Or so I thought. Someone asked me recently if I knew how large a piece of glass needed to be to resist a certain wind load. I don’t know why I remembered I had it, but a chart from the 1970 UBC used to be what was used to figure out if glass was strong enough for wind loads.To use the chart, you found where the the thickness of the glass shown along the top row intersected with the wind load column on the left. That number was the maximum square footage of glass allowed for that wind load. Example: ¼-inch glass, 40psf wind load = 27 sq.ft. maximum glass lite size, regardless of the height and width. These values represented a “fresh off the float line” strength of annealed glass.
Referring to another table, you could then increase the 27-square-feet allowable lite size by incorporating it into an insulated unit (27 x 1.5 = 40.5 sq.ft.), heat strengthening it (27 x 2 = 54 sq.ft) or tempering it (27 x 4 = 108 sq.ft). The factors could be combined, such that a heat strengthened, insulated unit could be, at 40 psf, 27 x 1.5 x 2 = 81 sq.ft. That’s assuming, too, if you heat treat one lite, both lites in the IGU had to be heat treated one way or another.
These charts were developed primarily by glass manufacturers. Easy, right? Not so fast, kemosabe!
The good folks at ASTM came along and developed ASTM E1300 “Standard Practice for Determining Load Resistance of Glass in Buildings,” which addressed more factors. According to Bill Lingnell, the old UBC charts didn’t take into account weatherability, and they used a statistical probability method following a normal distribution curve. By contrast, ASTM E1300 was developed using a failure prediction method, and also accounted for glass weathering. The IBC since has generally adopted ASTM E1300 as the glass strength standard.
Currently, there are a lot of companies offering ASTM E1300 software that will confirm the glass construction – insulated or monolithic, heat treated or not, laminated, etc. – and whether it will work for a given wind load.
Some programs will also calculate how much the glass deflects. If you’ve ever been in a completed building, and the occupants call in with a complaint that “the glass is about to fall out,” you know about this part of the discussion. Occupants aren’t used to glass that moves like a trampoline under certain exterior wind conditions. Glass deflects equally when annealed, heat strengthened or tempered. It takes approximately four times the force to break a tempered lite compared to one that was annealed, but both still deflect the same under the same load.
Some architects ask about limiting the glass deflection to 1 inch. This is a human comfort thing, as there’s nothing in the codes and/or in the glazing industry that limits deflection. In most instances, the means to limit deflection on larger lites or in areas with greater wind load is to increase the glass thickness. Most framing systems don’t easily accommodate varying glass thicknesses (i.e. 1” IGU in typical vision areas, 1¼” thick in higher wind load zones). Vision and spandrel glass thickness differences used to be a common practice, but have virtually disappeared due primarily to energy constraints. The software, too, will allow one lite to be different than the other, be it tempered, heat strengthened, or laminated while the other lite stays annealed. The UBC charts didn’t do that.
But that’s all the software will tell you. It won’t reveal whether the glass needs to be tempered, or if safety glazing is needed because it’s in a door, is adjacent to a door, or because it is adjacent to a walking surface where no handrail or horizontal is present. It also won’t tell you that safety glazing is required if the sill is lower than a certain height off the floor, or if there’s a horizontal or handrail 3’-0” to 3’-4” off the floor so that if the glass breaks, someone won’t fall through the opening.
Software also won’t tell you if the glass should be heat treated at inside corners for sun-facing surfaces. Glass doesn’t always allow all of the sun’s energy to pass through it, some is reflected, and if that reflection is onto another glass surface, the lite on that adjacent surface may need to handle more than just direct thermal energy. Reflective metal panels may have reflective effects on adjacent glass, also.
Other factors that affect glass substrate selection or make-up are reflective coatings and tintings, to name a few.
The lesson here is you can trust software, just know what it’s doing, and more importantly what it’s not telling you. Although many of the old rules don’t work as well as they used to, one that still holds true is: when in doubt, call the glass manufacturer. Their tech staffs are able to help with these types of issues. It’s nice to have a second set of eyes to help get your project right, both in the estimating and execution phases.
Lastly, PLEASE NOTE: Any comments on this blog that deal with my age, especially from Greg Carney, will be deleted with prejudice!