Sloped roof acting as a diaphram 1

[rittz]

How does a sloped,sheeted roof (roof trusses 4:12 +/-) relate to the classic diaphragm modle which is presented as horizontal beam with roof edges acting as chords and the sheeting acting as the web. The picture I see is a flanged beam with a large buckle in the middle and I have difficulty expecting that this condition can replace a flat horizontal diaphragm provided by sheeting the underside of the trusses.

 

[Lion06]

Are you asking about this with respect to a given function (e.g. bracing the exterior wall against out-of-plane wind loading)?

 

[rittz]

Yes Horizontal wind loading on the wall and roof resisted by the out-of-plane roof sheeting (acting as a diaphraghm). A classic diaphram is horizontal at the plate (chord) elevation. , chord to chord ... a roof diaphraghm is not

 

[MiketheEngineer]

Might be a bit of a an over simplification - but it is done all the time. The plywood roof plus drywall ceiling does act as a pretty good diaphragm. And it is allowed. I am more worried about blocking between the roof trusses in order to transfer those forces to the walls - seldom done....

 

[Lion06]

I agree with Mike. I think that a natural consequence of the sloped diaphragm is that the exterior walls are put into compression for wind pressure and tension for wind suction, but I do believe that it functions as a diaphragm and braces the top of the wall.

 

[dhengr]

Rittz:

Draw a cross section through the ext. wall and the roof truss, a free body diagram (FBD). How is the wind load resisted by the roof diaphragm? In any event you must get the wind load from the wall, through the double top pl. into each truss, and then into the sheathing diaphragm and maybe partially into a diaphragm caused by clg. sht.rk. We know the sht.rk. absorbed considerable load or energy before roof failure, but are just recently being allowed to consider the affects of osb sheathing, plus sht.rk., because their shear stiffness is so different. And finally, a well blocked and designed diaphragm takes these loads out the end shear walls or gable end walls.

You claim to understand the flat (horiz.) roof as a concept to do the above. The diaphragm doesn’t act differently when it is inclined, but the magnitude of the loads and the way the loads are input does. A 4:12 roof is still pretty efficient at transferring these lateral loads through diaphragm action, draw the FBD right at the double top pl. and truss seat and sheathing. What are the load components into the truss and sheathing and truss hold down, etc? Why do you need truss hold down hardware, other than for suction loads (uplift), what about blocking btwn. trusses and the top pls. and sheathing under the various slopes. The 12:12 roof is not nearly as efficient and the load components are much different; and a mansard roof is almost useless w.r.t. this lateral loading, where you would need (want) a moment connection btwn. the top pls./studs and the truss to assist in resisting this lateral loading.

The diaphragm acts the same, it doesn’t know the difference re: flat or inclined, but the loads and their magnitude and the exact way they are input changes; thus the need for added attention to hold down hardware, chord or shear transfer blocking at the top pls., trusses and sheathing. And, blocking and shear transfer and special attention to the framing and bracing on the gable walls may be important too.

 

[BAretired]

It is strange that this topic does not seem to be addressed in any of the textbooks, at least none that I have seen. Are we looking at a single diaphragm with a bend at the ridge or do we have two separate diaphragms, one each side of the ridge?

Sometimes, continuous venting is required at the ridge line. That means the diaphragms are separated by two or three inches. There should be a continuous chord member each side of the vent, but it does not seem to be common practice to provide one.

BA

 

[msquared48]

There are a couple of other threads recently discussing this same topic. You might try looking up "deep beam" and see what pops up in the search function.

Mike McCann
MMC Engineering

 

[BAretired]

There are quite a few threads of the past dealing with this topic. I searched for Psloped roof as a diaphragm" and found many references. The following is one:

http://www.eng-tips.com/viewthread.cfm?qid=178730

The problem is, nobody is quite sure of how the diaphragm behaves.

BA

 

[msquared48]

Seems to me that it should be possible to model the diaphragm at various pitches in RISA 3D both with a ridge beam, and with collar ties, to see what effect the diaphragm has on both vertical and lateral deflection. This is just a problem based on equal deflections of a combination of restraint systems.

Mike McCann
MMC Engineering

 

[BAretired]

Mike,
In order to model it, you need to tell the program whether or not the two roof plates are tied together at the ridge line.

In the absence of a ridge vent, I think you could tie each plate to a continuous ridge member. In that case I believe the two plates will tend to act as one and there will be no need for a chord member at the ridge.

If the two plates are separated, they will tend to act as two separate diaphragms and would need another chord member each side of the the vent opening.

