Particle Image Velocimetry (PIV) Explained - How do we see airflow in wind tunnels?

How do we tell what is going on in air, when we can't actually see it? How does PIV work in wind tunnels? Today, I explain PIV (Particle image velocimetry), a great way for determining the velocities and structures of airflows in wind tunnels.
Facebook: / kyleengineers
Patreon: / kyledrives
Instagram: / shadowraceengineering
Race car consulting: www.jkfaero.com
G+: plus.google.co...
RUclips: / kyleengineers

[@andjelkakosovac9979]

this is amazing. Thank you for showing us this. I am third year of mechanical engineering and we are talking about piv in the class called dynamics of the fluids, so it is really amazing seeing it like this!

[@beachboardfan9544]

How do you keep the particles from collecting on the camera lens, subsequently corrupting the data?

[@aydagr3219]

Thank you very much! This video was super useful as I am working on the
flow past a cylinder and really needed to know how the PIV set up is .

[@a2pany284]

Very informative and well-described video

[@jorgesilva5337]

Hi, how do you reduce the vibrations being transmitted to the camera (both from the wind tunnel structure and those induced by the flow)? How much would these vibrations affect the measurements? Thanks!

[@namHin22815]

Great explanation. Thanks a lot

[@a2pany284]

Is the camera used the setup also has a 50 micro-second exposure time (~20K fps) ?

[@espinetmatt]

So those of us who don't have access to stuff like this. How accurate are CFD software compared to analyzing like how you demonstrated in this video?

[@KYLEENGINEERS]

+shadowgoat Depends completely on the software and how well you set it up. For this case a 26 million element mesh running LES on Ansys Fluent for 2 months on a supercomputer still had some correlation issues, but was good enough for investigating large amounts of the slow structures. For full car CFD, you can usually get within 15% accuracy (in lift and drag) if you have a decent model, closer if you have correlated your model specifically to your scenario, like many F1 teams do. CFDs greatest strength is in showing trends though. When I do a full car CFD for a client, I typically use a turbulence model that promotes early separation, and thus produce a conservative aero kit that is less sensitive to real world disturbances. It's not on the knifes edge of performance, but better to be safe than be having separated flows in undesirable locations! CFD is a great tool, but the problem is that so many people can operate CFD software, but so few truly understand it, and that is the difference between 10% accuracy and 50% accuracy.
Thanks for watching!

[@espinetmatt]

+KYLE.DRIVES Followup question. Since you have the camera behind the object in the wind tunnel, this cant be used to look at flow patterns that may be obstructed, for example how a front splitter affects aerodynamics. Is PIV only useful for looking at air flow coming off an object?
Also, do you have access to one of these any time? If so, how?

[@qtheplatypus]

B Dykes The University of Sydney rents its wind tunnel at $190 US an hour with a minimum of 8 hours booking plus extra charges for calibration of the equipment.

[@erockromulan9329]

We just got one of these at work. I get to help set it up. Whee

[@ghmsad6119]

Can anyone help me how to arrange camera settings for particle imaging

[@aleksmaksimovic9948]

Oh my god, this is so cool.

[@yuvanscircle3437]

Amazing stuff! Really enjoyed what you did there. I see that you used MATLAB to find the velocity field. Any chance I can see your script?

[@BlahBlahBleh317]

Awesome video!

[@saikiran-zq7bs]

what is the intensity of light used?

[@m.sc.ahmedkmashan7282]

thanks ..can i show more ?

[@vagneroliveira4863]

Wonderful video! Congratulations and thank you for sharing. I am professor at mechanical engineering in Brazil and I am estimating a system of similar acquisition to that. Please to find out specification of the system for it estimates the purchase of one. Is it possible to send me your e-mail for larger details?

[@The419Productions]

is this similar to the Brownian motion experiment?

[@KYLEENGINEERS]

+The419Productions I'm not very familiar with Brownian motion, but my understanding was that it was to do with the random movements of particles in a stationary fluid, while in this case we are merely using the particles to track the moving fluid, and are assuming that they are not moving randomly within that fluid. At least, assuming the random motion is small compared to the prevailing motion. So from that perspective, I would say they are not similar.
Thanks for watching!

[@The419Productions]

Thanks!

[@rubenardesch]

magnificent

[@KYLEENGINEERS]

+Antonius Ardesch Thanks for the compliments!

[@samchen3820]

where can i get the goggles?

[@aydagr3219]

Thank you very much! This video was super useful as I am working on the
flow past a cylinder and really needed to know how the PIV set up is .