The Water Shops Project


The Water Shops Project

An HO scale diorama. This is not a kit.

This project started out as “layout party build” on a forum, which I eventually used for my opening picture on New England Brownstone’s main page. I’ve had a few requests for how I did it and I needed to put some instructions up on my site as to how to work with NEBS products, so I figured I could kill a few birds. I’ve made some painting changes since I first worked on the original tutorial, but the basic construction techniques remain the same. I’ve added and will continue to add more instructional info because the first edition concentrated more on after-the-fact than how-to.

I can say that the success of this project is proportionate to the amount of time I spent studying the prototype, even if this model is not an exact replica of the building. As long as it’s realistic and technically feasible, then it should be believable and acceptable. If there is something out of the ordinary that I’d like to model, I find it imperative to have the resources to back it up, like a photograph. This helps thwart the negative advances of the critics who will be more than happy to point out mistakes. Facts will always shut them up.

In the Beginning

On my layout with no name, I wanted to create a very New England look. As you can see in the pic below, the layout consists of a double track main, the “high line” and a lower “low line” section that runs down to an industrial area and to a hidden storage track. A branch that runs along the front of the backdrop, which is to be scenically blocked by some building flats and trees, connects the two. Although hiding the branch is not anything out of the ordinary, I wanted to try a little slight of hand to help create the illusion of not being there.


Copy of shelf_layout2-2-1 Copy of shelf_layout2-2

The Initial Mock-ups

Real tunnel

I wanted to put a passenger station with a pedestrian walkway where the river is indicated in the cad drawing, but I didn’t care for the way things were working out.   Pretty lame!  It didn’t have the impact that I was looking to create, so I started to think about a road with a large plate girder bridge with a set of stairs going up to the station from the road, like the one below.

Real stairs


Then I realized that I really didn’t need to have a passenger station, even though I would like one, I wasn’t going to have a lot of that type of service on my railroad and it would take up valuable real estate. I could get away with a small depot and still be happy, but I felt I could do better. So, ok, now what to do?

How about a river? I could have a mill flat sitting on the top of a stone foundation that has an arch with a river running underneath.  I’ve seen that before and knew exactly where there is a prototype for it, so why not?  I could cut out a portion of the hidden track’s roadbed and replace it with a bridge and could create the effect that the viewer would be looking through the archway under a mill, not under a hidden track.

The Subject

This picture is of the Mill River spillway through the old Water Shops complex up the hill in Springfield, MA. The bridge in this photo is not a RR bridge, but it gave me the idea for the scene I was looking for. I have little history on the building, but I didn’t want to model this particular building anyways, so I didn’t do any in-depth research on it. Preliminary Internet searches brought up very little pertinent info.
I was going to use some random stonewall castings, but when I went to take more pictures of the site I noticed that the builders had used cut brownstone (how convenient), which would make this project much easier for me. (I had already started on my layout project before I took this particular pictures, so you may see some random stonewalls being used in the mock-ups).



Copy of mill 004
The vehicular plate girder. Kind of rusty. I planned on replacing this bridge with a railroad plate girder bridge.
The Springfield Armory’s water Shops. This is where most of the forging was done in the early days. The building was added on to it’s left side with the addition being a mirrored image of the original.


The gabled roof right behind the towers was destroyed by a fire in the 80’s and never replaced.



amory 010
The towers were also reduced in height after the fire. The building still lives on as a multi-use industrial facility.


Copy of mill 002
Where the building crosses over the river there is a large arch, which is hard to see from the road unless you are looking for it. There is a steel plate girder road bridge in front that somewhat obscures the view, but if you look hard enough you can see the water rushing below.

I’d like to point out the coloring (weathering) of the stonework. The base color of all of the stone is reddish brown like the finished stone just below the brickwork. The stones above the arch had a bluish black color (I’m not quite sure what caused it) and the stones to the side are covered with moss/mold, fungus, hence the green funk color.

Copy of mill 001
I had to get out of my car to take this picture.  Notice the coloring of the stones.  They really are purplish hue.


Copy of mill 004
The vehicular plate girder. Kind of rusty. I planned on replacing this bridge with a railroad plate girder bridge.

Ok, now I had a prototype. Not that I was planning on modeling that particular building, I’m just using it for ideas.

