Video: tuft testing front wheel skirts (Geo Metro)
 

(See also: http://ecomodder.com/forum/showthrea...etro-3692.html

And -
Next step: Designing/building front wheel skirts v.1.0 (Geo Metro))

Basjoos, you're going to have company soon!

FYI, two EM members that I know of have installed front skirts: see http://ecomodder.com/forum/fuel-econ...cations.php#37

No video yet. I'll post it tomorrow probably when I've got the "skirt on" (cardboard) footage.

But a few screen grabs of the stock situation to whet your appetite:

0 km/h:

http://ecomodder.com/imgs/metrompg-t...d-cap-0kph.jpg

60 km/h:

http://ecomodder.com/imgs/metrompg-t...-cap-50kph.jpg

The video pretty much shows what you'd expect to see (detached flow in the first 2-3 rows aft of the wheel well, and noticeable wiggling in the tufts the rest of the way back).

What will be interesting to see is how much calmer those rear-most tufts will be with the skirts on.

You should see the camera rig I set up to capture this stuff. :)

Yes, we should. However, WE can't post pics of it. :P

Good stuff Metro! :thumbup:

Also, would you mind doing some even higher speed test (100 km/h / 60 mph) for us that drive faster?

Regarding speed: tuft testing doesn't require high speeds. It's just flow visualization. Speed matters more when trying to quantify some aero mod, where 100 km/h may reveal something you can't detect @ 60 km/h.

Also... my camera rig probably wouldn't permit very steady pics @ 100. Or stay attached to the car, for that matter! :D

Julian @ AutoSpeed seemed to like 70-80 km/h for his tuft testing experiments. I was aiming for 60 km/h.

-----

I captured the "skirts-on" video this AM, and again I was surprised by the results!

Just as I was surprised at what the tufts showed on the rear window/trunk of the Corolla after sticking AirTabs on the roof, sticking cardboard wheel skirts on the Firefly had a bigger apparent impact than I expected.

Here's a still captured from the "skirts-on" video (I'll post that later today):
http://ecomodder.com/forum/attachmen...1&d=1223218428

What I saw was ALL tufts streaming back cleanly/quietly, with the exception of the one on the bottom row immediately aft of the skirt (the curve at the aft end of the skirt was too great = flow separation). Occasionally the tuft aft of that one would wiggle too, but not as much.

The biggest surprise to me were all the other tufts on the door. Not nearly as much wiggling as in the first vid. The noticeable effects of an open wheel well extend much further back than I would have guessed.

More pics & vid coming...

I've been thinking that the best way to hinge a tight front skirt is from the top of the fender, so that it stays roughly parallel to the airflow when it opens. I think it would be adequate to let the tires rub on the skirt to push it open, although rolling elements could be added. At low speed, when you would usually want the skirts to open, only light retaining springs would be needed. To keep them closed at speed, it might be nice to supplement the spring with a large bellows fed by ram air. Speed-controlled latches would not allow skid recovery on ice. Mechanical interlinks with the steering system might have unfortunate feedback, but servos controlled from the steering wheel could be ideal. No rubbing; you'd just have to remember the extra width in tight spots.

In general, I think that fender skirts would benefit by overlapping the body somewhat, rather than filling the existing hole.

Bicycle Bob:

I'm planning to hinge mine at the top, as you suggest and as basjoos did on his Civic: http://ecomodder.com/forum/showthrea....html#post2112

Though once I start construction, I'll probably wish I'd followed the "rubber membrane" method Newton used on his Mira: http://ecomodder.com/forum/showthrea...kirts-398.html

I also like your idea of servos for hinge control (hadn't thought that there might be feedback if there were a direct link to the steering bits).

But I'm going to keep it simple and go with a low friction plastic stuck on the back side of the skirt which the tires will contact when they are steered. basjoos went with low profile metal rollers.

---

For chuckles, some pics of the camera rig, or more technically, "web cam on a metal stick, secured with duct tape, a nylon strap, rope and a bungee":

http://ecomodder.com/forum/attachmen...1&d=1223222514

The boom is an extruded aluminum U-shaped member (the track from a large vertical blind), secured to a roof rack. Between the "skirt off" and "skirt on" video sessions, I added the rope that goes from near the end of the boom to the bottom of the passenger door, tensioned with a bungee cord to the rear of the car. That put some pre-tension on the boom and reduced the camera bounce/wiggle compared to the runs without it.

http://ecomodder.com/forum/attachmen...1&d=1223222493

http://ecomodder.com/forum/attachmen...1&d=1223222493

It stuck out about 4 feet from the side of the car.

I was careful not to overtake cyclists or drive close to sidewalks or utility poles! ;)

Wow, thanks for the details. I had thought that Basjoos' skirts were fixed. So much to read. . .

Thank you for doing this !
I'm amazed that you did not get pulled over and ticketed for having something extending that far away from your car !

