Stair Design 1

[vmirat]

I'm reviewing an exterior stair design. The rise of the stairs is 16', the run is 25', and the stair width is 3'. There is no intermediate landing, so the stringers are continuous. This is an industrial stair, so the live load is 100psf x 5 per OSHA. The fabricator wants to use C12x20.7 stringers and put an MC8x8.5 channel across the midpoint of the stringers and support the stairs at midpoint with one column.

If you load the stairs uniformly, then I don't have a problem with the single column at midpoint. However, is it reasonable to load the stairs with an unbalanced condition of one side of the stairs? This would create a moment condition between the MC channel and the column.

 

[KootK]

However, is it reasonable to load the stairs with an unbalanced condition of one side of the stairs? This would create a moment condition between the MC channel and the column.

I agree with your reasoning here.

Can you point me to a reference clause for the 100 psf x 5 requirement? That's a new one for me.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Ingenuity]

OSHA does not state 5 x 100 psf, but rather 5 x 'normal' live load anticipated, which may well be 100 psf, I guess:

§1910.24(c) "Stair strength." Fixed stairways shall be designed and constructed to carry a load of five times the normal live load anticipated but never of less strength than to carry safely a moving concentrated load of 1,000 pounds.​

 

[TLHS]

That sure is an awful and ambiguous clause!

 

[Ron]

It is reasonable to use the 1000 lb concentrated load in an eccentric condition, but not reasonable to use the 100 psf x 5 for an eccentric condition....it's tough to get 100 psf on a stair tread, much less 500 psf! Getting that load on all treads simultaneously is impractical.

 

[JStephen]

The key is "anticipated", not "5 times the normal design load used by building codes". Make a reasonable estimate of load, multiply by 5, and use that or the code-mandated loading, whichever is higher.

 

[TehMightyEngineer]

Agreed with stephen; if this were mine I would take 20 PSF as a reasonable anticipated live load and multiply it by the 5 times. I suspect the intent of that OSHA rule was to mean they want a safety factor of 5 for live load. Either way, 500 PSF is clearly unreasonable.

To your original question, I would definitely consider all load cases, including one span unloaded. Don't forget the axial load in the stair stringers as well.

Maine EIT, Civil/Structural.

 

[KootK]

Does one also then have to factor the (5 x Reasonable Estimate)?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[dcarr82775]

OSHA loves its factor of 5. I tend to use it as the load factor, but then use ultimate strengths without the phi factors.

 

[SAIL3]

As Ron pointed out the 5x or 1000# conc load is ONLY for the tread design...I usually use 100psf on the stair as awhole..the one-col support should work but I would make it a moment conn as far as the stringers are concerned...if I remember correctly there is a 12ft limit on the rise between landings unless you get a code variance...IMO using an unbalanced load on the full stairs is unreasonable, however, using a mom conn @ the col will limit differential deflection...

 

[TehMightyEngineer]

I would actually argue that a unbalanced load would be highly expected. What if a fire breaks out on the upper floor and a rush of people stampeding down the stairs occur? Impact loads plus high live loads that move down the stairs will likely cause a code level load on one span only.

Maine EIT, Civil/Structural.

 

[vmirat]

This is an industrial occupancy with an occupant load of 3. An intermediate landing is not required and the stairs will never be fully loaded.

 

[canwesteng]

Just did a back of the hand calc here, but 25'/2x1.5'x100psfx1.5/2 (half run x half stair width x load x moment arm of half stair width) yields a moment ~1.4 kip*ft. Shouldn't be exceptionally difficult to achieve that.

 

[canwesteng]

Not to mention you should have some way of resisting notional lateral loads in that direction anyway. I'd think the moment connection is done in the shop and the whole thing comes to site as one piece, just leaving you to deal with fixity at the base

 

[glass99]

It would be interesting to know the basis of the 5*100psf requirement. Perhaps the thinking was that for an industrial structure they might use a stair to move around a large piece of industrial equipment like a boiler or something.

Also: you could pin the top of the column rather than welding it. A single large diameter bolt would transfer no moment.

canwesteng: I think the unbalanced moment is along the length of the stringer not across the width of the tread, and 500psf not 100psf. Meaning the moment is approx wL^2/10 = 25ft^2 * 1.5ft * 500psf / 10 = 46.9k-ft.

 

[JLNJ]

I must say that I don't see how you get any moment in a post under the stair, no matter what the (gravity) load.

Make the load as eccentric as you see fit, it shouldn't matter.

The OP should get some helpl with this beyond what he can find on the internet.

 

[JLNJ]

My apologies, I misunderstood that the question was regarding the eccentric loading of one span when adding a post created a two-span condition. Surely if you add a single-stanchion type post on the stair centerline you will have moments in the column. You can't count on a balanced teeter-totter. Place all the live load on one side and see what you get. If the stair is stiff enough in the out-of-plane direction the small horizontal force which would cause the column moment may be resolved that way.

 

[KootK]

I believe that the stair will act as a relatively stiff diaphragm in plane/plan. That ought to take care of lateral loads and effectively prevent lateral displacement of the top of the column. Eccentric gravity loads will tend to rotate the stair. The two modes of resisting that rotation are:

1) The stringers forming an up/down couple in bending similar to warping torsion in wide flange beams.
2) Flexure in the channel/column assembly.

The degree to which flexure in the column will resist stair rotation will depend on the degree of fixity at the column/channel connection. Most common connections would transmit substantial moment, even if unintended. For that reason, I believe that virat's concern is warranted. The connection and column should both be designed for some moment.

If I understand correctly, this is the situation:

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[glass99]

KootK - nice sketch! Unbalanced load across the width of the stair would indeed be possible, though the magnitude is less than the unbalanced load along the length because the stair is relatively narrow.

 

[KootK]

Thanks Glass. Part of the reason that I posted the sketch was so that we could sort out whether we are concerned about transverse imbalance, longitudinal imbalance, or both. Based on the OP's description, I'm assuming that the detailing is similar to what I've shown, at least conceptually. If that's true, the connections between channels are likely not great transmitters of torsion and thus longitudinal imbalance would not transmit significant moment to the column.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.