Thread: Roof Framing

I am curious what others report when they encounter a roof structure that was originally designed to support spaced plank lath and cedar shingles that has been re-roofed with decking and asphalt shingles.

Similar Threads:

• Question on framing in a header
• Steel Framing connections
• Roof Framing
• Deck framing
• Steel framing

Cedar shingles and asphalt shingles are very similar in weight. In my opinion adding a layer of sheathing over the spaced deck only makes the structure stronger. As long as the framing is capable of the load and the nailing of the new sheathing meets code it seems fine to me. I know it's a common practice. Cedar shingles scare me. It's like having kindling wood on the outside of your house.

Here in Las Vegas most of our new construction all has concrete tile. Re-roofers regularly replace cedar shakes with concrete tile roofs. The city or county requires an engineer's report saying that the the existing framing can handle the weight of the tile. Sometimes it can not and the trusses must be repaired. Sometimes they decide to switch to a lightweight tile.

Originally Posted by A.D. Miller

I am curious what others report when they encounter a roof structure that was originally designed to support spaced plank lath and cedar shingles that has been re-roofed with decking and asphalt shingles.

Based on what I see as something from the sixties on up has pretty much the same framing as your typical plwood and composite shingles roof.

Once you get into the older homes and the 2x4 framed and sparse support then I call that out all the time. Most homes from the sixties on up have the same framing techniques. I don't call them out.

As mike said It makes the roof stronger by tying it all in. Most I don't see a problem with.

I know that just the 1x are nailed to the rafters but looking at it in the other direction in the way the 1x go the whole lentgh of the home and are spaced pretty close together I think in that direction you have many more nails. Hopefully the roofers are making sure the 1x are nailed down well before putting the OSB or plywood over them. I guess there is no way of telling.

If this was a high wind area like near the coast I think it would be a much bigger call. But the one in how many chance that that street is going to have a tornado run down the middle of it I don't see any grave concern.

Originally Posted by Mike Truss Guy

Cedar shingles and asphalt shingles are very similar in weight. In my opinion adding a layer of sheathing over the spaced deck only makes the structure stronger. As long as the framing is capable of the load and the nailing of the new sheathing meets code it seems fine to me. I know it's a common practice. Cedar shingles scare me. It's like having kindling wood on the outside of your house.

Here in Las Vegas most of our new construction all has concrete tile. Re-roofers regularly replace cedar shakes with concrete tile roofs. The city or county requires an engineer's report saying that the the existing framing can handle the weight of the tile. Sometimes it can not and the trusses must be repaired. Sometimes they decide to switch to a lightweight tile.

MTG: The cedar shingles here must be different than what you have. No. 1 Blue Labels usually weigh in at about 35 lbs. per bundle. With 4 bundles per square you have 140 lbs. The lightest asphalt shingle being commonly used weighs in at 225 lbs. per square. The difference on a 50-square roof is 4250 lbs.

And, though the addition of solid decking does provide some stability for roof structures, it also has weight. 7/16" panels weigh 47 lbs. It takes about 156 of them to deck the average 50-square roof. That is 7332 lbs.

The total difference between cedar shingles and the lightest asphalt shingles on the same 50-square roof is 11,582 lbs, or 2.32 lbs. per square foot.

Now I am no mathematician, so someone check my math - decaf morning, so no guarantees. But, if this is correct, does this not usually require significant improvement of the framing. Or, are builders in other areas so conscientious that they overbuild their roofs to handle the additional 60% load?

Originally Posted by A.D. Miller

or 2.32 lbs. per square foot.

to handle the additional 60% load?

Aaron,

I'm not a coffee drinker, still on my Mountain Dew , so I did not run through the math to check it, but ...

... you are saying that 2.32 lbs / sf is an additional 60% load over the design load?

What is the design load ... 1.45 lbs / sf????

Originally Posted by Jerry Peck

Aaron,

I'm not a coffee drinker, still on my Mountain Dew , so I did not run through the math to check it, but ...

... you are saying that 2.32 lbs / sf is an additional 60% load over the design load?

What is the design load ... 1.45 lbs / sf????

JP: "Design load" was what you brought to the table. My statement simply indicated the increase in weight from the original roof with cedar shingles to the new roof with the addition of OSB decking and asphalt shingles.

Originally Posted by Jerry Peck

Aaron,

I'm not a coffee drinker, still on my Mountain Dew , so I did not run through the math to check it, but ...

... you are saying that 2.32 lbs / sf is an additional 60% load over the design load?

What is the design load ... 1.45 lbs / sf????

Originally Posted by A.D. Miller

JP: "Design load" was what you brought to the table. My statement simply indicated the increase in weight from the original roof with cedar shingles to the new roof with the addition of OSB decking and asphalt shingles.

