1:20 scale model wind tunnel build - help!

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic


   Intention: 
-	To build a 1:20 scale wind tunnel
-	To print models of HPV’s using a Makerbot Replicator 3D printer (I’ve just bought a s/hand one for work-ish )
-	To be able to test the relative CdA of various models using a repeatable, accurate test

Method
-	The wind tunnel is a straight-through wind tunnel using a ‘contraction’ section, a ‘test’ section, and a ‘diffuser’ section, with the fan at the rear to pull the air through
-	The whole wind tunnel will be made of 4mm Perspex, as I got a large sheet cheap at B&Q (damaged on one corner)
-	The design of the wind tunnel is as shown on the drawing below:

- 	The drag will be tested by mounting the model on a re-usable Lego ‘dolly’ and this will push against a digital weighing scale. I realise that I could use strain gauges instead but I don't want to get into electronics...

Issue
-	What wind speed to use in the tunnel?
-	My understanding is that there are two equations to use – a ‘drag force equation’ and a ‘Reynolds number equation’
-	Drag Force equation: Fd = ½ ρ v² Cd A
-	Reynolds number equation: Re = (ρ v L)/μ
My simple understanding (maybe incorrect?) is that Cd is related to Re, so in the Re equation, if L divides by 20, then v needs to increase by 20.
So, if the real world speed is 10m/s, then a 1:20 scale wind tunnel needs to be v x 20 ie 200m/s (or about 450mph). Is this correct?

Achieve-ability
Even at this small 1:20 scale the wind tunnel is 1.2m long, using a 380mm domestic box fan. I’ve yet to get hold of the fan, but I’d guess that the speed of air will be about 10m/s. Using a 4:1 ratio for ‘test area’ to ‘fan area’ means that I can only achieve about 40m/s, or 90mph.
I don’t know what effect that this will have on the validity of the tests. Will this make the results invalid? If it does, is there any way to correct them, other than:
-	Increasing fan speed
-	Using water instead of air (not practical)
-	Increasing the scale of the whole thing


Edited by russellbridge - 02 May 2015 at 11:38am

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
russellbridge Members Profile Send Private Message Find Members Posts Add to Buddy List BHPC Member Joined: 21 October 2013 Location: Manchester Status: Offline Points: 843	Post Options Post Reply Quote russellbridge Report Post Thanks(0) Quote Reply Topic: 1:20 scale model wind tunnel build - help! Posted: 02 May 2015 at 11:33am
	Intention: - To build a 1:20 scale wind tunnel - To print models of HPV’s using a Makerbot Replicator 3D printer (I’ve just bought a s/hand one for work-ish ) - To be able to test the relative CdA of various models using a repeatable, accurate test Method - The wind tunnel is a straight-through wind tunnel using a ‘contraction’ section, a ‘test’ section, and a ‘diffuser’ section, with the fan at the rear to pull the air through - The whole wind tunnel will be made of 4mm Perspex, as I got a large sheet cheap at B&Q (damaged on one corner) - The design of the wind tunnel is as shown on the drawing below: - The drag will be tested by mounting the model on a re-usable Lego ‘dolly’ and this will push against a digital weighing scale. I realise that I could use strain gauges instead but I don't want to get into electronics... Issue - What wind speed to use in the tunnel? - My understanding is that there are two equations to use – a ‘drag force equation’ and a ‘Reynolds number equation’ - Drag Force equation: Fd = ½ ρ v² Cd A - Reynolds number equation: Re = (ρ v L)/μ My simple understanding (maybe incorrect?) is that Cd is related to Re, so in the Re equation, if L divides by 20, then v needs to increase by 20. So, if the real world speed is 10m/s, then a 1:20 scale wind tunnel needs to be v x 20 ie 200m/s (or about 450mph). Is this correct? Achieve-ability Even at this small 1:20 scale the wind tunnel is 1.2m long, using a 380mm domestic box fan. I’ve yet to get hold of the fan, but I’d guess that the speed of air will be about 10m/s. Using a 4:1 ratio for ‘test area’ to ‘fan area’ means that I can only achieve about 40m/s, or 90mph. I don’t know what effect that this will have on the validity of the tests. Will this make the results invalid? If it does, is there any way to correct them, other than: - Increasing fan speed - Using water instead of air (not practical) - Increasing the scale of the whole thing Edited by russellbridge - 02 May 2015 at 11:38am

jdp298 Members Profile Send Private Message Find Members Posts Add to Buddy List Visitor (moderated) Joined: 24 February 2012 Status: Offline Points: 32	Post Options Post Reply Quote jdp298 Report Post Thanks(0) Quote Reply Posted: 03 May 2015 at 10:28pm
	You could increase velocity to make up for the lack of length, but that'd be placing you in unrepresentative velocities. It would be better to accommodate Re through temperature and or pressure (e.g. ¼-scale would need x4 pressure, or lower temps to increase viscosity). However neither are going to be possible in your setup. Cd is related to many things, mostly down to bluff-ness of the bodies, their length, lift characteristics and separation i.e. where the transition from laminar to turbulent occurs. There are also some engineering bodges which sometimes work in your favour. I mean that through a range of Re, Cd may not alter by a large amount, which reduces your need to scale pressure and viscosity as much. These are fairly well known for aircraft components, but not so for bikes. So you won't get absolute data. However, you can represent relative effects, and measure one design against another. Similarly you can seek reductions and improvements in designs , even if you won't have a truly scalable number to set against it.

