This is the first phase in a long (ongoing) process that will allow me to produce many copies of a ceramic lamp prototype. I started the prototyping process thinking I was simply going to prove out the feasibility and structural integrity of a much larger model. But I liked the smaller version so much I decided to create a new project around it. What follows is an accounting of that process.
For the basic principles of slipcasting, please refer to the following Instructable
Step 1: A Model is Born
If fired high enough, porcelain can become a beautiful, translucent material. With this in mind I developed a lamp composed of a radial array of ceramic members. I used Rhino and Grasshopper so I could quickly visualize a series of iterations with different dimensions, member sizes and numbers and with a range of aperture sizes. I settled on a thin, squat version that reminded me of George Nelson's Bubble Lamps of the 1960s.
Throughout the modeling process I was conscious of the constraints of slipcasting or casting in general, especially the concept of undercuts. In order to successfully de-mold from a non-flexible mold there cannot be any undercuts or the cast piece will be locked into the mold. Each of the arrayed finsin my 3D model are beveled to allow for an easy release from the mold.
Step 2: 3D Print Test
The relative merits of 3D printing in general can certainly be debated, but one of the biggest advantages of digital fabrication is in proving out conceptual designs. I plan to produce a lamp that is roughly six time the size of this this teeny, 3D-printed version, but it allowed me to get a sense of the proportions and the optical and material qualities of the final product, which is an amazing advantage. And I liked the small version so much I decided to make a huge batch of them, which will become an installation at Autodesk's Pier 9. Look for a future Instructable that outlines that process.
Step 3: Counterpositives
It's a fancy word. Basically it's a mold for making molds.
In the past I have directly milled blocks of plaster to produce slipcast molds, which is an incredibly messy and laborious process that leaves you with only one mold. This is unfortunate given that plaster molds can only produce 50-75 casts before they start losing a lot of definition. By creating a counterpositive you can easily produce many copies of your mold. For the larger version of this lamp I will mill a block of nylon on a 5-axis CNC router, but for the smaller version I 3D-printed counterpositives for each of the two mold halves.
Step 4: Mold Formwork
To house the counterpositives during the plaster casting process I used the lasercutter to make a tabbed box that I could easily disassemble and reassemble for future casts. Tape or rubber bands are sufficient to support the weight of a small cast like this one.
Step 5: Prep the Counterpositives
There are many theories and products to help plaster flow into all the nooks and crannies of your counterpositive. Over the course of my trial and error process I used three: a linseed-oil-based mold soap, talc, and finally a 50-50 solution of Windex and water, which worked the best.
The basic issue is to try to defeat water tension, which causes little pockets of air to develop in the smallest corners of the mold. Soap and Windex (both called surfactants) have the effect of reducing water tension. Talc is sometimes used to create a microscopic layer of air between the counterpositive and the plaster, acting as a sort of air release mechanism.
Step 6: Mastering Plaster - Adjust the Volume!
Calculate your plaster volume! There are several volume calculators on the web that simplify this process for you. It is also possible to get this information directly from your modeling software.
Step 7: Mastering Plaster - The Mix
Plaster is Calcium Sulphate. When mixed with water and agitated it quickly gets very hot, expels the water and hardens into form. While it is possible to get a good batch of plaster through several techniques, the only truly reliable and consistent way to produce excellent, bubble free plaster is to use a scale.
For slipcasting I use potter's plaster #1. The correct ratio for this type is 10:7. That means you multiply the weight of your plaster by .7. For example, 5 pounds of plaster needs 3.5 pounds of water.
Slowly sift handfuls of plaster into the water, letting it settle gently to the bottom. DO NOT MIX the plaster yet. Patience is its own reward in this step
By the time you've transferred all of the plaster into the water you should see it starting to peek out of the water a bit. Its fine to push that last bit into the water. But DON'T AGITATE IT YET!
In this state the unmixed plaster could sit for hours and not harden. Slowly dip your hand into the plaster to feel for any thick clumps. Break them up and start to slowly stir the batch. Take care not to introduce any air into the mix. Air bubbles will be easily trapped in the mix as it cures, which can be very problematic in the casting process, as we will see later in a different way.
Step 8: Mastering Plaster - Pour it On
When you're confident that there are no lumps in the plaster, begin to stir the plaster by hand very quickly, or at least as quickly as you can without introducing a bunch of air bubbles. When its consistency starts to approach that of buttermilk, POUR IT!
If you start to feel any warmth at all in the plaster mix that means it is very close to kicking, which means it will be too hard to pour in less that a minute. This is not just an issue of pouring; if you pour after the plaster is warm and the viscosity has increased there is a high likelihood that it will not reach the more delicate parts of the form. It also means you are less able to remove air bubbles from the batch.
I would rather pour an almost cold batch of plaster and wait a few extra minutes for it to kick in exchange for having confidence that the cast will be clean free of pockets and high fidelity.
Step 9: Sometimes You Fail, and That's Part of the Deal
By now you may notice my slight obsession with air bubbles. They are my sworn enemy. At least in the world of plaster casting.
In the first two images you can see the air pockets in the final product.
It took four tries with different techniques to produce an imperfect but passable plaster mold. I may make a design change in the scaled-up version that minimizes this issue.
Step 10: This is Only a Test
Stay tuned for the follow-up Instructable to see what these molds make!