Expanded polystyrene is recommended for applications below 160 degrees Fahrenheit, with this one coming in about 10 degrees below that mark. I started by covering the walls with aluminum foil, reflective side inward-facing, securing it with normal Elmer's Glue-All. I then drilled a 1/2" hole in the bottom center of the container, through which to pass the wires for the light bulb. The tiles were positioned in a diamond shape in the bottom, as depicted here:
The tiles have small felt feet on them in the form of a flattened "X", with top and bottom one inch in from the edge of the tiles. This keeps them from scratching the aluminum foil, while providing a solid base upon which to place mason jars, with enclosed garlic. While I have used mason jar canning lids in the past, I have decided to switch to using aluminum foil as the exclusive lid, as it doesn't need to be a perfect seal and I'm concerned about the BPA content of the canning lids. I have placed the reflective side up, as below:
I am able to fit 12 jars on each side of the cooler, for a combined 24 jars. This is my new standard, as of the batch that I just started, having used 12 jars total in all batches up to this one (no small wonder, given that mason jars are sold in 12s). For this and the last batch, I purchased my garlic from Sam's, which sells large bags of garlic for around $4. The bag provided about 22 heads of garlic, which I supplemented with two heads I purchased with the last batch. I should have noted which these were, as I'm less confident in their end quality, based on the fact that they're over a month old.
Note the stone filled jars from the last post. They provide IR absorption and thermal mass for the system. As a strongly IR absorbing material, stone also emits IR, bolstered and buffered by the high specific heat of the water. In my other experiences with fermentation, I have understood that light exposure can be detrimental to the fermentation process. Googling "Light effect on fermentation" seems to provide a number of articles, blog entries, &c. that suggest that both visible light and UV radiation can have a negative effect on fermentation. I genuinely don't know if they apply at 140°F-150°F--a perceived consensus indicates this range for black garlic (per this instructable and this make article)--but this mitigates the issue, while the dry heat resists mold and non-beneficial bacterial growth.
Each head is placed in an individual 8 oz. mason jar. Some of those that I have are normal and some are quilted, but I don't think it really matters which one is used, save the utility of having the measurements on the side of the normal mason jar. I leave the heads sealed in the jars for 30 days, at which time I remove the lids and let the garlic dry. This results in cloves that are dark, textured and have shrunk away from the skins of the garlic. Some in the last batch were drier than others, with some reducing to a powder when pinched between the fingers, while others were leathery and pliable. The harder examples were cut up and reserved for additional drying, for making black garlic powder.
To measure temperature, I punched a hole in the side of the cooler using a small meat thermometer, calibrated in boiling water (e.g. This Taylor product) around the middle of the height of the jars. Using this to monitor temperature, I used a number of different light bulb wattages.
For reference, I used a ceramic light socket I got from a cheap lamp I bought from Lowe's hardware store. There is one very near my home, and I found that they have ad hoc porcelain light sockets for between $3 and $4. I used a cord from another, broken appliance, and soldered some fork/U-shaped terminal connectors to the end of its wires. I bought the cooler at Academy, and I can't easily find the price, though I believe it was no higher than $30. Altogether, I remember calculating (some time ago) that the apparatus, plus electricity, with initial cost of garlic was less than the cost of twelve heads of black garlic online. For the record, this is an excellent use of incandescent light bulbs, as they put off far more heat and IR radiation than light. In the month it takes to do this, the oven costs on the order of $5-$8 per batch, including electricity and garlic, but not including the cost of manufacture.
Update (2012-12-3) - Research Data: Here is a copy of all of my research on cooler heating times and final temperatures for different wattages of light bulb. It contains the raw thermometer snapshots as well as a spreadsheet with the data organized into relevant pages. While I intended to do experiments for 11W and 7W, only the information for 60W, 40W, 25W and 60W w/water are complete.