curved stair stringers

huangyhg

curved stair stringers
hello, thank you for sharing your knowledge and experience!
i'm looking at stair stringer design for a curved stair, centerline radius 14'.
q1 - what are the design parameters, allowable bending, slenderness ratios, torsional properties, for a using a plate stringer rather than an mc12.  (i have the naamm steel stairs manual already.)
q2 - lateral bracing.  i agree that the risers stiffen the system to provide lateral bracing for regular bending, even though the compression flange (mc12x10.6) is still 'flying' in the breeze.  for torsion, it looks like the loading eccentricity is relative to the stringer supports (10 to 15 feet) and the torsional supports are the riser attachments (1.5'). so i get torsion based on 1.5' eccentricity between 10' stair supports and apply that torque value over the 18" distance between riser connections.
the owner doesn't want to use a closed section (add 3/16" plate, go to ts12x2x3/16).
find a job or post a job opening
you have to find someone with 3 dimensional analysis software, risa or something of that ilk... you can download a demo from their site and use it for analysis.
off the top, i would suggext that a mc 12 is light for the intended use and if it will support the loads, you may end up with the biggest helical spring i've ever seen.  for thee geometry, i'd be expecting an hss 4x12 type of secion.
dik
curved beams in general are not well covered by codes.  a helical beam is even worse.  so don't expect to find any kind of code or allowable stress that is specifically intended for the application.
as far as using a rectangular bar for the member, it should simplify the calcs.  not to say it will be adequate, but calcs will be simplified.  look for torsional properties in roark or other engineering references.  the 4x12 tube sounds like a promising   
best to use an hss section... the 4x12 solid would weigh about 150 lb per foot as opposed to 30 or so for the hss...  i don't know what size is required, but 4x12 seems like a good starting point...
dik
hss 4x12 seems a little heavy based on my hand calcs.
i think i'm going for a 3/16" cover plate on an mc12x10.6.
i downloaded the risa3d demo and am working on the model.
thanks for your comments.
might be a tad oversized, just a wag... but take a gander at the deflections... the 12x1.5 afforded by the mc has little torsional restraint... and may bounce like a son of a gun...
as for the model, you can do one in autocad and export a *.dxf file... kinda klunky but it sometines works...
dik
roger that!  no bouncing allowed!  i imported an acad model.  i'll let you know what i come up with.
glad to find out... have to realize that stairs often are subject to impact loading and not just the 100 psf live loading.
dik
stay away from the mc's, they are terrible to roll, the web becomes concave and the flanges go out of parallel.  plate stringers work well - develop the helix in 3-d to get the flat layout, them water jet or laser cut from plate.  the beauty of plate is that you can include in the flat layout the transitions to the upper and lower landing. boxing out mc's with a welded plate is a pain because the radius always gets altered by the welding.  trying to roll a bar into a helix is tricky; if you get a twist in the bar it is nearly impossible to correct and the bar offers no opportunity for the above mentioned tansitions.  tube stringers can be ok, but the outside and inside face of the tube rise at a different rate and you can get funny intersections - the top of the stringer does not stay flat and parallel to the plane of the floor.  also, if the stringer is supported at the upper landing and lower landing the "spring issue" is red herring if the plate is sized correctly.
imho, i would suggest that using an mc might creat a spring...a genuine concern... not a red herring... they don't have the torsional stiffness (wot makes a spring)...
in any event, rolling the hss tube will distort it somewhat.  you do, however, have a fairly large radius...
dik
thanks for the comments.  i should give you more information.  the spans are 13' or less, the curve only goes through a 90 degree arc and there is an intermediate support under the curved section.  so the torsion is indeed minimal, but still too much for an mc12 methinks (torsional constant = 0.06?? lol).
the shop said that rolling a tube was not an option.  from your comments i can see that the channel may also be a challenge.
i guess i was headed in the right direction with the plate stringer.  i'll have a look at that.