Tested: vortex generators (AirTabs) on a Toyota Corolla

[MetroMPG]

OK, pardon the poor image quality - because these are screen caps from some poor quality video.

Road speed is 70 km/h (43.5 mph)



Top = no vg's; bottom = vg's @ trailing edge of roof.

It's hard to see in the top pic, but the closest, middle three light colour tufts were swirling randomly, standing up at times.

FYI, the airtabs were duct taped to the roof of the car to facilitate installation/removal. The tape is flush on the roof & Airtab surfaces (the boundary layer here is significantly thicker than the tape anyway, so it would not affect the results).

Eight were installed (yes, one is black, the others are all white).



In this pic, the car is going about 50 km/h (30 mph).

Bad photography on the upper image. But you can even see the light coloured tuft on the right side of the center mounted brake light is standing up off the trunk lid.

[chuckm]

That's pretty dramatic. Where did you get the VGs?

[donee]

Hi MetroMPG,

I think the Airtabs are too tall for cars. My guess is they need to be about 1/3 the height. The boundary layer on the low energy vertical side of a truck near the rear is very thick. On a car, its much less.

You might try shaving down the central blade on them , and filling the gap with silicone sealer, epoxy or body filler. Then shape it smooth and blend in.

[donee]

Hi MetroMPG,

Looking at your pictures more carefully, you probably do not need the two outside Airtabs. Also, your spacing is much closer than recommended. So, loose two on the inside, and respace out the remaing four.

Remember each airtab creates two vortices. This is unlike the simple delta shaped VG's, which only create one vortice.

[garys_1k]

Nice job, they sure LOOK like they'd make a significant, positive difference in FE. The improvement in flow is so obvious that the lack of a clear, positive FE test result is very surprising.

The fact that the FE is unchanged, plus the extra noise, makes the pretty flow pattern pretty unimportant for everyday use. I guess I'd leave them off, or retry the FE tests at a higher speed to confirm the non-improvement.

[MetroMPG]

Hi Donee -

Mitsubishi measured the boundary layer thickness at the trailing edge of the Evo roof at 30 mm - just over 1 inch, which is close to the height of the AirTabs (1 inch = 25.4 mm).

In testing different VG heights, Mistubishi says:

...the drag reduction effects for the VGs of three different heights (15 mm, 20 mm and 25 mm) were all equivalent to –0.006.

That's not to say the AirTab is the ideal shape though. Again going back to the Mitsu example, they specifically chose a VG style (delta-wing shape) because it had a small frontal area. The AirTabs seem to have quite a large frontal area (even considering they create a pair of vortices), and that may be part of the reason why they didn't reduce overall drag & show any fuel economy improvement.

chuckm - an AirTabs distributor sent some to me.

[MetroMPG]

Originally Posted by donee

Looking at your pictures more carefully, you probably do not need the two outside Airtabs.

Agreed there, after seeing the non-vg tuft results on the sides of the rear window.

Also, your spacing is much closer than recommended.

Actually, they were slightly wider than recommended (which is 4 inches on center from the installation PDF). These were set at about 4.5 inches on center.

[MetroMPG]

Originally Posted by garys_1k

Nice job, they sure LOOK like they'd make a significant, positive difference in FE. The improvement in flow is so obvious that the lack of a clear, positive FE test result is very surprising.

I'll admit my jaw dropped when I saw the obvious change in flow shown by the tufts!

But I had to keep reminding myself: vortex generators themselves create drag. If the size, style & placement isn't optimized, you can end up with a net increase in drag (or zero sum gain), even while you apparently improve downstream flow.

We could just as easily stick a wind deflector at the trailing edge of the roof to force air down the rear window and tidy up the tufts. But we'd end up with turbulence aft of the deflector negating the gain of the improved flow on the glass/trunk lid.

[chuckm]

Well, these two videos are based on my 02 corolla...
First, without the vortex generator

http://s13.photobucket.com/albums/a2...nt=Vortex2.flv

[MetroMPG]

Chuck - is it really a vortex generator in that simulation, or just a 2d bump?

Videos coming shortly.