Out of plane wind on multi ply wood headers 2

A similar-ish discussion I posted that you may find some merit in

This topic is one of the banes of my engineering existence
The first question that always comes up is, how much wind load out-of-plane is actually taken by the lintel
I've spoken to many other engineers, researched our code basis, looked for research papers, spoken to glazing designers etc this seems to have a wide variance with no consensus
The best I've been able to come up with is that the lintel should, by engineering theory, take tributary loads in most situations but that some aspect of the real-world performance means that the loads we anticipate do not materialise

The second question is, as you've posed, how much composite action to take
Personally, I've always taken composite action but, running the shear flow numbers, the standard nailing patterns (something like 3 nails at 300crs) won't give sufficient capacity generally
My colleague here who I have discussed this with typically just takes a single member and doubles the capacity if he has a double lintel (so not composite action, just load sharing)
The problem with this is that timber lintel sizes get a bit ridiculous compared to what in-situ performance has shown works...and that's if you allow composite action...if you just take load-sharing, they get even more stupid and disconnected from reality

I don't know how you would expect to develop composite action in the weak axis, do you plan to epoxy each ply together? If considering a multi-ply header, I would add together the strength of each ply, as basically no composite action is going to be achieved by standard mechanical fasteners, and detail my performance checks on each ply acting individually. Keep in mind that wood shrinks/swells based on seasons due to ambient moisture and air temperature, and that causes expansion/contraction around the fastener dowels in the wood, which prevents a completely rigid, solid connection.

If the multi-ply assembly fails, I would suggest looking at a thicker section as you've indicated: glulam, wider LVL, PSL, or LSL would be good options to consider.

Honestly, I have done it on a few jobs but not typically as it is an unlikely point of failure IMHO. Also, my competitors don't so there is that

For windows over 6 feet I specify the sill plates and the connections to the king studs. Why 6 feet, just seems that after that width I get worried about it. Often a double 2x6 sill between the top of window and bottom of header will do what's required. Slap a Simpson Strong-Tie A35 angle at each end to the stud pack and everything seems to work. The largest opening I did was around 16 feet. I had to use muli-ply LVL sills.

Getting composite action with just the mechanical fasteners, would be a matter of having adequate fastener shear capacity, and enough deflection to engage that capacity. So, if you put enough nails in near the ends of the lintel, and bend it far enough, you could get composite action. I shudder to think what that would do to the finish materials, though.

Most likely, the satisfactory performance, despite inadequate strength 'by the numbers', is due to a combination of unanticipated restraint from the connected members, but mostly because the wind pressures rarely, if ever, gets anywhere close to the design value, since the wind pressure increases by the square of the wind speed, so if it's designed for 100mph wind, and the wind doesn't get over 70mph, the wind pressure never gets over half of the design pressure. Not to mention the considerable conservatism in calculating the wind pressure based on worst-case assumptions of how the geometry affect air flow, stagnation pressure, etc.

Rod Smith, P.E., The artist formerly known as HotRod10

To BridgeSmith's point, I'm okay with calculating weak axis bending assuming shear flow through the nails for full design loading - we don't care about finishes at that point so long as the wall doesn't cave in and cause a pressure transient that blows the lid off the house (or collapse the roof if it's a bearing wall).

For serviceability, using 10 year MRI, I would either ignore the shear flow and calculate deflection of a single ply taking its share of the load or specify adhesives between plies.

For serviceability, using 10 year MRI, I would either ignore the shear flow and calculate deflection of a single ply taking its share of the load or specify adhesives between plies.

Click to expand...

I'd be fairly particular that it needs to be a rigid adhesive, such as a PVA wood glue or a polyurethane glue. Construction adhesives, etc. are too flexible to provide any meaningful composite action at small deflections.

Rod Smith, P.E., The artist formerly known as HotRod10

In most cases the width of our built members does not equal our wall width, so we normally have a plate above and below for finishes like the sketch below. I guess the question I have here is, is this a problem? Have we noted problems that need attention?

I'm all about:

1) Extending any intermediate posts supporting the header up through the header to break up the lateral span

2) Having top and bottom plates there to do the work of resisting lateral loads because:

a) They are naturally the full width of the wall.

b) Their connections to the wind posts are (or can be) better than the header's.

c) You may well need them for LTB bracing of the header anyhow.

d) I get a lot of push back on attempts to use solid pieces rather than multi-ply given that my competitors don't do that.

I feel as though #1 and #2 have almost become forgotten wisdom. Wisdom that was prevalent before engineers got involved but, then, somehow got misplaced once we started trying to value engineer everything.

Brad805 said:

I guess the question I have here is, is this a problem? Have we noted problems that need attention?

Click to expand...

Probably not. But, then, I know that you know how this works:

3) There is now pressure on many projects to eliminate the plates to accommodate taller windows.

4) Our codes and standards are written in a way that suggests that an engineer ought to numerically assess the adequacy of all things. We're missing the requisite "don't worry about shit that hasn't been falling down" clause. Perhaps there out to be such a clause but, until there is, this kind of thing tends to create confusion within the design community.

