Glass shelf load capacity on unsupported ends

Are you asking at what point the glass will break, or at what point the glass will tip because you overload one side?

The technical term you’re looking for is “cantilever”. That’s the part of the structure that hangs out. Maybe Google cantilevered glass shelf load limit.

It’s complex for a few reasons, namely the fulcrum area needs to be considered in order to know what a stress concentration point looks like at the end of your bracket near the unsupported corner. A larger support area makes an immediate improvement on the failure load limit of the glass under compression (underside) and tension (upper side) towards the unsupported corner as does flexible padding which can compress and conform on the underside of the glass where it’s supported (load distribution). Then it’s a homogenous substance not like an I-beam’s separated tension and compression spans, but the principle is the same and I’m guessing that there are material limits available to plug into your favorite FEA simulation or to run a 2D mathematical analysis by hand.

Next, deformation limits of the material need to be known. Modulus of elasticity. This is knowable if you have data on your temper quality and glass specs. The glass will bend over the fulcrum until it reaches catastrophic failure at this limit, again variable based on contact dimensions with the nearest support and likely begin to propagate the explosion just above the bracket’s outer most dimension until it’s raining silicon. Sorry, I e forgotten most of my engineering physics calcs but these are good parameters to evaluate for a cantilevered load.

Side note: based on experience your glass won’t break even if the bar is fully loaded unless a bottle is slammed down. That shock could easily pop the tempered glass’ impressive resilience.

Some places to find load data at least for PSI and deforming limits are obvious in aquarium dimensioning on 3/8” glass, and then I’d also look at the civil engineering handbooks for skyscraper features.

Carhole said:

Side note: based on experience your glass won’t break even if the bar is fully loaded unless a bottle is slammed down. That shock could easily pop the tempered glass’ impressive resilience.

Click to expand...

This is also my gut reaction, but there are data tables available with differing degrees of precision and detail depending on the tempering (and/or finish) method (physical heat-quench, chemical, flame polishing, etc…). Surface defects and the presence of moisture is also a big deal.

The other thing about tempered glass is that it doesn't have a small break - or even crack. It's either whole or in a million pieces. I've smacked around tempered sheets in the shop and lab plenty of times, but nick a corner just wrong and the whole piece explodes.

Defenestrar said:

This is also my gut reaction, but there are data tables available with differing degrees of precision and detail depending on the tempering (and/or finish) method (physical heat-quench, chemical, flame polishing, etc…). Surface defects and the presence of moisture is also a big deal.

The other thing about tempered glass is that it doesn't have a small break - or even crack. It's either whole or in a million pieces. I've smacked around tempered sheets in the shop and lab plenty of times, but nick a corner just wrong and the whole piece explodes.

Click to expand...

Yep. Prince Rupert's Drops are an old demo and favorite of high-speed camera nerds. As soon as the internal stress is released, the integrity is zero.

Those things are so much fun.

If you want actual numbers have you tried using something like strian, if that's not too technical? I've not used it since I have FEA capability in the stuff I use but it seems to be a quick and plausible solution for simple cases like this.

NervousEnergy said:

I could put a very thin support along each side edge as they seem to have about 1" framing, but aesthetically it would be hard to match the brackets the wife picked out.

Click to expand...

They don't need to be all along the edge. You could just use a black round nub or point at the midpoint or corners to provide support without going all along and needing to be more matching to the brackets.