This is a shot of a mill turbine house in Hazardville, CT. It’s mirror image of my scene, but inspirational non-the-less. Notice the tail water discharge hole (the dark square hole to the lower left). This particular mill has a nice assortment of building materials and techniques used in its construction, which would really make a fine interesting model in the future, but not this time around.


Mocking it up

Some of the first arranging.


Some rock placement and the arch laid out on the stone wall


Here is the back wall with the arch carved into it. Hidden track would run right on top of the arch behind the brick wall. The wall’s windows were to be blackened out in-order to keep trains from “running” through the building.


Overhead view


The hidden track within the building.





Basic layout of bridge on top of abutments.

I kit-smashed my Concor/Revell/Heljan (?) engine house model for the for the mill complex and the “boiler room” for that kit would now become the turbine wheel room and will sit on top of a stone and concrete foundation that will have an opening, or tail water discharge from the turbine. I planned on working up the mold to build a better factory, but I ran out of time and that part of the build remains stock.
However, putting a cheap plastic building on a nice foundation turns the cheap looking building into a less cheap looking building almost instantly. Add a new paint job and some weathering and this scene would be complete.


The Tunnel Liner


Lining the Tunnel

I needed to put a lining of sorts inside of my river tunnel arch and I had no idea what the real life masons used on this mill, so I had to make an educated guess. I’d seen brick used as linings even though the arch’s face had voussoirs (arch stones).
I had an extra brick mold from a previous mold-making mission that was available for a project just like this one, so I was in pretty good shape, so far. I really didn’t take any measurements because I knew that I would have to rely on whatever cylindrical item I could find that would be close, would have to be good enough. So, after about ten minutes of searching and a handful of hopeful candidates I finally decided to go with the tube of silicon caulk.

This actually killed two birds with one stone because I was planning on using silicon for water effects and would have to have searched for it later on. As long as I don’t misplace it, I’m all right.


I found out that I would have ended up having to cut the tunnel after all.

I knew that it was going to be a pain to bend the mold, wrap a piece of styrene over the open side of the mold and then tape it. I did it, but it wasn’t easy. Curving a mold is one thing, bending it 180 degrees and forming it into a horseshoe, well that’s another story. Semi stiff silicon rubber, small diameter plastic surfaces and greasy styrene (Vaseline from previous molding job) don’t stay still. Even if I had two extra hands, I don’t think it could have been any easier. All I could think about was a greased pig. I will have to do something different next time.

I mixed and poured the casting and let it sit for about an hour. I started to de-mold the casting when I suddenly realized that I was playing with a bear trap. I was able to keep it from flying everywhere, but the casting ended up getting broken. However, I had originally figured that I could walk my way around the fact that I wasn’t going to pour a perfect diameter tunnel to match the wall arch and could cut the cast tunnel down the middle to fudge it, so I wasn’t disappointed.

Looking at the pic below you’ll notice the cutout in the mold adjacent to the styrene. This is the sprue hole to pour the plaster. The mold was originally designed for flat-back service, not for bending around caulk tubes, so when I cut out the sprue hole I compromised the mold’s integrity and ended up with the mold kinking at the tube side. I should have cut the sprue hole in the mold 90 degrees from where I did and poured as if filling a U from one leg. The stresses on that spot would be more natural and conducive to keeping the molding form correct. Hindsight is always 20/20. Next time.

I used an old brick mold and wrapped it around a tube of caulk because it was the right diameter for the arch. The cut out in the mold next to the sheet styrene was the sprue/gate for me to pour the plaster into the mold.

I ended up breaking the casting even more, by accident, but it wasn’t anything super glue wouldn’t fix. It didn’t come out perfectly, but it’s impossible to see from the fascia, anyways. In fact, no matter how hard you try it is impossible to see the brick detail with out direct lighting.

Sadly, the casting came out broken. Luckily, it really didn’t matter since no one can see it. In fact, I could have simply used another material for the arch. This is a case where one should not waste time on unseen things. Even if you look at it at a lower angle, you still can’t see any bricks.


This is a picture that tells the truth about the distortion caused by wrapping the brick mold around the tube.

