switch back stair analysis/design 1

[matrixeric]

Dear All:
I am a structural engineer who has never done stair analysis/design ...for one of the projects I am currently working on the client wants us to design a "switchback" stair.The spans are such that deflections are becoming a big head ache esp when I apply the uniform live load on one flight of stairs only and nothing on the other flight or landing ...Upon searching IBC etc I got the design loads to be the worst effect of a uniform load of 100 psf or a conc. load of 300 lbs; but these do not tell me if I can assume a percentage of this 100psf load to help with the deflections esp. when I apply the load ONLY on one flight of stairs in the unbalanced load case....
Please let me know if there is a reference/code out there that addresses this issue. Any advice would greatly help me in designing this stair.

Thanks.
MatrixEric

 

[jike]

Are you doing a stair where the intermediate landing is unsupported?

 

[Lion06]

I have never designed stairs, but have checked several sets of calcs that have come in for jobs we are doing. I have never seen a stair stringer treated as a continuous span (as you are suggesting). Maybe this is a function of the switchback stair (I have never heard of this). If, however, you are designing continuous spans I have to think you must consider skip loading to provide the most critical conditions similar to any continuous beam.

 

[UcfSE]

What member is seeing the unacceptable deflections, stringers or column? How are you modeling the stringer connection at the top (to a slab or deck angle?) and at the bottom?

 

[matrixeric]

I will try to answer all of your questions at the same time:
Jike: I have one column in the centre of the intermediate landing which is my only support in addition to the ends of the stringers at the foundations and deck connection.
StructuralEIT: You are correct. The geometry with which I am dealing with meant that I need to have the stringers analyzed as continuous to make it work.
UcfSE: It is the stringer that has deflection problems. This being a AESS stair we are trying to limit the deflections to a 0.5"(live load deflection) at this point of time. Do not know if the constraints given to us will allow that.I am assuming stringers to be assumed fixed at the bottom(fdn.) and pinned at the level where it is connected to the deck.

Just so that it is easy to visualize, can some one let me know a way to post an image/pdf out here.

 

[jike]

If you have a centered column at the intermediate landing, you should be able to reasonably control deflections. This column should be a T support with a fixed base.

It is advisable to have the full LL on one side to obtain the worst possible situation. It is not a common situation, but I can envision the possibility of a group of people getting on the stairs to take a group photo.

 

[UcfSE]

Click on "Process TGML" right above the "submit post" button. Scroll down to the image instructions.

 

[wikidcool]

AND...You'll have to upload that pic to an image hosting site (like

http://www.photobucket.com

for example), to get a tag like the one referenced in the "Process TGML" instructions.

 

[Lion06]

Alright, the stair I am envisioning has an intermediate landing and a landing at the actual floor level. The stringer is running up to the intermediate landing then reverses directions to run up to the actual floor level such that the intermediate landings are over top of each other and the actual floor landings are over top of each other (but the intermediate landings are not over top of the actual floor landings).

------- actual floor
-
-
-
-
------ intermediate (pin support)
-
-
-
-
------- actual floor (fixed support)

With this being the case, and you designing for fixed at base (foundation connection) and pinned at the landing I am failing to see how this would be considered a continuous beam.

 

[jike]

The stair would have a pinned base (not fixed) and each stringer would be designed as a simple beam spanning from the floor to the end of the landing (cantilevering past the T shaped column support depending its exact location).

 

[Lion06]

jike-
I agree, but it sounds from the OP that these stringers are being designed as continuous for more than just the intermediate support. Continuous over the intermediate support still means it is supporting only one flight of stairs, but the OP says that one flight of stairs (as the unbalanced case) is the worst.

 

[jike]

I suspect that the unbalanced LL is for one flight to obtain the maximum deflection at the intermediate landing and the maximum bending in the T support column.

 

[Lion06]

matrixeric-
Can you shed some light on this?

 

[DarthSoilsGuy]

To start off, i've never designed stairs or thought about doing it. but i've inspected the welds and bolts on a bunch of them in NC and VA. I'll tell you what i've seen go in public bldgs.

very generally speaking: the base is pinned to the slab with anchors. The landing is supported by posts (also anchored to slab) and is anchored to the wall, usually grouted(or ungrouted) CMU wall. The 2nd floor landing edge angle is welded to the stair stringers in some fashion after that. i have seen large bldgs where the stringers were also anchored to the CMU walls at every other riser. i have seen large bldgs where the landings are CIP conc. keyed/reinforced into the CIP walls and the stringers span w/o anchoring to walls. Couldn't tell you about span lengths though.

The other comment i have is that the vast majority of stair welds/bolts that i've inspected have been approved submittals from the fabricator and not designed by the structural engineer or the SE company.

 

[matrixeric]

thanks to wikidcool/UcfSE i learnt how to post images ...
Please see links below for a rendering and stem model of the stairs

hopefull the sketches answer why I am doing what I am doing ... please note the location of the only T column support at the landing

 

[Lion06]

matrixeric-
are you concerned about deflection of the member supporting the landing?

 

[jike]

I assume that you are having trouble controlling the deflection at N19 and N5. Is this correct?

I would suggest using a tubes for both the column and cross piece. The column size has a great influence on the deflection out on the end of the landing. The cross piece of the T has to resist torsion and I would recommend a tube to minimize torsional rotation which should also help!

I would also recommend a continuous member from N17 to N3.

The connection at the bottom of the stair stringers should be pinned.

 

[matrixeric]

thanks for input guys ... keep it coming ...

Jike:
You are spot on 5 and 19 being the nodes ...your point of having a tube column is well taken ... but with the size limitations I have for the column I may not get a tube section to work ... i sure will try ...i will make those WF beams of the T column to a tube ... i do have one member from N17 to N3 ... However, i still do not understand why you are particular about the bottom of the stringer to be pinned... If i provide a base plate and weld the stringer all around and this sits on a pier ... I think I am ok with assuming it to be fixed ...

Coming back to my original question - Any suggestions on the deflections limits? Also, what would be the "l" in a "l/X00" deflection limit when I have a continous stringer like this case ...for example stringer is supported at 16 and is cont. thru 17,19 and is supported at 15 ... shoudl I be using the length 16-15?

 

[jike]

It is not normal to have a fixed base on a stairs. Are you providing a footing that will minimize rotation? This can be done but is not the usual practice. I did it once with a concrete stairs cantilevered out of the ground and then turned 90 degrees up another flight and I convinced the Architect that we could do this without a post. He located a reflecting pool underneath and it looked pretty sharp.

I wouldn't worry too much about the proper L/xxx value but more about the absolute value of deflection at the corner. I am assuming that you don't have glass handrail. Can you control the deflection to 3/8 to 1/2"? The handrail will probably help stiffen the corner even though you have not taken it into account.

 

[csd72]

matrixeric,

In order to get a truly fixed connection you need slip critical bolts and this is not easily achievable at a base plate. At best this is partially fixed.

But this is meaningless as the upper span appears to be longer and it is pinned.

I think you have a scheme problem here rather than an individual member issue.

I would suggest:
1. Move W beam at top of column around 90 degrees so it cantilevers out to the edge stringers. Leave column oriented as it is and provide fixed connection at the top.
2. cantilever the outside stringers past the end of this.
3. run the inside stringers into the side of this.
4. other 2 landing beams are only there to support the edge of the landing.

I think you will remove your deflection issues then.

csd