BA

 

[msquared48]

True, but that can be handled with wood blocking and continuous steel strapping. Shouldn't really be a problem BA. The only thing is the combination of beam sizes, diaphragm sizes and slopes, collar tie levels, etc., etc., etc. It would take a lot of time to do a proper analysis.

Perhaps this would be a good topic for someone's Masters Thesis.

Mike McCann
MMC Engineering

 

[dhengr]

The thread I was thinking of, as I posted above, was Struct. thread #507-301144, “Rafter Thrust,” by Lake06, 20JUN11; wherein one of the suggested solutions was to consider the sloped diaphragm action to help rationalize (solve) a lack of rafter ties and to pick of the rafter thrusts. BA and MikeMc and I participated in that thread, and it would seem that Mike and I are picking up right where we left off on that thread.

The need for proper diaphragm chord blocking and also blocking to distribute the shear flows over shear walls is absolutely essential, and Mike give a good explanation of this in that thread, as I recall. Proper attachment at each detail level is also essential. The effectiveness of the diaphragm in reacting lateral loads was left somewhat in question as it relates to the slope of the diaphragm. If someone has access to RISA 3D and would model this problem and run it, I for one would love to see the results, at a few different slopes. I believe that what I said in my post above would hold true. I think, on a typical residential roof, the rafter and certainly the truss is the primary bending resisting member and that the osb sheathing is the diaphragm element, as long as rafters and blocking keep the diaphragm from buckling, and properly transmit the load components up into the diaphragm. What I have always had trouble putting my finger on is; when the beam/column action of the rafter starts to become dominant, and when the diaphragm action becomes ineffective, or too small to count on, and instead you need the moment connection btwn. the rafter and double top pl./stud to pick up the lateral load at the top pls. At lower slopes or a flat (horiz.) roof we have no problem with this, at slopes around 10:12 and greater, I start to have my doubts, even though the rafter thrust is decreasing.

I think the reason BA’s text books, and mine, don’t cover this sloped diaphragm issue is because our text books are so darn old. Our text books were made of virgin wood as was our sawn lumber. They are from an era when engineers, architects and builders thought rafter ties were a good idea and only the architects didn’t understand why. Furthermore, some of our texts mention plywood as a new fangled manuf’ed. product of dubious parentage. We understood that diagonal ship lap sheathing was better than horiz. laid sheathing. I’m partly kidding, because my earlier texts do cover plywood diaphragms, but we sure weren’t stressing them like we want to today. Maybe I should not be speaking for BA, but I’ll bet I’m not too far off. Then, in the earlier thread BA suggested that the sloped roof diaphragm system might be looked at like a folded plate system, but now he puts the lie to that by asking the question of what happens when is often the case, that the sheathing isn’t nailed to the ridge beam/board or, worse yet, when you have a 3" gap at the ridge for venting. And, nobody wants chord blocking up there either because it interferes with the venting. As has been said a number of times, the blocking at the ridge and double top pl. at the ext. wall isn’t even provided. And, I guess the explanation for that working is that on a typical residential roof the diaphragm proportions don’t require serious sheathing joint blocking or chord design to function under reasonable loadings.

While we don’t encourage student posting of homework problems on this forum, I would encourage a grad. student who was looking for a thesis problem to tackle this assignment with our help, and publish the results. I suspect some of the wood products people have looked at this if we would dig deep enough. And, as I post I see that Mike has stolen my thunder, here.

 

[abusementpark]

The diaphragm acts the same, it doesn't know the difference re: flat or inclined, but the loads and their magnitude and the exact way they are input changes; thus the need for added attention to hold down hardware, chord or shear transfer blocking at the top pls., trusses and sheathing. And, blocking and shear transfer and special attention to the framing and bracing on the gable walls may be important too.

Flat or inclined diaphragms (i.e. monoslope) I can visualize. It is the gable roof with the extreme pitch break at the ridge that is foggy to me. How does in-plane load in the sheathing transfer across the ridge without inducing some vertical load in the rafters or trusses?

The problem is, nobody is quite sure of how the diaphragm behaves.

That sheathing has a lot of mysterious effects on the behavior of wood framed roofs - particularly stick-framed roofs.

 

[MiketheEngineer]

Must agree that the ridge vent always bothered me - and I never show one.

But I am sure they work good enough - the forces transverse through the top chords of the lumber. Not exactly allowed - but probably OK.

Probably could never show it good on paper....
But then I remember an old experiment where they took a brand new house - loaded it up to code and started cutting.

They removed almost two-thirds of the roof members before they got a failure. Interesting....

 

[a2mfk]

If we assume for discussion a very simple rectangular ranch style house with a gable roof, and the wind loading scenario is perpendicular to the ridge- some of the upward vertical component at the ridge from one side of the diaphragm is negated from the downward "pull" of the opposite side of the diaphragm. The windward side will create a positive or compressive force into that side of the roof, while the leeward will create a tension in the opposite side of the diaphragm.