Very impressive.


( What surprises me is that your hubcaps are totally smooth, and your wheel well gap is tiny compared to most cars, yet there was such a large effect to be seen ! Imagine the effect on other cars ! )

While this smooths out the side more is it possible more turbulence is created underneath due to air not escaping from the rear of the well?

:D:D:D

Funny - a roof rack being used to help FE (in the long run). Nice setup.

I havent started designing my car yet, but I gave some thought to the wheel skirt problem for the front. I was thinking of having a flexible fabric of some sort and have a guide bar that was attached around the wheel to the steering mechanism, that would push the fabric clear of the wheel during turns.

Once I get my copy of SolidWorks I will try and draw up something for you to see. (Probably about two months out still, upkeep on my car just set me back $900)

Hello,

Have you heard of SpaceClaim? It is a solid modeler that costs much less than SolidWorks and it very intuitive to use -- a lot like SketchUp:

SpaceClaim | Paperless Parts
Paperless Parts

I don't think it would be a problem. You suggest that some significant volume of air is entering the well from somewhere and needs an escape. I don't think that's the case.

Video ready...
 

Better late than never...

http://www.youtube.com/watch?v=Lo3EEHDoUgc

Wow. That's pretty darn conclusive. It would be tuft to beat that mod for turbulence reduction.

Thanks for that. Very professional looking. My wheels are turning to figure out a set for my van.

Nice.

MetroMPG -

Great Job! It will be neat to see your execution of the solution.

CarloSW2

Awesome video MetroMPG! Dramatic difference I say. Keep up the great work :thumbup:

Very nice. I was surprised to see the improvement even above the wheel well. That is quite the difference.

Yes, but doesn't flow seperation happen earlier/easier (you know what I mean) at higher speeds?

perhaps it would be interesting to use the same setup to test vortex generators behind an open front wheel well.... front wheel skirts are not really an option for everyone,

but perhaps some vortex geneators just aft of the wheel could improve the situation over stock.
some people claim they should be placed in front of a wheelwell so that might be tested as well.

the full wheelwell cover will likely be superior, but it would be interesing to see where vortex generators are positioned betwee these two extremes

Oh the puns! :D

I was too, until I re-read a bit about wheel well/tire aerodynamics:

"The top of the wheel is moving into the air stream at twice the driving speed, and is trying to drag the air forward against the stream." (Contribution of different devices to the total drag - Chalmers University)

So the air being dragged up & forward at the top/outside of the tire collides with the normal air stream flowing aft ... and makes a mess. I think that may be what we're seeing above/aft of the top of the wheel arch.

I think VG's are a waste of time. They have to be incredibly carefully designed (access to good CFD or a wind tunnel/smoke test) to deliver results that in the end, even when optimized, won't even be detectable in the real world.

No argument there!

What I'm not sure is whether it's significant when we're debating a 25 mph difference (between 35 mph and 60). I'm guessing it's not significant, but I'll defer to someone who knows more than I do on this question.

Great video! That is quite the difference. Next you'll have to do some FE quantification without the rack and boom :thumbup:

I wouldn't think it is significantly different at that speed considering the angle of the skirt. It might show, but it looks shallow enough to not (but looks can be deceiving). Another test! ;P

If only my cruise control would work reliably, I'd be out there ASAP to get some ABA. :) I need to trouble-shoot it some more to see if I can figure out what's wrong with it (after replacing it once already).

Any chance you can repeat the tuft test with a partial cover? There may be significant improvement just by lowering the wheel arch a bit. That would be really helpful to less ambitious folks like myself. Thanks for the good work.

Hi,

Did you solve the wheel clearance on turns; and if so, how? Or was this just a proof of concept?

I believe Ford did something similar with the Probe V concept. Seem to recall Phil (aerohead) talking about it somewhere...

More video...
 

Here's the raw video, for the hard core aerophiles...

Check out what happens to the airflow around 1:10 (open wheel well) when I turn the wheels slightly to the left (about 1/4 turn of the steering wheel).

http://www.youtube.com/watch?v=xn8bEkriXq0

This was just proof of concept. The tire pushed the skirt off each time I turned the car around to repeat the run, and I stopped at the side of the road & re-taped it. But basjoos already solved the clearance issue: top hinge.

I think you're right: there's likely a good amount of improvement with a partial cover. I'm guessing approx the top 1/3 or so of the wheel. You'll see this on a number of aero cars (on the rear wheels) like the Mercedes Bionic car (you have to look closely; it's a clear plastic skirt) and their big 4-cyl diesotto concept saloon; even the GM EV1 only covered about half of the rear wheel.

Not going to repeat the test though. It's a lot of time & work! Plus, I'm more interested in the full meal deal.

Darin, awesome work as always.

Regarding partial cover, I'm confidend that anything you do to close the gap between the wheel and fender will benefit the aero of your car.

