Timber handrail factor of safety

[glass99]

I have a round timber handrail which will be attached to a glass balustrade, and I need to determine the bracket spacing.

The material is white oak, and the architect would like a 1.5" diameter handrail. I have seen a bunch of these around, and the bracket spacing is approx 4ft. However, if I use the allowable stresses in the NDS of 1200psi and assume simply supported, the distance between brackets can only be 8" (!). The modulus of rupture of the material is ~15,000psi according to the supplier, so the NDS has a super conservative allowable. Is it possible to justify a simple factor of safety of 3.0 in a situation such as this?

 

[boo1]

IBC 2407.1.1 Loads. The panels and their support system shall be designed to withstand the loads specified in section 1607.7. A safety factor of 4 shall be used.

 

[glass99]

boo1 - the IBC 2407.1.1 provision you are looking at is under glass and glazing, and is actually a FOS = 3.0 because there is an allowable stress increase of 1/3 for handrail loads. Normally handrails are stainless steel, and are designed for the steel factors of safety, which is approx 1.67. -> my assumption is that I should design the handrail to be in accordance with the timber code, though it gives ridiculous results. My other hesitation in using the glass FOS for timber is that the modulus of rupture value for timber is a function of specimen size, so is not straightforward.

I checked in with two timber handrail suppliers, and they typically place the handrail brackets at 4'-0" O.C. for 1.5" handrails. All I need now is a calculation to justify it!

 

[Splitrings]

It appears you are mixing ASD and LRFD design methodologies. NDS has LRFD values listed now.

 

[Splitrings]

I just looked up the modulus of rupture of white oak and it is listed as 15,200 psi at 12% moisture. The test specimen is a "small clear specimen". Obviously this is an ultimate strength. I believe NDS lists ASD values for the 95 percentile. That is why the values are so low.

 

[Splitrings]

And we can't forget the NDS values are for "visually graded" lumber!

 

[glass99]

Splitrings - thanks for the NDS insight.

Am I going to get significantly different results if I do the calculation in LRFD rather than ASD? If I use the ASD numbers, the factor of safety between the average modulus of rupture and the allowable is 15200 / 1200 = 12.7! The 95th percentile is probably the issue as you say - is there a way of avoiding it? The built handrail material will be similar to the "small clear sample" being that its not that large of an object and it won't have knots.

 

[Splitrings]

I quickly glanced at the LRFD bending stress for oak and it was about 3800 psi. The other issue I see is the listed values are for 2" and wider sections not round handrail. You might look to see if there is a stress rated handrail product out there. If you are forced to use visually graded allowable design values there are going to be low.

 

[glass99]

Lets say I could somehow make an argument that visually graded stresses are excessively conservative for a handrail with no knots - what would be the NDS required factor of safety? Is it 2.25 on the 15200 psi? Convert to an LRFD equivalent if you want.

 

[Splitrings]

(LRFD )NDS lists a conversion factor K, which is equal to 2.16/theta for Fb. Theta for bending is 0.85. They also list a time effect factor of 1.25 for impact. The 1.25 is not to be used for connection design. The time effect factor for other than impact is less than or equal to 1. The load combinations are what are typically seen for LRFD; 1.2D + 1.6L +....

I am not as familar with the IBC. They may require different factors for the load combinations.

 

[glass99]

Handrail loads are live loads, so would be 1.6X. I am going to assume its not an impact load.

So the total factor of safety is 1.6 * 2.16 / 0.85 = 4.07?

 

[Splitrings]

The K factor simply adjusts the ASD design values to LRFD design values. Your ultimate safety factor would be the ultimate stress level (15,200 psi - if this is truely representative for the handrail you are using) divided your service load stress level. This may not meet code requirements. Assuming every piece of handrail you get will have an ultimate stress level of 15,200 would be pretty risky IMO. Wood is highly variable.

You might check out ASTM D5457. This standard allows calulation of LRFD design values by a reliability normalizing method.

 

[LorenAWC]

I don't think this is a ASD vs. LRFD issue but a visually graded lumber vs. milled piece issue. Yes, the value from the NDS is taken as a clear specimen for white oak, but then it is reduced to compensate for defects, grade, and general MOV including the 95% offset to get a design value.

The question for you is if the 15,200 psi is an average or a design value. If it is an average, I wouldn't use it.

When railings for AWC's Residential Deck Construction Guide were being tested, a factor of safety of 2.5 was used. Also, a load duration factor of 1.6 is included.

 

[msquared48]

You guys need to get a grip on things here - other than the handrail.

Mike McCann
MMC Engineering