedure itself, of course, remains the same. Woodland Scenics "fine" grade light grey ballast (B74) is gently scattered from a small teaspoon, leaving a small pile which is then spread out using my index finger.
Once tamped down this way, more ballast can be added where needed (which is a lot easier than getting rid of too much ballast). It is impossible to avoid getting at least a little bit of ballast where it shouldn't be (e.g.on the sides of the rail), but all it takes to remove this is a gentle approach using a fine brush and a toothpick along with some patience.
Once everything looks okay, scenic cement is applied from a dropper to the edges of the ballasted area, letting the ballast and scenic material soak it up thanks to capillary action. Adding more scenic cement to areas already soaked in it will make it "spread" to areas that are still dry.

Again, some patience is required, but it doesn't take too long and causes minimal disturbance to the unsecured ballast grains (glue applied directly to dry ballast will cause it to float and drift around). Since taking advantage of capillary action for ballasting, I have in essence switched from using a syringe to a simple drop dispenser with a rubber teat, since the latter provides a more controlled and gentle application.

Again, the scenic ground cover is built up in steps by first covering the painted styrofoam bases with white glue, which in turn is covered with a mix of Woodland Scenics "fine turf green grass" (T45, light to medium green) and "weeds" (T46, medium to dark green) in order to avoid a uniform look. Adding different colour shades of Woodland Scenics "bushes" (FC146 medium green, FC147 dark green, and FC149 forest blend) then results in another layer, which is secured with scenic cement.
  In order to provide a smooth transition to the road crossing scene, a short tree line needs to be built up along this basic scenery - which means more tweaking and pruning of some "cheap model trees from China sold on eBay".
Since this is strictly background scenery and therefore explicitly not supposed to attract attention away from the foreground, it's not about detail but rather the overall impression.
Tree stems therefore do not need to be visible, even more so since the trees are heavy and dense in foliage (a real life example of such a tree would be the Arborvitae Tree).

This in turn means that I really cut into the "plastic skeleton" of these trees, leaving only a minimal number of unconnected "branches" - they won't be visible, but this provides just the right amount of structural cohesion and flexibility at the same time. The cut-up structure of the tree still shows off some of the very bright green the tree originally comes in, but once it's all turned around the Woodland Scenics foliage hides practically all of that.

Since the layout itself only has a very shallow depth of 6 inches (15 cm), the background scenery needs to be extremely compressed. It would be almost impossible to fit an actual complete model of a tree into a space of 2 inches (along the single track), let alone into the cramped quarters of just one inch (along the segment with the mainline and the spur). Cutting up the trees is therefore very much the equivalent of having super-low-relief structures (as, indeed, the storage building on the spur is).
This also means that the tree line needs to be dense; too much light foliage which would allow the onlooker to see through the trees would be a dead giveaway.

Stage designers have developed many different tools to creating the illusion that a scene is much deeper than it actually is, and this is just one example of how applying theatrical stage design to a layout works. Another trick of the trade applies to hiding the true dimensions of low-relief buildings. As with the warehouse on the left-hand segment, an unobstructed view of the storage building on this segment will clearly show that the side walls are very shallow since they meet up with the backdrop almost immediately.

In order to create the illusion of an actual building that reaches further into the background (as a real structure would) the view of the side wall needs to be obstructed, This is exactly what stage designers do both in theatres and on film production sets, and in this case the effect is easily accomplished by again adding some sprawling vegetation. On the more visible right hand side of the building the tree line meets up with it, along with some shrubs and bushes. For the left hand side of the building - which is in the far left corner of this segment and thus less open visually - I opted for a more refinde model of an individual tree.


Unlike the cheap quantity-over-quality plastic examples from China, this one was hand-made in Vietnam using strands of wire to simulate the bark and branches more realistically (as is the foliage). Also available on eBay, these trees are of a far superior quality, but obviously also carry a higher price tag (you get what you pay for). The corner seemed a good spot to "plant" such a model - visible enough, but not too much. The tree is used as is, with no cutting, and it does protrude slightly out into the spur a bit, but since its location is the end of the line for that siding, this isn't a problem.
Building up the scenery in layers like this also has the distinct advantage that it's always possible to go back and add more shrubs, bushes, weeds and other items later on.

But already at this point, the background scenery fulfills its purpose. It provides a setting, disguises the otherwise obvious spatial limits of the layout, and overall sets the mood of the scene. And of course it works best when there's something in the foreground that shifts the focus of attention.


Work in progress, more to come...



Text and pictures are (c) 2022-2023 Adrian Wymann.


page created 12 March 2022
last updated 23 March 2023