a great run!

There are a couple major take home things for this time

• When you are filling up the shaker with dry pigment, have a paper funnel ready to go. A nice big one.
• There were problems due to the first batch being too thick. It slowed down the milling, but it also really slowed down the dispensing: the dispensing took twice as long!
  
  On the second batch I added about 250ml of extra water. It was too much, but it really sped things up. I am trying next batch to add an extra 125 ml. This should be a more happy medium.

This was a largely good run.

Using the shaker bottle to mix the pigment into the medium went pretty well. The process was quite fast. However, it was also very messy, and most of the time was spent cleaning up.

To improve this, I bought a digital kitchen scale, and I will be pouring the pigment into larger containers. Most of the mess came from having trouble pouring the pigment into the weighing containers, and then into the shaker bottle. I may also want to get a scoop from Ikea.

After that, there was a brief time spent mixing the paste on the drill press for pre-dispersion while I fill up the pans for the run. The pre-dispersion does not take long, but it does help get rid of some of the bubbles.

The runs through the mill went fairly well. For the first batch, I forgot to add the clove oil. This actually made the last two runs less effective. It also made the paste settle out in the bowl overnight. It seems that adding the clove oil has the pleasant side effect of acting as a kind of emulsifier. It seems to help keep the particles suspended. It may even be a slight deflocculant, although I am less sure about that.

I used a 10ml pipette to dispense the paste into pans in series. That worked fairly well. However, I think that using automation for the dispensing will be cleaner and less laborious in the long run. My order of electronics has arrived! I will be creating a automatic dispensing machine. The first stage will be to simply create the dispensing head. Then I may set up some kind of way to navigate to seperate pans.

Today I used the blender ball water container to shake the cyan pigment. It worked so very well. Normally, the cyan gets everywhere, but today I can happily say outside of a little bit of staining on my fingers, I am pretty darn clean.

I still did the weighing and mixing outside, and at one point I spilled a small amount of the pigment. I washed it off of my workbench, and it stained the leaves of the gout weed, so I washed it ever further. The clean up time for using the blender ball cup was larger than the time to actually mix it.

I basically divided the pigment into three parts. I poured some dry pigment into the blender ball, and then added about 250ml of medium. I then sealed up the container and then shook it back and forth 60 times. That was enough to mix it into paste.

I saved about 250ml of extra medium to use as a rinse for the shaker before cleaning.

I mixed the paste on the drill press, and this evened out the paste, and actually removed some air bubbles. It was worth it.

Milling went well, but took a long time. I have built a wooden stand to hold a conical hopper. This hopper has a plunger that I can life to start and stop the flow of paste onto the mill. This worked well, but I think I will use a nozzle one step larger in order to avoid more clogging. Ergonomically, using the hopper was much better, and my back is not killing me.

Clean up and set up are taking a long time. It think it is time to start doing double runs of colours in order to reduce the time spent on both.

Dispensing is painful and time consuming, though. I have been looking at finding a way to automatically dispense the paste using some kind of peristatic pump and controller. I found a really nice one that I like on Aliexpress, but it is over $1000. The good news is that I may be able to make one for much less using a microcontroller.

I weighed out the pigment powder, and then I went outdoors. I added 1ml of ox gall and whisked the cyan into the medium.

It still took one hour to do, and the mess was pretty rough. It was better than the mess I made in my basement, but I still had to shower right afterwards.

There was cyan in my hair, ears, face, and on my neck and shirt. I needed a good long shower.

My wife suggested a possible alternative. Tomorrow I am going to buy a protein milkshake mixer from Canadian Tire. It has a stainless steel ball in it that helps to mix the powder. I am hoping the enclosed container will help make it more managable.

Dispersing on the drill press went very well, and there were no issues. Before milling, I did lift the crucible to dislodge any paste that might have formed.

Milling was straightforward. The rollers were set to 6mil, which was slim but reasonable. The milling was slowish, and the rollers could have been opened up a little. However, a little bit of paste did make it through to the bottom tray, so I am happy with the overall speed.

The paste thickened as it milled, but also flowed more consistently. It did make sense to scrape the thick gunky paste into the slow rollers at the end of each cycle. It certainly helped to scrape the apron down at the end of each run with a spatula.

