Torsion on LVL beam supporting brick 2

[structurebeton]

I recently did an inspection on an 18 ft span beam lintel, most likely made of an LVL beam (Drywall is still in place). The beam is supporting half of a 2nd story, roof and brick. The brick is supported on a steel angle which in turn is bolted to the LVL beam. The beam is currently at a deflection of L/300 and the brick facade has cracked. There is also slight twist of the steel angle.

The questions are:
1) Is it allowed by code to support brick with a Pre-engineered wood beam? (Despite the fact that the criteria of L/600 deflection is respected). I could not locate on IBC2006 were this is addressed.

2) Do you have a good reference on calculating strength of a wood beam due to excentric loading, such as brick?

3) In repairing the beam, should we simply replace the wood with a steel tube beam, which then would require steel columns...

Any comments would be appreciated. Thanks,

ST

 

[KBVT]

No, brick can not be supported by a combustible material (including wood). This was often done at setback gable roofs where the lintel was bolted to the bottom chord of the gable truss. This violates the IBC.

 

[IceNine]

R703.7.3 Masonry veneer above openings shall be non-combustible

 

[structurebeton]

So, no LVL beam is allowed at all on residential, to support brick? I suppose the steel angle would have to be designed so it is self supporting (i.e. not attached to the LVL)????
A fire treated wood would then be allowed?

 

[DaveAtkins]

In response to your question 2, I have designed LVL lintels to support brick veneer. More correctly, I have used PSL lintels, because they come in a 5.25" width.

The torsion on a rectangular shape = cM/At, where c is a coefficient which depends on the aspect ratio of the rectangle, M is the torsional moment, A is the area of the rectangle, and t is the length of the smaller side of the rectangle. This shear stress is added to the shear stress due to direct shear, and the total shear stress is compared to the allowable shear stress (for example, 290 psi for PSL).

But I have found the difficult part of designing a header for torsion is the end connection. Usually, several lag screws are required to transfer the torsion from the header to the king stud.

DaveAtkins

 

[hokie66]

Dave has nailed the answer. Look closely at the end connection. The torsion has to be developed.

 

[structurebeton]

So the question is, from a Code point of view, is it allowed to use a PSL? How did you justify using a PSL vs the fact that combustible material is not allowed as lintel material?

 

[LobstaEata]

There is nothing in the IBC that I can see that prevents masonry from being supported on a steel ledge angle which is bolted to an LVL beam. (except fireplace mansonry) Solid blocking at the ends of an LVL lintel should be provided regardless of whether the LVL is eccentrically loaded or not. Torsional restraint at the ends of beam can be easily resolved into a force couple at the top and bottom of beam. Blocking may be evaluated based on the calculated restraint force.

One caveat is that there are LVL manufacturer recommendations that there be a moisture barrier between the masonry mortar and LVL. I suppose this is to prevent degradation of wood fiber and lamination bond at the point of bearing.

Hope this is helpful.

 

[structurebeton]

Lobstateata, that is what I was inclined in believing, but I could never find anyone to point out clearly an answer.

So, do you all know how to evaluate the torsional deflection resulting on a wood or PSL beam?

 

[LobstaEata]

By torsional deflection, I assume you are referring to the resulting rotation at mid-span? Torsional stresses due to warping requires information about the full length of the member, the boundary conditions on the member and the variation of loading along the length of the member, as outlined by Dave above.


Try the following link for a basic treatise on the subject.

http://www.ae.msstate.edu/~masoud/Teaching/SA2/chA6.4_text.html

 

[JAE]

Wait a minute - keep in mind that the IBC doesn't apply here.

This is a residential house - therefore IRC applies. Also, the IRC allows brick veneer to be supported on steel angles attached to wood studs for instance (see Figure R703.7.2.1).

Supporting from a PSL (LVL or other) composite lumber should be OK provided the following:

1. The wood beam is designed for long term deflections and limited to L/600 or 0.3" (per ACI 530).

2. The brick sits directly on a steel angle and not on the wood itself.

3. The resulting torsion is taken into account (per DaveAtkins above)...and also considered for the design of the supporting end posts.

 

[LobstaEata]

Quite right about using IRC JAE. My mistake.

 

[hokie66]

structurebeton,

In trying to calculate the expected rotation of the beam, remember that it is a summation of the amount the beam twists between the supports and the amount the ends rotate due to bending of the supporting posts. You can probably get a good idea of which component is most important by observing the way the brick is cracked.

 

[structurebeton]

There is one vertical crack right in the middle of the span, and one diagonal crack on the right corner, and another diagonal crack on the left corner, higher on the wall. It clearly shows a failure due to bending of the beam, but it could also show failure due to torsional stress.

 

[hokie66]

My guess is the supports have rotated. Be careful, because if that is the issue, the whole thing could suddenly collapse.

 

[structurebeton]

Attached are the pictures of the cracks...I didn't feel they were wide enough to be classified as an imminent catastrophic failure.

 

[hokie66]

Thanks for the photos. I would tend to agree that it is not about to fall down. Probably has somehow achieved equilibrium.

I think you have to now remove the interior finish from the beam enough to expose the beam section, the end connections, and the connections of the shelf angle.

You said the deflection is L/300, so it looks like that is around 20 mm. Certainly enough to cause the cracking without any accompanying rotation, as there are no vertical control joints.

Part of the shelf angle deflection could be differential to the beam deflection, depending on the connection between the two.

 

[msquared48]

Rather than taking out the torsion at the end of the beam, if it is possible, I would remove the torsional rotation with diagonal struts back to the floor diaphragm from the top and bottom of the beam at regular ingervals along the length. This could be accomplished with tension strap connector ties and diagonal compression members.

And, yes, I have a 24" deep brick fascade over my garage doors that is supported on 12" deep wood headers and 4x4 steel angles, but no kickers. I have no structural problems, but the span is only 8 feet too.

Mike McCann
MMC Engineering

 

[structurebeton]

I don't understand how or where you would use the tension straps?

 

[msquared48]

Depending on the details of the section involved, the top of the beam wants to rotate outward from the structure with the torsion. The tension link would come in restraining this top rotation with a connection to the top of the beam.

Mike McCann
MMC Engineering