Gas station without pumps

I’m still catching up on my posts about my pottery. Here are more pots that were made and glazed in Fall 2024. All of them were made with Bravo Buff clay, as I did not use any other clay until the Fall 2025 class.

There were 4 pots glazed on 30 Oct 2024, two of which (the small pitcher and the striped bowl) was posted about last time. Here are the other two:

A pitcher

A black-and-white bowl

There were two pots glazed on 20 Nov 2024:

A green tea cup

This may be my favorite of that semester, though I’m not usually fond of green glazes.

A blue-and-white bowl

It will probably be a while before I catch up to my current pots, as my camera failed and I don’t like it well enough to get it repaired—I’ll have to make up my mind about what new camera to get before I can finish photographing the 2025 pots. I do have a few more photos I can still blog about from the first half of the Fall 2025 semester—I’ll try to get to those soon.

My Halloween costume for 2025 was as Jack Frost. I had originally bought the costume (suit, tie, and socks) for a role as Jack Frost in Electra, Daughter of Claus, a play written and planned for the 2024 Christmas season by the Cabrillo Performing Arts Collective, but that show was cancelled, because the organizers did not understand all the limitations on Cabrillo clubs and the show could not be scheduled at a time when the club could legally do it.

Because I had the costume, I decided to design makeup for it for the “fantasy” makeup assignment in the Makeup and Masks class that I’m taking at Cabrillo this Fall. I ended up doing the makeup three times: once for the class, once for a party with former colleagues of my wife, and once for giving out candy on Halloween.

The in-class assignment used clown white, Ben Nye MagiCake Aqua Paints, and Ben Nye ivory hair color:

For my second attempt (for the party), I used my QIDI Plus 4 3D-printer to make half icicles for my forehead out of transparent TPU, used mehron white hair color, and some water-based cream makeup from Amazon. I also used blue and silver acrylic paint to decorate an old pair of shoes (to which I added shiny silver laces):

For the third attempt, I printed very thin 6-pointed stars out of TPU (0.2mm at the edges, thickening to 1mm in the center), rather than using a paper pattern and painting with makeup as I and my wife had done for the first two iterations. I used the same hair color and makeup as for the second iteration, but added Mofajang white hair wax to the beard to make it whiter and to shape it into a pointy beard.

The water-based cream makeup does not need setting powder, but it does crack after a few hours. If I do the Jack Frost makeup again, I’ll probably 3D print the eyebrows out of TPU (thin, like the stars). I might also try painting the TPU with acrylic (rather than using makeup to color them).

I’ll probably make some TPU prosthetics for my final project for the Makeup and Masks class, but I’m not sure what—the assignment is to build a “monster” character that includes almost all the techniques from the course, including a latex mask built on a clay sculpture added to our plaster life mask.

For the Cabrillo College Fall play (Metamorphoses) I was asked to 3D-print the Corinthian capitals for the tops of two columns on the set. The capitals needed to be fairly large (sitting on top of a 9.5″ diameter column), and even my new QIDI Plus-4 printer is not big enough to print them in one piece. I was given a scanned model of the desired capital, but I had to do a fair amount of work to make it printable. First, I had to clean up the model with Formware’s free online STL-repair tool.

I planned to print the capital upside down, so that it got narrower rather than wider on the later layers, but the ring at the base of capital would have needed a lot of support, so I modeled a new ring with a sloping, rather than horizontal surface, and superimposed the ring using OpenSCAD. This eliminated the need for almost all the support.

I scaled the X and Y dimensions to make the capital have the desired size for the connection to the column, but I scaled the Z dimension separately to make it fit in the printer. The printer ads claim it has 280mm available in the Z direction, though QIDI Studio insists that it be strictly less than 280mm. The slicer accepted 279.9mm, but the top couple of layers had the nozzle dragging on the print—I think my printer only goes up to about 279mm, so I rescaled the second capital I printed to 278mm to be safe. This meant that the two sets of pieces were not interchangeable, but the pieces for the first capital all had a marred top surface from the nozzle dragging, so it was easy to tell them apart. The marring did not matter as it was at too small a scale to be visible to the audience, and would either be covered by paint or not visible from the house.