Now since windward pressure is more than negative leeward, there will be an imbalance, which I would opine is resisted by the trusses. Even though they were not designed for this type of force, trusses in general are well suited for this type of brace since they usually have a flat bottom chord with a diaphragm of sorts (ceiling), braced to chord, and internal webs. This will be a short duration load, and in my guesstimation would result in less internal stresses than a full code dead + live load combination.

When wind loading is parallel to the ridge, in theory the ridge will be at the zero-shear area of the diaphragm. So a ridge vent in this loading scenario shouldn't have much affect whatsoever.

Of course in practice, nobody builds ranches anymore. Having done LOTS of roof inspections in Florida, anything built in the last 10-15 years has a slew of hips, valleys, dormers, etc. forced upon it by pseudo-designers, not even architects, who think that makes the house pretty... So how do these roof diaphragms perform? Well, most of them are not designed by an engineer so who knows what people are doing in these instances. There are so many unintended redundancies in a residential structure, like Mike said. I have seen fire and wind damaged buildings whose roof systems were still largely intact despite several damaged or missing trusses, walls, etc. Like Dave said, an FEA model of some roof systems would be very interesting, amazingly all of the parts and pieces probably "just work", even if we can't sharpen our pencils and show where every last pound of force is resolved.

And frankly, until all residential building construction is up to snuff (ie load path and connections), I don't think we will find out, because a roof diaphragm is the last thing I expect to fail in WIND country. Too many other things will fail first- windows, doors, sections of plywood, uplift connections, etc...

Seismic is another issue, and I wouldn't be surprised in the right scenario if this controls, since the other lateral and uplift forces are largely not a factor unlike in a wind event.

But I doubt we will see any research into this until we get a body count, hate to say that but...

 

[hokie66]

Agree with a2mfk about the practicality of roof diaphragms in most residential construction today. The roof profiles are so complicated that they defy analysis. In Australia, we don't use solid sheathing on roofs, and rather place the bracing or diaphragm at the ceiling level where it is flat and simple. In most cases, a ceiling is the stiffest element anyway, and until it fails, the roof level bracing is superfluous.

 

[bigmig]

This is an interesting post for several reasons....1st the ratio of sloped roofs in North America v.s. the number of sloped roof diaphragm problems in ANY text book or code in the US is a little low in my opinion. There are no flat roofs where I live (mountains).

2nd, the amount of fog surrounding 98% of the engineering community on the topic of how roof vents effect sloped roof diaphragm behavior is amazing as well. Ventilation is a real deal. Many of us have seen moldy, saturated i.e decayed roof systems because of lack of proper ventilation. But how many of us have seen a roof diaphragm fail because a ridge vent was installed?

Sorry for lack of input. Just some observations. Good post.

 

[a2mfk]

Those Aussies, making things simple and easy. Well, here in 'Merica, we like our ceilings as useless insulation holders and our diaphragms to be mysterious puzzles only some wizard of FEA could solve.

I have used a ceiling diaphragm a couple of times. Once in a gable roof with probably a 12:12 pitch, once I drew a couple of diagonal braces to scale and then explained this conundrum to the architect, that had no problem with it.

Hokie- Down under do you use gyp board as your diaphragm or sheath it first and then cover with gyp? And I assume this is done before the interior walls are framed, otherwise that would be some very tedious carpentry ala plywood jigsaw puzzle...

I won't use gyp board as a diaphragm after doing hurricane forensics in 2004 and saw what moisture does to gyp board. In the extreme case that summer some areas got hit with 3 hurricanes within a couple of months and numerous power outages. A small leak in your roof can ruin your gyp board diaphragm.

Bigmig- ZERO roof vent diaphragm-related failures, EVER! Well, show me the pictures if they are out there

 

[msquared48]

I have to take issue here with the ridge vent issue, for as I see it, there are two design scenarios here.

The first has to do with horizontal shear as from EQ or wind forces. With a diaphragm of equal size to either side of the ridge, the shear forces in the diaphragm should be zero at the ridge anyway. But for unequal diaphragms, there will be a shear value to develop which can be detailed, even with the venting.

The second scenario has to do with the sloping roof diaphragm supporting vertical loads. In this scenario, it makes no matter whether or not there is a ridge vent that breaks the continuity or not. Each diaphragm to either side of the ridge will take it's own portion of the vertical load depending on their relative stiffnesses. It's just as BA brought out earlier that the chord force heeds to be developed through blocking and strapping.

Thoughts?

Mike McCann
MMC Engineering