Uplift on Wood Roof Trusses 2

[abusementpark]

I've seen where some pre-engineered wood roof truss engineers will provide uplift reactions based on MWFRS pressures, and consequently, the hurricane ties at each truss support are sized based on MWFRS pressures.

What you do you guys think about that? I really wish ASCE 7 would further clarify the differences between MWFRS and C+C. I think this is one of those gray areas.

 

[a2mfk]

ron9876- I'm a FL engineer too, and I am with you and your experience. Especially if the wall gets up to 10-12ft tall, and down in S FL those pressures can get HUGE. I seem to remember having a job in V=130mph with pretty tall walls where we dedicated a strap to uplift and then a rigid HGAM or similar to lateral, then put blocking in for shear transfer. I am sure the GC was thinking what the heck are these guys doing here...

I have this gut feeling very few people do that combined check but I hope I am wrong. I can see a CMU wall blowing in because it sheared off an inadequate uplift strap that was only designed for uplift, or maybe lateral, but not both. But I think there must be some good engineers at Simpson, it seems they have every detail covered. Well, its good business for them too.

I've done forensics after hurricanes and tornadoes, and therefore I am overly conservative in my roof pressures and connection design.

Horse dead now...?

 

[abusementpark]

Abusement- I do very little wood framed design anymore, but when I did I would not recalc for wall design. Just use C and C. Its a pittance in the overall cost and the extra insurance is worth it. And guess what, nobody will ever know but you and whoever (if anyone) that reviews your calcs.

Architects and contractors in my area take notice when you put a strap at the top and bottom of every stud.

This is a common discussion on structural message boards and to me its and academic one. I am not sure why someone would argue very strongly for MWFRS unless that is the way they have always done it and they want to justify their method. Economically its negligible in the cost of a structure...

I just wish ASCE 7 would provide better guidance on whether or not situations like this should fall under MWFRS or C+C. The codes hints at the "layers" factor and the "multiple surfaces" factor. I wish they would clarify these considerations. I think the cost of all the additional strapping could be somewhat significant on a large project. I'm sure if you asked an owner whether or not he cared about a extra couple grand, he would say yes.

Some engineers may argue for MWFRS because they think the modern wind code is conservative. In my short time in this profession, I've seen so many poor designs with regard to wind that have surprisingly lasted many years. I'm talking about things so underdesigned per modern code that you could not even prove it would hold under the unfactored loading - nevermind a factor of safety. I know these cases are anecdotal, but I've seen enough of them in a short time to make me scratch my head.

Also most structures I have designed are in Florida where CMU is predominant, so wind uplift transfer is usually an afterthought. In a really high uplift situation that effects walls and foundations I would recalc for MWFRS.

Even if the tributary area of the wall or foundation is less than 700 sf??

Not everyone specifies blocking for shear transfer, but I usually do. Unless you have a small enough heel height, then there are connections that can transfer the roof diaph. shear load into the top of the wall.

Right, file that under another thing a lot of engineers neglect. Even if the heel height is small, in order to transfer the shear you still have to make some argument that the truss connection is taking some moment. How do you prove that?

Different regions, different methodologies, as long as it is thought out and somebody uses engineering judgment... Feel like I am beating a dead horse here. Sorry.

Don't worry about it. I enjoy the discussion.

 

[abusementpark]

I agree with JAE. What would be the logic for using smaller than C&C loading for truss connections to walls. If both the wall and the truss are designed for C&C then why would the connection between the two be designed for MWFRS.

What about the rest of the uplift load path? Top plates to stud, stud to bottom plate, bottom plate to anchor bolts? Should that be C+C as well?

 

[southard2]

The way I was taught in school was that components and cladding were used for elements loaded in a single direction such as roof deck, steel joists, and columns. MFWRS are to be applied to all members that take loads in two directions or more. This guidance is more a rule of thumb as all structural components should actually be checked for both all the time. The primary purpose of the components and cladding load cases are to capture the effect of the PEAK gust that only apply to small surface areas. In a hurricane the wind is not a uniform pressure. It has all kinds of small and large gust. So MFWRS loads are the average pressure applied to your lateral resisting system. Components and cladding are applied to small areas. So you don't mix and match the two as they represent different things unless you are doing so in a conservative way.