Yes, I intentionally added that in there because your statement reflected that an * increase of 60% * was an increase of 2.32 lbs / sf.

That means the "original load" (based on your calculations) would have been only 1.45 lbs /sf. That ain't much at all. Take the original 1.45 lbs / sf and add on the "60% increase" or 2.32 lbs /sf and you get 3.79 lbs / sf, and that's not much either.

Just pointing out that something is either amiss in the calculation, or, the trusses are likely designed to handle it.

Maybe MTG will address this?

Originally Posted by Jerry Peck

Yes, I intentionally added that in there because your statement reflected that an * increase of 60% * was an increase of 2.32 lbs / sf.

That means the "original load" (based on your calculations) would have been only 1.45 lbs /sf. That ain't much at all. Take the original 1.45 lbs / sf and add on the "60% increase" or 2.32 lbs /sf and you get 3.79 lbs / sf, and that's not much either.

Just pointing out that something is either amiss in the calculation, or, the trusses are likely designed to handle it.

Maybe MTG will address this?

JP: I'm not talking trusses here, but rather rafters.

Originally Posted by A.D. Miller

MTG: The cedar shingles here must be different than what you have. No. 1 Blue Labels usually weigh in at about 35 lbs. per bundle. With 4 bundles per square you have 140 lbs. The lightest asphalt shingle being commonly used weighs in at 225 lbs. per square. The difference on a 50-square roof is 4250 lbs.

And, though the addition of solid decking does provide some stability for roof structures, it also has weight. 7/16" panels weigh 47 lbs. It takes about 156 of them to deck the average 50-square roof. That is 7332 lbs.

The total difference between cedar shingles and the lightest asphalt shingles on the same 50-square roof is 11,582 lbs, or 2.32 lbs. per square foot.

Now I am no mathematician, so someone check my math - decaf morning, so no guarantees. But, if this is correct, does this not usually require significant improvement of the framing. Or, are builders in other areas so conscientious that they overbuild their roofs to handle the additional 60% load?

Everything I am talking about addresses mostly trusses. Plans usually specify a dead load in PSF that the roof framing is to be designed for at a minimum. The project engineer usually does a calc of the actual dead loads and then rounds up to an even number - not only that they typically include a "misc load" in the calc that adds to the total. It has been a very common practice to use a 10 PSF dead load on the top chord of trusses for many years (or 14 or 15 for tile). If they designed the structure for the bare minimum load then yea, there would be a difference. Engineers usually never design anything for the absolute minimum load. For one thing it covers their backside and for another it usually does not add greatly to the overall cost. For example from a truss perspective - it doesn't cost much more to design something for 20-15-10 than it does to build it for 20-10-10. That's why most trusses provided for the southwest by reputable companies are already designed for tile whether the project has tile or not.

Also you are looking at dead loads on the top of the roof only when you come up with a 60% increase. For decades they have designed framing for simple design loads that sometimes just came from tables in the code. An example would be a 20 Live + 10 Dead top chord load. When you are talking trusses, they also also usually supported loads on the bottom of 10 PDF Dead (plus an additional 10 PSF live load check that was non-concurrent with the top chord live load). So if you look at a truss that goes from say a comp shingle to a lightweight tile you might only increase the total load by 2 PSF.

The original truss might have been designed for 20 TCLL + 10 TCDL + 10 PSF = 40 PSF total. If you add another 2 PSF you're talking perhaps 2/40 or a 5% increase. You have to remember that I am seeing this mostly through the eyes of jobs I have seen in the southwest.

I've yet to see a truss fail because you added a layer of plywood to the top chord. Look at it this way. A truss is designed to support itself without regard to the materials tha connect to it. When you add plywood to the top chord of a truss you are probably doubling the strength of that top chord. That added strength is totally ignored in calculations. It adds a degree of redundancy to the design. Here's a little sketch I just drew to illustrate this concept. Granted it ignores some things like continuity of the plywood, connections, etc...

Also consider that we design the whole roof as though it has a live load equally distributed over the entire top of the building. This live load is typically either snow or construction loading. Construction loading is the weight of men working on the roof with their tools, materials, and staging. If you are talking about a project in a moderate climate you're probably OK for 20 PSF all over the entire roof. Here's the thing. Picture a crew working on the roof of a single family residence. How many times to you ever see more that 5 or 6 people on a roof? The chances that you will have 20 PSF over the entire roof is almost nil. Now add to that this fact - trusses when tested on a test rack often fail at 2 to 3 times their design load.

If it is trusses then by all means recommend an engineer look at it, but from experience I don't see it as much of a problem to add a single layer of decking to an existing building in terms of it's ability to support the vertical loads. IMO, it probably adds to the strength. I hope that clarifies my earlier statements.

