Roof Paver Uplift & Parapet Walls

[asabender]

I have been trying to check an uplift scenario for a finish-work installing client. In this case, they are trying to put terrace patio pavers on a high rooftop in a hurricane prone region. The IBC specifically says that aggregate is not allowed on roofs, and that ballasted roofs are not allowed in this application (1504.8). The roof isn't ballasted (it's structural, this is just a wearing surface) and ANSI/SPRI RP-4 specifically reference 22psf or greater pavers (which we are) as an alternative to ballast stone (RP-4-4.1.3.1), so one could infer that we are in line with IBC, so long as we meet the uplift requirements of ASCE 7 on the pavers. My first question is just, "Does anyone see a major problem with that logic?" Obviously, it has to be designed so that a bunch of concrete squares don't go a-killin' in the next hurricane.

Assuming pavers on roofs are generally copacetic, IBC has a specific requirement for "ridgid tile", which includes concrete, under 1609.5.3. This code basically says that uplift is qh*Cl where Cl is .2 for concrete, then goes on to use it in an overturning calculation. The code also says that I have to be in accordance with ASCE 7 ch 6 (1609.1.1). ASCE says that components and cladding are basically qh*GCp (since GCpi=n/a); since it's a stepped flat roof, GCp should be from 6-12&6-11b as required. ASCE 7 could also be read as allowing the blocks as "other structures", which would make sense to me, since they aren't fixed to the building. Following these methods, and going with the worst case, I'm killing them in zone 3, and really making life hard in zone 2.

At my request, my client has provided parapet information. The parapet is more than 3 feet tall (usually 5), is designed sturcturally by others, and SHOULD have a large effect on suction since my maximum height is something like 300 feet. All the research that I have read shows large decreases in zone 3 and decreases in zone 2, but I can't find anything in ASCE 7, which is, in this case, IBC's go to guy (and, in general I would agree that it is the book to use... but where is the effect of the parapet?!). Does anyone know how to take the parapets into account?

Finally, the code combinations dictate that I use .6D, which I'm all for in a structure where design dead load is conservative and loads may vary, but these are pre-cast blocks with dependable weights and no room for variation. Gravity will not get weaker during the hurricane. Does anyone know if, in a case like this, .6D is still a requirement?

 

[IsaacStructural]

I can't speak to all your points, but I've worked with companies before that adjusted the .6D for cases where the dead load was very reliable. For example on heavy industrial where the bulk of the deadload was our heavy structural framing, which consisted of our structural beams, columns, braces and gussets. We would increase the .6 to not overly punish ourselves. In this case our computer programs typically didn't include the weights of gusset plates, which could be quite substantial, so we were still being conservative.

M.S. Structural Engineering
Licensed Structural Engineer and Licensed Professional Engineer (Illinois)

 

[ron9876]

ASCE7-10 Figure 30.6-1 addresses parapet effect on roof uplift.

The code requires 0.6D. In my opinion that is the end of that discussion but you can find detailed discussions about that by searching this site.

The intent of the code is to not allow anything on the roof that can blow away. So if you put tile on the roof you will need some type of attachment or some type of product approved system that has been tested for this condition. It is unlikely that the tile weigh enough to resist the uplift on a 300' building based on dead load only. Have you checked for any local amendments or requirements?

 

[asabender]

IsaacStructural, thanks for the reply. I see your point, and agree, but I also see where ron9876 is coming from. Unless there is some way that the code itself lets you loose, I'm going to consider myself trapped.

Ron, the intent of the code is certainly that, and I really do not want to step around that intent, but they do not require physical connection. I am using a copy with amendments. In this case they are certainly a bit light to be used without physical connection, but I'm going to see if I can get them working (at least on the lower floors, which there are 2 of below 70'. I only have ASCE 7-05 in front of me, but I'll see if I can get my hands on the -10. Thanks for the point out, it's exactly what I'm looking for.

Thanks for the quick responses guys.

 

[Ron]

A 22psf paver does not provide adequate uplift resistance, alone, to be used in high wind zones. You will have to provide supplementary attachment. Without regard to reduction factors, 22psf will be exceeded on most roofs in wind zones greater than 100 mph.

 

[asabender]

Understood... the paver is considerably in excess of 22 rather than exactly 22.

 

[Ron]

So you are going to add 22+ psf to your dead load and still potentially compromise the uplift issue with the pavers? Hope your insurance is paid up. You'd be better off to provide a waterproofing membrane with a topping slab.

 

[RFreund]

I believe the only change for a taller parapet is that:
Zone 3 now matches zone 5 pressures.
Zone 2 no matches zone 4 pressures.
I would double check that though.

I had some questions on this in the past:

http://www.eng-tips.com/viewthread.cfm?qid=326478

EIT

http://www.HowToEngineer.com

 

[asabender]

Well, thanks for the response all of you. It was a rush check, and I explored all avenues which I could reasonably explore before telling the client that they were SOL. Even on the lowest area, they were only good in zone 1, and only barely, at full DL. I recommended that they use an interlocked paver system, checked in flexure, to achieve the required aesthetic without creating a hazardous situation (it's amazing how much of a reduction you get from the sq ft increase). The last thing the city needs is a bunch more severed heads on the streets.

For the record, if anyone comes here for reference: I spoke with a member of the board and a few grizzled old engineers I know, and they were all of the mind that the code is a guideline only. If you use sound engineering judgement, which you could back up in court, and design a safe system, you can step around something like that .6DL (which they all seemed to imply that they had done in the past). In this case, the DL is KNOWN, not some variable, conservative assumption. I recommended that the contractor verify weights in the field and use 95%.

Ron, it's not my design, it's architectural; I'm just checking that it is (or isn't) safe against pedestrian death by paver. I will act responsibly, and would never, ever put people's lives at risk while paying up my insurance.

Rfreund, thanks for the re-direct, very helpful.