Because the model was too big for the printer bed, I had to cut it in quarters. I was worried that the original scanned model was not exactly the same on all 4 sides, so slicing it would have required keeping track of which piece went where. As I was not going to be assembling the pieces, I thought that this would be rather risky, so I symmetrized the model by superimposing 4 copies rotated 0°, 90°, 180°, and 270° about the Z-axis. Because the original model was not exactly centered,I had to tweak the center axis a little bit before doing the superposition. To make the tweaking easier, I colored each of the copies differently in OpenSCAD, so I could see where one stood out more than others, and either tilted or translated the original model to compensate. I also trimmed a tiny amount off the top and bottom, so that those faces were level .

Once I had the model as perfectly centered as I could get it, cutting it down to just one quadrant was trivial in OpenSCAD. I hollowed out a cylinder almost the full length of the capital, to reduce the amount of filament needed. I also hollowed out a triangular cross-section ring from the most solid part of the model, so that there would be some solid walls, rather than just loose infill in the interior. Because of the amount of filament and the time it would take, I reduced the gyroid infill from my usual default of 15% down to 8%, and I wanted a little extra stiffness in the vertical direction (probably unneeded, but it was fun to add the extra feature).

The quarter-capitals took almost 10 hours each to print and used 450g of PLA, so the pair of them took 80 hours to print (well a little more, because of time needed to change spools or clear filament tangles) and used 3.6kg of filament. Surprisingly, there were only 2 or 3 tangles in the entire print run—the filament had been properly wound, even though I was buying the cheapest high-speed PLA I could find (iSANMATE from Amazon at about 12ドル/kg including tax and shipping)

I’m pleased with the way they came out, and I think that the set designer is pleased also.

I know that it is 2025 and I am in a Fall 2025 pottery class, but I never got around to blogging about the pots I made last year, so I figured I should catch up on those in a couple of posts before blogging about this Fall’s pots. All the pots from last year were made using Bravo Buff clay.

The molded tea-ball dishes came out well, but all the thrown pieces are quite heavy and feel clunky and amateurish.

I wanted to try printing a new Shakespeare cookie cutter with my new 3D printer, to see if the better print quality would result in a better cutter. The CAD tool I use, OpenSCAD with the BOSL2 library, has been updated since I designed the cookie cutter, and the old code no longer ran, due mainly to changes in the BOSL2 library.

Because limitations in that library had prevented me from doing what I really wanted with the design several years ago, I decided to try my original thoughts again, but with the updated library. The new design process required me to take the SVG design, convert it to Beziér paths in BOSL2 format, and then use path_sweep and path_sweep2d to make the embossing and cutting edges. I added a backplate to keep the embossing edges where they needed to be without flexing. The path_sweep now works well to make the edges (that is what I could not get to do the right thing years ago).

I also had to come up with a new way to get the SVG design into BOSL2 format. What I do now is to use inkscape to output the svg with absolute (rather than relative) coordinates, and edit the “d=” path into BOSL2 format, partly by hand and partly using emacs macros.

The new cookie-cutter design has tapered cutting and embossing edges, rather than linearly extruded ones, which I thought should help extract the cookie cutter from the dough more easily. I printed the cookie cutter out of PETG (which should be less brittle than PLA, and so less likely to break despite doing narrow 0.8mm edges). I also printed in 0.8mm layers, to make a smoother surface that was less likely to stick to the dough.

I had made the backplate with holes to make it easier to clean out dough that stuck, but I think it would work ok with a solid backplate. A solid backplate with a rib for a handle would make a design that I think could be injection molded out of ABS (though the tooling cost would be pretty high—one would need to make 1000s to justify the cost).

To test the new cutter, I made some more cookies today using my standard shortbread recipe:

1⁄2 cup European-style butter
1 cup pastry flour(I lied—actually I used 1⁄2 cup cake flour and 1⁄2 cup all-purpose flour, because I was out of pastry flour)
1⁄4 cup powdered sugar

Sift the flour and sugar together. Soften the butter and cut it into flour-sugar mixture with a butter knife, and shape the dough into a smooth ball by hand.

On a silicone baking sheet, roll out dough to 6mm thick (using the new cookie sticks to set the thickness). Cut the cookies and remove dough between cookies. Put the silicone sheet on an aluminum baking sheet. Bake at 300°F for about 35 minutes.

I made 9 cookies with this recipe (plus a little bit left over to make a small cookie). The recipe has 480 calories for the flour, 880 for the butter, and 120 for the sugar, totaling 1480 calories, which comes to about 160 calories per cookie.