W section w/ bottom plate lintel

[PG35]

I'm designing a lintel to span 28'. It supports 5' of 8" blk. centered on the beam and 5' of face brick at an eccentricity of 6". There is some roof load but very little. I want to use a W16 section w/ bottom plate.

I find that a cont. 3/8" plate is adequate to support the brick if the plate is designed to act as a cantilever (fixed end at the connection to the bottom flange). If welding is frequent, is it unreasonable to assume that this an adequate assumption.

The torsion calculation apprears to be very lengthy. I got the AISC design guide 9 for torsion. 5' of brick acting at 6" doesn't seem like it will be a big problem.

Is it normal practice to provide brg. plates with anchors at the brg. locations to help resist the torsion at the ends of the lintel? I will provide some rebar welded to the top flange at 4' o.c.

Does anyone have experience with a similar situation? I don't feel comfortabledealing with torsion.

Any help will be appreciated.

Thanks.

 

[JAE]

A wide flange will twist quite a bit (very little torsional stiffness) unless you brace it externally to some other element.

We have even gone to using tube sections with the bottom plate to increase torsional stiffness. You still have to torsionally restrict twist at the ends.

 

[jike]

As JAE states that you still have to take out the twist at the ends. You might want to consider framing into steel columns at each end.

 

[271828]

I think JAE's point is that you should calculate the twisting angle and resulting vertical brick deflection and that you'll probably find that it's no good. Folks often calculate stresses and stop there. As he hinted, a tube will work better. I'd actually be very surprised if you can get anywhere close with any W16.

 

[JAE]

Correct-o-moto 271828. I was in a hurry. Watch the twist on WF shapes as they would create high bending in the wall - and produce cracking in the masonry.

 

[271828]

Hey, that reminds me of something.

I think in some situations, the CMU can keep the beam from twisting. I usually put 6" headed studs on the top of the W-shape. With the right boundary conditions (that don't always exist), I think wall bending could provide twisting restraint.

No clue if the OP's situation does this, but it's something to think about.

 

[hokie66]

I agree with all the advice above. Wide flange is no good for torsion. You have to either get the beam centered under the load or use a closed section to resist the torsion. Any chance of using a concrete beam?

 

[JAE]

Hey - I've done that before - used a series of vertical, deformed bar anchor or weldable rebar welded to the top flange and extended up into grouted cells in the CMU.

This was an attempt to engage the wall out-of-plane stiffness. The couple formed by the rebar and the face of the block wall would tend to prohibit twisting of the beam. Hard to measure its stiffness and the resulting bend in the wall so I try to avoid having to do that.

 

[PG35]

Thank-you everyone for your responses.

The Architect prefers that W sections be used so that it can infilled with block.

I will add 16" long rebar welded to the top flange (at 32" o.c.) of the beam to restrain the twisting.

I'm concerned because the span is so great (28').

Would stiffeners at frequent intervals (40") on each side of the web work?

I can use plates on the outside to create somewhat of a box shape.

 

[PG35]

Any comments about sizing the plate? Fixed end cantilever o.k.? I'm not considering bending in the long direction.

 

[hokie66]

The plate cantilevers off the bottom flange. That will not be your problem.

Stiffeners perpendicular to the web will do little to increase the beam's torsional capacity.

Don't be too quick to jump at the advice 271828 and JAE that you may be able to depend on the wall to prevent twisting of the beam. As 271828 says, it depends on the boundary conditions. I think that would require a substantial load on the wall from above, and you don't have that. The rebar on the top flange may help somewhat to limit the lateral movement of that flange, but I think in your case it might even accentuate the lateral movement of the bottom flange. You need to consider that the Architect's preference at this stage won't help you if the beam rotates and the brick falls off.

If you can use plates to create a box shape, then I think you are home free, provided you have a way of resisting the twisting at the ends of the span. But I thought the Architect was insisting on stacking block between the flanges.

 

[csd72]

If you are stuck on the wide flange idea, how about using a channel.

The web and the shear center are closer to the load center and therefore it does not have the same level of torsion.

It is also narrower and fits into a brick width better.

