Welcome to the Coal Gondola page. This page documents the process of scratchbuilding, molding and casting a late 1800’s gondola for use on my turn of the century model railroad.
Why do this? Because there is very little in the way of useful models, either ready-to-run or craftsman kits, that are available and appropriate to the early 1890’s period. This requires almost everything needs to be scratchbuilt. If all the models to be used on the railroad had to be scratchbuilt to a fine detail level, it would take a very long time, longer than most people working on a layout can afford to devote to the model building process.
So in order to maximize the time spent on model building, the casting process is employed to quickly duplicate the effort involved in making one high-quality master model. A single model is built and a mold is made of it. As many as 20-30 or more copies, each equal in quality to the original model, can then be made in a fraction of the time required to build similar models one at a time from handmade parts. If the mold lasts, extra copies can be sold to other period modelers, in order to help defray the costs of the materials involved in the casting process (rubber for molds, resin or metal for the castings, etc.). But the important thing is a sizable, well-detailed and unique car fleet can be built in far less time than would otherwise be necessary.
The railcar I have selected to model is a Missouri-Kansas-Texas gondola built around 1879. It is similar to a partial photograph I have of Danbury & Norwalk gondolas sitting in Danbury yard in a late 1870’s shot. Unfortunately, the whole car cannot be seen, but they look similar in size and shape to the MKT gondola drawing. There are no existing photos or drawings of actual Housatonic gondolas as far as I know, so until better information is uncovered this design will have to do.
The model will represent an older 28’ gondola belonging to the Housatonic Railroad, used to haul coal from the tidewater port at Wilson Point, CT, to inland destinations in Western Connecticut and Massachusetts. Gondolas were preferred by the railroad for this type of traffic as they could haul other types of loads when not in coal service, which a hopper car could not do. (Yes, I know about the convertible hopper designs, my railroad did not have them.)
The prototype Housatonic RR gondola had a capacity of about 15 tons, not very much. By 1892 railroads were developing larger wood cars that had capacities of 30 tons, with the majority at the time holding about 20-25 tons. Wood cars wore out quickly, usually within 12-15 years, so in 1892 these 15 ton gons would be approaching retirement. By the turn of the century steel was starting to replace wood for underframes, and capacities were increasing to 40 tons. By World War One the capacity was generally up to 40-50 tons with steel parts and panels taking even more of a leading role. The short lifespan of wood cars, coupled with the quick advancements in capacity, means during this period the look and size of railcars is a strong signal about what era you are modeling.
After scanning the illustration from the book “The American Railroad Freight Car” by John H. White, I placed a printout of it on a reducing photocopier and played with the size ratio until the gage between the rails in the picture scaled out at 4’ 8.5” with an HO scale ruler. At that point I knew the railcar image was locked into HO scale. I made about 20 copies to make sure I wouldn’t lose it. (I could have done this in the computer, but at the time I did not have good photo or illustration software.)
Next, I went shopping for supplies. I bought scribed sheet styrene that was pretty close to the decking spacing, and strip styrene in sizes that were pretty close to the sizes of the beams used in the underframe. A bunch of detail parts were purchased too, including Grandt Line Nut-Bolt-Washer (NBW) castings and narrow-gauge D&RGW Gondola side stakes with stake pockets.
Before starting I took some 120-grit sandpaper and roughed up the strip and sheet styrene a little to give it some wood-grain texture. I felt this might also help hide any errors I made and had to sand a bit. This turned out really well, and later I augmented it a bit by scribing in some random grooves with the back point of an X-acto # 11 knife blade.
Starting with the decking, I cut out a sheet matching the size of the car deck. Next came the end beams and side beams, cut from strip styrene. I marked the location of the taper on each board (at the ends) and sanded them down to match the profile. The notches in the side beams to clear the end beams were cut and filed, then the box frame was assembled and cemented to the underside of the cars’ deck.
Next, the bolsters were cut and fitted. I didn’t model the fine appearance of the prototype’s metal bolsters, but built up several thicknesses of styrene to the depth of the side beams, and made a rectangular pad for the truck to sit on. The truck was balanced on the bolster as it sat in the car, and the bolster adjusted back and forth until the truck looked about in the right place (The truck used is larger than the prototype truck). Once this looked OK, the location was marked, measured and transferred to the opposite end, and both bolsters were cemented in place.