Notice how the casting is thin in the middle. This was due to the mold getting distorted while wrapping around the tube of caulk. The mold was not designed to be wrapped around anything, so it didn’t like getting bent.

I added a strip of scribed siding to the backside of the tunnel to help reinforce it and to help seal out stray light. The hidden track will ride right on top of this.

Here I added some scribed wood sheeting to give it some support.


The underside. It’s not a perfect fit, but who cares since no one will see it.


Final assembly.


The Wheelhouse Foundation

…kind of sounds like a charity for widows of steamboat captains.

I needed to make a mold for the cast concrete section of the turbine house foundation, or wheel pit, so I whipped this up out of some scribed sheeting I had lying around.

The first concrete form. this is the onethat was not sealed correctly and caused the casting to have the wave in it.

Fair Warning, if you build a mold out of wood, make sure you glue stiffeners to the backside of the scribed sheeting and seal it. Otherwise, you have things like this happening.  Notice the nifty curved section in the casting? It must be the one place I didn’t glue. It wasn’t a big deal. I could cut the casting below that point and work from there, or rebuild my mold. I decided to do the latter.

FAIL! NOTE: Must seal all wood faces in future molds.


The New and Improved Wheelhouse Foundation

I really didn’t want to try and make the first warped foundation casting work by hacking on it, so I decided to make a new form and pour a new casting. Since I was starting from scratch I decided to do it right this time around. I spent the time to add draft (taper) to parts of the mold to make it easier to remove the casting without cracking it, particularly on the core that was to create the rectangular tail water discharge. I also beefed up the form to prevent anymore warping during the pour.
I had added some vertical scribes in the original mold to simulate separate boards, but they didn’t come out that well, so I really laid into it this time by spending a lot more time on it. I also added nail holes and knotholes. The nail holes were added with a fine pointed burr and the knotholes were burned in with a red-hot sewing needle.

New mold.


Version 2.0 I added nail holes and knots in the wood. I used a pin that was heated in a candle flame to burn the knots into the wood. On a side note and in reality, most board formed concrete is not this detailed as most was smoothed over after the forms were removed.


The details came out great.  I have to admit that it does look cool and nobody other than me will ever realize it isn’t quite realistic and that’s why I used it.


A Fitting End to the Stonework

I had to finish up on the transition from the fill, the abutment and the natural features of the river. I really don’t care for the stone abutment sitting right on the bedrock look, even though that is exactly what would have been done, but rather the look of years of dirt, crud and vegetation creating the transition, which I would be taking care of shortly, but I had to deal with holding back part of the ROW fill.

On the lower left of the scene I had a 45 degree slope of dirt from the fill that would have ended up farther into the river than I intended it to go, so I cut out a small piece from a random stone retaining wall casting I had laying around and placed it to help keep the fill back. I cut the wall to fit the stone outcrop and could have stopped right there, but I didn’t.

Not being one to let it go right there, I decided to “fit” the wall to the rock outcropping. With the help of some “articulation paper” I was able to get a good tight fit. Articulation paper (AP) is a common item in a dental office. The doctor uses it to check contact when inserting crowns, bridges and such, between adjacent and opposing teeth. It is nothing more than double sided carbon paper. The kind I use is black on one side and red on the other, but you can get them in all kinds of colors. It doesn’t really matter, as long as you use it correctly, which means “rubbing in” (see pic) then cutting or grinding away the high spot indicated by the colored mark (see following pic). Same way the Egyptians made razor tight joints between the stones in the Pyramids, except I don’t think they had rolls of (AP).

As you can see from the following pics, the more I marked and trimmed, the more marks appeared.  Note: I didn’t mess with the marks on the bottom (the flat area, which is shown on the top in this pic), just the sloped part of the casting.

Rough layout


Rub-in articulation paper
Turn piece over.  Notice mark, remove mark.  I did not worry about the marks on the bottom, only the ones on the cut out section.
Flip over, remove marks.
Rub-in again.


Flip over, remove marks.


It takes longer than cutting and slapping it up in place and it not really necessary, but it’s worth the effort. This was more of an exercise in technology transfer, from dental lab to model railroad.

Hey, someone might find it useful.

The Other Side of the River


I wanted to see if I could duplicate this.

