One of the reasons I bought a 3D printer was to start printing my own clay extruder dies, for a variety of purposes. I've made my own extruder dies before in my father-in-law's metal shop, but I'm not that good at milling and as a result the shapes I can make are quite limited. With a 3D printer I can make any plate I can reasonably design.

So, when I realized that I was going to need a lot more test tiles to create a reliable seashell pink in cone 10 reduction glaze, it became time to print some plates. Two plates -- one for the test tiles, and one for the test tile holder. Why I need a holder will become a lot clearer once you see the design. But first, I needed to design the plates.

I have a TA Metalworks extruder. It's a similar design to the popular North Star Extruder (and my plate designs will probably work for both). This model of extruder has a long square tube, and normally the extruder die is clamped between this and a square retaining bracket below.

There are two problems with this design. First, you tend to get a lot of clay squishing out the sides between the end of the tube and the die. More importantly, though, this means putting a lot of clamp pressure on the die, which would force me to make my 3D printed plates with 100% fill to make them sturdy enough, and even so the PLA might crack. If I made a die that fit inside the extruder tube, instead, I wouldn't be clamping the die between two pieces of metal, and I could then use a gasket which would limit clay squeeze-out.

So, I first made a metal retaining plate with a square hole that was 1/4" smaller than the tube. I then screwed this together with the retaining bracket for the extruder. You could also do this by welding 1/4" metal plates to the inside of the retaining bracket if you wanted, or you could use heavy wood instead of sheet metal. I put some gasket rubber over the top of this to make for a good seal.

Next I carefully measured the inside of the extruder tube and designed a blank extruder die that exactly fit, as tightly as possible. I made it 1cm thick so that it would be strong enough, and then went to work on my designs, using TinkerCAD.

Designing the die for the test tile was easy enough; it's an 8cm long flat tile, with two raised semicircular bumps in order to check glaze texture breaking and pooling. This is a shape I've been using for ages, and this 3D printed plate replaced a rougher metal plate I made on the milling machine.

I printed my dies using Crealty Silky PLA, with 50% infill printed in a gyroid pattern for strength. So far, I have yet to have a die crack, even when I used clay that was a little too hard. The main funny thing about 3D printed dies is something you'll notice when you clean them: they float!

You'll notice, though, that my tiles are not self-supporting. That's on purpose, because after I'm done firing them I like to screw them onto a board for display and quick reference. But it does mean I need a holder to keep them standing upright in the kiln.

So, I created this extruder plate for holding test tiles upright in neat rows in the kiln. Theoretically you can have three rows of tiles in it, but because none of them can touch it's more often two.

Time to do some extruding! I assembled everything and pushed some Tacoma MAC10 white stoneware through. Works great! And notice -- no clay squeezing out of the sides.

I then took the long strips of test tile, cut them into individual tiles, and added holes in the bottom for eventual screws.

Turning them over, I use a numbering stamp to number each of them so I can track my tests.

Here's the full deal, awaiting bisque. You can see how the test tiles will go into the kiln once I'm ready to use them for glazes.

My 3D designs are all up on Thingiverse if you have a TA or Northstar extruder yourself and want to try out making your own dies. The software I use is TinkerCAD and OpenSCAD for designs, and Prusa Slicer for slicing. I do my printing on a Creality Ender 3 Pro with a WhamBam print surface.

So I've posted a bunch of new items for sale in the FuzzyChef Online Store. This is my first major firing with Lava Street on lots of my pieces, and I'm really loving it.

Pictured above is a "cassole", the traditional baking dish for the French classic, cassoulet. Made in the traditional conical shape, it's about 13" in diameter and holds a gallon and a half. You won't find this at Sur Le Table. For this and other stuff like salt pigs, I've created a new store category of ChefWare.

The kiln overfired slightly; I'm going to have to work on recalibrating it. However, the overfiring was great for the pieces it didn't ruin due to glazes running. For example, these two mugs turned out amazing:

Anyway, lots of new stuff up, and more to come, so go shopping!

Have you wanted to get started making your own glazes, but aren't sure quite how?

For the second time, I will be teaching my "practical glazemaking" class at The Clay Underground in San Francisco. This two-evening, hands-on class focuses on basic glaze mixing and measuring techniques. Students will make their own glazes from recipes and learn to produce variations.

The class will be on the evenings of August 13 and 20. Registration closes August 3, but it's likely the class will fill up before then.

More information on the registration page.

