Tuft Testing - "Vortekz" vortex generators on 2003 Jetta Rear Window
 

I've seen a few people attempt to find fuel economy improvements using vortex generators. However most folks were using "Airtabs", which are a larger flow disturbance vs. a smaller delta-wing style VG. In addition many folks were also placing them at the very rear edges of their bluff body style vehicle, or at the trunk-lid edge of sedans/hatchbacks. That doesn't seem like the ideal application to me.

I went with the "Vortekz" delta-wing type: Welcome to Vortekz.com

Their website is not too impressive to say the least.

Nevertheless, I bought some. The only other person I saw using these post on a forum such as this was a thread over on gassavers with a Mazda 6 sedan. I think he's getting the best documented mileage of anyone with this car. Otherwise, these VGs mimic the type used on Mitsu Lancer EVOs and Subaru WRX STIs to direct air down and around the rear spoiler. I don't think anyone could deny that they work in this context. If we can visualize them bending down the airflow, and accept the Mistu paper on drag reduction (with the spoiler), then it can't hurt to give this a try.

Those looking for measured fuel economy impact of my application will be disappointed here. Sorry. Too many mods going on, and 700 mile tanks just last to long to do them one at a time. In any case, I did what I could with some tuft testing.

The array:
http://i287.photobucket.com/albums/l...z/IMG_1731.jpg

Stock roof

http://www.youtube.com/watch?v=0DBnJn1MWQ8

http://i287.photobucket.com/albums/l...ckTuftTest.jpg

So there is obviously detached flow in most of the center of the car. I was a little surprised with how well flow stayed attached down the sides. Or at least visually "attached". You can see one of the center trunklid tufts just dancing straight up into the air though. Other window glass tufts turn front to rear. But it appears to me that the flow definitely comes over the roof and partway down the rear glass before severely separating. This was a good find for my selected VG location.

Installation:
http://i287.photobucket.com/albums/l...z/IMG_1734.jpg

http://i287.photobucket.com/albums/l...z/IMG_1736.jpg

Back onto the road for a quick test at speed:

http://www.youtube.com/watch?v=98gsuMJRQMk

http://i287.photobucket.com/albums/l...GsTuftTest.png

The still image with the VGs looks a little blurrier as the sun was setting. Obvious flow attachment improvement down the rear glass and deck. Cool. Whether this saves me fuel or not may remain difficult to ascertain. I think I'll refill 'er with bio asap so I can try to asses them on the next tank vs the current one.

These things really don't add much in the way of frontal area. Technically none, but the if the airflow is attached when it hits them, there is obviously going to be an impact on drag.

http://i287.photobucket.com/albums/l...z/IMG_1747.jpg

http://i287.photobucket.com/albums/l...z/IMG_1748.jpg

View from interior:
http://i287.photobucket.com/albums/l...z/IMG_1749.jpg

http://i287.photobucket.com/albums/l...z/IMG_1746.jpg

http://i287.photobucket.com/albums/l...z/IMG_1754.jpg



I also threw some tufts onto my passenger side glass where I recently did a flush mirror delete. I'll post up the results of that vs. a small safety mirror over in my general build thread.

Man, over 40 views and no comments. :rolleyes: Here I thought I was doing something useful, too...

I remembered Trebuchet's CFD thread from a while back and decided to compare my results. I'd say he was doing some pretty good modeling!

I drew a line of my estimated flow separation point on this shot:
http://i287.photobucket.com/albums/l...TuftTest-1.jpg

Vs:
http://ecomodder.com/forum/attachmen...1&d=1226637053

I'm also now more encouraged to do my decklid extension, since he got these results with one modeled:

http://ecomodder.com/forum/attachmen...1&d=1226793943

If the VGs are drawing more flow down the rear glass and onto the decklid, further airflow smoothing via an extension should help minimize the total wake, too, no?

So what do you guys make of all this?

Actually, it kinda looks like the decklid extension is creating a much larger wake, but then I could be wrong.

Definitely improved flow to the rear window there. I wonder if there's a way to figure out how far ahead of the flow separation you need to put the VGs for optimal benefit.

I have to wonder if you could use fewer VGs and get close to the same improvement. Your original air flow near the sides of the rear window was pretty well attached.

I believe it was MetroMPG who did a similar test on another car using Airtabs. Seperation results were the same as yours.
He also tested airtabs on a minivan of some persuasion. In both cases there was no difference in mpg.
Putting the VG's in an area of seperated flow (Maybe the back edge of your deck lid) would be useless.

I cannot tell a difference between extension & not from the pictures.
Are you planning on making the extension adjustable for testing or putting VG's at the rear of the extension?

I notice similar flows on my 07 Civic. We often get misty rain in the Pacific NW, and any wind starts to produce drops which then flow in the path of the wind. There's always a "dead spot in the lower middle of the rear window.

One thing I was wondering about was if the wind noise would be reduced if the airflow is smoother. Did you notice any difference?

Very interesting thread.
Nice cfd, however i think you actually need to make everything correct with wheels and flow into engine since that changes alot.

From the decklid extension I dont see that he got any improvement. Most likely alot of air will pressure bellow the decklid. However with an diffuser you might have another result

Its intresting with these vortexes that separates air and also with different irregularity that causes air to slow down so it more easily follow edge etc. It works for golfballs atleast:)

Yeah I wondered that too right after I stuck the last VG on. doh. I am contemplating pulling the very outer ones back off. As you say the airflow out there was already fairly well attached, so VGs out there can only be causing extra drag...

As per the deck lid extension, well, I guess I'm not so sure. But the CFD pics above were done with NO vortex generators. I pleaded with Trebuchet to run some of those in a simulation but his interest waned before it happened. Not sure whether the CFD could properly model those, but it would be interesting.

Only got one rainy day of commuting with these on, but the drive to work was my best ever scan gauge reading. Also a downpour that forced slower driving. Drive home was clear, and only average mileage. Have not noticed any less wind noise, but seems like that would only be noticeable in the back seat anyway, since my airflow up front is unchanged.

What you are seeing on the sides of the rear glass are C pillar vortices IMHO.

Good experimenting, Deezler.

I wouldn't expect to see any MPG improvement show up - it'll probably be too small to see above the normal statistical noise of tank-to-tank variation. Recall that Mitsubishi estimated a drag reduction of just -0.006 in their study.

I like your idea of putting the VG's aft of the physical "peak" of max frontal area projection. And they still clearly influence airflow placed at the top of the window!

The $60,000 question is whether their additional induced drag (from their vortices) is greater/equal/less than the potential for drag reduction because of a smaller wake.

In each case where I tested VG's, I felt there was more noise when they were in place. Subjective though. Take 2 grains of salt and call me in the morning.