Stair Stringers

[broekie]

I am designing an aluminum stair stringer that extends from landing to landing. The tread for the stairs is typical aluminum bar grating. The bar grating is attached to angle which have been welded to the channel stringers. The stringers are 3' apart and are oriented as such ] [.

My question is: What should I assume for the unbraced length when designing the channel for flexure. It seems as though the angles and bar grating would help tie this thing together and reduce the unbraced length. Using the entire length of the stringer seems very conservative. If I do this, I end up with allowable stresses under 1 ksi.

Does anyone out there have any advice from what you have done in the past? Any codes (NAAMM, etc.) that could help me with this?

 

[epr]

brokie - a similar topic is under discussion in the following thread:
thread507-86850

 

[SlideRuleEra]

Interesting problem. I would say that the unbraced length is the distance between the welded angles brackets, as measured along the length of the channel (maybe about a foot). This assumes that each angle support bracket goes diagonally across the full width of the channel's web. The diagonal placement of the angle/grating should provide significant bracing. Concur with epr, check out his recommended thread. In my opinion the diagonal placement of the treads makes your's a different situation.

 

[Lutfi]

Let me warn you about using aluminum for stair stringers, it is vibrations due to its light weight. Just make sure that it is braced adequately and the span is short else you will have stair users who will freak out if it flexible. I am familiar with the type of treads you are talking about and they will not provide adequate bracing alone!

Strength is not everything! Human comfort is a serious consideration.

Good luck

 

[IFRs]

To get an allowable stress of 1ksi you must have a very long stringer. I have made many aluminum stringers, both of flat bar and channel. The treads do add a measure of lateral support and stabilize the stringer. I ususally analyze them using the entire length as the unbraced length. Your situation sounds unusual - with a very long span. Aluminum may still be a good choice but I would try some full scale testing and also check the deflections. If you use the distance between treads as the unbraced length, also check the entire structure full length for problems.

 

[EIT2]

My understanding is that you must brace, or come in direct contact with the top/compression flange. Coming near the top flange does not qualify. Right or wrong?

I use the entire length of stringers for my unbraced length. But, I design steel stringers so this is seldom a problem.

When considering deflection in a stair, I use a minimum of l/480. But to allow this in a long stringer would result in excessive deflection, bounce, and nervous people. I like to limit stringer deflection to 1/2".

Why aluminum? Corrosive environment?

 

[broekie]

Thanks for your responses. Tee length of the stringer is approximately 24'. Aluminum was chosen by the owner because of the corrosive environment (WWTP). I am using a 12" channel for the stringer and have a deflection of approximately L/800.

 

[IFRs]

When considering "bracing", I look at two modes of failure and bracing.

If a structural shape is to fail due to compressive loading member will fail by bowing laterally, then I want the bracing to keep the channel from bowing. In this case, I don't think you need intimate contact with the top flange.

If you are analyzing the top flange in compression and determine that as an outstanding leg it will itself buckle, leading to a more global failure of the beam, then I think intimate contact is needed.

 

[broekie]

Thanks again for everyone's input. Our office (and this forum) was a little divided on what to use for the unbraced length. The architect allowed us to put another column in the stairwell, reducing the moment by a factor of 4 or so. Now the aluminum channel works without a problem, even assuming the compression flange is unbraced along its entire length.