CBC Handrail Factor of Safety

[surjaan]

CBC section 2407.1.1 calls for factor of safety 4 to be used for all glass handrail support members. We are already using 200 # point load or 50 # /ft load for designing the support members. Is it not FS is already built in these loads.

 

[structSU10]

FS is not built into these loads. Lets say you have someone trip and grab the handrail to steady themselves. That can easily get to a 200# load.

Also remember glass is brittle, therefore little 'warning' before failure. You want to make sure it would take something crazy for the railing to shatter.

 

[Michel60]

Good morning surjaan,
One thing to keep in mind when applying the FS=4 is that it's calculated against the capacity of the element. I know it may seem obvious to some but I can't tell you how many times I've seen it misapplied as a load factor or taken against an allowable stress. And as strcutSU10 points out glass is a little bit unique as a structural material. There was a thread a week or so ago for a "cantilever glass wall". You might search that one out. I posted a few more comments about how glass is designed that may help.

regards,
Michel

 

[MiketheEngineer]

You must apply the 4:1 against the loads.... BUT, we usually check that against the ultimate capacity of the structure. Don't forget - most "allowable" loads already have a 1.5 - 2 SF!!

 

[Michel60]

Good morning Mike,
I do agree with your approach for most materials, but glass is different. It's ultimate failure capacity doesn't have a direct proportional linear relationship to load. So using a using a 4 as a LF just isn't right. Glass fails as a result of small defects that can only be addressed in terms of the probability for their size and distribution. Glass is then designed based upon the "probability of breakage"(Pb). Normal glass design assumes a Pb=0.008 and for glass handrails the Pb=0.002. ASTM E1300 Appendix X3 provides procedures to calculate Pb for a given load and configuration. Appendix X8 provides procedures for calculating an "allowable stress" for a given Pb and glass configuration. Results of these will show that to gain an added FS you may only need to reduce the stress level by a smaller amount than you would find by using the 4 as a LF.

Regards,
Michel

 

[surjaan]

I do understand using FS 4 for galss, it is a brittle material and FS of 4 is full justified, but using FS of 4 for the supporting steel post is overkilled it will result in a very large section.

 

[Michel60]

surjaan,
It sounds like you may have a different condition than what the Code references assume. If the glass is attached to a vertical post or baluster which includes a top cap and/or handrail then the glass is just an infill panel. In that condition I believe the 200# point load would be applied directly to the supporting steel member and doesn't require a FS=4 since it's not a load applied to the glass (you have a steel framework supporting the 200# load). However the component load of 50# on the glass as a "panel filler" as referenced in 1607.7.1.2 would have a FS=4 on the all connections and supports back to structure. That should change the statics a bit and provide some relief. Just be certain that the glass doesn't support the handrail or act as a top-cap/guard.

regards,
Michel

 

[surjaan]

Thanks Michell, I do have a pipe rail on the top of steel post where 200 # load is applied and glass in ifill panel.However OSHPD wants FS 4 for all glass supporting members.

 

[Michel60]

You would still design the glass for a FS=4 but I think you could argue that the 200# load isn't applied to the glass according to to the Code provisions. I've dealt plenty with OSHPD so I appreciate how hard it is to get them to change their minds, but if you're getting unreasonable results with the procedure you're currently using it could be worth having another conversation with them. The 200# load applies to "Handrails and guards..."(re:1607.7.1.1); the 50# load is sufficient for "Intermediate rails, balusters and filler panels..."(re:1607.7.1.2).

Good luck,
Michel

 

[surjaan]

Michel, I fully agree with you and thank you very much for your response.