Drag reduction modifications in wind tunnel: Audi A2 from Cd 0.288 to 0.204
 

Hi,

Thanks to Ausias who tipped me off with this Cd study:

http://wwwm.coventry.ac.uk/researchn...%20Vehicle.pdf

Lupos 0.25 Cd target starts to feel pretty plausible without chopping the roof ;).

Enjoy.

----

SUMMARY...

(added by MetroMPG, because these online files have a bad habit of disappearing...)

• START: MIRA wind tunnel showed a stock co-efficient of Drag (CD) of 0.288
• COOLING: vents, intakes, panel gaps taped on front end - Total Contribution: 9% of total Drag
• Underfloor:-Foamboard and tape -All components covered -Total Contribution: 7% of total Drag
• MIRRORS - door mirrors removed: -Total Contribution: 5% of total Drag
• Wheels and Wheel arches: Wheel Spoilers / spats; smooth wheel covers; Wheel Arch Blanking (rear fender skirts); Underfloor Wheel Arch Blanking
• Base pressure recovery (tailgate "box cavity")

http://ecomodder.com/forum/attachmen...1&d=1332944398

0.25
 

Cd 0.25 should be a walk in the park.I'm certain of it.Even lower if you like.:)

Awesome paper. Makes it easier to quantify every mod accordingly.

Good simple paper to guide mods. I'm a wind tunnel engineer at work and like seeing studies like this with measured data.

Thanks for sharing. :D

I had seen mention of the " box cavity" method of reducing drag at the rear in an old paper* from the UCDavis " future car " hybrid Ford Taurus .
It's good to see results quantified here .

* no longer online.

Vekke - nice find! I added the summary to post #1, since these files often disappear from the web after a while.

If you notice the caption on the bottom of the picture of the A2 at the top of the paper, there is mention of the A2 being introduced at .25 Cd .
This number is often given as the official Cd in published literature ( I've also seen .27 )
This makes me wonder what would have caused such a drastic rise in drag to .288 .

The only visible changes are the style of wheels, and surely this would only account for a count or two .

( And it also makes me wonder if they did the same mods to the .25 version, just how low of a Cd they would have achieved ! )

The A2 was available in several different trim levels/models , one of which was a high-efficiency version ("3 L" diesel). That's what the .25 figure refers to. It had a different grille, wheels, tires (size) and partial underbody panels (if memory serves), and possibly also a lower ride height.

Ride height is one mod our experimenters didn't attempt here.

Too bad those slides didn't give more info about the effects of the rear box cavity versions.

A2 3l Had 0.25
145/80R14 tires
Lower ground clearance
smooth hubcaps
almost fully blocked front grille
smooth underside to some extent
plus lots of small modifications

Lupo 3L has smaller mirrors I am not sure if A2 has them on also

More interesting info ..
 

This was just one presentation from the "The Aerodynamics Challenge Dissemination Event on 19 October 2011"

Some more presentation can be found here :
Coventry University - The Aerodynamics Challenge

I have often wondered if there is a "synergy" effect when combining different drag reduction mods. Would there be a perfect combination ?

This is a gold mine of information !
Thanks so much for sharing .


Is any of this presentation available on YouTube or elsewhere as a video ?
I would have loved to hear what the speakers had to say .

Plus a lighter rear seat. Did it also have lightweight (magnesium) suspension elements? This, of course, doesn't effect aero.

...this presentation (aft body taper) looks interesting too: http://wwwm.coventry.ac.uk/researchn...ple%20body.pdf

...and, here's Jaguar info on wheel design: http://wwwm.coventry.ac.uk/researchn...l%20Design.pdf

Lots of useful data which need lot of thinking to be able to use that when engineering/making parts. I liked those moving wheels, because I had one similar idea that was there also.

So is their 2 inch concave tail section showing me I should fit the skin 2 inches long to form a concave at the back of the Kammed off tail I am building for my Vetter style motorcycle?

I'm sure with a motorcycle, your Kamm will be much more pointed/tapered than a subcompact car. You could probably get away with six inches of cavity if it gave you much less added weight and if the Kammback was only a few square feet in area.

On a motorcycle, your Cd is basically just a raw factor of your frontal area. That's why a recumbent riding position is one of the most effective ways to bite a chunk off your fuel consumption. As long as your fairing is rigid in the wind, your mpg will kick butt. The rear back inches of the Audi fairing would end up being negligible on a motorcycle when compared to weight gain/frontal Cd.