The trouble with the shear connections required to make a multiply beam composite are:

5) Testing as shown us that slip in these fasteners tends to neuter the composite behavior.

6) Better fastening, like SDS screws installed on an angle, are cost prohibitive.

My understanding is that other parts of the world currently have provisions for working out composite action of dowel fasteners with efficiency factors included to deal with slip. I'd like to see that stuff make its way over to North America so that we can have a tool for use in dealing with issues like this.

My understanding is that other parts of the world currently have provisions for working out composite action of dowel fasteners with efficiency factors included to deal with slip. I'd like to see that stuff make its way over to North America so that we can have a tool for use in dealing with issues like this.

Click to expand...

I suspect that the current labor to material cost ratio in North America makes it more economical in most applications to use larger members rather than paying the labor costs to attach them sufficiently to get significant composite behavior.

We have certainly seen a shift toward relatively cheaper materials and more expensive labor in the US. Bridge girder fabrication is a great example. Years ago, it was common to have thin webs with vertical stiffeners every few feet. Today, all that welding would cost a fortune, so we just make the webs thicker and eliminate the stiffeners. It's more material, but cheaper overall cost. Other places in the world, labor is relatively much cheaper.

Rod Smith, P.E., The artist formerly known as HotRod10

Thanks for the inputs. I could not find a single worked example or reference to checking oop loads. It sounds like it's not generally checked, except maybe in longer spans and even then it sounds sporadic but it also doesn't seem to be causing any problems. Forte does allow for lateral loads on a header but if you need 2 or more plies it craps out and gives you a message that wind checks aren't performed on more than 1 ply (i'm surprised it doesn't just add up the contribution of each ply). I'm looking at a 16ft span and to get plates to work it'd be multi ply lvl plates, just as jayrod had noted. I went back and checked several shop drawing packages from cf steel jobs and they all include a check for oop loads on the headers - i'd guess that traditionally wood jobs had smaller spans and it was never an issue but by the time cf came around bigger windows were the norm so it got baked into the routine checks.

bookowski said:

i'd guess that traditionally wood jobs had smaller spans and it was never an issue but by the time cf came around bigger windows were the norm so it got baked into the routine checks.

Click to expand...

I think this is a big part of it. And a big part of why we all need to be careful working on wood structures (especially residential) as architectural tastes change. There's a lot of "we've always done it that way" out there and the residential world has more than its fair share of it. But we haven't always done it this way. We're only a couple decades out from really chopped up interior spaces with lots of bearing walls and relatively short spans (or at least extra redundancy). As designs change, the way we build them needs to change to ensure they will perform.

This hasn't been a big issue, but I think that's mainly because 1) design level wind events are extremely rare and 2) if a design level wind event happens, chances are the rest of the structure is going to be gone or so severely damaged that the inadequately fastened multi-ply header is just another piece of rubble to be cleared away. As more and more of the housing stock is replaced or retrofitted to resist higher wind loads, I think some of these smaller details will begin to be the issues that bubble up. And that's a good thing - it'll mean safer structures overall.

16'-20' spans are common in our area with trib loads btw 20'-40'. Below is an example of the type of thing that pops up in our area. I am not sure the engineers designing these have ever thought about out of plane loads on their headers. I am reluctant to add more details until we have some test data or a guideline from APA or the CWC.

Brad - That'll be an overhead door in guides attached to the jambs, thus placing the vast majority of the load on the jambs and not on the header, correct?

Brad -

I see many large openings where it does not appear to be considered as well. Although your picture above concerns me much less than some. In that case the header may see very little out-of-plane loads - assuming an OH door is going into that space. As long as the door spans primarily horizontal most of the door wind will be carried in the jambs. I will say that looks like a pretty tall pole barn with what looks like post boots at the base of the posts instead of embedded posts - I hope someone looked at that one prior to erection.

I do tend to look at out of plane loads on headers, yet try to be reasonable with assumptions about composite action. I am sometimes a bit less conservative when it comes to deflection checks on such things and allow a bit more "partial composite" action, but try to get stress calcs to work out with individual plys.

edit - @phamENG you beat me to it..

By the way - This issue (composite or partially composite action of multi-ply wood members) is not limited to this header condition. Double or triple top plates where roof members do not match the spacing of wall members - say trusses @ 24" o.c. and studs @ 16" o.c. - is another time where this comes up from time to time.

Pham, yes, that would be an overhead door and the header would not see a large out of plane load. We have seen them in many different arrangements. If you wanted a 12'-0" wide window at the main floor level they would not hesitate. That was just a quick image I found online. I do not get involved in pole bldgs, but I have to deal with framers that do. I am not suggesting any of that is correct, but as koot mentioned, details do get noticed nowadays by the trades on site.

RW, These pole bldgs get stood up all the time. A modern combine needs an 18'x18' door to fit easily, so 20'-0" tall walls are very common.