Slip Casting 3D Printed Vessels

Slip casting is a field of ceramics which utilizes plaster molds to make ceramic forms. The molds are made from existing objects, and then are reusable for the fabrication of many ceramic pieces. The molds are filled entirely with ceramic slip, a processed liquid mixture of clay particles in water. The water in the slip is absorbed by the plaster, causing the clay particles to condense into solid clay, but only around the walls of the mold. The longer the mold sits, the more water is absorbed, and the thicker the walls of the ceramic piece will be. When the desired thickness is reached, the remaining slip is poured out and the mold is removed. The ceramic positive can then be further worked, or left to dry and then kiln-fired.

For this project, I utilized Rhino 3D and a desktop 3D printer to fabricate two custom ceramic vessels for the churning of butter. The objects were made digitally, cast in plaster, and then ceramic casts were taken from the mold. I faced a number of challenges along the way—some of which were shortcomings of technology, others were a result of my own poor judgement....

This project is part of a larger exhibition entitled "Butter Boy & Bros", facilitated by MASS Gallery in Austin, TX.

Now let's dive into the details!

The initial concept for this project was deeply underway before I even sat down to model these objects. In fact, I had quite unwisely made a component of the project prior to the ceramic pieces, which the dimensions of the vessels were then contingent on. This object was a fist-shaped butter churn "dasher," which I CNC milled from a scan of my own fist and placed on a pole. On paper, this was the most challenging element of the project, so I elected to tackle it first and attempt to fabricate the vessels later in response to the product. This was overall unwise, seeing as I should have either fabricated them simultaneously or fabricated the vessels first and customized the dasher to their dimensions. The reason for this is I had complete control of the scale of the product when working with wood, but when the end result of the vessels was a ceramic object that would take on much of its own character through the process, there were many more variables to juggle. We'll dive into that later...

The basic composition of the vessels was quite a simple process. In Rhino, I drew out a vector line of the silhouette of the vessel, accounting for the shape of the dasher by importing the original model. Next I selected the Revolve function under the Surface tab, which allows me to make a solid form by rotating this line 360 degrees.

Now that I had a full object made, I could compare it with the original model of the fist and analyze the outcome of the dimensions. The overhead view shown above allowed me to estimate how the diameter of the vessel would translate when made from ceramic. There are some variables when working with ceramics that needed to be accounted for. One of which is the fact that the final ceramic form will have a thickness, but this model is an infinitely-thin surface. I needed to account for the presumed thickness of the walls—at lease a quarter of an inch. I also need to account for the shrinkage rate of the clay, which for most low-fire slips could be around 5-10%, but you will have to check specifically for each individual clay in order to ensure proper results if the scale is important.

I evaluated these factors and scaled the churn to a proper size that would allow some wiggle room for these variables. The dasher needed to fit through the top hole of the vessel, so I analyzed the walls and shrinkage to determine a proper diameter.

For this print job I would be utilizing a Lulzbot Taz, a midsize 3D printer operating through the PC application Cura. Due to the limited printing dimensions of the Taz, I'd have to print each vessel in two pieces and then assemble and polish them later. To Achieve this, I split each model into two segments in Rhino. The larger vessel had a diameter and height of almost exactly the maximum of the printer; about 10.5 x 9.5 inches.

In order to reduce the cost of this print, seeing as I only needed the exterior of the object, I selected the Spiralize function under the "Expert Configuration" menu. This would print only a single layer, only a thin shell, by printing in repeated spirals upward. Most 3D printed objects are constructed with a matrix of material on the interior to support the flexible plastic. Though the cost of my print was greatly reduced, the end result would be very fragile. There would be a lot of polishing and layering of additional materials, such as epoxy, putties, and fillers in order to make this object durable enough for casting.

In order to merge the two halves of my objects, I would be using a number of materials, but the most important would be Loctite plastic epoxy and Bondo all-purpose putty. The epoxy would weld the pieces together, while the putty would reinforce these joints. I would also use the putty around the bases of the vessels to add both strength and form. A downside (and the inherent nature) of the spiral printing is that it has no additional support, so objects can sometime ooze and sag mid-print. This occurred with the bottom curves of both objects, so I applied the putty around the interior and exterior of the bottom curve with a pallet knife and silicon spatula. A side note; silicon kitchen spatulas are perhaps the most useful tools I use around the studio. Their abilities, when used responsibly, are truly incredible.

After the Bondo putty fully dried in 15-30 minutes, I began sanding the course surface with 60 and then 120 grit sandpaper. In most cases, one would want to continue upping the sandpaper grit to achieve a super fine surface, but I would be applying one more substance to the outside, so perfect smoothness wasn't yet needed.

