Consider it as MWFRS or C&C? 2

[jiamin]

I am recently participating in a job to estimate a building roof wind damage (gable roof). The roof is caved in at windward and ballooned out in the leeward and the roof rafters and joists are buckled. The estimation report will directly affect the determination of which side insurance company will pay for the repairing. The calcs show that if the roof is considered as a Main Wind Force Resisting System and use ASCE7-05 Figure 6-6 to determine the wind pressure coefficient Cp, the conclusion will be that the roof design meets the code requirement. But if use ASCE7-05 Figure 6-11 to determine the wind pressure coefficient GCp (consider it as component and cladding), seems the conclusion will be the opposite way.

My question is for the roof rafter and joist buckling failure, should I consider it as Main Wind Force Resisting System to do design checking or should I consider it as component and cladding to do design checking?

Appreciate your oppinion, thanks.

 

[JAE]

It depends upon the tributary area of the rafter/joist members.

MWFRS is typically used for the overall building system design - the roof diaphragms, shearwalls, collectors, wall X-bracing, and since it is for the whole building per se, the area of building associated with the wind load is quite large. Thus, local peaks and valleys of wind pressure are averaged out over the whole area and this results in a MWFRS pressure.

For individual members, who take wind over a smaller area, the peaks and valleys don't "average down" as much as a larger area would. Thus, C&C wind is generally higher for those small area elements.

ASCE7 states that if your tributary area gets to 700 s.f. and larger, you can use the MWFRS wind instead of the C&C wind.

 

[Jack1977]

Thank you very much for your information, JAE. It is really very helped.

But now one more question comes.

The scenario is like this: the tributary area of one rafter is smaller than 700 s.f., but the tributary area of the multi-buckled rafter is larger than 700 s.f. So one can argue that the whole failure area is larger than 700 s.f., and all those buckled rafters failed at same time by wind load exerted on this larger than 700 s.f. area, therefore it should be considered as MWFRS. On the other hand, one can also argue that the failure occurred at one rafter first, then it caused a “domino” effect resulting all rafters to fail, therefore a smaller than 700 s.f. tributary area should be considered and the design checking should be based on C&C.

For above two argues, what are you guy’s thought?

Thank you very much.

 

[JAE]

Each rafter would have the wind load based on the rafter area (less than 700 s.f.) so the domino effect may have occurred....or at various times, each rafter's applied wind peaked at the higher amount due to its smaller area.

 

[SteelPE]

I was always under the impression that MWFRS members were receiving load from more that one surface. Granted there are exceptions to this definition.

When I think of a rafter, I think of a member that is spaced at 16” or 24” o.c. if this is the case that you are talking about then how far would you be distributing the load with your blocking to get an area > 700 square feet?

Also keep in mind the definition of wind area… the area is “the span length multiplied by and effective width that need not be less than 1/3 the span length.”

 

[JAE]

Here is a thread from some time ago that hashed this issue out.

thread507-181974

 

[Ron]

One way to think of C&C vs. MWFRS is what's visible and what's not visible.

Components and cladding are generally visible...they're the outside portions that transmit loads to the "inner" structure. To give an example, the sheathing on a shingled roof is a cladding (yes,it's "clad" with shingles, but the shingles serve no purpose other than watershedding...like paint). The sheathing serves other purposes such as lateral bracing for the top chords of trusses or rafters, and the sheathing can be used as a diaphragm, when it can become part of the MWFRS.

One key, as JAE has noted, is the tributary area. For fasteners, the tributary area is quite small, so the factors go up. As you increase the tributary area, the localized influence decreases...as JAE says, it "averages" out.

In my opinion, trusses are clearly MWFRS, even though the influencing tributary area might be quite small.

One other point of C&C vs. MWFRS....the C&C function is to tranfer load....the MWFRS function is to share and distribute load.

