Glass Annealing Schedule Calculator
Tell us what glass you are firing and how thick the thickest part is, and we hand back the whole cool-down: what temperature to hold at, how long to hold, and how fast to drop the kiln after that. The numbers come out in the same rows your kiln controller asks for. Everything runs in your browser, nothing uploaded.
Type these into your controller
| Segment | Rate ? | Target | Hold | Takes |
|---|
What thickness costs you in kiln time
Every bar is a full cycle at your soak temperature for a flat slab of that thickness. Your own row is filled in.
Learn more: annealing glass by thickness
What the schedule is actually doing
Hot glass that cools unevenly ends up with the outside set hard while the inside is still contracting. That locked-in fight is residual stress, and it is what makes a piece crack on the bench days after it came out of the kiln.
Annealing fixes it in two moves. First a soak: hold the whole piece at one temperature long enough for the middle and the surface to reach the same number, which is where the glass is still soft enough to let the stress bleed away. Then a controlled fall slow enough that no new gradient builds up on the way down.
So a schedule comes down to a soak temperature, a hold length and the rates for the staged fall below it. The hard part is that none of those can be guessed from the others, and every one of them moves with thickness.
Where these numbers come from
The soak length and the three cooling rates in this tool are read straight off Bullseye Glass Co.'s Annealing Thick Slabs chart, publication #8100, which Bullseye states is derived from Corning's method as published in McLellan and Shand's Glass Engineering Handbook, third edition. The chart covers 0.25 inch through 8 inch slabs and stops there, and so does this page.
Two patterns run through the whole chart, and they are what lib/schedule.js rebuilds. The soak is four hours per inch of thickness, exactly, at every row from 0.25 inch through 8 inch. The three cooling rates always sit in a fixed ratio to each other of roughly 1 to 1.8 to 6, so once the first one is set the other two follow.
The first rate itself falls with the square of the thickness. At 1 inch it is 27 degrees per hour, and 27 divided by thickness squared reproduces the chart from 1.5 inch up without a single exception: 12 at 1.5 inch, 6.8 at 2 inch, 3 at 3 inch, 0.42 at 8 inch. Below half an inch the published figures flatten off, because no kiln needs a cap of 432 degrees per hour.
You can check the arithmetic yourself on the 1 inch row. Four hour soak at 900 F, then 100 degrees down at 27 per hour is 3 hours 42 minutes, 100 more at 49 per hour is 2 hours 2 minutes, and the last 630 degrees at 162 per hour is 3 hours 53 minutes. That totals 13 hours 38 minutes, which is the rounded 14 hours Bullseye prints.
Why thick glass gets so expensive so fast
The soak doubles when the glass doubles, but the cooling rates quarter. That is what makes the total climb far faster than the thickness does.
In real figures, 1 inch is 13 hours 38 minutes and 2 inch is 46 hours 24 minutes, a factor of 3.4. Double again to 4 inch and it is 170 hours 5 minutes, a factor of 3.7. Double once more and 8 inch takes 653 hours 40 minutes, over 27 days of kiln time for a single block.
This is the part worth knowing before you design a piece rather than after. Shaving a cast billet from 3 inch down to 2 inch is not a 33 percent saving in kiln time, it is 99 hours down to 46.
The rule that catches most people out
Bullseye's chart only applies to flat slabs of uniform thickness, set up so they can cool evenly from the top and the bottom. Anything else, which covers most cast work, bowls, slumped pieces and anything with a thick rim, takes the cycle listed for at least twice its thickest area.
That rule is why the uneven toggle on this page moves the answer so much. A 0.5 inch uneven piece does not follow the 0.5 inch row at 4 hours 36 minutes, it follows the 1 inch row at 13 hours 38 minutes. Picking the wrong one of those two is a three-fold error in the direction that cracks glass.
Bullseye also warns that even its most conservative cycle can fail if the kiln itself cannot cool the work uniformly, which no schedule can fix. A cold draught across one corner of a shelf will beat any number in the table.
What changes between glass families
The soak temperature is the part that belongs to your glass rather than to the chart. Bullseye's own COE 90 glass soaks at 900 F on every row. COE 96 sheet glass runs hotter: Youghiogheny's Y96 firing schedules begin the controlled cool at 950 F for slump, tack and full fuse alike.
Borosilicate is hotter again. TimeWork Kilns' borosilicate reference gives annealing points of 1058 F for Kimble KG-33, 1049 F for Corning 7740 and 1040 F for Duran, and tells you to set the dial at 1050 F. Soft rod glass around COE 104 sits in between, with a published range wide enough that your own supplier's figure beats any default, which is what the custom option is for.
What this page does with that is shift the whole schedule up or down together. The soak moves to your glass's temperature, the two intermediate breakpoints stay 100 and 200 degrees below it, and the rates do not change. On borosilicate that final ramp is more conservative than the glass needs. The last stage runs below the strain point, the temperature under which the glass is rigid enough that no further stress relaxes out of it, and boro expands about a third as much as soda-lime glass so it shrugs off cooling that would crack a fused plate. Slow costs you time and nothing else, so the tool leaves it slow.
FAQ
How long does it take to anneal 1 inch thick glass?
About 13 hours 38 minutes from the start of the soak, for a flat uniform slab of COE 90 fusing glass. That breaks down as a 4 hour soak at 900 F, then 3 hours 42 minutes cooling to 800 F at 27 degrees per hour, 2 hours 2 minutes to 700 F at 49 degrees per hour, and 3 hours 53 minutes to room temperature at 162 degrees per hour. Bullseye rounds that to roughly 14 hours on its own chart.
Why does twice the thickness take more than twice the time?
Because the cooling rates fall with the square of the thickness while the soak only grows in proportion to it. Going from 1 inch to 2 inches takes the cycle from 13 hours 38 minutes to 46 hours 24 minutes, a factor of 3.4. From 2 inches to 4 inches it goes to 170 hours 5 minutes, a factor of 3.7. From 4 inches to 8 inches it reaches 653 hours 40 minutes, a factor of 3.8.
What schedule do I use for a piece that is not a flat slab?
The one listed for at least twice its thickest section. The published chart only covers flat slabs of even thickness set up to cool from both faces, so a bowl, a cast billet or a slab with a thick rim takes the doubled row instead. A 0.5 inch piece that is not uniform therefore follows the 1 inch cycle, which is 13 hours 38 minutes rather than 4 hours 36 minutes.