Next I lightly sanded a few areas on the objects where the printer had left marks or trails. This happens often in the spiral printing—the location where the machine shifts to the next layer often is left with a drag mark. This plastic shell is very fragile, so I just barely rubbed the surface until the severity of the imperfection was less pronounced.

For connecting the two halves, I mixed my two-part epoxy and applied a small amount to the fusing edges. then I adhered some tape strips to the top half and lowered it onto the bottom segment. I alined each of the edges and pressed on the tape to keep it in place. The two edges didn't line up perfectly—likely a result of some imperfections in the printing—so I would be using the Bondo again to reinforce the seam and to smooth out the uneven joint.

As a last stage before casting my molds, I decided to attempt to coat the objects in Durham's Water Putty. I was hesitant because I was unsure if the putty would bond well to the plastic. Durham's is a powdered wood filler that is made by combination with water. I hoped that coating the object with reinforce the surface—because the printed plastic was still slightly flexible—as well as remove the ridges of the printing and replace it with a unique painterly texture.

I applied the Durham's and let the objects sit over night. But as I had feared, the putty cracked off in large chunks, leaving a vastly uneven surface. My initial response was aggravation, but I quickly realized that it was actually a beautiful instance of chance in the artistic process. The 3D printed surface wouldn't allow me to lie about its nature, it needed to be revealed. So I decided to cast them as-is.

Plaster is a challenge—especially in a large quantity. I was going to need a lot of plaster for these molds, and I would need to pour it in many batches, which left a lot of room for error and lack of consistency between different mixtures. This was by far the most labor intensive stage of the process.

In order to cast these molds, I'd be using clay, foam, and plywood to position the objects and pour the plaster for optimal clip-casting. I could've done a lot to reduce the quantity of plaster by lobbing-off unneeded sections, but I choose to forgo these steps in order to get the molds made quicker. I would regret these decisions when it was time to lift and pour the molds in the slip casting process.

I situated the objects to allow for at least 1.5 inch walls around, set up the plywood from for the exterior of the mold, and filled and potential holes with clay around the seems and corners. The plywood forms are called cottle boards—they proved an adjustable exterior box for the plaster to set in, and are secured together with wood clamps. In order to provide a solid base to the mold, I used a sheet of plywood with the silhouette of the object cut out to allow me to avoid using more clay than I wanted. Generally in making plaster molds, half of the object would be sunk into a block of clay to act as the other half of the mold. I elected to use the plywood as a solid, flat base on which to cast the first half of the mold, which I screwed in to the exterior frame for extra reinforcement. Next I filled all seems around the edged and corners with clay to avoid seepage of plaster. When the whole setup was secured, the last step was to lube-up the frame and object with a thin layer of Vaseline to allow a clean and smooth removal of the object from the plaster.

Now it was time to pour. For info on the plaster-mixing process, I recommend referring to a more specific source. Its a delicate process that requires a much more in-depth explanation. But the greater premise is to find the right balance of plaster and water. More plaster is better, if the mix is too wet, it will stay moist and powdery. I needed a very dry plaster in order to properly cast my ceramic pieces.

After the first side of the mold was poured, I flipped over the plaster block—without removing the interior object—removed the plywood silhouette board. Next I repeated the same process of setting up the frame, filling cracks with clay, and applying Vaseline. Then I poured again. The plaster of the second half would meld perfectly to the other side of the mold, and then cleanly come apart thanks to the lubricant.

Each half of the mold required about 3 batches of plaster. The total quantity weighed close to 200 pounds. I now would have to fill the mold with heavy slip, and then lift and pour it out. This would be a whole new challenge.

After the plaster had set I still had some polishing to do before they were ready to cast. Firstly, the Durham's and the clay had lodged in the nooks and crannies of the plaster, so I had to tediously pick out the remaining chunks. The plaster was still soft, so I used a wooden clay tool and a compressed air sprayer to gently dislodge the fragments.

Next, I had to do some manual carving to make sure the mold didn't have any undercuts which would keep me from cleanly removing the object after it was cast. Because the printed object were made from flexible plastic, the bottoms bulged in slightly. I hoped the Durham's would prevent this, but there was still a slight give. Using a wide scraping clay tool, I carved out a flat bottom on each side of the mold and then put them together to make sure it was an even plane.

Because the molds were so dense, and because I was on a time crunch, I needed to speed up the drying time of the plaster. For slip casting, the plaster needs to be completely dry so that the water in the slip can absorb properly and solidify the form. I popped each of the mold pieces in a kiln for almost 48 hours, rotating them a few times a day for even drying. The kiln ran at around 175 degrees Fahrenheit—I probably could have heated it a little higher, but I was being cautious. If the molds surpassed boiling point (212 degrees Fahrenheit) the water would boil out of the plaster and it would crumble....

