Frequency for Wind tunnel

[StrucutreDr]

thread507-242316
Dear Experts,
Could someone shed light on the matter that the time period to be sent for wind tunnel should be based on uncracked (SLS) structural frequency or based on cracked section (ULS) to determine the Wind forces.
If it should be uncracked (SLS), is there any reference other than the ACI 318 commentary R10.11.1 which hints of using SLS?
What is the logic of using uncracked Time period for estimating Wind forces from Wind tunnel since we know greater time period will yield greater wind forces.

 

[bookowski]

I just went through this so here's my 2 cents. I don't know any official guidance but I spoke to the wind tunnel about what they deem correct and what they typically see. Rather than cracked vs uncracked I will refer to strength vs service model, since either may have varying degrees of cracking. The wind tunnel (I won't say who but there aren't many in the game) told me to send the results of the strength model first, i.e. longer period since more cracked, and that they can later do a desktop modification of the results for a different period (from service model). I'd try asking your wind tunnel guys to see if they have the same response or any addl info.

As an aside I asked them if they have any good data about actual vs reported periods. He told me that they only really get that on buildings where they are doing supplemental damping and need to verify/calibrate it. On those he said that the estimated/analysis period is not that close to what they see.

 

[Roy88]

Hi StructureDr,

There is a paper on this. Please see link below for the paper. Refer to Table 2 of the paper, there is a recommendation of structural properties for various limit state.
We will provide the wind tunnel specialist with 2 set of dynamic properties.

For Service - Partially Cracked (SLS), 25 Years Return Period, Damping about 1% to 2%, Applicable for drift and acceleration check
For ULS - Fully Cracked, 50 Years Return Period, Damping about 3%, To be combined with code defined combinations for ULS design.

Link:

http://www.cppwind.com/wp-content/uploads/2014/03/240-6.pdf

 

[StrucutreDr]

@bookowski,
You mean that the Wind tunnel guys asked for ULS periods?
I think the practice is to provide SLS dynamic properties for Wind tunnel. But i wanted to ask for reference and the logic behind using SLS properties.
The posts in the thread "thread507-242316: Dynamic Properties - Wind Tunnel" do not provide convincing logic.
Could you please share your ideas?

 

[StrucutreDr]

@bookowski,
Our wind tunnel report includes the forces corresponding to the SLS dynamic properties (uncracked frequency) along with forces corresponding to 5% higher and 5% lower than the SLS frequency.
Our SLS model loads are ok as it is not changed during detailed design, however the ULS model shows frequency much lower than the frequency provided for Wind tunnel. It's about 10% lower than the SLS frequency.
Could someone shed some light that forces for ULS calculations be based on what frequency.

 

[bookowski]

Yes, that's what I was saying. I just reread their original email and they said that ideally they are provided with both service level properties and strength level but if for some reason you can only provide one then it should be the strength level (uls). On my recent job we provided both and in return they provided service loads (for wind drifts) and strength level loads. For damping they recommended 1.5% for service and 2% strength.

When I say service vs strength that's not necessarily uncracked vs cracked, depending on your building you may still have cracking at service level - so it might be cracked vs more cracked.

 

[Struc-Dr]

@bookowski,
Then what is the intent of ACI 318 commentary R10.11.1, if Wind tunnel forces should be based seperately for SLS and ULS.

ACI 318 commentary R10.11.1:-

"Section 10.10 provides requirements for strength and assumes frame analyses will be carried out using factored loads. Analyses of deflections, vibrations, and building periods are needed at various service (unfactored) load levels to determine the serviceability of the structure and to estimate the wind forces in wind tunnel laboratories."

 

[StrucutreDr]

I noted that in ACI 318-14 the ACI 318-08 commentary R10.11.1 has been removed. Although the code is completely restructured, but the commentary is nowhere in the code.
It is usually considered that the commentary is not code.

Keeping in view the responses from wind tunnel experts I came to know that the ACI commentary R10.11.1 is outdated and removed in latest version ACI 318-14.

It turn out that it is usual practice to provide one set of frequencies for wind tunnel test. This can be based on SLS frequency.
The SLS wind tunnel analysis (forces, acceleration etc) should be based on SLS frequency. However it should be requested to the wind tunnel testing company to provide sensitivity analysis for the variation of frequency and forces. Based on this the ULS model forces must be determined corresponding to the ULS frequency (Wind speed and damping).

 

[azza833]

During my time at Windtech (

http://www.windtechconsult.com),

I have found it quite common for the structural engineer to send us two sets of data for SLS and ULS. This may come in the form of a direct export into an excel template or ETABS model, which we can extract the data from. We are generally happy to conduct our analysis for SLS and ULS design using the data from both sets (Cracked or Un-cracked/Partially Cracked) and there is some dialogue with the structural engineer on a case-by-case basis as to the appropriate direction to take. We also provide sensitivity, checks on both damping and natural frequency. We do not conduct corrections to the results based on a desktop approach. The actual structural data which affects the response is fully integrated into our analysis.