Desktop Wind Tunnel

Inspired by Windsible, I built a desktop wind tunnel for 1/43rd scale model race cars. I use it to visualize the flow of air over them with a small smoke machine - I’m not looking to take measurements off of it (yet!)

I went through a bunch of iterations on this - initially I had contraction / expansion cones at each end, but I couldn’t get a stable (“laminar”) flow through the tunnel, so I dropped those and packed the inlet with drinking straws to prevent minor disturbances outside the tunnel from introducing instability inside, 

Then I had a problem with pulses from the pump in the smoke generator - so I “buffered” the smoke under the “road”, letting the low pressure in the chamber “suck” it out - this produces best results, damps out the pulses from the pump in the smoke generator and giving you nice clear smoke lines.

Finally I added a HEPA filter before the fan - it does introduce some blockage, but allows you to run the smoke without filling up the room and setting off the fire alarms.

Its best run at speed ~10 (about 60% power) without the smoke actively “pumping”, just letting it be drawn out the chamber.

Bill Of Materials (BOM)

Wiring Diagram

Wiring is pretty straightforward - connect up the Blue PWM signal on the fan to the yellow wire on M5Stack Dial B connector, and connect everything else to power

Software

The software just controls fan speed for the moment - I want to put in load sensors, control the LED colors etc when I get a chance.  I’ve used UIFlow for now, although I intend to rewrite it in C when I get the time.  The software is here: https://uiflow2.m5stack.com/?pkey=4e813d3c3ed54bbd9518e13fe99b5789

Parts Production

Some work is needed to produce these parts:

• I didn’t model the honeycomb on the grills or flow straightener, so you need to remove the top+bottom layers, use 10% honeycomb infill.  Make sure the screw holes aren’t floating!
• Print the back piece with thin, floating boxes up the side of it for supports to latch on to, otherwise it moves around a lot and is a messy print.
• Print the road on a textured build surface to make it look and feel like a road; you can also use a MMU (or AMS, or 3 toolhead XL) to print the white and read grid box markings.
• Laser cut acrylic sheets (use the DXFs provided).  Use 2mm clear acrylic.  I initially ordered them online but eventually brought a Glowforge to do this.
• Glue together the various components for the door - theres a handy jig STL to clamp everything square.  Use plenty of glue and clamps.
• Cut drinking straws to 148mm
• Glue in the o ring to bottom cover (to stop smoke leaking)
• The smoke rake and smoke elbow are best printed on an SLA machine - they will come out okay on FDM, but you can order them online from PCBWay, or use this as an excuse to get a SLA machine…

Assembly Steps

1. Screw together base
2. Build output fan section
3. Route the wires for the LED strip through the back and screw on the LED holder.
4. Route the wires for the LED through the base, but don’t bolt on the back yet
5. Assemble input and output sections on to the base
6. Assemble back
7. Assemble door
8. Screw on bottom cover

Future Plans

• I now have an XL (yay!) so I plan on making the base printable in one piece - this will mean shortening the input and output sections, which I don’t think need to be that big.  This should make assembly a bit easier and reduce the amount of printing time.
• I also plan on adding more o ring material around the door, particularly at the hinge - I think air is getting in there and disturbing the flow.  I want to experiment with multi-material printing and print the o rings in place using TPU…
• I think I’ll add some load cells and servos so I can print a little wing to sit in there, vary the rake and wind-speed and look for stalling.
• I’m considering moving the light back towards the back and making the door cover the whole top, so you can see the flow structures across the car and not just from the side.  This will also mean the back is easier to print…

Research

I don’t know how to build wind tunnels, so I did a bunch of research on how to get a stable flow.  Here are some of the helpful links:

Reddit

• https://www.reddit.com/r/3Dprinting/comments/vsr7r3/i_made_a_mostly_3d_printed_wind_tunnel_idk_if/
• https://www.reddit.com/user/shanchou
• https://www.reddit.com/r/3Dprinting/comments/1c4tvue/hi_everyone_weve_made_some_tweaks_based_on_your/
• https://www.reddit.com/r/3Dprinting/comments/z9fca0/i_3d_printed_a_wind_tunnel/
• https://www.reddit.com/r/engineering/comments/jwz55j/desktop_wind_tunnel_one_hell_of_a_desk_toy/
• https://www.reddit.com/r/3Dprinting/comments/1bvtdux/my_friend_made_this_desktop_wind_tunnel_and/

Similar Products / Projects

• https://www.interactiveinstruments.com/educationalproducts/jetstream/
• https://engineeringlibrary.org/reference/laminar-and-turbulent-fluid-flow-doe-handbook 
• https://www.grc.nasa.gov/WWW/k-12/airplane/tunnozd.html 
• https://www.printables.com/model/849713-modular-wind-tunnel-for-stem-education/files 

Youtube Playlist: Desktop Windtunnel