Meet "Lava Street", my newest favorite glaze. This was originally called Orange Street and I got it from the Ceramic Recipes group on Facebook. I've renamed it after adjusting the formula, since on my ware and with my stuff it looks more like lava. Above is Lava Street over Faux Fur, which does a kind of mocha diffusion.

I love it, but it needs more testing.

For example, if I want to use a glaze on eating ware, I need to test it for leaching, to make sure it won't poison my customers and friends. I say "leaching" rather than "food safety" because the FDA has been rather cavalier about standards for ceramic food safety, making the term meaningless. Whereas leaching specifically refers to the tendency some glazes have of letting toxic metallic oxides bleed into your food or drink, something we don't want.

I do have glazes which leach out, and use them exclusively on decorative pieces or on the outsides of pieces and not on any surfaces expected to come in contact with food or hot liquids.

Now, the really scientific way to test for leaching would be to send out samples to a chemistry lab. But as an artist, I can't afford that. Fortunately, there are some household tests which work pretty well; in fact, I use a four-fold method to determine if a glaze is "safe".

1. Glaze Ingredients

Here's the recipe for Lava Street/Orange Street again:

• Silica 13%
• Talc 11.5%
• Dolomite 7%
• EPK 4%
• Gerstley Borate 15%
• Minspar 39.5%
• Bone Ash (Tri-calcium phosphate) 10%
• Red Iron Oxide 18%

The most toxic chemicals used by potters in glazes today include barium, manganese, and uranium. Somewhat less toxic, but still a problem if they leach into food, are other chemicals like copper, cobalt, chrome, and nickel. This glaze contains none of the chemicals which are marked for special attention to prevent leaching; the large amount of iron it has is fairly harmless even if it does leach.

2. Unity Formula

In order to be sure we're not going to leach, we want to make sure that we'll have a good amount of silica melted into glass to hold all of our other chemicals. So the Silica:Alumina ratio is one of the things you want to check for a safe glaze.

So here we have a Silica:Alumnia ratio of 9.9:1, which is high in glass. Unity test: passed.

3. Lemon Test

One thing which can cause leaching into food is acids. Lemons are one of the most acidic things you're likely to have around the house. So a standard testing technique is to put wet lemon slices on a test tile, like so:

Cover to keep moist, and leave for a couple of days. Then lift off the lemon slices to see if the glaze is marked or discolored; such would indicate leaching or other issues. Results?

Lemon Test: Passed.

Oh, and apparently Lava Street does really well on tiles with impressed designs.

4. Dishwasher Test

We also want to test vulnerability to alkaline substances, and hot liquids. We can do both at the same time in the dishwasher. Take two test tiles fired together so that they're nearly identical. Put one in the dishwasher like so:

Then run it through for 5 to 10 cycles. And if the holiday season is good for anything, it's the opportunity to run a lot of dishwasher cycles in a short time. Seven cycles later, let's look at the result:

Can you tell which one went through the dishwasher? I've asked three other people, and most of them got it wrong. So, Dishwasher Test: Passed.

Conclusion

I can now be confident that by using Lava Street on my dinnerware, I'm not poisoning my customers. Which makes me very happy, because I love the glaze and want to use it inside and outside my pieces.

Now, I just need to do something about its tendency to run when layered over other glazes. I had to spend some time grinding down the foot of this mug.

All ware and test tiles above are on Leslie's Vanilla stoneware, fired to Cone 7 in oxidation, using a simple medium-fast cone fire program with a 10 minute hold

... continued from part 1

One glaze I used to have when I fired cone 10 electric which I really miss was one I called Verdigris. It was based on a Val Cushing Cornwall stone matt glaze recipe, with tin and copper so that the glaze was actually bronze:

I've been trying for a few years to "down-temperature" this glaze without success. Recently someone recommended a Cone 6 recipe called Pinnell's Matte, which is widely popular. The recipe I was given used copper and titanium, which didn't work out so well, but based on experience I also tested copper and tin. This is a lot more promising:

Please note that the picture is mislabeled; it should be Tin/Copper, not Copper/Tin. The tile on the right is 8% copper, not 8% tin.

I need to flocculate the glaze (it's 60% Nepheline Synenite), so that it will go on thicker and stay suspended, and test further. But that's promising.

Pinnell's Matte Bronzed?