I wanted the riverbank area directly in front of the viewer to look something like this. I knew I could get it to look fairly close to this using gloss medium, but I felt it would look better if I used the Enviro-Tex water instead. I needed to work on the transition between the abutment and river and figured that sand would have eventually settled into this area either by the erosion from the ROW fill, or by the eddy action of the water over time.

I mixed a thin mix of Hydrocal® and poured a small amount in that area and then with a pencil I pushed it around into the corners. After I was happy with the arraignment I rapped the worktable with a hammer to settle the plaster down into a more natural shape. I added some WS medium size brown ballast to the plaster bank while it was still wet and then after a few minutes I pushed the ballast down into the plaster, so it looked as though the rocks were filled in with accumulated sand.

I gave everything some time to dry then applied a diluted coat of aged concrete with a dab of dirt colored paint. It actually looks better in real life and should look much better once I set up for a “real” pic. The other pictures show the other areas where I repeated the process.
The lousy transition from the sand to the deep dark water will be remedied with the airbrush before any “water” is added.


Tad bit closer view


I poured a thin layer of plaster and tapped the dio’s base to cause the plaster to settle and feather itself with the river bottom. I then added some tiny cobbly looking rocks and pressed them into the plaster once it began to set.

Painting the Stone Foundations and Abutments

 I wanted to give the impression that the mill and the railroad stonework were built at different times, so I gave each their own paint job. I painted the mill foundation by mixing a glob of ‘crimson” acrylic paint from the tube with some WS burnt umber and stone gray, which did a poor job of making it look like a red sandstone. It just didn’t look right. Almost like it was transparent with a lot of white showing through causing it to look too pink.

The paint job didn’t look right.


Copy of mill 015
I really wasn’t expecting to get the coloring perfect, but somehow got it right. I had to use some Floquil Boxcar red with a dab of grimy black to cover over the original attempt.  I then used a wash of grimy black that brought out the details in the stonework.



It was looking good, but something didn’t quite look right. So, then I went back and added more A/I wash and dusted the who thing up with the airbrush loaded with dust colored paint.

I then added some orange, yellow, white and black chalk powders to give it the really dry look. I have found that this is probably one of the best things to do when it comes to weathering rocks and stonework. Stones, rocks and concrete are very light in color when they are dry and become darker in color when they get wet.

The orange and yellow were used to simulate rust staining and were concentrated more under the bridge shoes. I didn’t make big orange stains, even though I have seen them like that on proto abutments, but I didn’t want them here. The chalks were used sparingly as to not to cover up the base coloring. Less is more here and just enough to do the job. Notice how it doesn’t look all that orange or yellow.

I added the black chalk to simulate cinders and soot. I used a big fluffy artist brush to apply it using a very light downward stroke. This was a general application ,but was a tad bit more concentrated in the areas just below the bridge to simulate rain carried coal crud.
The white chalk was an attempt to give the stone that efflorescence look. It didn’t work that well for that, but it still came out looking decent.

Painting the River bed

Since there is no real depth to the water I fell back on a trick I learned from Dave Frary that uses darker colors for depth. By using an airbrush and some sand colored paint I was able to blend up to the sandbar. The technique works great.

This pic shows what everything looks like after some painting.
I liked the results, but I tweaked it some more. Here it is after the attack of the airbrush. Even though I planned to use the Enviro-Tex I still used the blending from light to dark to simulate depth technique. It is by far one of the best ways of simulating depth in a water scene. Seldom does fake water (if you use resin) need to be any deeper than 1/8” and a lot of the times it can be less. In that case, use the gloss medium.


Rough painting the sand



The River

Well, not quite, but getting closer.

I’m was finally starting to see the light at the end of the tunnel in respect to adding water to the scene, but I kept finding things I needed to do before I actually added the water. I wasn’t quite sure what water modeling technique to use, so I did a bit of final searching through some past threads and determined that I wasn’t going to use one exclusively because I think each material has it’s own merits and should be used accordingly to give the desired effects.

This particular scene is rather flat due the fact that the hidden track that I’m trying to hide is set at a certain height, so a super cascading waterfall it was not going to be (that’ll be in another build). With this in mind, the highest point of the modeled water will be no more than 1” (25mm). I wanted the water to be interesting, but more importantly, realistic. So, back to the resource files.