The inside beams were installed next. I measured the distances from the outside edge on the drawing, transferred the dimensions to the underside of the car with a pencil and started cutting beams from strip styrene. A little sanding to fit and each went in nicely. I had to do the ends beyond the bolsters as separate parts.
A thin plate of styrene sheet was used to cover the cavity where the draft gear box was to be mounted, followed by a Kadee coupler box top trimmed and cemented to the pad. The hole was filled with putty and sanded flush. I’d probably leave the hole intact next time.
Now it was time to start detailing the car. I got out the NBW castings and started trying to apply them. It took a long time and a lot of patience, I can tell you now that it’s a good idea to leave the NBW castings on the stems, and to drill out a hole for the stem to go into on the model. I didn’t do this and it took a long time to apply all the details.
“Metal” bracket plates were applied to the sides of the car where the bolsters were supposed to be mounted on the inside, these got a pair of NBW’s each. The truss rod NBW’s in the ends were very simple, the smaller ones representing the bolts into the side and center beams were very tough. Too small to be held with tweezers, and static electricity will make the $#&*@ part stick to the end of a knife blade! I also added a couple of small semicircular bits of styrene to the B end of the car, to hold a brake staff.
I put the car assembly aside now, and started on the gondola sides. I didn’t have strip styrene that matched the side boards, but the scribed sheet styrene was pretty close if I used two ‘board’ widths. I cut off a strip and cut it to length for the side and end boards. Now, because I’ve been building this model with the intention to mold and cast it, I understood I only needed one side board and one end board. These would go into a separate flat mold and be cast twice for every complete model needed, along with a couple of other separately applied parts.
The Grandt Line side stakes with stake pockets were long and a little heavy but I decided to live with it. I cut then down to the right length, trimmed the bottom of the pocket to more closely resemble the stakes and pockets shown in the drawing, and marked the boards where the stakes had to be glued on. I carefully cemented the stakes to the side board, and had a fit when I was done - the second stake from the end overlapped the bolster plate on the side of the car body. I was able to carve them off, sand down and reapply them a few scale inches to the inside. The appearance really wasn’t affected, thank goodness.
The end board was done similarly, except the stakes are supposed to extend through the car floor and be fixed to a beam underneath. I simply used some of the cut-off ends of the other stakes and cut them flush with the bottom edge, they will look fine once installed. I also glued a corner nailer to the edges, these will provide some gluing surface to the edge when the side have to be glued together in final assembly.
Needle beams were next - these are the beams that sit across the bottom of the car near the middle that the truss rods push up on. The needles, or queenposts, were a problem - Grandt line makes fine detail parts, but I was worried they wouldn’t cast well without injection molding, which I’m not set up to do. In the end I decided since the prototype car’s queenposts were not much more than flat saddles, I’d just glue on some fine half-round strips on either side of the truss rod location, and sand them to the proper shape. I’ll worry about finer queenposts on the next model I build. (This worked out just fine, BTW.)
That about covered all the parts I needed to cast. Some detail parts like brake wheels are simply better to just buy from Grandt or Tichy. And so we were ready for the molding stage of the project.
So now I was ready for the molding process. The add-on parts were glued to a sheet of styrene large enough to support the lego dams I was going to use. This was the first time using legos and I was a little nervous about whether they would hold back the liquid rubber adequately. After building the dams I sealed the outside edges with yellow glue, to keep the dam tight against the styrene base. (Today I would use low-temp hot glue instead, it sets up faster and is just as easy to remove.)
It was time to mix the rubber and pour the mold. Before pouring, all surfaces got a coat of mold release, which is a 1:10 mixture of Vaseline and Paint Thinner (Mineral Spirits). The Vaseline is carried in the volatile thinner, which shortly evaporates and leaves a thin coat of Vaseline on the mold dam and model. Simple but it seems to work well. It doesn’t hurt that the rubber doesn’t stick to Styrene or the plastic the lego blocks are made from. Unfortunaely, it likes to stick to rubber (itself) really well, so you have to watch that.
