Fastening to LVL Below the Neutral Axis 7

[XR250]

It is my understanding that you were not supposed to attach a vertical load to an LVL below the neutral axis. I have a situation where I have an LVL (4) 1 3/4x18 supporting framing on top and 2 ft. of brick on the face with a lintel. If I use a an L5x3 1/2x1/4 lintel fastened at 16 inches on center, the shear load is 90 lbs. per connection. The fasteners would be located 3 3/4" from the bottom of the beam. Is this enough to worry about or should I weld some tabs to the lintel to get the bolts above the neutral axis?

 

[dik]

Thanks

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Eng16080]

XR250,
For some alternative ideas:
[ol 1]
[li] Could you fasten a sheet of plywood or OSB over the face of the LVL beam and then attach the angle to that? This would be detailed to transfer the load from the angle to the ply/OSB and from the ply/OSB to a location above the neutral axis on the LVL. There's hardly any load on this thing. I would think nails spaced on a grid at 6" or maybe even 12" would work.[/li]
[li] Same idea as above, except if the extra thickness of plywood or OSB is an issue, then use flat metal straps oriented vertically every 12" or 24" (whatever it needs to be) to transfer the load above the n.a.[/li]
[li] If the header beam happens to be oversized for the load from above (excluding the angle and brick), could you analyze it as if there's a 14" LVL above (for example) just supporting the load on top, and then a separate 4" LVL below only supporting the angle and brick? It would be as if there are two stacked non-connected beams. If that works by analysis, I'm not sure I would be too worried about tension perp. to grain since the load on the lower 4" beam would be above the n.a.[/li]
[/ol]

 

[jerseyshore]

Craig, what is cross-banded LVL? Never heard of it.

I feel like we're discussing two similar, but somewhat different situations. We have hanging a load from the actual bottom of a beam or supporting something below the neutral axis, but with little to no eccentricity vs a situation like OP has where you can induce real torsion on a member because of an eccentric load. Both perp to grain tension, but different axes.

I think what JAE says above covers that first part perfectly about the concerns perp to grain for glulams and sawn lumber. For the second situation, it's still perp to grain, but on the other axis where LVL's are probably more susceptible to issues because of how they are manufactured.

Now XR250 doesn't have a very heavy load here, but I'd still want to see some bracing or blocking behind just to prevent that beam from rolling over time.

 

[Craig_H]

what is cross-banded LVL? Never heard of it.

Neither have I. That paper is partially written from a New Zealand perspective, maybe someone on here from that neck of the woods can comment. I suspect that most of the LVL in the North American Market is not cross-banded. The referenced paper indicates that standard LVL is more susceptible to splitting under tension perpendicular to the grain. I was surprised by that.

 

[ChorasDen]

Lots of posts here to catch-up on, I have a bit of reading to do, but addressing the question in the most recent posts (I have not read the paper to confirm) but cross banded may refer to LVL that has several veneers turned 90 degrees during manufacturing. There is at least one manufacturer in North American that I know of that does this, often for rim board applications. LVL generally makes a poor rim board due to dimension stability with moisture, but turning the veneer is one method I've seen used to try and address this concern. The loss in strength is less of an issue in this use because rim board in usually resting on plates below.

 

[XR250]

Now XR250 doesn't have a very heavy load here, but I'd still want to see some bracing or blocking behind just to prevent that beam from rolling over time

It's a window header. How would you go about that without having the architect fire you?
In my mind, the composite action of the lintel bolted to the fat LVL gives pretty good resistance to a weak axis couple.

 

[jerseyshore]

How big a window are we talking about? You can't make it a loose lintel?

 

[jerseyshore]

And to answer your question, if it's a header like this, I may try to increase the depth to make it partially flush and/or use a larger vertical leg on the angle so that the bottom of the header doesn't have to be flush with bottom of angle. Can gain back a few inches with these options to give it a bit of bracing from the joists.

 

[XR250]

20 ft

 

[DoubleStud]

Between the plywood nailed at the edges, anchor ties for the veneer, and the arching action by the veneer, I have not worried about it... or shall I? I think people have been doing this forever to support veneer above windows. I probably follow that notes for other applications though.

