[manmeet singh]

Hi dear,
could u please help me to scale down model. what is the exact procedure. there are few things to control, Reynolds number, geometry scaling as per wind tunnel size, blockage ratio, wind velocity etc. so please let me know how to start. it is for build environment. thanks in advance

[fluid23]

I will assume that this is subsoninc as opposed to transonic or supersonic flow...

1. Start by selecting a wind tunnel.

2. Next, set a limit on blockage ratio (typically 5% is as high as you will want to go). Then determine a maximum allowable size from this limit. Either your wing chord, cylinder diameter, or whatever other value you are using for your Reynolds number reference length.

3. Now you determine a minimum size based on the Reynolds number you are testing at, max tunnel speed, and expected atmospheric conditions.

4. From this range and the tunnel specifications, you can select a size/speed combination that will give you the proper Reynolds number and sufficiently low blockage ratio. Pay attention to the required tunnel velocity so that you don't end up in a different flow regime.

Typically you will want to end up closer to the max allowable size rather than the minimum allowable size to keep the velocity differences to a minimum.

[manmeet singh]

Dear Sir,

Thanks for your reply. Actually still confused with reynold numbers. My wind velocity is 12 m/s and height building L-60m. I got reynold number 4.44*10 power 5. If i keep same renolds number than wind vellcity for wind tunnel is very high, which is impossible to produce. So my questionis how i can select reynold number which is viable to produce in wind tunnel. Wind tunnel has max wind velocity 30m/s. Thanks

[fluid23]

Sounds like you need a different tunnel. What are the dimensions of the test cell? Is it open-circuit/closed-circuit?

[manmeet singh]

My full scale is 60m height and 800 m width. I scaled it down to 1:200.

[fluid23]

Well, first things first. Your Reynolds number is not correct. If Re = V * rho * L / mu = 12 * 1.225 * 60 / 1.98e-5 = 4.45e7 not e5.

If we say that we want to keep tunnel flow incompressible (Mach <= 0.3) then you will need a tunnel speed that is around 103 m/s and a building height of ~7m. However, this puts your width at 93.3 m. I don't think you will be able to find a wind tunnel anywhere near this size.

I think your only option is to use a water tunnel. However, this is still going to be difficult given that your width is 13.3 times your height. Even if you can find a tunnel that goes to 100 m/s (~0.067 Mach) you are going to need a tunnel larger than 6m wide which is going to require a flow rate approaching 170 million kg of water per minute which will require a massive pump and power supply.

You may have to consider truncating your geometry. Can you elaborate a bit on your model and goals for the analysis?

[manmeet singh]

I think, i have to consider critical Reynolds number where we can get/choose low Reynolds number to find realistic wind velocity for wind tunnel. actually i need these results for validation of my numerical results. I have a case study of building environment, in which height building is 60m and whole area radius is 800m.

[davidwilcox]

Manmeet,
Building simulations in atmospheric flows are Reynolds number independent. The flow is fully turbulent.

[manmeet singh]

Hello David, Thanks for your reply. Could you please make it more clear. Like Independent Reynolds number? If i want to know Reynolds number to be use for wind tunnel then how i can calculate. I think it should be lower than the actual calculated one at full scale. Because at lower Reynolds number we can only test model in wind tunnel, to bring down the wind tunnel wind velocity. thanks

[davidwilcox]

Manmeet,
Atmospheric flows are fully turbulent meaning it does not transition to turbulence. Hence there is no critical Reynolds number. You mentioned earlier in your series of posts that you have some simulations of building flows. In those simulations, did you use the log law as your velocity profile at the inlet? Also, what is generally done in wind tunnels to simulate atmospheric flows is to thicken up the boundary layer so that it resembles a log law. This is through roughness elements. The size and displacement of the elements must of course be such that the roughness reynolds number be fully rough. If you have access of Schlicting's book on boundary layers, you can find more info there.
Look up this paper:
Effect of roof-mounted solar panels on the wind energy exploitation on
high-rise buildings

There are some references in there that you can chase after.

[manmeet singh]

Thanks David for your reply.

[Lucas_Gasparino]

Originally Posted by davidwilcox

Manmeet,
Atmospheric flows are fully turbulent meaning it does not transition to turbulence. Hence there is no critical Reynolds number. You mentioned earlier in your series of posts that you have some simulations of building flows. In those simulations, did you use the log law as your velocity profile at the inlet? Also, what is generally done in wind tunnels to simulate atmospheric flows is to thicken up the boundary layer so that it resembles a log law. This is through roughness elements. The size and displacement of the elements must of course be such that the roughness reynolds number be fully rough. If you have access of Schlicting's book on boundary layers, you can find more info there.
Look up this paper:
Effect of roof-mounted solar panels on the wind energy exploitation on
high-rise buildings

There are some references in there that you can chase after.

Hello!

I'm having a similar problem to Manmeet, although my dimensions are not so big.
Our uni wind tunnel is 1.5m square section and my estimated Re was 7.8e6. The original airspeed is already 20.125m/s, and scaling it based on Re always results in the Mach number getting too high.
My model is composed of 2 buildings, where my interest is focused in the gap between them (although not so narrow, it still increases wind speed considerably). We'll use a log profile for the CFD, but i wanted to know if I can take down the Re for the wind tunnel, as long as it remains turbulent.
We have nothing larger or more powerful than this wind tunnel, and I want to avoid the water channel facility as it has no PIV system yet.