Good luck with your motorcycle, sendler. We'll all love to see what you'll come up with!

As opposed to just squaring it off at half chord I should say.

An extra two inches wouldn't hurt. You'd probably see improvement. Beyond two inches of whatever the study Audi used was not stable enough to hold it's shape in the wind.

It's not a 3L version they started with, and a different tunnel.


What surprises me is the effect, considering it's only a small modification - Certainly compared to full boat tails or Kammbacks.


The A2 and Lupo 3L both had some expensive weight reduction measures incorporated, while still looking the same as the regular versions.
This led to their excessive price and the weak market response.

They don't seem to have a mechanism for allowing the wheels to turn though ?
Closing the gap is easy, making it actually work is something else again. ;)

What they didn't do is fit covers on the insides of the wheel, or easier, fit a vertical blanking plate "fencing off" the inside wheel cavity.
This would need to fit close to the wheel, and be fitted on the suspension so it moves along with the wheel.

What is even more surprising is the fact that the 2° diffusor helps.

On this kind of station-wagon-like rear ends - no diffusor at all ( 0° ) was found to offer the lowest drag, with a cavity, the diffusor helps another 6%.

A rear cavity is something that I could easily fit on Hägar.
A bit of foam and double-sided tape will do.

It is not very box like when you look at the roofline, for me it looks like more of hatchback etc from rear roof section.

With addition of rear cavity it is more like boxed from rear. Before cavity they used 7 degree diffusor, but with cavity 2 degrees was optimal, so if I'm not horribly misreading that would also tell that they did move from coupe/hatchback to boxed end, however it is much smaller boxed end.

High side window line and blunt front are perhaps what makes illusion of car being box shaped, sides are also rather straight down and it is relatively tall compared to width, but if one photoshops everything under window line away it is not boxed station wagon shape at all and what affect to diffuser angle is how upper side, roof, trunk lid etc are angled from top, afaik.

This should make it easier to see how it is not station wagon kind of shape at all (extract from document mentioned in 1st post):
http://jtbo.pp.fi/images/road_pics/A2_roof.jpg

That is one of the pits easy to fall with aeromodding, they use lot of tricks with cars that cheat eye, to see true shape and form sometimes one needs to take photo and edit it a bit.

I missed that this had been added to this thread and added different thread to share the larger list of links when I found them: http://ecomodder.com/forum/showthrea...ies-21204.html. Moderators seem to think there's enough difference to keep them separate. Thanks Vekke for your original post. I have a new mod that I can consider for my coupe instead of the Kamm.

The roofline drops on most cars these days - even on my boxy Volvo V50 - though not always as much as on the A2.


The box cavity should be an easy mod on a square-backed wagon though.
It won't be as efficient without a bellypan though.

Roofline indeed drops on many new cars and as air has no idea of knowing if car is wagon or not, then it might be misleading thinking that all wagons should have diffuser at 0 degrees, if roof drops, then it is not wagon in sense of aerodynamics and I think that one must pick more appropriate form from this image Sven posted to other thread:
http://ecomodder.com/forum/showthrea...tml#post296642

Most of such examples were made at time when wagons were truly boxed shape, so that is why wagon is shown to have 0 degrees at optimal diffuser angle, but it might mislead modder to think that all wagons should have that as new stuff indeed has angle, so it can't be true anymore with all wagons.

So when one puts on boxed cavity to rear, one need to put diffuser, that is closer to 0 degrees, but without boxed cavity diffuser must be more angled if roofline drops, when roofline does not drop 0 degrees is best, how about boxed cavity on boxed rear end without dropping roofline? I guess 0 degrees still?

This leads to thinking that boxed cavity turns dropping roofline to be more towards boxed shape, but as it makes it smaller boxed shape it will be improvement to efficiency.

Or something like that. Most important to know would be just that there are no absolutes and it is impossible to generalize, every situation must be examined separately, but understanding why some forms require certain angle is helpful to make right judgement of what angle is going to work or which things will never work.

The 100mm extrusion is interesting along with the rear diffuser in that they are 'behind' the wake yet effect it, the 2" diffuser is particularly effective , I am assuming it could be used with other Kammback designs with similar results / success.

