Saw-Tooth Beam

[BSVBD]

Please see attached...

I have a saw-tooth beam to design.

I believe it is a simple-span beam, full-pen welded all-around at each joint. It will bear on bearing plate and grouted / reinforced 8" CMU each end.

Other than the moment to occur at mid-span, i do not believe there would occur any moment concerns at each joint.

Any thoughts?

Thank you all!

 

[hokie66]

The forces in the flanges require development, so you probably need stiffeners in the corners.

 

[KootK]

Any thoughts?

1) At each knuckle, you need to consider how the flange force is turning the corner. You can address this by a)providing web stiffeners to turn the flange forces around the corners or b) proving that the web alone could handle the bending stresses at those locations without the flanges helping out.

2) You'll need to calculate deflection in a manner that pays homage to the fact that deflection will be more closely related to the stretched out length of the beam rather than the clear span between supports.

3) Ensure that the framing tying into the beam will brace it torsionally. Trying to assess this beam for lateral torsional buckling would be a nightmare.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Guastavino]

KookK and Hokie66,

So I'm thinking (maybe errantly) that if you use the same size beam and full pen weld it on the miters, you are essentially developing those flange forces, albeit, you are relying on the web and flange of the adjacent leg. If you think about it (ignoring the web), the forces in the flanges produce a couple that has equal and opposite forces, so you could replace that with an equivalent moment that you transfer to the next leg. Albeit, there would def. be some localized effect.

Just thinking out loud hear. Don't take this as gospel.

 

[rb1957]

why wouldn't you make this a decorative feature (like a plaster overlay) rather than a working structural entity ?

maybe have some pegs sticking up from the brickwork and a re-bar spine that would align the feature to the building ?

another day in paradise, or is paradise one day closer ?

 

[hokie66]

njlutzwe,
Not sure I understand what you are saying, but a member just mitred and welded back together definitely does not have the same strength as the straight member or a member with stiffeners. Maybe it will still be strong enough, but that requires analysis.

 

[KootK]

In addition to the methods that I mentioned above, I do know of a third:

c) Instead of turning the flange force with a stiffener, use the transverse bending strength of the flange for this purpose. You can use an angle shaped section formed by the plates coming together which would have some strength to it.

So I'm thinking (maybe errantly) that if you use the same size beam and full pen weld it on the miters, you are essentially developing those flange forces, albeit, you are relying on the web and flange of the adjacent leg.

I think that what's missing here is a competent path for the vertical components of the flange forces as they make their way around the knuckles. The tension and compression flanges will both tend to either pull away from the web or push towards it depending on the scenario. That action needs to be resisted by stiffeners, a segment of the web acting as a column/tension tie... something.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[dhengr]

BSVBD:
I’d make that crazy canopy as follows: The front beam or fascia member might be an ST8x4 or a ST10x4 with its 8 piece’s ends plumb cut at each change in direction of the saw tooth shape. The two end cuts nearest the end bearing walls would bisect the angle change. At each of these direction changes, I’d have a .25" or .375" thk. vert. pl. a little higher than the plumb cut length on the ST. These vert. pls. would span back to the brick wall and also act as butt weld pls. to full pen. weld at each prep’ed. end of the ST pieces, for a continuous fascia member. At the brick wall I’d have .25" vert. pls. which matched the saw toothed shape and bolted to the brick wall; actually bolted through the brick and into the conc. blk. wall. These vert pls. would be capped by .25"x6 or 8" formed pls., “V” shaped t&b flgs. going back to the brick wall and .25"x4" pls. at the vert. pl. at the brick wall plane, this later forming a “C” section. The “V” shaped flg. pls. would be welded t&b to the vert. web pls. and lap over the t&b of the ST fascia member, as a reinforcement pl. at the butt end joints. I’d have this shop fab’ed., and have the fab’er. punch or weld studs on flgs. so you can install treated 2x nailers to all the flgs. for your roof and soffit sheathing application.

For analysis, take a look at cranked beam or folded plate design. Of course, you have the advantage of running this on a computer. But, as others have mentioned pay particular attention to the stresses and quality welding at each one of the change in direction joints. At each of the joints the flg. forces have components (tension or compression) in the plane of the flg. and perpendicular to the flg. as they try to move around that corner. As Koot said, they push into or pull away from the web. That’s one reason why an ST, with two webs, might be better here. The fascia member will have some lateral thrust at each of the side bearing walls. I imagine under the right conditions you could get 4-6' of snow and ice build-up on this little bugger, given its shape and side and back wall heights. Be sure to vent this canopy enclosure, and pay very special attention to the flashing at the existing walls and roof drains. Is this a South side Milw. church?

 

[Brad805]

How about concrete if your stresses are not too high? Someone will have quite a few finishing details to make a steel assembly pretty, and a precaster could set up a form and cast that will all the reveals the arch likely wants quite easily.

 

[rb1957]

well that was a dumb suggestion from me ! I only looked at the main view, and didn't see the section.

I wouldn't try to make the beam work the way being discussed. I'd cantilever the peaks and troughs from the building, and let each panel work pretty much in isolation.

another day in paradise, or is paradise one day closer ?