View Full Version : Color via Batch How To

Dave Bross
12-20-2016, 09:29 PM

Continuing on that train of thought....the first (and sometimes many more than the first) melt is often a throwaway. Plan to make something one color with it. I say throwaway because you want to check the expansion on that first one (it will be off) and them tinker with the batch to get a good match for your clear.

If you're wanting special effects then gird your loins, reread the Josh epic up above, and be sure you have the patience and desire to do that. It's definitely a "don't try this at home" until you get familiar with more simple melts and get your process and knowledge down anyway.

Getting back to basic colors....you will come up with an expansion number unique to that color in that furnace, and sometimes special effects (silver glasses) depend on things like humidity and atmospheric pressure. After all that, your hard-earned recipe will not "cross the street" without re-doing the learning curve.

You are going to track what you do with E&T calculations, right?????
If not, expect frustration, insanity and (most horrifying of all!) lack of repeatability. You're already nuts if you're doing glass.

Ghetto time: I mix batch in a bucket or small plastic drum cut in half. I use a drywall mud or paint mixer and a drill, cardboard with a hole for the shaft on top of the bucket to keep the dust down. Fast and some of the most complete mixing so far. If you want super-sanitation on dust issues the industrial supply houses sell plastic bucket caps that look like a womens showercap. You'll have to add the hole for the shaft. I mix outside so the dust just fertilizes the grass.

uspigment.com has the black tin, silver nitrate, and most all else you'll need at good prices. Excellent service too, although they'll do crazy things like ship very toxic things in a single plastic bag and a priority mail envelope configured to look like a python swallowing a small animal. The postal sort machines do their very best to demolish this.

One of the largest range of colors possible is combos of copper and cobalt. Turquoise through blue.

Don't start with silver glasses (with the possible exception of the chalcedonies) for the benefit of your mental health and wallet.

Reducing melts are a total pain in the ass in an electric. Check some of my old posts on making copper ruby for the details on some of what you'll encounter. Cob up a quickie gas furnace and you'll suffer less.

Start well below the 1% threshold (.5% is good)to get a feel for how much density you get for how much colorant. You'll probably also get something that will fit the clears better. Some colorants are much more powerful than others. Find out which ones. The archives here will be some help with that if you're good at reading between the lines.


Before I go into the copper ruby recipe I need to explain oxygen content in glass. I won't go into the detailed chemistry but the amount of oxygen in a glass determines the color in many glasses, and copper is one of the most radical of these.

With oxygen you get turquoise blue, without oxygen you get copper ruby, too little oxygen and you get overstriking on the ruby. Some glasses must have exactly the right amount of oxygen to shoe color, selenium glasses being an example.

Most glasses are melted in what everyone calls "oxidised" state. There are nitrates in the batch that create a high percentage of oxygen in the bubbles. There is also antimony in oxidised batches, what antimony does is change valence around 2170 F. Over 2170F nothing happens, but when you lower the temp. below 2170F the antimony goes looking for oxygen to complete it's valence switch. It absorbs the oxygen bubbles into the melt, making them disappear.

The next step in the decreasing scale of oxygen in a melt would be a "neutral" melt, which would be no oxidisers, no antimony, no reducing chemicals. Here's where the fining problems begin. Without the oxygen from the nitrates, the bubbles you see are all carbon dioxide. The only way you're going to get rid of them is to let them float to the top, which they are quite slow about. Thinner viscosity glass helps with this but the tradeoff is that the thinner glass eats the pot more, so, more cords, less pot life. Basically the glass has to sit at melt temp until all the bubbles float up. This can take days, or sometimes not work at all if the bubbles are too small. The CO2 bubbles don't rise nearly as quickly as ones with more oxygen in them.

Then there are the reducing glasses, which want so little oxygen that we add things like black tin or other chemicals with lots of carbon in them to
eat up more of the oxygen in the melt. The problems here are not going overboard on the reducing chemicals (affects the color in negative ways) and getting them to fine. With even less oxygen than neutral glasses the fining is even tougher. You almost have to drop the viscosity of the glass chemically or with temperature and wait for a while. Again, it may never fine is the bubbles are small.

I'll use a lot of not-quite-completely fined color glasses like the chalcedony and phosphates at the point where all the big bubbles are gone but the tiny bubbles are still there. Makes for a bit of sparkle to them.

Copper ruby is a reducing glass, so if you were going to try it with Spruce Pine premixed batch (I think that was what you were asking?) you would need the unoxidised version of that batch. The weight percentages on the colorants are .3 copper, .6 black tin and .2 antimony.

The glass strikes slightly during working and completely in the annealer. It should be cooled down to where there is no color showing from the heat at some point in the working. The color will develop in the annealer at somewhere around 1-2 hours, you have to keep checking and see how it's striking and then lower the annealer temp. to stop it when you have enough strike. The times will probably be a bit different if done in Spruce Pine batch, you would have to watch and adjust. You can also change the tin percentage to alter the striking, less=slower. If you go too low on the tin it won't strike. Read the copper ruby directions on Gaffers website for a more detailed explanation of the working process but ignore the temps. they give, this glass will go at different temps.

That amber I showed you is fairly easy to melt. Just add .6 weight percent molybdenum disulfide to an unoxidised batch.