So for example lets take steel roof deck connections. You must actually design the deck connections for two different load cases.

For the components and cladding load case you design the deck and the deck connections for the uplift components and cladding load based on an area of less than 10 ft2. (span * span / 3)

Then you must also check the deck connections for MFWRS. Here you would check the deck for MFWRS uplift in combination with your diaprhagm shear (this checks the deck for overall effects of wind on the entire building. It is based on wind loads spread over the entire building envelope walls and roof) The deck at the boundaries must transfer the diaphragm shear to the shear walls similar to how a steel beam connects to a column with bolts. I find that the final line of welds often must have shear collectors which many engineers are failing to do. One puddle weld to a joist every five feet sometimes doesn't cut it. You can do two welds and often and that will work but then the joist seat must be designed for rollover.

Lets take timber truss connections now. Again there are two load cases. In one load case you have only uplift components and cladding force acting. When the wind hits the building your are checking that one truss that gets higher wind because a gust hit it and only it in isolation. The pressure depends again either on the tributary area of the truss or span * span /3.

A second components and cladding load case check would be the out of plane shear transfer from the wall. This is for high isolated gust hitting a wall. Again you check this in isolation. So if the connection of that one truss can tranfer that small gust area it is OK. It would be very unlikely that the same truss would handle both a high concentrated gust on the wall and the roof at the same time.

For the MFWRS load case you must combine both MFWRS uplift and MFWRS out of plane and MRWRS in plane pressures. This would be a correct methodology according to how I was taught and from everything I've read since. For trusses dumping diaphragm shear out of the deck into the shear walls or for trusses receiving shear from the walls being sucked or pushed on by wind this is a macro event. Kind of like the repetive member factor in wood design. All the trusses act in unison to transfer these lateral loads. Therefore you use the MFWRS shear loads in combination with the MFWRS roof laods.

In practice it takes time to come up with all these loads. I usually will combine the components and cladding loads and connect for that to be conservative. I will only usually seek out the MFWRS loads when checking combined loads if I'm really having trouble getting something to work at a reasonable cost. For example if I have to use multiple connectors or something really impracticle I'll sharpen the pencil and use the MFWRS loads in combination and the components and cladding loads in isolation. If you place shear collectors between the truss seats than you can completely eliminate the in plane diaprhagm shear transfer from the combination. Often I will pour concrete tie beams to support steel joist since you can get much stronger connections into concrete versus masonry.

So in summary you really must always check everything for both the components and cladding load case and the MFWRS load case. But by definition components and cladding is a singular load because its intent is to capture the effects a concentrated but high pressure gust that effects only a small area. When checking the MRWRS systems load path you must check the truss connections for shears in combination with uplift.

You will notice however that as areas increase the components and cladding pressures (for roof and walls) will approach the worst case MFWRS pressures. So again for saving time its usually practicle to work with the components and cladding loads for both your isolated checks and your combinded shear and uplift checks. If you get into trouble you can use the MFWRS loads as that is the intent of the codes. It can reduce your loads sometimes by 10 - 20% which can help when checking the combined shears and uplift.

Roof deck connections have become kind of a specialty of mine so I've become very intimate with ASCE 7-05, etc... I'm pretty darn certain what I've written above will never guide you wrong.

John Southard, M.S., P.E.

http://www.pdhlibrary.com

 

[southard2]

Abusement Park

Top plates to stud will be controlled by your components and cladding uplift load from your trusses. I would based this on the trusses area since the top plate has some redistribution qualities as well. And lets face it that is the component the load is coming from. The load is from the truss not the wind hitting the top plate connector directly.

The same thing applied for the stud to bottom plate.

However for the bottom plate to your foundation I find it is usually controlled by the combination of MFWRS out of plane shear in combination with MRWRS uplift. In practice I combine the components and cladding loads instead since they are higher and easier to grab. I think most engineers are simply sizing the bolts here for diaphragm shear transfer. But again in reality I find that the bottom plate often shears due to the out of plane wind pressures. I had to program a spreadsheet for houses cause otherwise designing all the parts and pieces becomes daunting. Most people aren't actually engineering houses. They just do the same thing every time. I try to avoid designing houses anymore. Of course right now I take what comes.

John Southard, M.S., P.E.

http://www.pdhlibrary.com

 

[southard2]

I want to comment one more time about the "multiple surfaces" aspect of wind loads. I don't believe multiple surfaces applied to components that have several layers of roofing etc.