 

[strguy11]

I think the commentary says trusses can act as both. I would say the truss chord between the panel points is C&C, while the entire truss and uplift reactions are MWFRS

 

[StructGen]

Lets put it this way, rafters and joists are not part of main wind force resisting system so they automaticaly fall in to components and cladding category. But ASCE-702 allows to design any cladding element with area tributry area greater then 700-Sqft for MWFRS loads. Please see 6.5.12.1.3

 

[JAE]

StructGen's statement is what is correct. It's all about the tributary area.

 

[JedClampett]

When I first saw this post, the question that came to my mind was, when was the building designed and what code was criteria? If the building is less than 6 or 7 years old, than the "I" codes were probably (but not certainly) the current design. But if building is ten years old and the code was the old UBC (SBC, CBC, etc.), why should the ASCE 7 be the gauge? If the design was close enough that MWFRS vs CC makes a difference, then maybe it's fine per UBC.

 

[JAE]

JedClampett you bring up a good point. If this is a study to determine culpability in the original design, then the original design engineer should be held to the standard of practice at the time of the design. So determining the applicable code would be a first step.

Using the ASCE 7-05 for a building built many years ago isn't right. The original post doesn't seem to state the age of the structure.

 

[abusementpark]

I don't tributary area is the end all check. I think you have consider the "more than one surface" aspect as well.

 

[JAE]

abusementpark - I don't think so. The surface thing is a good general rule to test the member for C&C, but the C&C vs. MWFRS wind loads are truly based upon statistical variations based on area and location on the building.

One example I posed in the linked thread is that of an interior footing under a column that supports load near the corner of the building. If the column only supports a very small area of roof (i.e. less than 700 s.f.), then the C&C wind load uplift on that roof should be used to check against uplift.

This despite the fact that the wind load passes through many elements (roofing, deck, joists, beams, connection plates, columns, anchor bolts, etc.)

There is no way you'd use the MWFRS wind for something like a small area corner column/footing that isn't part of the wind x-bracing.

 

[strguy11]

Here is what the SEI has posted on their website regarding this issue:

http://content.seinstitute.org/static/faqwind.html

16. When is a gable truss in a house part of the MWFRS? Should it also be designed as a C&C? What about individual members of a truss?

Roof trusses are considered to be components since they receive load directly from the cladding. However, since a gable truss receives wind loads from more than one surface, which is part of the definition for MWFRS, an argument can be made that the total load on the truss is more accurately defined by the MWFRS loads. A common approach is to design the members and internal connections of the gable truss for C&C loads, while using the MWFRS loads for the anchorage and reactions. When designing shear walls or cross-bracing, roof loads can be considered an MWFRS.

In the case where the tributary area on any member exceeds 700 ft2, Section 6.5.12.1.3 permits it to be considered a MWFRS. Even when considered a MWFRS under this provision, the top chord members of a gable truss would have to follow rules of C&C if they receive load directly from the roof sheathing.

I also looked into what the standard practice of truss designers. Here is what alpine says on their website:

http://www.alpeng.com/index.php?option=com_content&view=article&id=74&Itemid=9

Trusses have always fallen into a gray area regarding use of the appropriate analysis method. By definition, a truss is an assemblage of structural elements, which would put it into the MWFRS category. But a truss also receives wind load directly from the roof sheathing (i.e., cladding) and therefore acts as a component, which puts the truss into the C&C category. Roof trusses can be found in the Commentary for ASCE7 as examples of both MWFRS and C&C.



It is common practice to use a combined analysis method when generating wind pressures for a truss, using both the MWFRS method and the C&C method. Most two-dimensional software analysis programs offer a choice of wind analysis methods when applying wind loads. The preferred method depends on the building designer’s specifications. The standard practice of component industry is to design the truss for a combination of MWFRS loads and C&C loads, and to specify uplift connections based on the MWRFS loads. However, design specifications may require trusses to be analyzed using the MWFRS method but with the uplift reactions based on the C&C method, or analyzing the entire truss using the C&C method.

 

[abusementpark]

JAE,

My comment is based on a discussion I had about this exact issue with one of my college professors who happens to be one of the guys who is on the ASCE-7 wind committee.