 

[jerseyshore]

People have been doing it forever, definitely. And brick veneer on residential is not very popular in NJ, but when I do see it, quite often there are cracks over openings because of shrinkage and/ or bad lintels. It's something I hate seeing so I try to do everything I can to avoid it if possible.

One thing I saw on a multi-story building with this exact situation (long wood headers supporting the brick angles) was that within a few months after the brick was installed, the combo of engineered lumber headers & posts vs dimensional lumber everywhere else caused some shrinkage issues at the jambs, cracking the brick pretty bad. Obviously when you have a multi-story building those issues add up the higher it is, but always something I keep in mind working on a project with brick & wood.

 

[Euler07]

I haven't read the whole thread but I find this NDS quote snippet particularly amusing:
"Designs that induce tension stress perpendicular to grain shall be avoided whenever possible…".
Obviously, any load applied to any member will induce tension stress perpendicular to the grain. This is basic Mohr's circle theory from university. This is why a timber post loaded in compression will have splitting along the line of the grain (perpendicular to the direction of load).

Personally, I would never support brick using timber since the timber will shrink/creep over time and this will eventually crack the brick. But no one supports brick on timber around here so it's not difficult to convince the builder not to do it.

Having said that, I can't imagine that it's an all or nothing depending on whether the load is above or below the neutral axis. Surely it's more of a linear relationship depending on how high up the load is. Therefore, you're probably better off going back to first principles and calculating the actual perpendicular stress capacity of the beam for the proposed fixing location.

 

[Euler07]

In addition to the above post and as an interesting side note, Mohr's circle theory is also why a post loaded in compression may fail in diagonal shear as this is the weakest plane for the perpendicular/parallel stress relationship.

 

[RipPECity]

NDS commentary 12.5.1 specifically defines the "light load condition" for below the neutral axis in the tension zone as 100lbs at 24" o.c. min.

It's interesting they define this only in the commentary.

 

[dik]

... is less than 3 psi tension for 2" dimensioned lumber. Doesn't make any sense... best to know the actual failure and stress modes. More testing has to be done.


-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Smoulder]

Neutral axis requirement is related to fravture, no? Like how notches on the compression side are only checked for reduced section because the stress field doesn't promote fracture.

 

[dik]

I haven't looked at it, and I will when I get time, is to check Mohr's circle using the tension parallel to the grain and the tension perpendicular to the grain to see what the combined effects are.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[phamENG]

Is Mohr's circle applicable here? Isn't it intended for analysis of a finite element of an isotropic material? It may be fine for judging a finite piece of wood, but off hand it seems it would break down at grain boundaries which is our primary area of concern.

dik, your 3psi assumes a uniform stress distribution, right? My guess is that you'd have a significantly higher peak stress at point of load tapering off on either side.

 

[dik]

so, it's locally 20psi... it's still very small. I replied badly to that and apologise. If you have a 1/2" dia bolt in 1-1/2" dimensioned lumber that's a stress of 133 psi perp to the grain... which is in order. This could give rise to local tensile stresses of 200psi, or 300psi at the edge of the hole (I don't know what the stress distribution is). It doesn't matter where the bolt is located, above or below the NA. It's necessary to know what the actual interaction at the fastener hole is.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[phamENG]

I think the location relative to the neutral axis is meant to be a good rule of thumb based on relative stiffness. The issue with tension perpendicular isn't that it doesn't exist, it's that it's low and variable. Enough so that we can't reliably quantify it. I would guess that the variation is enough to drive the low end with safety factor into the negligible digits (I haven't looked for actual data on it - I have too much to do and usually start at the assumption that the code writers didn't just pull it out of their arses).

So if you go into it assuming you have a failure plane parallel to the grain through the point of load application, placing it above the neutral axis makes the 'loaded' side of the split the stiffer side. If you place it below, you're more likely to be dependent on the tension perpendicular to engage enough of the section to resist the load locally. So a good detailing practice more than a truly quantifiable check.