All the calculations in the world only give you a "ballpark" number for glass melts. The truth is in melting it and seeing what you get.

There are too many ever-changing variables to get a truly accurate number on paper.

If the first melt missed, then you go back to your E&T calculations to get a ballpark number of what you'll need to add on the next melt to raise/lower expansion.
Then you melt that and see what you have.

This is where the E&T numbers really shine, they seem to be even more accurate for this "adjusting" phase. Sometimes it takes a third melt to dial it in where you want it, but rarely more than three unless something else among the variables has gone wrong and is messing with you. Stuff like thermocouples, melt times/temps., hyroscopic moisture in the chemicals, etc. etc.

I always assume I'm in for three melts to dial something in and I'm often pleasantly surprised when I hit it in two.

The more you can standardize your process and sources of supply the fewer strange surprises you'll have.

The main thing Pete will teach you next summer is process.

Pay particular heed to the comment about melt times/temps.
Wire melters "slow down" as the elements age, leading to higher expansion numbers.

My antennae shoot up like a pop tart out of a toaster when I 'm forced to switch suppliers on something. Fortunately, switching sodas had no repercussions for me either.

I don't think the chemically combined water bakes out either.
Volf made the comment that some water helps melting a lot.
Water, or, more accurately, the steam from the water is a problem with anything like selenium that wants to vaporize readily.

Be careful to balance that with empirical experience...or in simpler terms don't let too much analysis get in the way of hands on experience.

This is particularly important with glass formulas/melting because everyone gets unique results depending on their location, equipment type, phase of the moon (kidding) etc.

Something I learned the hard way myself...

As for sand, see if you can find Short Mountain silica there. It works better than most. Usually found at pottery suppliers.


In the case of batch it depends on which furnace you melt it in. The expansion varies a lot depending on speed of melt.

This really is "your own little universe"

You need to be testing because your situation/furnaces are not the same as the ones across the street. Homemade polariscopes are easily made and can save much embarassment:


You can do a lot with thread testing too, my sequence is thread test then confirm via polariscope.


As far as nuggets and other cullets...I would be inclined to be lazy and depend on them being the same expansion all the time...but sooner or later there may be a surprise, so best to test there too. It only takes a minute or two to test and you'll be the one who caught it before the damage was done.

It's really not a "set in stone" immutable number situation, it's a "fits as compared to what"
You can get a "number" with a dilatometer but what you really need to know is will it fit...and it's really not about the expansion number so much as if the glasses cross the set point threshold at about the same time.

Viscosity plays into this big time and there's no easy way to track that. You can track and change expansions (assuming you're batching) to change the set point so that's how you end up with a number of different expansion numbers to get glasses that fit each other.

Of course, if you're using cullet and color bar it's just a yes/no on the testing with no way to fix it later. Learn how to batch and you can fix the mismatches.

Time once again to read this...it's a hoot as well as educational:



OK, the short version....

Very basic - sand, soda, lime glass

As close to 70% sand as you can get, usually in the high sixties due to necessary compromises.

At least 8% modifiers, usually lime because it's inexpensive but could be almost anything in the 2nd column of the periodic table.

Try to keep it under 18% total alkali (see 1st column of periodic table for alkalis), usually sodium for economy but it won't polish well or be as pretty without some potassium. If you have to go over 18% alkali (matching Moretti?)you can do it if you use...

Enough alumina to keep it durable. Nick Labino's rule...divide total alkalai by 8 and that's your alumina percentage. This applies to all glass.

OK, now let's spend a little more money and get some way nicer glass.

Again, potassium. Yeah, it's a bit expensive but 3-5% will make a noticeable difference in appearance and polishing.

Zinc will brighten up even your basic cheap-o soda lime alkali only glass at around 2% with a big bonus of much added durability.

Lithium will make it melt like gangbusters even at very small percentages. The price is spiraling up daily but worth whatever you have to pay if you're melting in something under powered. Bonus features...improves glass strength and reduces thermal shocking. Also drops the viscosity and extends the working time of the glass.
Need to go the other way on expansion (decrease)? Add more alumina or zinc in tiny percentages.

Strontium is a wonder modifier. Gives some of the benefits of lead without the toxicity. It's way "cleaner" than most forms of calcium.

Here's some from Volf from his Strontium chapter:

"Because of its identical charge and similar effective radius, r1, Sr is sometimes compared with Pb, despite the differing electron configurations of the two elements. The latter are reflected in the different polarizabilities of the two elements and their properties in glass. Pb as an ion of a B-subgroup element with a high atomic weight is twice as polarizable as Sr. In contrast to Pb, Sr with its lower atomic refractivity neither decreases the surface tension nor concentrates in the glass surface. Electrostatic charges therefore are not created on the surface of Sr glass by friction, as is the case with lead glasses. The closeness of the radii of Sr and Pb is responsible for the possibility of replacing considerable amounts of lead oxide in glasses containing 30% of PbO; in this way, Partridge reduced the PbO content from 30 to 20% by introducing 2-15% wt.-% of SrO."

There's more about how Strontium fines/melts better because it fluxes at a lower temp. and puts less gas into the melt than Barium.

E&T and scales