• Nepheline Syenite 60%
• Ball Clay 8%
• Bentonite 2%
• Strontium Carbonate 20%
• Flint 9%
• Lithium Carbonate 1%
• Copper Carbonate 5%
• Tin Oxide 5%

The other glazes I've been trying to reproduce at Cone 6 are a set of glazes made with raw wood ash from my in-laws' fireplace and furnace. This is a mixture of madrone, pine and cedar ash. I'm looking to produce both a clear yellowish glossy glaze with impurities which looks like the piece came from a wood firing, and a second runny glaze which I can color and put at the tops of pieces for interesting effects. I got some promising leads on both.

My big trial was a simple ash glaze recipe made with RedArt clay. I got this idea from one of the potters at the annual ClayFolk show, who has some quite nice ash glazes based on it.

As you can see, this ranges from a shino-like satin glaze all the way to a glossy slightly runny glaze.

The 30% or 40% would be good candidates to replace my "fake wood firing" ash glaze; I'll test further.

RedArt Raw Ash Glaze No. 1

• RedArt Clay 40%
• Ferro Frit 3124 10%
• Raw Wood Ash 30% to 60%

I also did a test of a slightly different glaze based on a recipe from Ceramic Recipes group. This recipe uses a much higher percentage of Ferro Frit 3124, and much less ash. I wasn't impressed.

There are a lot of old recipes on the internet and in old books which use raw wood ash and Albany Slip as a glaze. While Albany Slip is long gone, Laguna mines a California clay known as Arroyo Slip. You'll notice the middle tile stuck to one of the other test tiles; such are the perils of testing glazes. This produced a much more promising result for a runny ash glaze; it bunched and dripped like I want. You can see this more on the backs of the tiles:

Based just on this test I'd say the 40% is promising, except that I know the kiln was overfired, so maybe I should go with 45% instead. The 2% Cobalt is also a bit dark, and I'm wondering about using 1% instead.

Arroyo Slip Raw Ash Glaze

• Arroyo Slip 40%
• Ferro Frit 3134 10%
• Raw Wood Ash 40% to 50%

Anyway, a mere three dozen test tiles (including some failures I didn't show), one good glaze and four worth further testing. That's an exceptional return for a glaze testing batch; usually it's more like one good tile out of all 36. I'm excited to get glazing in 2015.

Folks who have visited me at a booth or Open Studios know that I make my own glazes. I find commercial glazes to be both inferior and exhorbitantly expensive, and I learned pottery back when they simply weren't available. So it's all glazes I formulate, or mix based on others' recipes.

All of this means doing lots and lots of glaze tests in order to get a small palette of good glazes: on the order of 30 to 40 tests per keeper glaze. I generally put a few glaze tests into every glaze firing I do. What follows are the tests for this month.

First things first: our Toki Kiln is overfiring. For some reason the folks in Berkeley think that cone 6 is 2232F, which is about 30F over what I'd put it at, especially on a 13 hour firing, and the cones agree with me. So this was a Cone 7 glaze test.

Let's start with the big success. This is Orange Street, a glaze which has been sweeping the folks on the Ceramic Recipes Facebook Group because it is so awesome. The blog title picture is Orange Street on a test cup. It's more red than it was for other folks who have used the recipe; that could be because of the overfiring, or it could be differences in the materials, especially Talc.

Interestingly, the test with 15% Red Iron Oxide looks pretty much indistinguishable from the test with 18%. While iron is cheap enough I don't care, it would be worth reducing the iron in order to reduce leaching.

Red-Orange Street

• Silica 13%
• Talc 11.5%
• Dolomite 7%
• EPK 4%
• Gerstley Borate 15%
• Minspar 39.5%
• Bone Ash (Tri-calcium phosphate) 10%
• Red Iron Oxide 15% to 18%

Randy's/Toby's Red (this glaze has had a lot of names), though it has the same chemicals in slightly different proportions, didn't do nearly as well. As you can see it's mostly black; this could be due to materials, or to the overfiring, or something else. I've been told that if I refire it to bisque temperatures more of the red will come out; I'll try that, but with Orange Street looking so similar and already working, I'm not sure it's worth troubleshooting this glaze.

Far less successful was my attempt to create a chrome green, so that I would have a deep forest green in a glossy translucent glaze. I tried adding Chrome in different concentrations to the two Glossy Bases from Mastering Cone 6 Glazes, but ended up getting a semi-matt glaze out of it; somehow the chrome changes the glaze so that it doesn't get glossy. Will have to start over with a different base, or maybe try the reformulated bases from the website (due to materials changes).

More tests on this blog tommorrow ...