If you encounter a roof with decking and asphalt shingles, how would you know that the roof originally had cedar shingles, or what the roof was originally designed for? All you would be looking at would be a rafter / collar tie roof (per your description) that was decked with shingles.

I guess you could tell that the decking wasn't as old as the rest of the house, but that still wouldn't tell you what the roof was originally designed for.

We see the skip sheathing (lath boards) with open spaces that has the decking above that. Around here, they never strip those 1x's before sheathing (not that I would want them to.)

If you encounter a roof with decking and asphalt shingles, how would you know that the roof originally had cedar shingles, or what the roof was originally designed for? All you would be looking at would be a rafter / collar tie roof (per your description) that was decked with shingles.

I don't know of any roofer's that remove the skip sheathing as it would be too much work.

Also, it should be pretty obvious as there would likely be a bunch of debris left in the attic from the tear off.

"Also, it should be pretty obvious as there would likely be a bunch of debris left in the attic from the tear off."

Not around here they always clean up all that mess in the attic


On the subject, it always amazes me how some of the 80 yr old homes with 2x4 rafters with no ridge board, no collar ties, and no purlins have been holding up a layer of wood shingles and 2 or 3 layers of asphalt for all these years. I saw one with two layers of wood shingles and three layers of asphalt. When I figured out how many there were, I wanted to get the hell out of that attic before it all caved on me. And if anyone is tempted to tell me that more than two layers is not allowed, I know that. Although some places here allowed three layers for years.

Originally Posted by Jim Luttrall

We see the skip sheathing (lath boards) with open spaces that has the decking above that. Around here, they never strip those 1x's before sheathing (not that I would want them to.)

Originally Posted by Brandon Whitmore

I don't know of any roofer's that remove the skip sheathing as it would be too much work.

Also, it should be pretty obvious as there would likely be a bunch of debris left in the attic from the tear off.

IMO, the problem with removing them is twofold.

First, removing nailed on boards with a 20 oz framing hammer can easily damage the structural members below. That lumber is usually very dry and hence easy to split or crack.

Secondly, when you remove all of the lateral bracing of the top chord, they lack any lateral restraint at all. They can easily buckle under even a small load unless they are braced both laterally and diagonally in the roof plane. Trusses with no sheathing or braces are a dangerous condition even when they are not loaded at all. In this case they are loaded with the full weight of the ceiling, attic mechanical, insulation, and possible live loads.

Originally Posted by Mike Truss Guy

Secondly, when you remove all of the lateral bracing of the top chord, they lack any lateral restraint at all. They can easily buckle under even a small load unless they are braced both laterally and diagonally in the roof plane. Trusses with no sheathing or braces are a dangerous condition even when they are not loaded at all. In this case they are loaded with the full weight of the ceiling, attic mechanical, insulation, and possible live loads.

When removing them, one would (I would think) start removing them from the bottom at one end and install sheathing as they worked across and up the roof, making the roof system stronger as they go.

When strip sheathing has been installed in the past and structural panel sheathing is to be installed, leaving the spaced sheathing without blocking in the open spaces on the top chords/rafters is adding limited strength to the roof system and is not installing the structural panels as required. One would need to block in the spaces between the spaced sheathing to be able to get the required nailing patters for new structural panel sheathing.

I'm not an engineer and not sure what weighs more and what causes a greater load, but having both installed shingles and sheathing I can tell you that sheathing *weighs a lot less than* a second layer of shingles would - at least that's the way it feels when you carry the two.

Originally Posted by Jerry Peck

When removing them, one would (I would think) start removing them from the bottom at one end and install sheathing as they worked across and up the roof, making the roof system stronger as they go.

I would think so, too. Then again I have seen an entire roof system which collapsed during construction because of inadequate lateral bracing. I've seen an entire building have a collapse because they were using short scraps of OSB as bracing material instead of 2x4's. I try to never underestimate how badly somebody out there could mess things up.

Originally Posted by Mike Truss Guy

Then again I have seen an entire roof system which collapsed during construction because of inadequate lateral bracing. I've seen an entire building have a collapse because they were using short scraps of OSB as bracing material instead of 2x4's.

I have seen many of them myself, a couple while I was standing there watching, one while I was warning the framer supt. that the trusses will likely collapse like dominoesssssss ... and there it went ... never got to finish my statement about installing bracing. I saw one collapse as I drove up, no bracing except a couple or rows of 2 foot long 2x4 connecting the trusses together, allowing for a hinge affect at the nails, and before I could stop and warn them ... down it came.

I'm putting more faith in the people taking the strip sheathing off because it is *already there*. But who knows.