Design Criteria/Code for Rooftop Pergola 1

[bookowski]

I'm not looking for how to calc wind loads or design a pergola. I'm wondering if anyone can point me to anything that clarifies what types of accessory structures are required to be designed to what level of loading.

Backstory is that I have an incredibly stupid design for a rooftop pergola frame. The designer developed the concept of an aluminum hss frame. It all needs to fit up an elevator and be assembled on the roof and they will not weld on site. So they wanted concealed hss moment connections. They are achieving this with thick cap plates on the hss (8x8) and access holes to bolt it up (it's like your worst nightmare of what an arch dreams up and thinks is ingenious). When I first saw this I told them over and over that this was going to result in a very heavy design. Now that they've digested what it looks like (a monstrosity with internal diaphragm/stiffener plates etc) they are asking what it's designed for, i explained that it's designed for wind. They asked a semi-reasonable question which is how I decided that this needs to be designed for the same level of performance as the building. Assuming this is a ductile failure and I can keep it from flying off the roof is there anything that clarifies what performance level such things must be designed for? What if they're ok with this thing yielding and flopping over after 75mph wind? I assume that if i buy a backayrd pergola at costco and it flops over in a hurricane I don't have much recourse - can I reasonably have similar "may flop over in high wind" language?

What if this was a rooftop sculpture and it was well anchored down but would go floppy at some non design level storm? The closest thing that I could think of that we deal with locally is large landscaping trees on rooftops. Owners on those projects choose if they want them to stay standing during a design event or they accept that they may flip over and the trunk is only restrained to ensure that it can't go far.

 

[driftLimiter]

Can't you just point to ASCE 7 Components and Cladding requirements for the minimum wind design load according to code? We could play what ifs all day long but generally if there is a line in the code about it I choose to stick to that.

 

[EZBuilding]

I think you are getting painted into a corner that puts you into a precarious situation. There is a different standard of care between you as a Structural Engineer determining that a structure can be designed for less than code winds, versus a landscape architect specifying a tree on a roof. I have been involved in projects where tie downs were provided for such trees on a roof - but I expect I have been previously involved on projects where no one identified the size of tree and it's possible need for it to be anchored during a wind event. If your backyard pergola blew over in a hurricane, you might not have much recourse against Costco - although if it ended up through the window of your neighbor they may have recourse against you. In a commercial setting, however, you have to expect that the owner's insurance company will find every opportunity to recuperate costs should something like that happen.

I practice in South Florida, and when I have worked in other regions I have found that my expectations for appurtenances sometimes goes above the standard of care which is seen. In South Florida, anchoring down exterior mechanical equipment is a constant struggle for wind loads. In other areas I have had people look at me like I was crazy when discussing the same concern. I expect that for the purposes of your pergola, yielding is analogous to failure - although you could certainly justify loosening up serviceability criteria.

I think you do have some building-code outs that could give you some help, though they still rely on some level on engineering judgement.

Consider the pergola as a risk category I structure.

Review FBC Chapter 2002.3 - allows for removable panels for screen enclosures. Maybe you can reduce the tributary area for your pergola in reference to this chapter section.

Review FBC Chapter 3105 - awnings and canopies. Similar to above, you may be able to reduce the tributary area

I'm not sure how many panels you have which could be considered removable - maybe the only thing of value would be the risk category here. Always beneficial to go back to ASCE 7-16 and make sure that you wind design methodology was the most appropriate one.

 

[phamENG]

Remind them that when this fails at a lower than design level event, even just a serviceability failure of too much deflection, they’ll have roof system failures and leaks at every column. That alone is reason for me to maintain the risk category and design wind loads.

 

[Aesur]

Typically such structures would need the approval of the building official to say they are a different risk category than the building. I would start there and see if the building official will agree it can be designed as a Risk Category I, if they come back and say no, then it's simple, design to match the building and tell the client it's a requirement of the building official who has the ultimate authority. So best case you can design for Risk Category I (which I'm not a fan of for even canopies in a parking lot because during a storm people end up under them which puts their lives in danger should a failure occur), worst case you can point out that the building official is requiring it to match the building and you are not to "blame".

 

[JStephen]

If an item like that goes flying off the roof, that presents some hazard to people on the ground. I was thinking that was one reason tin roofs were frowned upon.
I was also not aware that you were allowed to designate different parts as different risk categories.

 

[bookowski]

EZbuilding - Thanks for the ideas. This was the train of thought I was hoping to poke at - risk category, any code outs, etc.

I generally agree that ideally it should be designed for full wind and inherit it's parent risk category, which is how I ended up at my current design. I'm trying to play devil's advocate because I can see the logic behind asking why there is no distinction here and didn't want to brush off their question. Like most things if you play it out to the logical extreme it's clear that at some point it would be reasonable (reasonable being a cost/risk assessment) to accept more risk of a failure, particularly when that failure will almost certainly not result in any bodily injury. I wanted to poke at it a bit so I could go back to them with a reasonable answer. I hate to give people a meaningless answer as driftlimiter proposed, essentially jamming the code in their face and saying 'this tell me to do it'. This plays into an annoying stereotype of our field. Your comments about FL are interesting here because this is actually a client from FL that is building this in NY and gave me the "in FL we do this stuff all the time and it's never this heavy".

Just to clarify (not that it changes much) my use of pergola here is probably different than imagined. This is a ~60ft long x 4.5ft wide x 9ft tall tube frame with zero other coverings - no fabric, curtains, infill panels or framing, etc. It is literally just the tube frame sitting along the edge of the roof as an architectural feature. The forces involved are tiny but the nature of the connection concept is what makes it garbage.

 

[Aesur]

Have you considered looking at a steel option and showing them the difference steel can make compared to aluminum? I assume your current moment connection involves some welding which for aluminum reduces strength significantly, whereas with steel it's usually not a big issue.

 

[JStephen]

I was just looking at another question, and ran across the Commentary in ASCE 7-16 Section C.29.3.1, which basically puts the kibosh on the idea of using a different Risk Category for a rooftop structure, as opposed to the building itself.