I have used this detail many times and never had a problem.

 

[PMR06]

I really don't like the wide flange braced by wall idea. It is almost impossible to quantify the strength or stiffness for just the masonry gravity load case, let alone the behavior when you throw in of plane wind or suction.

I've had a similar situation to this in which I fit a tube between the exterior brick wythe and an interior face shell. A continuous plate on the bottom supported the masonry each side, and columns in the door jambs spanned the torsion between the roof and floor diaphragms. It worked great... the Architect got his face shell and I got my torsional stiffness.

 

[271828]

"Don't be too quick to jump at the advice 271828 and JAE that you may be able to depend on the wall to prevent twisting of the beam. As 271828 says, it depends on the boundary conditions."

Absolutely. I was typing about a general idea, not about the OP's particular situation. There are times when this would work fine, but it requires judgment.

Not you, but some here seem quick to dismiss an idea PERIOD if they think of specific cases when it won't work. I don't see any need for that. There are lots of ideas out there that are wrong for one case, but great for another, requiring judgment.

I'm with everybody else, though, that an HSS is preferable.

 

[csd72]

to take the twist out, continue the plate at least 6"to 8" past the edge of the opening (line it up with a grout course) this will take the torsion out by bearing on the masonry.

 

[hokie66]

Some of the ideas here work fine for shorter spans, but remember PG35 is spanning 28'. PMR06 has summed up the problem and given the best advice for this specific situation, provided you have enough width to incorporate a tube which will take the torsion. Suggest your bottom plate should take the out of plane bending.

 

[ChipB]

PG35,
Enercalc will do torsional analysis for you.

I'm assuming this beam is carrying a parapet wall. Perhaps over store frontage.

Is it possible for you to run kickers @ 5' centers from the bottom flange of the lintel beam, to the top flange of the next adjacent beam. This will place the horizontal force into the roof diaphragm, where it should cancel out with the tensile force of the diaphragm attempting to restrain the top of the lintel beam, and resolve your torsion into a force couple between the two beams, thereby taking a significant amount of torsion out of your beam.
Chip

 

[MDukh]

PG35,
By now, just Design Guide No.9 provides some recommendations to count masonry restraining ability. In turn Masonry Code and Technical Notes are shy about the case. For practical reason read answer by William Bretnall (from 11/01/97 at Steel Interchange, AISC MSC) regarding lateral masonry bracing ability to beam top flange.

ChipB advice to use kickers has a good practical reason. Our office uses the same solution (w/ 10’ kicker spacing!) for years. These kickers will take care about torsion and wind loads.

There are some other considerations:
1. By observation, to obey deflection criteria and 16” depth limit you shall use W16X89 (bf =10.365”), which mean;
- Beam/Lintel shall be shored during construction because 8” CMU backup will act eccentrically to beam;
- Beam flange will cross insulation and vapor barrier into airspace which will create problem w/ durability because of high corrosion environment and thermo bridging (see Brick Technical Notes No.47);
- On one of our projects, because of thermo bridging, architect demanded to use shelf angle instead of soffit plate to create separation between outer and interior portion of lintel;
- There will be difficulties to use regular adjustable anchors (because of space between veneer and beam web). Angles or WT placed vertically and spaced 2’-0” o.c. along beam length would help. These shapes will act as Beam/Lintel stiffeners and as support for shelf angle and insulation. CMU fill I consider as additional load and labor with no practical reason.

2. Considering all above, to decrease beam flange width you need to increase beam depth. I would still recommend, as other did, HSS section shifted 2” away from interior face of wall to provide CMU finish. I would keep bracing @ 10’ O.C. to avoid excessive Veneer rotation due to combined local soffit plate/shelf angle deflection and beam rotation between bracing points.
3. If thermo bridging is not a matter of concern, soffit plate shall be the same length as beam and continued over support for full bearing length (see Example DPC-12, Masonry Design Guide). Beam/Lintel will be supported by CMU backup and Veneer return.
4. Secure beam ends by wall reinforcement, which shall be continued thru holes burned at beam flanges and welded to top of beam.