GeoffBird Members Profile Send Private Message Find Members Posts Add to Buddy List BHPC Member Joined: 20 September 2005 Status: Offline Points: 2396	Post Options Post Reply Quote GeoffBird Report Post Thanks(0) Quote Reply Posted: 04 May 2015 at 3:08pm
	Hi Russell - yes your understanding about wind speed required for the same Reynolds Number is correct. As Jon says, having a different Re doesn't invalidate e results - we used to test 40% or 50% models of Le Mans cars at about 30 m/s, which is equivalent to about 30 mph full scale. With experience, the tunnel results could be correlated to reality, but it was known, for instance that diffuser ramp angles could be greater on the real car as separation did not occur as it did on the model. I would think that 'purer' streamlined shapes would be less affected by such phenomena though. You will get better results the less 'blockage' there is from the model, so the percentage of the tunnel test-section cross sectional area filled by the model frontal area. IIRC, they usually tried to keep this below 10% but I could have remebered this wrong. High blockage causes a venturi effect that speeds the air up over the model. Again, this is probably less critical for streamlined shapes - an Formula One car rear wing sends a rooster tail of air about 2 metres upward, so require tall test sections. Wind tunnels, like CFD are best for comparisons unless you have lots of data to get a correlation between reality and model testing
	Right Time - Right Place - Wrong Speed

Yanto Members Profile Send Private Message Find Members Posts Add to Buddy List Moderator Group Joined: 11 July 2005 Status: Offline Points: 1521	Post Options Post Reply Quote Yanto Report Post Thanks(0) Quote Reply Posted: 06 May 2015 at 12:34pm
	Russ, I can't help with the theories. However, i have an observation, using the digital weighing scales at 90 deg to their designed orientation may introduce all kinds of variables that you cannot quantify, these may also vary with changes in applied force.

russellbridge Members Profile Send Private Message Find Members Posts Add to Buddy List BHPC Member Joined: 21 October 2013 Location: Manchester Status: Offline Points: 843	Post Options Post Reply Quote russellbridge Report Post Thanks(0) Quote Reply Posted: 16 May 2015 at 8:32pm
	Well, the wind tunnel is all built now and the first run is done, with a small cube in the tunnel to test it, but the fan isn't strong enough. The 'desk fan' Draper box fan is only producing 9mph wind speed in the smaller test section, giving tiny readings of about 2g of sideways force. (the digital scales on their side seem to be functioning OK so far FYI Ian, and are self zero-ing upon turn on - can't grumble for £5 off ebay!) Time to find a new fan - I think I can squeeze a decent 350mm axial extractor fan in there which should give 2600m³/hr airflow. Will report back with some photos / videos when it's working a bit better. Russ ps Geoff I didn't get that .stl file - let me know if you tried to send it.

Yanto Members Profile Send Private Message Find Members Posts Add to Buddy List Moderator Group Joined: 11 July 2005 Status: Offline Points: 1521	Post Options Post Reply Quote Yanto Report Post Thanks(0) Quote Reply Posted: 16 May 2015 at 9:10pm
	look out pedant attack: when i were a lad grams weren't a force, can i have my force in Newtons please? something like 0.0196133N

GeoffBird Members Profile Send Private Message Find Members Posts Add to Buddy List BHPC Member Joined: 20 September 2005 Status: Offline Points: 2396	Post Options Post Reply Quote GeoffBird Report Post Thanks(0) Quote Reply Posted: 17 May 2015 at 4:17pm
	Sorry Russell - slipped my mind. You should have it now.
	Right Time - Right Place - Wrong Speed

GeoffBird Members Profile Send Private Message Find Members Posts Add to Buddy List BHPC Member Joined: 20 September 2005 Status: Offline Points: 2396	Post Options Post Reply Quote GeoffBird Report Post Thanks(0) Quote Reply Posted: 17 May 2015 at 4:27pm
	You follow in illustrious footsteps - here is the wind tunnel built by the Wright brothers in 1901: Exciting project!
	Right Time - Right Place - Wrong Speed

AlanGoodman Members Profile Send Private Message Find Members Posts Add to Buddy List Admin Group Club Chairman Joined: 04 March 2005 Location: United Kingdom Status: Offline Points: 8036	Post Options Post Reply Quote AlanGoodman Report Post Thanks(0) Quote Reply Posted: 17 May 2015 at 5:10pm
	Pah! That'll never get off the ground...


GeoffBird Members Profile Send Private Message Find Members Posts Add to Buddy List BHPC Member Joined: 20 September 2005 Status: Offline Points: 2396	Post Options Post Reply Quote GeoffBird Report Post Thanks(0) Quote Reply Posted: 25 May 2015 at 2:01pm
	Russell has very kindly made me a !:20 scale rapid prototype model of my trike:
	Right Time - Right Place - Wrong Speed