I always try to mix the rubber without stirring air bubbles into it, but it never seems to work. I always get air into it. Fortunately I usually get the biggest bubbles out, and the small ones don’t get caught next to the model. I usually mix a very small batch and paint the rubber on, ensuring good contact with the model, but this time I just poured it and agitated the rubber with a toothpick to get it into all the crevices. It worked fine for the add-on parts mold, unfortunately I should have been more careful with the car mold, I lost a little detail around the draft gear. Nothing important, though.
After waiting the requisite four hours for the rubber to set up, I gently removed the legos. They worked out really well, with just a little bleeding at the seams. Then I gently pried the rubber from the master. It was a little difficult, because of the undercuts on the side pieces (stake pockets), but eventually I got it off successfully. I immediately bagged and stored the master model against the day this mold tears and I may need to pour another.
Being an impatient person, I immediately mixed up some resin and tried to make a casting. The first two didn’t work out well, lots of trapped air bubbles and I wasn’t able to force the resin into the undercuts. I quickly cut a slit in the mold over the undercut stake pockets, and then I was able to force resin into the cavities with a toothpick. Also saw where I had to tease the other cavities to release trapped air. Some extra resin poured before adding the press plate helped to float off the air bubbles, and by the fourth casting I was getting some nice parts.
(Note: The next time I mold parts from this mold, I'll use a glue syringe filled with liquid resin to literally blast the air from the undercut stake pockets. Thanks to Terry Wellman for this tip.)
As you can see in the photo below, the parts come out of the mold attached to a large piece of flash, or extra resin that is forced out of the mold. The pieces have to be carefully removed from the flashing with a razor blade or x-acto knife, and trimmed and sanded. This is neither difficult or time consuming but it is necessary to be careful when doing it. The car casting also has flash, though I'm usually able to break off most of it after releasing the model from the mold. A few quick swipes with the razor blade and it too looks pretty good.
About this time the first part of the two-part car mold was set up. I carefully removed the legos, then very gently peeled back the clay the car had been embedded in. So far so good, there wasn’t any undercutting of rubber from the first pour. And, while I cleaned the clay spots from the model with a Q-tip and mineral spirits, I did not disturb the model in the rubber. After flipping over the first part of the mold to face upwards, I rebuilt the lego dam and applied two coats of mold release. After the spirits had evaporated, I mixed and poured the second part of the mold.
This time, to save precious rubber, I included small chunks of an old mold I cut up after the first layer of rubber was poured. This is OK to use as filler and helps save a lot of money, as the rubber is really expensive. The freshly poured rubber bonds right to the cured rubber without any trouble. I figure I saved about $10.00 by doing this, and I still have more rubber chunks to use in my next mold.
Long story made a little shorter, the two part car mold was also successful, though I lost a tiny bit of detail on the B end of the car where there was some clay I missed cleaning out. Fortunately a couple of swipes with an x-acto knife on each casting and it is back in decent shape. The first couple of castings didn't come out too well, until I figured out how to tease the resin into the details with a toothpick, and put together the two parts of the mold without trapping any air bubbles. (Use some extra resin and let it ooze out the sides of the mold -- taking the air bubbles with it.)
Now I finally had enough good parts to put together at least one full car kit: The car itself, the sides, ends, and needle beams. While the pieces were still apart, I drilled out and tapped the coupler box and truck bolster holes on the bottom for 2-56 screws. Being the impatient person that I am, I took all the parts to the bathroom and washed them all with dish detergent (always a good idea when dealing with new resin castings), dried them off and got to work assembling my first kit. (Note that some commercial resin kits need to be cleaned in paint thinner because of the mold release compounds they use.)
The glue of choice for resin kits is CA, or Cyano-Acrylate cement - or just call it Krazy Glue like everyone else. Regular plastic welders like Tenax or MEK won't touch this stuff. I started with the sides, holding them in place while I applied CA on the back where the side boards meet the car decking. Capillary action drew the glue down into the stake pocket in most cases. I worked from the center out to the ends, cementing one stake at a time. Once both sides were attached I test-fit the end boards in place, sanded them down on the edges for a snug fitand CA-ed them in place. Finally the needle beams were located and glued down too.
You can see the assembled car in the photo to the left. At this point it is still lacking fine details like a brake staff and wheel as well as truss rods, couplers and trucks, but you cat get an idea of how the car is going to look already. To say that I'm tickled pink at this point is an understatement. In my own humble opinion, I think it looks just great! Better than I was expecting it to. Compare this photo to the line drawing at the top of the page and see how it looks.
Continuing in my impatient madness, the next step was to add the fine details and get the model painted, so I can finally see what it is really going to look like. I got some carpet thread from my sewing box to make the truss rods from, threaded on a Grandt Line turnbuckle casting, and CA-ed it in place. Four times and that was done. A 3/4" piece of .015 piano wire and I had my brake staff, a Tichy brake wheel went on to finish it. One more quick bath in soapy water and the model was towel-dried and left overnight.
The next morning I got out my airbrush and gave the model a coat of black paint with just a little drop of white in it, to make the paint seem a bit older and chalked. I used a water-based acrylic craft paint I picked up at the local craft store, it covered very well but it doesn't seem to have 'bitten' into the surface of the model. Hopefully a coat of Dullcoat after decaling will hold it together.
Once the paint dried, it was back over to the workbench for couplers and trucks. I used Accumate couplers, which I really like, but after seeing them on the small car they seem even more huge than usual. I'll have to investigate the new scale-size accumates, I guess. Trucks were Kadee 101 Archbar sprung trucks, with a red Kadee washer between the bolsters. Once of the track, coupler height was perfect (I assure you I lucked into it). (Future models will probably have Walthers Archbar delrin trucks with Reboxx code 88 metal wheelsets.)
After screwing these items on, the model is done except for decals. I hope they will be arriving soon, a friend with an ALPS printer is making me up some sets with the appropriate reporting marks and number series for these cars on my railroad. Here are a few photos of the completed Coal Gondola before decaling:
And here is a photo of a couple of these gons on the layout of my good friend Don Ball's Moraga Springs Northern:
I'm sorry to report that time has played a cruel joke on me. The model you see in the builders' photo at the top of the page was made about 2 years ago using a polyurethane casting resin from Thermocast International called RP-40. For over a year the castings were solid and held their shape, but recently I found a lot of the models I made with this resin have begun to curl and twist. Observe below the prototype gondola today, as well as another I had assembled but not painted:
As you can see, both carbodies now exhibit a swayback that John Allen would be proud of. Some of the castings made for the bobber caboose have curled even worse. Conditions where these models were stored were pretty optimal, in a cool NJ basement. I will probably continue to use the prototype car, weathering it to look really old, but the other model is just too far gone. I will not use RP-40 again for any type of casting and I can't recommend you do either.
I've recently started looking for an alternative and Alumilite's resins look promising. I will try it out, and with any luck the models I make will be straight and square for many years to come. There is a link at the bottom of the page to their site, please check it out.
I was able to get a supply of Alumilite's Slow-set polyurethane resin, and I've had mixed results with it. The slow-set resin requires you to heat the molds to approximately 140* to assist the resin in it's set-up (it can't generate enough heat on its own during the catalyzing stage). This is not so bad, but I have also had problems with getting good, bubble free castings with it. Alumilite's rep told me a few weeks ago the Slow-set isn't good for the type of work I'm doing (small, thin parts) and I will be getting some of their regular resin soon to try out. So the saga continues...
I hope that you have enjoyed reading about my experiences building, molding and casting this freight car. I hope you'll consider taking on a project like this for your own railroad, it is not very hard (though I'd be ready to make a couple of mistakes while learning how to do these things. If you are interested in learning more about moldmaking and casting models, I invite you to join the YahooGroups Casting List, a discussion group of artists, modelers and others who are interested in this fun and interesting activity. You may visit the home page for the casting list at:
The YahooGroups Casting list
Or, click on the picture below:
Use this link for good information on the casting process and quality resins and RTV's:
The Alumilite Home Page
This is a link to an excellent three-part tutorial on the casting process:
Dan Perez Studio's Moldmaking and Casting tutorial
Thanks for taking a few minutes and looking over my work. Please feel free to send me a message if you have any questions or comments.