None of those four are available. COME ON SON!

Thank you, Metro, for saving some of it. UGH.

I found this, but it seems like it's more of an overview of the study in the first post? I haven't seen the original.

http://www2.warwick.ac.uk/fac/sci/wm...view_may11.pdf



I'm sure they're stored somewhere. If someone wrote them a nicely worded email I'm sure they'd send you the files.

Being quoted by the Finnish Master of the Universe Ecomodder.
:D

It is amazing the power of the duct tape.

Thank you.

Maybe, I'll have to look for an e-mail address.

This link is working:
Information: How To Reduce Drag From 0.28 to 0.20 (Scholastic more than anything)

Direct donwload from the Audi A2 forum is: http://www.a2oc.net/forum/attachment...9&d=1360233797

To read online in Scribd
Session 4 - 1 - Hussain Ali - Drag Reduction on a Production Vehicle

<p style=" margin: 12px auto 6px auto; font-family: Helvetica,Arial,Sans-serif; font-style: normal; font-variant: normal; font-weight: normal; font-size: 14px; line-height: normal; font-size-adjust: none; font-stretch: normal; -x-system-font: none; display: block;"> <a title="View Session 4 - 1 - Hussain Ali - Drag Reduction on a Production Vehicle on Scribd" href="http://www.scribd.com/doc/231804529/Session-4-1-Hussain-Ali-Drag-Reduction-on-a-Production-Vehicle" style="text-decoration: underline;" >Session 4 - 1 - Hussain Ali - Drag Reduction on a Production Vehicle</a> by <a title="View ALP1981's profile on Scribd" href="http://www.scribd.com/ALP1981" style="text-decoration: underline;" >ALP1981</a></p><iframe class="scribd_iframe_embed" src="//www.scribd.com/embeds/231804529/content?start_page=1&view_mode=scroll&access_key=k ey-Vy2o5Gpf6pRAadeWcKNU&show_recommendations=true" data-auto-height="false" data-aspect-ratio="1.3323485967503692" scrolling="no" id="doc_47833" width="100%" height="600" frameborder="0"></iframe>

perfect combination
 

Hucho tells us that for the lowest drag we'll need airfoil or half-body shapes.
*The 2005 NUON NUNA,@ Cd 0.07 would be the 'airfoil' benchmark.
*The 2013 Cambridge University CUER would be an okay 1/2-body benchmark,@ Cd 0.10.
Mods which push a body in those directions would be winners.
The key is the high sectional density and elongation which allow streamlines to converge without separation,leading to a higher base pressure.

Cd values from bar graph
 

I did a photo-enlargement of the graph,measured each 'bar',then did linear interpolation to develop Cds for each 'bar.' With Cds attached to each mod,it's easier to evaluate their individual contribution.
Reading from left-to-right:
*0.2881
*0.2818
*0.2656
*0.2383
*0.2353
*0.225
*0.2214
*0.2162
*0.2143
*0.2100
*0.2044

Last night I came across a paper in Sustainable Vehicle Technologies: Driving the Green Agenda (2012) by a Tata Motors engineer involved in this study. In addition to much more information on the project, they also took the car to the Soufflerie 2A aeroacoustic tunnel in Paris to test the various configurations with a moving ground. Here's what they report:

https://ecomodder.com/forum/member-v...re8411-s2a.jpg

Interesting. Any idea of what 'suction upstream of the belt' means? Aren't the belt and the air moving at the same speed?

I would bet that the belt doesn't encompass the entire floor inside the wind tunnel. So that would mean there must be an area ahead of the car and head of the belt but in the windstream where the air is passing over the stationary floor. The point at which the air flow encounters the belt would be interesting in terms of turbulence, wouldn't it?

Above the fixed floor in front of the car, a boundary layer builds up; the suction slot dissipates that boundary layer so it doesn't skew the results over the moving belt.

suction
 

On the highway,the road is stationary and it doesn't have any boundary layer,just the car's body.In the wind tunnel,the moving belt may be at a different velocity as the airstream,creating a boundary layer which must be vacuumed off,so as not to create a condition a car wouldn't see in the real world.In some stationary-floor tunnels,they just lift the car up enough to clear the thickness of the boundary layer which forms.The uprights at DARKO, elevated the vehicles a bit,as there's no suction slot there,nor at the A2 tunnel.