This is the continuation of an earlier Instructable that covered the creation of 3D-printed counterpositives and plaster molds for a ceramic chandelier module. It might be helpful to review that process if you want to understand how it relates to this one.
This process will result in a 256-piece ceramic chandelier. The only way to produce that many pieces in a relatively short period of time is to create multiple molds. Counterpositives theoretically allow for the production of an infinite number of molds. I produced 14 molds, but only eight were good enough to produce high-quality positives.
Armed with eight molds I was able to produce 256 positives in 32 rounds. Each round takes roughly 1.5 - 2 hours to complete, from pouring the slip to pulling out the final piece. 60 hours is a lot of time in the casting room, but it certainly beats 500+ hours!
What follows are the finer points of that process.
Step 1: Pour Slip, Wait
Slip. What the hell is it?
Well, there are a couple of kinds of slip. If you've ever taken a ceramics class you've undoubtedly used slip to glue pieces of clay together. This slip is typically made by adding water to clay until it liquefies. The slip used for slipcasting is somewhat different in that a deflocculant (typically a mixture of Darvan and bentonite) is added to ensure maximum flowability. I purchased pre-made slip in gallon containers from my local clay retail shop, but you can also make your own.
How the hell does this process work?
Plaster and clay resist one another. In a nutshell, that is the key to understanding this process. A dry plaster mold will absorb the water in slip at a fairly uniform rate. You fill a plaster mold with slip, let it sit for 15-20 minutes, pour off the excess, and what remains is a .25" thick shell of clay in the form of your mold. Many, many objects are made using this process, including cups, plates and even toilets.
Since I am interested in producing a translucent ceramic object, I want my final pieces to be as thin as possible while retaining enough structure to stand up under their own self-weight. Through a series of tests I was able to shorten the mold-time to four minutes. This produces a shell slightly thicker than 1/16", which is perfect for producing a warm, diffuse light.
Step 2: Demold!
Finding the right time to demold your pieces will take a bit of trial and error. If your mold time is near the typical 15-20 minute mark, you should have success demolding around the 80-90 minute mark. But this is dependent on the density of plaster in your mold, how dry your mold was when you poured the slip, relative humidity, etc.
Because I was casting small pieces with very thin shells I found the perfect time to demold my first cast of the day was 50 minutes after dumping the excess slip. Any shorter than that and the clay was a little to wet to demold easily, any longer than that and the piece would be overly dry and potentially even cracked a bit due to shrinkage.
Any additional casts during the course of one day tended to need more like 60-70 minutes due to the molds being slightly wet.
This is art, not science. Generally it's a good idea to cut the clay away from your pour hole. Anything you can do to reduce friction during the demolding process will reduce distortion of the piece and prevent a lot of heartache. In the case of my molds, I could tell they were ready to demold if there was a hairline gap between the plaster pourhole and the adjacent clay (aka plug.)
I used a clay tool to push the clay away from the pour hole and then to clean up the upper edge of the clay, using the mold as a guide.
If all goes well, your mold pieces will separate easily and removing the clay piece will be a breeze. If not, you can try gently massaging the piece using equal pressure on both the interior and exterioir faces of the piece. I've also had some success using compressed air to expel difficult pieces.
With luck and persistence I produced 270+ pieces using this process.
Step 3: Into the Fire...
The final, and least directly controllable aspect of this process is firing the pieces in a kiln. It takes a bit of courage to close the lid on a kiln full of and allow them to be subjected to 2000+ degrees for 12-16 hours. This was my first experience using an electric (as opposed to gas) kiln, and the results were extremely uniform and predictable. After the first test batch fired perfectly I felt totally confident in letting the machine just do its job. Autodesk's amazing shop at Pier 9 in San Francisco doesn't have a kiln (yet. I'm working on it.) so I brought the unfired pieces to my friend's studio (Happy Potters) in Berkeley for the final phase of their fragile life.
I was able to fit around 85 pieces in the kiln for each round of firing. Four rounds later, it is impossible to see any difference between pieces fired in different loads.