The example we discussed was similar element (like an interior column) that collected load from a monoslope roof compared to a gable roof. In the specific example we were discussing, he deemed the column under the monoslope roof as C+C, while the column under gable roof with same exact tributary area as MWFRS by virtue of the more than one surface verbage in the code. He had further insight that I cannot precisely recall, but the overall point was clear to me.

 

[ChipB]

From the description of this failure, it sounds like it doesn't matter how it was designed, it was constructed poorly.

For the windward side to be "caved-in" and the leeward side to be "ballooned out", the roof displaced laterally.

Where the rafters (or trusses) are supported on the outside walls, solid blocking should be installed between each one. The plywood should be attached (via nails) to this blocking at a maximum of 6" o.c. as it is a boundary edge of the roof diaphragm. Depending on your design wind speed, this spacing may be closer.

Contractors often neglect this, thinking the eave attachment is enough to stop this racking, but it's not. How does the wind shear from the roof, get into your shear walls? There is no clear load path without this blocking or some other means (Simpsons, but you want lateral capacity, along with uplift capacity).

Without the blocking, the rafters can roll over, which sounds likely to be your case. Go back and look at the end supports. I bet the blocking isn't there.

FWIW: I typically detail the solid blocking in addition to whatever uplift connection I need. Architects complain b/c of air flow, but I usually make them cut 4" dia. holes at mid-depth centered on the blocking. The contractor complains no matter what.

 

[JAE]

abusementpark,

Thanks. This topic has been hashed and re-hashed numerous times and I have to admit, the ASCE 7 and other documents seem to include in their commentaries the "idea" of whether an element receives load directly vs. through cladding.

I've always seen that the "direct surface" concept is easy to utilize for engineers, but totally inconsistent with all the other published data about how wind loads were derived in the first place. And as you say, some committee members and FAQ's seem to support some use of the direct surface idea.

The different wind loads: MWFRS and C&C, were presented in the codes to simply reflect the different statistical peak loadings of wind over small areas vs. large. This is fact.

I have seen no reasoning, logic, or experimental research that shows how wind is lower (MWFRS) because more than one surface or element is between your structural piece and the wind itself....my interior column/footing example per se.

Whether this interior footing has one, or twenty, structural elements between it and the wind makes no physical difference to how much wind is actually delivered to the footing....the wind just doesn't "know".

So I guess I'm frustrated that the ASCE committee 7 doesn't do a better job of defining this - although they do include lots of commentary on it - it still seems to confuse engineers (thus all these posts on MW vs. C&C loading).

 

[abusementpark]

JAE,


I agree with what you say and find it insightful. The code is too nebulous on this distinction.

However, I can see where the "more than one surface" logic comes from. Perhaps they are saying it is too conservative to take more than one statistical peak when designing an element that collect loads from two or more surfaces. In the case of the interior column, maybe the argument is that you shouldn't be taking the components and cladding statistical peak load from each surface because the odds of this occuring simultaneously is very slim??

Also, since wind loads are actually dynamic loads that have been simplified into static loads, is there any reason why a column might not ultimately see the same magnitude of load as a roof rafter (due to energy dissapation) since the column is much further along in the load path?

 

[JAE]

If an element takes load from more than one surface (i.e. a column on the corner of a building that sees wind from two directions) then yes, I'd say that statistically you might have a case where wind peaks occur on one face but never simultaneously with the other.

But for a column that takes a specific area of roof above, that roof does get its peak wind over the tributary area of the column. Thus,the column WILL see that peak wind..and the footing below also. You still include all the associated dead loads when checking 0.6D+W combos, but the wind statistics don't change with various dead loads.

Just a gripe with me as I like to have engineering always logical and rational and this "number of surfaces" doesn't fall into that category that I can see.

I see the surface descriptions simply used to help clarify in peoples minds the difference between how an X-brace in a building sees wind vs. a roof purlin. But in using that descriptive logic, I think people take it too far and forget the reasoning between the distinction of MWF and C&C wind.