OK, so I'm sold on the tuft testing.

Next step: http://ecomodder.com/forum/showthrea...etro-5462.html

.

Or, maybe some jumbo zig-zag tape fore-&-aft of the wheel wells??
(but not this JUMBO)! :p
http://i46.photobucket.com/albums/f1...r/CRV/h012.jpg


I can see where adding VGs/ZZ tape forward of the wheel well might help some if you had moon disc wheel covers..
But, I wonder if they would help the turbulent air coming off the rear of the wheel well??
Does turbulent air get settled down when it hits a VG?

Xringer: do you have any info about the effectiveness of that kind of tape or VG's ahead of wheel arches? I know the AirTabs web site recommends that particular positioning, but I'm skeptical it does anything. They certainly don't offer any quality information about them.

I'm going to stick with things I know that work.

I think that Zig-Zag tape works, after reading the sailplane info/study and seeing what Neal Blanchard did with his DIY ZZ tape.

The wing-oil-flow pictures were impressive. It's hard for me to understand how something that thin can have such a positive affect. Preventing flow separation/ maintaining smooth air flow. (Like golf ball dimples?)

Since it seems to be doing something down about 1mm off the steel, I'm not sure exactly what it would do 'ahead of wheel arches'..
It might cause the flow to hang on to the metal and flow around inside the wheel well. (Making things worse?).

I think you are right about the airtabs. Sticking them 'ahead of wheel arches' would likely just cause more drag. Since they are so large..

I wouldn't put anything in around those arches. (Except maybe a wheel skirt).. :p
http://i46.photobucket.com/albums/f1...filledin-1.jpg

My crazy thinking about suppressing turbulence around the front wheel,
is to use a skirt (like in the above pic), or try a moon disc and some heavy duty ZZ tape 'ahead of wheel arches'.. Nothing too big. Maybe 1 or 2 MM thick. Nothing that will add more sq inches to the frontal area..
A ZZ patten that would generate a curtain of small vortexes, one every few mm.

In my day dream, these small vortexes off the ZZ tape would flow over the wheel well gap to the top of the moon disc. Spinning across the disc just above it's surface. Smooth!!
Then, there's the problem of the rear wheel arch gap.. :confused:
Maybe confused air at that point could be helped out by some more ZZ tape.

Like Neil used on his A-pillar.. But on the leading edge of the arch..?.
http://ecomodder.com/forum/showthrea...ds-2969-6.html

Maybe some before-and-after yarn testing just forward of the door would tell you something. I'm a novice at this stuff. Just trying to learn the ropes..

Good luck!

That tape is what i was looking for AFTER the wheel, it would help A LOT. But I may use dimple tape instead since it would look much better on the car.

Oh and I have VG's on my car ("Vortekz?) and I beleive that they do work. NOBODY with my car has been able to even get close to my mpg highway rating, even hypermillers. I did some testing (--- "vortekz" Generators Tested!! --- Pictures!! - GasSavers.org - Helping You Save at the Pump) and it showed MUCH better flow attatchment on the window which only leads to a better simulated teardrop shape.

BUT, i dont think airtabs work because they have way to much frontal area, as "vortkez" are extremley slim and have probably 1/20th the frontal area, so the net aero gain is negative as compared to the "Vortkez". Now if think ZZ tape is even better as it has practically NO added frontal area (But I didnt know about zz tape when I put my VG's on).

Hi,

Since zigzag tape is so thin, it cannot help reattach turbulent air. The pictures of the zigzag tape in the wind tunnel have it in the high pressure air on the leading part of the wing, and it helps create a slip layer near the surface.

It can't pull turbulent air in from a distance away.

separation
 

Seems like the big-dogs are going to argue that flow separation ,if it occurs,will be constant over the entire speed range above 20-mph,as at that velocity,Reynold's Number is fixed along with Cd.

wheel arch
 

I don't have any independent mpg numbers for partial covers but what I will tell you,is that if they're more than a couple inches deep ( 50 mm+) they'll tend to rub the tire,as I discovered on my Volkswagen, Honda,Dodge,and Toyota, and will need to be articulated as basjoos and MetroMPG have done.------------------------ For all the trouble,seems like we're better off doing the full,articulated skirt.Some early high-mpg concept cars began with fractional wheels skirts only to be superceded by full skirts in later endeavors.I think there's a big lesson there.

That's the thing about any sort off VG - which includes zigzag tape, dimples, etc... The whole theory behind it never involves turning turbulent flow into attached flow - it's taking turbulent energy and transferring a bit into flow about to detach.

Furthermore, I'm not convinced comparing a sailplane to a car is of any use. The ASK 21 sailplane has a max speed (where this tape would have most effect) of 175mph (if you can find minimums, that'd be great) and is a streamlined body to begin with. Maybe the lower pressure offsets the velocity discrepancy - I'll have to do a quick calc a bit later :)