?Causes of flashing buckling on new roof? 4

[EngineerofSteel]

I'm invited to look at a new roof, recently installed by a large, experienced contracting firm. For some unknown reason, the flashing is "buckling".

I have not seen this before. What are some of the potential causes?

pre-manufactured trusses, 24" o.c., 5:12 pitch, 40-foot span

We have had a lot of rain recently.

What should I look for during my site visit?

 

[hokie66]

Temperature is my first thought, with inadequate provision for thermal movement. What is the flashing material? A photo would help once you have seen it.

 

[Ron]

Agree with hokie66...probably thermal, but could be wood shrinkage as well. Is it truly buckling or is it "oil canning"?

A photo would help tremendously.

 

[Ron]

Something else came to mind as I pushed "submit"....

Is this a valley flashing or an "L" flashing (or "J" flashing in some circles)? If "L" flashing, those are often installed "step" fashion to prevent buckling and oil canning.

...again, photo please.

 

[MiketheEngineer]

As noted - a picture would help -

But a drastic moisture change that you alluded to might cause problems...

 

[EngineerofSteel]

Leaving now... will get pics and will answer your questions. Thanks for input. Appreciated.

 

[EngineerofSteel]

Upon inspection, the warping is not at the flashing at all. All flashing is in excellent condition.

The roof spans are long: 140 feet. The largest wrinkle is on the portion built in August/September. The wrinkle appeared a few days: when temperatures dipped below freezing (uncommon in recent years).

Inside the attic, in the areas beneath the distortions, the construction is pristine. There are very few shiners, even. There are no signs of movement or displacement. Even the blown insulation is evenly spread (if something is slopped in an attic, that is usually it).

The contractor used a type of tar shingles they have not previously used. The supplier gave them a free upgrade.

The significant wrinkle occurs exactly at the marriage line where new roof members meet the previously existing hip trusses. The entire building has all new roof sheathing. This was spec'd in calcs as C-C or C-D, but OSB was used.

 

[bimr]

APA has a publication named Buckling of Wood Structural Panel Sheathing D481 complete with recommendations to minimize this problem. APA recommends the use of metal clips to create 1/8" gaps beween panels. The problem is caused by the edges of the sheathing swelling due to high moisture environment or when the sheathing is installed without leaving expansion gaps between panels. Refer to

http://WWW.apawood.org

The attic is properly not ventilated and when cold, moisture will condense on the underside of the OSB, which is cold. This causes delamination of OSB edges by swelling.

In addition, OSB will not shrink back to original size so it should be easy to tell if it had been wet.

 

[Ron]

bimr nailed it (bad pun intended!). This is a common problem with sheathing and OSB in particular. It's dimensional stability is variable.

You have a rather severe case of it...it is usually a bit more subtle. This can be caused by placing the OSB in a very dry condition, without proper spacing, then the OSB gets wet or the attic humidity stays high and it buckles.

It will eventually fracture the shingles as this is a repetitive stress issue as well.

 

[EngineerofSteel]

It appears that the OSB has shrunk relative to the tar shingles.

if the plane of the shingles has a ripple,it has greater surface area... and is longer than the OSB. If the OSB had expanded due to moisture, I would expect the opposite.

hmm... Perhaps if the tar shingles have stretched in expansion, and then the OSB relaxed to a smaller form?

The panels are definitely snug... with no room to expand- at least in the joints parallel to the ridge line... which are visible from inside the attic. (It is an unblocked diaphragm.) The joints perpendicular to the ridge line are not visible... as they are at the truss top chords.

 

[Ron]

DD..if they are completely snug now, after the attic has probably reached a reasonable humidity equilibrium, then they likely expanded in place prior to this. The problem is usually in the long direction of the sheathing, not the short direction. Did they use ply-clips in the joints parallel to the ridge?

Shingle strips don't usually expand enough to buckle themselves. They are not the tail wagging the dog. As they heat up, their asphalt becomes more pliable (thermal rheology of asphalt), thus can't easily support transferring compressive stresses.

 

[EngineerofSteel]

An important question is this:

In the west wing attic, the HVAC unit was heating the space. In the east wing, there is no such condition. In the west, it was warm and we worked up a mild sweat walking along the trusses. In the east wing, it is cool.

What effect is expected from the action of the HVAC unit?

 

[EngineerofSteel]

Ron,

I anticipated the failure of the shingles through cycling. The ripple in the east wing roof diminished in the few days from the construction supervisor's first visit to our visit on Tuesday.

Reviewing mentally, and looking over my photos, I noticed that I could not see the soffit vents from inside the attic space.

It appears the blown-in insulation has covered the soffit vents. According to the APA report mentioned by bimr, this can stifle circulation and contribute to a problem such as is present on-site.

 

[EngineerofSteel]

Also, I noted in reviewing the photos:

Near the ripple in the west wing, plywood is used. In the rest of the diaphragm, OSB is used.

In the east wing (built second, in cooler weather), it is all OSB. The ripple is less here. This ripple is the one the supervisor said reduced in a few days. The west wing ripple seemed to be the same, in his opinion.

Note: there are one major ripple in each roof. But, there are some mild ones at other locations.

 

[bimr]

In cold climates, the primary purpose of attic or roof ventilation is to maintain a cold roof temperature to control ice dams created by melting snow, and to vent moisture that moves from the conditioned space to the attic.

If the moisture is not vented, it will condense on cold surfaces in the attic.

http://www.buildingscience.com/docu...D-102_Understanding Attic Ventilation_rev.pdf

 

[EngineerofSteel]

Outstanding, bimr.

Page 11, paragraph 3 explains exactly the conditions in the attic: outside temp below 45, inside temp above 45.

Contributing factors are therefore the open HVAC unit and the closed eave vents.

Solution: dry out attic (looks dry now, but should be dried out as evidence of moisture and conditions to produce it are evident & known); clear eave vents & add baffles; repair HVAC.

Thanks, all.

-DD

 

[Ron]

DD...there are no ply-clips on the joints. That contributes to deflection between supports.

 

[boo1]

and improper spacing leads to buckling

 

[EngineerofSteel]

The panels themselves do not appear to be warped. They are planar.

The tar shingles have a ripple. This ripple is about 6 to 8 inches wide, 2 inches at the apex. It looks almost like the diaphragm shrunk somehow, and a wrinkle resulted.

If the diaphragm was installed wet, and then shrunk as it dried, this is what I would expect it to look like.

To see a visual representation:

Place a standard 8.5x11 paper on a flat surface (desk) in landscape orientation. Hold left end firm to table top. Push right end toward left one-quarter inch. The resultant arch looks like what is on the roof.

 

[EngineerofSteel]

The supervisor pushed one part down by standing on it. It rebounded when he removed the pressure.

It is only the tar shingles which have warped.