Roof trusses receive their load from the roof. So does the roof deck. Most of the time the columns are also receiving their wind load from the roof.

Truss connections receive their load from two surfaces. The walls and the roof.

If you design almost everything with components and cladding based loads you will be fine. But when doing combined checks of shear and uplift you can use the MFWRS loads. The worst case MRWRS load however is usually not that much smaller than the components and cladding load. In fact the diaprhragm shear is always a MFWRS load.

John Southard, M.S., P.E.

http://www.pdhlibrary.com

 

[ron9876]

I fully agree that ASCE7 should provide much more guidance for these types of conditions. But it doesn't. From reading this discussion it is obvious that there is no clear answer to as simple a question as truss to wall connections. Shouldn't be that way.

Based on the current ASCE7 I don't see how you could argue that the wall to truss connection would be anything but C&C loading. So it follows to me that the components of a stud wall would be based on C&C loading. I also can visualize a localized peak gust that would envelope both a portion of a wall and roof.

We live and practice in a world where lawyers rule. It sucks but it is the real world. If a person was defending their design in a hostile environment I don't find anything in ASCE7-05 that could be used to defend MWFRS loads for this condition.

 

[JAE]

I just keep repeating - it is all about the exposed wind area that the element you are designing takes. It is just that simple.

 

[JennyNakamura]

ron9875 and others,

The attached technical note from the CFSEI (cold-formed steel engineering institute) basically says that you design rafters, trusses, and the individual truss components for C&C but you design the uplift connections to the stud walls for MWFRS. Of course, it is acceptable and up to the designer if he chooses to use C&C for uplift tie design as it is more conservative and saves time in not having to re-calculate for MWFRS forces at the uplift connections. I think the article gives a pretty good rationale for this and references some other design standards that substantiate this position.

 

[JAE]

JennyNakamura,

I still think C&C loads should be used for truss hold-down connections at their ends. To me that is because the "typical" failure of roofs in high wind events is either due to deck/diaphragm uplift failure (causing instability in other members and then collapse) or due to connection failures at roof edges from roof framing separating from the wall.

However, the document you posted brings up an interesting aspect of the condition of a truss hold-down on a roof...the last paragraphs of your posted document suggest that due to the long, slender area of a roof truss member, the statistical variations over the area do actually average out more like MWFRS wind pressures.

I can see that aspect of it.

 

[NITTANYRAY]

Components and cladding design for roofs includes zones 1,2&3 as well as roof overhangs. Depending on where the truss is located it could be subject to all four of these conditions. MWFRS does not consider different loadings at the ridge,eaves, corners and overhangs and usually involves smaller uplift forces than C&C design. The Main Wind Force Resisting System is the total system that resists wind and does not include higher forces that may occur on a localized portion of the roof. If the truss is part of a frame such as in a pole building it should be checked as both a component and as an element of the frame since the loads could be very different.
It is important that the trusses be checked for C&C not only for the uplift connectors but also for possible stress reversal in the webs and chords. Typically they don't brace webs and chords in tension but if uplift causes compression in the bottom chord or web members it could result in a failure in longer members. The steel joist industry provides uplift bracing for joists when uplift is a problem but I rarely see this in the wood truss industry. I also rarely see truss programs that properly design gable end trusses since they don't consider loads on the side of the trusses. The EOR needs to make sure that all possible load cases are properly addressed by the truss manufactures. I frequently require the shop drawings to be revised and resubmitted because the technicians that typically plug and chug don't know how to handle special cases and many engineers that stamp the drawings seem to be on autopilot.

 

[abusementpark]

Jenny,

Thanks for posting that paper. That is the first time I have ever seen this issue discussed in a technical paper. The references cited in the paper certainly indicate that MWFRS pressures are the basis for roof to wall anchorage in prescriptive methods of wood design. Very interesting.

I would still like to see something definitive from ASCE 7 on this matter.

 

[abusementpark]

Also most structures I have designed are in Florida where CMU is predominant, so wind uplift transfer is usually an afterthought. In a really high uplift situation that effects walls and foundations I would recalc for MWFRS.

Is it that hard to make wood framing work in really high wind speeds? Or is the CMU more of a preference for durability?