Most rivers (around here) are a deep brown to black. The white is just turbulent aerated/bubbled water and can be generated by a stick if the velocity of the water is high enough. The bubbles dissipate rather quickly and there is usually darker pooled, or low velocity water right next to it. Of course, this is all dependent on the amount of water flow, too. During the spring melt there is so much volume of water flowing that everything is whitewater. Where as during low water or dry conditions there may be stagnant pools in natural potholes that may be disassociated with the original river, which may be just a trickle of water barely creating any white water.

I didn’t want to create a raging spring river. But elected to go with a mix because that would allow for me to try a few different techniques.

What I had in mind was to come out of the tunnel with a slightly turbulent water flow that quickly dissipated into smooth water. From there it would ducloseup_millrunmp into a larger area, or pool thus reducing its velocity. The tail water from the turbine house will dump into this pool, also.

To give credence to the pool I placed a plaster stone, that had it’s back flattened by rubbing it on a piece of sandpaper laid out on the table, between the two major rock outcroppings that opposed each other from opposite banks. This creates a double weir effect. Instead of one big flow I divided it into two smaller flows, which delivered their contents into the area at the front of the layout. Since I have to follow some basic physical properties like “water flows downhill”, I decide to do the two sections in front of the tunnel in two pours. The first pour would cover both areas. The second pour would be from the tunnel to the double spit.


Base painted. I added the white where I wanted some turbulence with the idea that it would appear like underwater bubbles.

The water leaving the turbine creates a lower pressure on the downside of the turbine and affords the full use of the total head of water regardless as to where the machinery in placed and because of this the water does not come flying out from the discharge tunnel, but instead creates a rolling/boiling effect.


Silicone rapids. When I poured the epoxy I dipped a brush into the pour and then painted the rapids with the epoxy.



Dam to keep the resin in the scene and not on the floor.




Making use of a goodly amount of silicone sealant to make sure the epoxy wouldn’t leak.
Another shot of the silicone. Notice the irregularity in the stone wall like there is a piece missing. That is where I didn’t have a piece to fit and decided that I could cover it up easily with a stone casting, so that’s what I did.


Being On The Level

Next thing was to make sure the whole thing was leveled. The level was cut from an old aluminum tri-square and works rather well in tight quarters. Luckily, I didn’t have to make a bunch of adjustments. Notice how I leveled the base out in both a North/South and East/West directions. Just because it’s level in one direction does not mean that it is level perpendicular to the first sighting.

Freshly poured resin and smooth as glass. When they say this stuff climbs the banks and rocks, they mean it.


Leveling for the resin pour.  Notice the small amount of silicone that I stippled right in front of the turbine discharge.  Once the silicone started to skin over I flattened out the little wisps that looked more like splashy water than rolling water.
Leveling both ways.

Pouring the resin


Resin poured. I used Envirotex casting resin, but there are others out there. I have to say that it worked well and that I had no problems with it.




To give the water some ripple you need to pick at it, wait, pick at it, wait… You can pick at this resin for hours thinking it is just perfect only to come back a few minutes later and find it absolutely flat, like glass.


mill 054
Honestly, I should have stopped right here. This is what real water looks like. I over did the white water.

Final touches

Like I said, the white water is too much for the flow. Not complaining though because I learned a lot from this project.


The Final Shot



I’m not finished!

Last Word

As I had mentioned earlier, I overdid the white water. After observing the real thing more, I started to pick out things that didn’t look right on the model. One is that white water dissipates quickly if the flow is not heavy. I on the other hand I added too much aeration, which seems to flow on for too long.

There’s really not too much to the whole diorama. As I began taking it to shows as a display I thought about finishing it up so that the ugly backsides of the abutments couldn’t be seen, but then I realized that it was a great way to show how the product line can go together. I’ve had a few show attendees tell me that it was a good idea, so I figure I’ll keep it that way.

Well, that’s about it. I hope some of you can use my tutorial to help you out with future water projects. The most important thing I can think of is OBSERVE your subject, whatever it may be. Assuming will catch you every time, but if you have your guns (camera) loaded, you can’t go wrong, even if your model is not 100% prototype.

Happy modeling and feel free to contact me with any questions you may have.

Thanks again.


Share This: