Making Plaster Casting Molds

I hadn't made plaster molds in over 30 years so it was like learning new tricks all over again.  I bought some liquid latex called "Mold Builder" (32 oz, $22) from Castin' Craft and a box of 24 1" chip brushes through Amazon and borrowed some gauze pads from our 1st aid box.  I used one of the extra walls I had gotten ten years ago to practice on in case I screwed something up.  My first attempt was usable but with only five coats of latex it came out too flimsy.  I did confirm that I could spread the latex directly on the bare plaster without creating problems.

Here's my technique.  I laid a sheet of parchment paper on my workbench.  My wife uses this for baking and it prevents food from sticking to cookie sheets.  Waxed paper works well too.  I use inexpensive "chip" brushes because the latex builds up in them and they only last for two or three coats.  I trim the bristles with scissors so I'll get a nice even coat on the mold.  To start I spread a thin layer of latex on the plaster wall section.  This first layer is critical and will define the quality of your castings.  Make sure the latex is spread evenly in a thin layer and then go back with your brush and gently work it into all the joints, cracks and crevices.  I use the brush this second time to make sure the coat is spread evenly over the entire surface and to remove any bubbles that may have formed.  If the first coat is applied too heavily it might hide bubbles so it's more important to ensure full and even coverage.  The first three or four coats should be kept thin and bubble free.   With each layer I also build up a 1/2" wide flange around the outside edge.  I lay on four coats before adding a thicker layer with the gauze.  The gauze pads I use come folded over in three sections and when I unfold that it's still three layers thick so I use it like that without separating it any more.  I lay the gauze on the casting and am careful not to overlap it to prevent unnecessary lumps in the mold.  In the picture below you can see where I overlapped the gauze and then had to cut recesses in the support frame to account for the unwanted lump.  The gauze gets held down with a heavy layer of latex that's been worked into the fabric with the brush.   I then add three more layers of latex over that for a total of eight.  In between each coat I let it dry until all the latex has turned a light yellow color and no longer shows any white which would indicate it hasn't fully cured yet.

When I was cleaning up the original wall castings I realized that the walls were not exactly straight and tended to bow out in the middle a bit.  I suspect the molds weren't properly supported when the plaster was poured.  This caused me to think long and hard about a better way to go about this.  Back when I did this 30 years ago I was creating molds of rocks to be used on mountain sides and it didn't really matter if they were cast to lay flat.  Making walls is a different story.  They have to be straight and flat or they don't look realistic.  I decided to make a wood frame to hold everything flat and square during a pour.  I built the frames on a backing panel of 1/2" plywood.  Each wall section has windows or doors and cap stones running along the top edge.  They also have stone buttresses and heavy stone frames around the windows.This made making the support frame complicated because all these details are at different levels.  I glued wood shims of various thicknesses in places where they were needed and the end result provides a good flat support for the mold (see picture above).  With that done I decided to go ahead and make molds for six other wall sections just so I'd have them if I ever needed them.  It's a pretty quick process to make the molds so why not?

OK, here's a good reason why not.  Of the six extra molds I made only three came out perfect on the first try.  I suspect I was rushing things.  Because of all the detail in the rock walls the molds were very difficult to release from the original castings and parts of the first layer separated and smeared, ruining the details of the stone wall.  Bummer.  For my second effort I really took my time and allowed plenty of drying time between coats.  To get eight coats of latex takes at least two days of work so it's worth while not to rush the process.

Once I had a good mold cured and released from it's original casting it had quite a bit of flashing that needed to be trimmed off.  For the outside flanges I trimmed them with scissors.  For any flashing inside the mold I carefully trimmed it off using wire cutters that have a good sharp cutting edge.  I stretch out the flashing while cutting to help prevent cutting too deeply into the mold.

For a really excellent explanation of mixing and pouring plaster check out New England Brownstone's web site (click here).  They also have some great painting techniques for plaster cast stone walls (click here).

Another problem I ran into showed me the necessity for applying thin, even first coats.  When I started pouring plaster it took several tries before I found a good consistency.  My first pour was too thin and the casting came out pock marked with bubbles and grainy looking.  My second try used a thicker mix, like a thin milkshake, and the casting came out perfect.  For these first two tries I used a smaller wall section but now I felt ready to take on the larger roundhouse side walls.  I used the same thicker mix but when I removed the casting from the mold I found it was full of air pockets again.  ???  A closer look at the mold revealed small air bubbles had formed in the first layers of latex and I suspect this created air pockets during the pour.  I realized these two molds had been the first ones I made and I wasn't careful enough to get thin, bubble free coats.  I had to throw out my two largest molds and re-do them, this time with thinner first coats.

Another problem that was revealed after pouring the larger wall sections was an uneven wall surface.  I thought I had created enough support for the molds but the areas around the windows looked wavy.  I needed to add about an extra 1/8" of support in this area.  On some of my molds the outside edge is built up too much in some areas and the mold won't sit squarely in the frame.  The inside edges of the frame need to be rounded off to allow the mold to sit down flat on the bottom.  The best way to verify the fit of the mold in the frame is to press your finger around the inside edge and see where the mold doesn't sit flat.  Re-work the frame to allow the mold to sit firmly flat on the bottom of the frame.  This will ensure a good square and flat casting.

I now use a spray-on mold release and this makes the whole casting operation much easier with less chance of damaging the mold when it's removed.  I also found that when I pour the wet plaster into the mold I should only fill it about 1/3 of the way.  Then I take a ball peen hammer and tap the wood frame from underneath and on the sides to drive out air bubbles.  When this is done I'll pour in the rest of the plaster and repeat the tapping.  This will cure most of the problems with air bubbles.

I also found out that you can buy high quality latex for making molds.  Make a water tight 5 sided box (no lid) and place your wall section in the bottom.  There should be about a 1/2" gap on all sides to form a thick side wall on the mold.  Then pour in the Latex to fill the box to the top.  I think tapping the box with a hammer or using a mechanical shaker will drive out any air bubbles from the mold.  I haven't tried this yet but it's the way all professional Latex molds are made.  More expensive but better molds.

Building a Doodlebug

On the layout my main goal was point-to-point operation but I added a return loop on the lower level between the stamp mill and the roundhouse.  The purpose of this was to run a railbus on autopilot to create interference for yard and switching operations.  I found a "Doodlebug" kit on eBay  that fit the era, 1920's, but would need some modifications to suit my eye.  The running gear for this kit was supposed to be an HO scale Bachmann trolley car but I felt it looked kinda wimpy and the overall profile was too high above the tracks.

The first thing I did was to carve out the frame and floor to allow use of an HO SD40 diesel switcher. The power trucks from this switcher looked much better than the puny wheels of the trolley and the drive motor from the switcher was already set up for DCC.  The wheelbase measurement was spot on too.  I used a Dremel to carve out the base of the Doodlebug and did it in stages so I wouldn't go too far too fast.  My first attempt looked good but the body sat too high.  I lowered it an additional 12 scale inches and then added running boards to visually lower it even more.

The cow catchers that came with the model was small and clunky so I made my own.  I used scale 1" square and L plastic shapes and formed them up using tape to hold the rounded shape while I glued it together.  I added grab irons next to the doors made from brass rod.  I used thin clear plastic sheets made for my printer and fogged with flat clear spray paint for the windows.  I plan to use most of the interior for electronics so fogged window panes allowed light to pass but didn't provide a clear view of the interior.

I couldn't find suitable headlights so I made my own using plastic straws and 0.010" styrene shapes.  I placed LEDs inside the new housings and wired them up to the diesel switchers control board.  I also added a brass bell above the windows at one end of the Doodlebug.

I painted the body Pennsylvania RR maroon with black trim to give it a sharp, early 1900's look.  The prototype for this model was an awkward looking homemade affair with a center control cab and a small steam engine for propulsion.  My modified version has the cab at one end sharing space with the baggage compartment and using gas/electric propulsion.  I installed a bulkhead at midpoint to separate the driver from the passenger compartment and to add some needed bracing to the car body.

This came out pretty sweet.  The SD40 mechanism runs quiet and smooth and can do creepy-crawly very well.  I'm pleased with the final paint colors and I think it gives it the right look for the 1920's.  The cowcatchers also give it the right look, at least to my eye.  I'm sure someone out there will point out that this combination isn't correct but I don't even care.  It gives me what I wanted for my layout and it runs great.  I still need to convert it over to deadrail operation and add sound but for now it's complete.  Fun project.

Turntable Pt 4 - Installation and Final Adjustments

With all the bugs worked out of the program and the model completed it was time to get the thing off the bench and onto the layout.  This should be a simple matter of dropping the module into place and plugging in the power.  It turned out to be sooooo much more.

The physical positioning was the first thing.  I had recently added the final top layer of Homasote for the yard section of the layout.  When I went to drop in the turntable module I found two problems.  The biggest was that the framework for the layout was in the way and had to be modified.  The second was that I wanted to rotate the module to fit the gear hanging under it for the easiest access for maintenance.  When I rotated the module it wouldn't fit in the hole.  I ended up running the base through the table saw again to square it up.  I also had to clean up the edges of the hole where the Homasote didn't match well.

With that completed I got the module in place and the turntable bridge reassembled.  I now had to deal with several inaccuracies I had ignored during construction.  One end of the bridge was 1/16" higher than the other.  I thought this might be caused by the dolly truck installation where I had added some shims to get the rail up to the right height.  I ran the wood buck that the bridge is built around through the table saw and squared that up as accurately as I could.  The I used my digital calipers to measure the thickness of the bridge ends and found a 0.030" difference there.  It turned out that the dolly trucks had a large variation in the thickness of the castings.  Rather than trying to correct this I used a hand file on my shims to make up for the difference.  The bridge now sits flat and is square with the world.

My next problem was that the pit wall isn't perfectly round in one spot and one end of the bridge is too close when driving past this area.  I want to have 1/16" clearance all around.  I took another look at the bridge and found that the wood framework extends just a hair further on one end.  I filed and sanded that and now have those problems taken care of.

My next issue was getting the bridge to go through a full rotation smoothly.  I wanted to have as little weight as possible on the dolly trucks to prevent unnecessary loads on the stepper motor due to small variations in the ring track height.  I added a second bearing at the base of the turntable shaft and this solved the problem by removing any side-to-side movement.

Now the turntable operates smoothly and the module can be taken out and replaced easily.  My next problem was with repeat-ability when going from one track to another.  I found there was an error of about 0.030" in alignment when reversing direction of rotation.  The drive belt had too much slop in it but if I tightened the idler too much it would load down the stepper motor.  I found that I could add two more idler pulleys between the stepper motor and the turntable shaft pulley.  These can be adjusted to add just enough tension to remove any slop in the belt.

To finish off the turntable I painted the pit walls a concrete color and used Bragdon weathering powders to add streaks and stains to the concrete.  In the floor of the pit I added a layer of dirt and then a layer of cinders that will also be used on the rest of the yard area.  The dirt I got from my sister's back yard in Pennsylvania.  I couldn't use local dirt around my home because I live in Florida and what they call dirt is mostly sand.  The cinders I got from the Arizona Rock and Mineral Company (click here).  These guys have a great selection of scenery material.  I used HO and O scale cinders because I wanted both fines and granular.  To finish it off I added small clumps of weeds here and there.

With these these things taken care of the operation of the turntable is reliable, repeatable and really enjoyable.