Gas Station Canopy 7

[tw]

I've been asked to help an architect design a gas station canopy for a typical 2 pump island that will be an addition to an existing station. The canopy will be about 38x15 ft in plan with 2 columns.

Neither he nor I have done one before. It appears that the foundation is typically a drilled shaft similar to a large sign. The columns support the canopy laterally by cantilever action. The steel framing for the the ones I have seen under construction appears to be pretty light.

I would like to know what is typical for the job and how one would apply wind load. It seems to me that uplift would be significant (110 mph ASCE 7 wind speed) but the light framing I have seem suggests otherwise. I also have heard that in an actual hurricane these structures don't perform well.

Thanks,
tw

 

[Ron]

These structures must be designed to the wind loads of the code for the area. You are correct about the foundation system (though a large "block" shallow footing is sometimes used instead of a drilled shaft/pier). The most significant stress on the footing is the overturning moment due to wind load. Dead Load and Live load are usually smaller than wind load.

This would be analyzed as an open structure, usually as a monoslope roof, using the provisions of ASCE 7's "Force Coefficients for Monoslope Roofs over Unenclosed Buildings and Other Structures, Cf". I usually distribute the resulting load rather than apply as point load as noted in the table notes.

 

[tw]

So this would not have uplift & steel framing/roof would only support gravity loads.

Seems to me there will be some differential pressures around roof in high wind and this may be why canopies get torn apart in storms.

tw

 

[Ron]

tw...No! There will be uplift.

Check it both ways...Cf method and by open structure method (no internal pressure, all external pressure)...in some cases the result is significantly different. Be sure to check it with wind perp and parallel to long axis.

 

[thaidavid]

tw,

One reason these types of canopies don't perform well under extreme wind loadings is that, despite designing them to the ASCE wind load provisions, there are peculiar dynamic forces acting on them that are not usually taken into account. Some of these unrecognized force criteria include: a long natural period for wind-induced vibration (due to their flexibility), vortex shedding around external corners with small aspect ratios in plan, and uncalculable micro-gust effects. These would advocate a more conservative (robust) design than is usually seen. The end result is this: what is the consequence if an inexpensive canopy is damaged or torn away in a wind storm? Are there many adjacent structures in the path of damage? Can the Owner tolerate the repair costs, based on the probability of failure? These are all judgement calls for the Engineer and the Owner to make together - make the Owner share in this decision.

I hope this helps a bit -

David

 

[tw]

OK, now maybe I get it, but the code could be clearer. What about this?:

Say my canopy is 38 ft x 25 ft with a 3 ft tall vertical surface all around (parapet & structure depth):

Since structure is open, use ASCE 7 eq 6-20 and make two calculations.

(1) Roof Design for wind -- Use Cf from Table 6-9 (Roof angle actually zero but use 10 degrees per Footnote 2) based on Af being the roof surface area, not the projected area normal to wind. Apply this pressure/force to roof up and down.

(2) Max Horz force -- Max horizontal force for design will come from evaluating parapet as solid free-standing wall using Table 6-11, applied per Footnote 4.

If the above is correct, couple of follow-ups:

I assume that conditions 1 & 2 would be combined.

What about applying Max Horz pressure (from no. 2 above) to projection of parapet above roof on leeward side?

 

[Ron]

tw...yes, 1&2 would be combined, as well as the leeward pressure. Commonly, you would be looking at a factor of at least 1.25 for the windward wall face and 0.75 for the leeward, all acting in the same direction. The result is going to be a very large overturning moment, thus a significant foundation.

Typically, the highest member stresses are going to be at the beam/column interface and at the base of the column. Unless you have a large column with lots of weld area, you will likely need a knee brace or gusset at the beam/column interface.

 

[tw]

Yeah, that all makes sense to me.

I stopped by the jobsite of a similar structure under construction here in Baton Rouge and introduced myself to the contractor and asked if I could look over the contract drawings. He let me and I don't believe the engineer took all these factors into account.

This one under construction has a similar 3 ft tall parapet with single 10x10x1/4 tube columns at 24 ft o.c. (canopy is multi-bay, maybe 5 bays long) The column base contains no gussets and is simply fillet welded all around. Further, his foundation is a 6.5 ft sqr, 18 in thick. No way either one of those works for Baton Rouge wind and soil.

I fear/suspect that many of these are done this way. Contractor told me these designers do this type of structure all over the country.

tw

 

[LPPE]

tw- You ever see those shows on TLC about hurricanes? Whats the first thing they show? Thats right, a gas station canopy being blown away. Tell that to the contractor.
I'd design the columns as cantilevers, fixed at the base. I asked a similar question to yours a while back, but I dont remember the thread name. Try a search. If I remember correctly, dik or Ron had some good comments about Cf.

 

[Ron]

tw...pylko is right (or at least he uses the same approach that I use!!)

Even neglecting uplift and considering only the parapet load (with 15' clearance under canopy), your base moment will exceed 65 k-ft. That's a pretty good moment for a chunk of concrete to withstand in overturning. It takes a lot of concrete to hold those "sails" down! A drilled shaft would be more appropriate here unless you could go to a full mat to accommodate the overturning and the pavement (just completed one like that..no clearance to put shaft in because of tunnel beneath...had to use mat). Your problem is compounded when using a mat because of all the piping you have to contend with for gas station.

Also, for that level of moment, an HSS 10x10x1/4 wouldn't work.

I think you are right to be second-guessing this. Those canopies that are "done all over the country" might have been designed one time for one location and now the owner is using the engineer's original design to fit all. Not good. Most municipalities require site specific calculations. I would hate to think that some engineer is "rubber stamping" this design for expedience. It happens, but it isn't good for our profession and certainly doesn't protect the public as we are required to do!

