The Test Panel

I liked the look of the large candlebox and wanted to do another one.  I bought a beautiful deep blood-red glass that I wanted to use for it, but unlike the yellow it had no milky elements to it, it was perfectly clear with just a mottled surface texture.  This meant that if the glass cracked it would be particularly visible.

In my construction experience I'd heard of non-shrink concretes and grouts and decided to use one of them for the new cast.  I built the mold, more or less identical to the yellow-glass one, although the pattern of webbing between the glass panes is different.  I deliberately made it simpler with fewer strands, using the classic artist's mantra, "It's not about what else you can add, it's about what else you can take away".

Before I poured the non-shrink material in the freshly built mold, I first built a mockup of a single side of the new mold, one panel, one face.  My plan was to cast this one side out of the same material and watch it to be sure it didn't crack.  Only then would I cast the larger piece. 

I waited a full four weeks and no cracks appeared so I thought I was safe and went ahead with casting the large red candlebox.  The next day the sample cracked.  Over the next few days, the glass cracked substantially, and even broke through and cracked one of the thick side elements.  I was disappointed, but it was a good experiment.  Even non-shrink concrete apparently shrinks a little bit.  On the upside, the cracks are finer, the concrete finished smoother, and it generally has a nice surface appearance.

Also as an experiement that I haven't yet repeated, one section of the panel was cast without glass - hence the recessed concrete shape in the center.  I like the idea of not having glass in every panel.

This test inspired a design for a full-sized concrete-encased stained glass window.  This design has not yet been executed.