Wind Tunnel testing data

[omarEIT]

I am trying to compute forces on a solar panel racking system that is made of aluminum. I have followed methods per ASCE 7-05 to calculate the uplift, downforce, drag, etc loads on the open racking system. I have also performed wind tunnel testing on the racking system to achieve more accurate results since we are manufacturing this racking system and wish to patent it soon.

My question regards the data I obtain from the wind tunnel testing. My understanding is that the ASCE 7-05 methods 1&2 take wind speeds from the maps provided that are based on 3 second gusts. However, in the wind tunnel testing, the wind speeds are held at a constant length of time greater than 3 seconds (say 1 minute) in order for the tunnel to record multiple data points per each wind speed.

When comparing the final wind tunnel data loads to the method 1&2 data loads, I am getting significant differences in values. I am almost certain the difference has something to deal with regarding the 3-sec gust (method 1&2) versus the constant 60-sec gust (wind tunnel). Is there any factor or note in the ASCE7-05 or IBC that accounts for altering wind tunnel testing data to account for this?

I was told my someone they thought the wind tunnel testing data should be multiplied by 0.70-0.75 to account for the gust length difference but I don't want to just take someone's word for it without reading up about it in a code or manual. Any help or input would be appreciated.

 

[ChadV]

I assume this is a full-scale model that you're wind tunnel testing? Are you using the wind speed given on the ASCE maps for the tests? How much discrepancy is there between your calculated and measured loading? For reference, the 3-second gust is generally (but not always exactly) equal to the mean wind speed plus about 3 standard deviations.

My understanding is that whatever wind speed you're testing at, you can determine lift, downward pressure, drag, etc. coefficients which you can use to scale your loads to the design wind speed (keeping in mind that the loads are proportional to the square of the wind speed). From what I remember from working in ASCE 7, there's a gust factor G which can be taken as 0.85 under certain circumstances. Assuming you're wind tunnel testing at the 3-second gust wind speed, you might be justified to reduce your coefficients by multiplying by that factor, but I'd like to wait and see what other responses there are to your question tomorrow before I start making definitive statements.

Remember also that it's not uncommon for wind tunnel testing results to be more than a bit different from the loads calculated by code - the code has to be calibrated to be useful for a wind array of circumstances, sometimes at the expense of pinpoint accuracy.

Disclaimer: if I'm inaccurate in anything I say here, somebody please correct me - I'm fresh from a pretty detailed course in wind effects and wind tunnel testing, but have only practiced a bit under ASCE 7.

 

[Ron]

Wind tunnel testing typically prevails. Don't try to correlate...it's a separate method.

 

[cvg]

shouldn't you go and ask the person who performed the wind tunnel testing to clarify this for you? Did they give you an engineering report? Did you ask them to address this issue in their report to you?