After they came out of the kiln and cooled to room temperature, they were ready to pour!

The casting process began by using ratchet straps to tightly secure the two halves of the mold together. The pressure of the slip would be immense when the mold was full, so I had to tightly secure them together. Now it was time to fill em up! I poured gallons of terracotta slip into each of the molds—the larger of which required 6 gallons to fill. The slip than sat for almost an hour and a half as it solidified around the surface of the plaster. As it sets, the level of the liquid lowers as the water is absorbed.The thickness can be estimated by observing the top lip of the mold. The longer it sits, the thicker the object will be. When the walls reached about 3/8 of an inch thick, I poured out the remaining slip. Typically this should be done pretty slowly and in a circular motion to ensure no uneven accumulations on the inside, but this was a major challenge with such cumbersome molds. The larger mold required the use of a mechanical gantry lift to get it off the ground.

Next I left them to dry for about 24 hours and then slowly and gently wedged open the molds with wooden shims. Generally, smaller slip casted pieces would only need to sit for a few hours or over night to dry properly for removal, but these were very slow in the drying process due to their size and the bulkiness of the molds. The following afternoon, I carefully popped the pieces out of the plaster and got to work polishing up the seams and any imperfections. The top rims were pretty rough, so I cleaned them up and added some details. Then I left them to dry for roughly 36 hours until they were ready to kiln fire.

Due to my tight schedule, I needed to transport the vessels to a friend's ceramic studio to use his kiln. The one I have usual access to is highly insulated, so a full kiln cycle takes about 3 days. At this point I was just 3 days away from the showing of these pieces! The smaller kiln I'd be using would fire them overnight, taking roughly 24 hours to heat and cool properly.

For transport I had to tuck the pieces in for a safe ride across town, securing them with blankets and seat belts. Though hard and dry, the greenware was still quite fragile to cracking and breaking. Unfortunately the top rim of the larger vessel was quite fragile and it was bumped during travel. A few chunks fell off, which I would fire and glaze separately, and then adhere with an epoxy later.

As a final measure before firing I picked out some little chunks of plaster lodged in the rough surface of the vessels. I used pointed metal clay tool to lightly scrape them out without damaging the surface of the objects. Since we were firing them so quickly, any leftover plaster could be a potential hazard. Though unlikely, there was a possibility that they would pop or explode in the kiln, breaking the whole piece.

For the bisque firing— the initial hardening of the clay—the kiln reached cone 04 which is 1940 degrees Fahrenheit.

The following day I was back to do the glazing. Because the pieces were made of red clay, I would need to coat many layers in order to get a solid color of glaze. I coated the interior with a clear glossy glaze by pouring in a slightly watered-down quantity of glaze and swirling it around to coat the entire inside. As a final measure, I used a dry rough sponge to sand off areas of glaze along the textured edges around the vessel to capture a faux antique feel.

For the glaze firing, the kiln reached cone 06 which is 1830 degrees Fahrenheit. The cone scale is slightly confusing, seeing as 06 is cooler than 04... Check out a cone chart to see how this works out.

The following evening, it was time to crack open the kiln and see the final results!

As a finishing touch, I used some wire to attach some mahogany handles that I had milled previously.

The next night was the big event. I would be attempting to churn butter, start to finish, and then serve it to the exhibition attendants. I filled up the churn from the milk jug with 2 gallons of heavy cream, and began to slowly agitate the contents with the dasher. After 10 minutes or so, it began to get quite frothy and was sloshing up to the brim of the churn. It began to splatter out around the base, creating an incredible halo of striking white speckles. Then, perhaps halfway through the churning process, I accidentally struck the bottom of the churn with more force than I intended. I heard a muffled crunch and then watched as fragments of ceramic drifted out across the floor on a sea of pure white liquid. The entire contents of the churn poured out across the gallery. At first I was shocked and aggravated, but this unexpected result was far more stunning than what I had initially planned. This surprising accident felt incredibly right and nearly intentionally. The physics of the performance and my props took their own course, deciding the conclusion of the event for me. And now we were left with this stunning centerpiece of the show. This experience inspired a new degree of contemplation on performance and the importance of chance in the creative process. Though it was fundamentally a failure of my creation, the object dictated the conclusion of the exhibition for me, and the end result was for more profound than anything I could've prepared for. I'll be revisiting this project again in the very near future, with a stronger ceramic churn and more of an open mind to possibilities of chance.

Check out the full documentation of the show here.