Cantilevered Glass Design, Balusters, loading

[Bowsers]

Hello fellow applied physics nerds. I've been scratching my head over a design concept that I was tasked with demonstrating is 'safe' (to the standards of the ibc and asce 7).

The design is with a glass on the edge of the building. This is a location that needs fall protection per IBC. I generated my wind loads (parapet), and am applying them with my live loads using ASD load combo 6, which is 0.75LL +0.75*(0.6 WL). I've completed a RISA model, and analyzed the stress. Generally speaking the glass fails (is over 6ksi stress) in this load combination.

My PM is telling me that I don't need to be applying the loads concurrently as the 200# point load/50 PLF is really an impact load, and has provided justification from a wood deck book.

My concerns are twofold. 1) wood experience ductile failure, where glass failure is brittle. 2) wood is an anisotropic material, that may undergo some form of strain hardening (thinking about this further, laminated glass is also anisotropic, so this may not hold water).

I've seen other posts justify the load as impact load also, but I am still not fully convinced what is acceptable in wood would hold true for glass. Could anyone please help shed light on this issue, so that I may rest easy that I'm not endangering anyone.

Thanks

 

[jjl317]

Having designed systems for conditions like this in the past, I believe (if memory serves) that the load combination you mentioned was considered. However, I have always thought if someone was outside on an elevated balcony at 75% of the design wind pressure, they should have their head examined.

However, depending on how the glass is being supported (i.e. continuous shoe vs. point supported), and what the exact glass make-up is, there are certainly a number of options on how to analyze the glass itself. If you can provide some additional information, perhaps some other options could be suggested.

 

[dik]

for guardrails... most codes do not permit outside force only...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[NorthCivil]

if deflections are giving you grief, model the glass as non-linear

glass is non linear geometrically when under large deflections (large being greater than half the thickness of the sheet)

ensure you evaluate the equivalent thickness as per ASTME1300 for laminated, as laminated thicknesses will not be as stiff as a single ply sheet

 

[Bowsers]

Thank you all for contributing to this.

More specific information on the project itself:
Cat II design, Exposure B, V = 110 MPH, parapet wind loadings @~sea level. C&C wind load = 49.4 psf (ult).

Glass is 5/8" laminate, located on building edge w/ 30' drop. Dims are 38" wide x 58.5" tall glass, supported by (2) "Glass Vice" glass support system (ESR-3563). The glass as designed includes no handrail or toprail.

The design for concurrent wind and live loads is failing as stress < 6 ksi.

Even if I were to accept that wind and live load do not need to be concurrently applied (see item 2 below), I still believe this system would not be adequate (even though "this design has already made it through permitting in two identical designs"). It does not have the required guard protection for this hazardous location. (fails IBC 1015.2.1)

I have two additional things to contribute.
1)Load duration factors do exist for glass (astm E1300 table x6.1), with 3 seconds being 1.0 (where it is 1.6 for wood) and 10 years being 0.31 (is 1.0 for wood).
2) I found the CR lawrence engineering report that they used to qualify their glass shoe system. In their calculations wind load and live load are not applied concurrently. Because of this, I'm willing to accept it as "standard of practice," but I would love to see it explicitly stated somewhere. I still have not found any justification for this in any code book (or commentary), ASTM E1300(which is for windows, not balusters), or ASTM E2358.

 

[dik]

I cannot imagine a lot of people being out with winds of a magnitude to produce such pressures.

One thing (of many) that I've been curioius about with glass is that tempered glass derives its strength from internal stresses, with the outer fibres in compression. Do these stresses ever relax... over a period of several decades?

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[StrucPEng]

I have always designed assuming the concurrent Live & Wind load. I have not heard anyone argue that the live loads are "impact" before. Usually Have designed these for 10-60 minute durations (think a group of people outside watching fireworks and all leaning on the glass. With that said, what are you assuming for your interlayer modulus? Since wind is the principle force duration you can assume a higher interlayer stiffness (@ 3s) than you would if you were designing for only the live load. Also temperature will matter, with winds that high it is probably reasonable to assume that it is not a hot sunny day which would give you a better stiffness value also.

 

[NorthCivil]

"It has made it through permitting before"

Just because it has slipped by someones desk once who didnt know what they were looking at, or didnt notice that there was something that should have been looked at closer, doesn't mean its OK.

I have seen a lot of cringe worthy glass balustrades all over the world. Doesn't mean I will sign off on one.

I have heard a story from an old course instructor once about being in the hospital next to a guy who had fallen through a dodgy balustrade. Won't forget that one.

 

[JLNJ]

In my limited experience with glass guards, the anchorage of the system to the underlying structure is often difficult to achieve and may be your weak link.

The other question, as with all guards, is: where do you apply the load and how does the load get shared across the width of a panel?

For the glass itself, I would not apply wind and the guard load both. They are both small probability events and in my mind the statistics just don’t warrant combining them. I recognize that not everyone agrees with this. I’d rather be conservative in the application and location of the load and the system anchorage than gloss over the application and hone in on load combinations.