At the end of run #6, the grind tested at 18-25um. I kept running for two more, and the grind tested the same. I added 4 light shakes of clove oil at the end of run #6.

Clean up was reasonable. Cyan is very powerful, but the mess was OK.

Dispensing was time-consuming: about an hour. I set the pipette to 2.3ml to fill the pans. The watercolour paste was very viscose, and had to wait for the pipette to fill. I had 73 pans from before that needed topping off, I a gave each one about 1ml. So I am using 2/3 as the equivalency rate for pans, considering that I will have to top off hundreds of pans later.

140g of cyan ($10.30) + 4 cups of medium ($9.58) = $19.88 / 350 pans = $0.057/pan

I ran the cyan through the mill 8 times, and it made about 4 cups of paint. I filled about 350 half pans with it, or about 1 ½ cookie sheets. This will be enough for about one year of watercolour for Alison and I. Re-using a Kewpie mayonnaise squeeze bottle for filling the pans was a great idea. Using a nail brush to help with cleanup was useful, too.

For the first three runs through the mill, there were some pretty thick chunks of loosely dispersed cyan. These ended up drying on the slow rollers towards the end of each run. I was able to clean this back with a nail brush and adding water to the slow rollers. For the remaining runs, the paint was getting better and better dispersed, and I was able to do the final three rolls back to back. I used the two largest bowls from my mixing set to catch the paint.

Cleanup was pretty decent. I got a couple of mysterious dots of cyan on the floor, but very little. Blue shop towels proved to be very helpful again. I threw my clothes and apron into the laundry afterwards. It might be a good idea to put a towel on the floor before starting. All told, the process of milling and cleaning took about 3 hours.

TODO: In the future, load the mixing bowl fully up with medium first so that the dispersion blades are covered. This will make sure the mixing is less messy and much faster! Go ahead and mix the whole batch - it will all get used later. Wear bad clothes, and goggles. Perhaps run the mix on the drill press a little longer for smoothness. Put a towel on the floor when using a mill. Perhaps get a second squeeze bottle ready to go. Start testing with the grind tester and making colour swatches.

Dispersed a new batch. I used the watercolour medium that I brewed up earlier that afternoon, which had clarified nicely, leaving me about 4 cups of medium. I set things up in my basement, chucking the laboratory mixer into the drill press with a mixing bowl underneath it. I used a ½ cup measuring spoon for medium, and a ¼ cup measuring cup for the cyan pigment, an N95 mask for myself, and gloves. I loaded the mixing bowl up with 4 ½ cups (2 cups) of medium, and then slowly shook 4 ¼ cups (1 cup) of cyan pigment into the bowl. The pigment and medium did not quite cover the mixing blade, and it took a fair amount of time for the material to mix, about 20-30 mins. It also sent a fair amount of cyan dust into the air, getting all over the place. Later on, I added the remaining 4 ½ cups (2 cups) of medium, and this covered the blades of the dispersion mixer very well. I was able to add the remaining 4 ¼ cups (1 cup) of cyan to the medium very quickly, and with very little dust flying in the air. Clean-up was fast on the tools, but slow on myself. Total time, including cleanup: about 1 hour. Used about 140g of cyan, leaving 2.13kg remaining.

Mixing: I placed the medium first into a mixing bowl, then placed the cyan pigment on top. I used a whisk, and this went quickly, although the paste was pretty rough and there was a lot of dry powder floating on top.

Milling: I set the roller gaps originally to 15 mil each. I had some trouble with paste falling through the rollers. A fair amount of the pigment dried on the rollers as well. I put the paste through three or four times. Afterwards, I learned that the dispersion was not good enough: there were quite a few small clumps.

Mixing TODO: Mixing a larger batch would make a lot of sense. Use a whisk right away.

Milling TODO: try starting the mill off at less than 15 mil gaps on the slow side, and 10 mil on the fast side. Using a spatula to scrape out the mixed paste helped a lot. I will have to put the paste through the mill more times.

Next: try milling cyan again. - Try using a 2:1 medium to pigment ratio