Also, don't forget that the fasteners for the sheets and other miscellaneous attachments can carry a fairly high load in these applications. When those sheets fly off, they can do some damage to other structures/property. At 110 mph you are on the edge of the "wind-borne debris" criteria. (you are also close to the "edge" of the 110 isotach....is it possible you are in a 120 mph range?)

 

[exchange]

Just don't forget to account for the rainwater also. That's how the rest of them fail. Also, there is usually apretty big hole for the downspout!

 

[tw]

Yeah Plyko, I found it by looking under threads you started. It did not show up in my original search for gas station canopy for some reason so I am pasting it in here for more info:

Original message from Plyko:

You've all seen these types of structures; The monosloped roof supported by 2 columns at the center (Like a T).
I'm using BOCA for this particular job, and I am coming up with rather small wind loads on this type of canopy. The mean roof height is under 15' and the limits in plan view of the canopy are 28'x33'. Using the "Other Structures" formula (not adding in the Af, so I can get a psf wind load) for components and cladding I get 9.5 psf. This seems ridiculously low. Job is in CT on the coast, Pv=18.5 psf, I=1.08. Cf = 0.45.
Anybody have any suggestions of why this seems low, or perhaps I'm using the wrong Cf (Monosloped roofs over unenclosed buildings)????

Response from Ron:

pylko...I've done hundreds of canopies similar to this. I typically use Cf=0.85 for free standing canopies and Cf=1.33 for those attached to or immediately adjacent to buildings. Yes, that exceeds the table, but justifiable for overhang conditions.

I approach design in same way as you have done, by spreading over entire area and not using the concentrated approach. I check separately for concentrated loads if necessary.

Assuming I=1.08 and V=110 mph, I get factored q=20.5 psf, with Cf=0.85.

Ron

respectfully copied by tw

 

[v2]

I do not see why the uplift pressure should be considered, since the wind speed at the bottom and the top of the roof would essentially be the same, because the structure is open and not part of a building.


V2

 

[Ron]

v2...if that premise didn't create uplift, planes wouldn't fly.

When the wind blows laterally across a sheet, turbulence is created at the leading edge. The turbulence changes the velocity at different points and changes the direction of the air flow at different points. It creates a "hump" in the streamlines of airflow at any horizontal line. Secondarily, another area of turbulence is created downstream. This causes "flutter". That's why flags flutter.

Another source of uplift is when wind blow into a structure it is compressed by the confines of the structure. In the case of an open structure, only two planes of confinement occur, thus the uplift is smaller than for a partially open structure. Since air is compressible, the confinement allows some pressure differential, thus uplift.

 

[NigelB]

We have designed fuel bays for truck facilities in many locations across the country. We found that the canopies are typically bought as a pre-engineered package. The supplier provides the full canopy, including structure and finishes. They also design the foundations. I suggest that you specify the canopy this way, and request a set of sealed structural drawings and calculations for the structural frame and foundation.

Nigel

 

[jcox]

I work for a company that designs and builds several gas station canopies. You would be amazed at what is and has been done that we end up fixing. We have lost several jobs because the TS6x6 column that were originally used that were wet set in 50 gallon drums of concrete. The owner would request that we use the existing columns and put an even larger canopy on top and please provide sealed drawings confirming that existing conditions are acceptable. Usually, this is where a less reputable company comes in and builds out of the back of there truck, and if you ask enough engineers, eventually one will say OK.

As for wind design. I don't use ASCE 7 too much. But with other codes, I use the wind loads prescribed for open structures. I find that in high wind areas that end zone uplift will control deck support spacing. I design the lateral load to be windward and leeward on both fascia's. I do account for drain holes in the bottom of the column. And I try to limit my column deflection due to wind to a "reasonable" amount. You can use your own definition of reasonable. For the columns, the moment is the killer. The gravity load usually isn't very high for a typical situation and wind will only produce tension on the column which normally helps. I tend to ignore this benifit in design. Footings do end up being large blobs of concrete. I prefer a drilled shaft. I find that in order to accomodate anchor bolts for most canopies you dont' want to go smaller than about 3' diameter.

In hurricane regions I have seen vortex shedding be a problem for fascia, but I have also seen wind borne debris knock the fascia off. I don't know how to reasonably accomodate that. After one major storm on the east coast, out canopy was the only one without significant damage out of four gas stations on a corner. The only comment the owner made after minor repairs is that it looked like we over designed the canopy, and it looked like he wouldn't be able upgrade his image because he wasn't going to get any insurance money. You can't win.

 

[jheidt2543]

jcox,

Look at the bright side, you didn't get a call from his attorney or insurance carrier!!

 

[carlputnam]

This has been a very interesting thread and exactly what I am looking for except I think there is a mistake that everyone has accepted. ASCE7-98 equation 6-20 is for horizontal loads only. "Af" is defined as the projected area normal to the wind not the pressure.
I have been scouring the code and there does not seem to be any mention of uplift forces on open structures exceptTable 6-7 which sets GCpi =0 but no tables give values for GCp for open buildings.
Obviously there is uplift but I can't find it in the code. I dd not believe this until an ICC structural engineer pointed it out to me.

Side question: When you distribute the point force what distribution do you use uniform?, linear?

 

[tw]

From ASCE 7 under eqn 6-20:

Af = projected area normal to the wind EXCEPT where Cf is specified for the actual surface area

See my May 21 post, that's the way I see to do it.

TW