REAR SPOILERS: Length-'Angles'

[aerohead]

There's been some interest in mods for notchback cars,especially rear spoilers.
I've gone back in time,looking over early wind tunnel investigations and will share the following.If you have any of Hucho's books you should already have this data.For those who don't yet have a book,here are some talking points for design.
----------------------------------------------------------------------------
From SAE Paper# 760185,by Hucho,Janssen,and Emmelmann,they tabulated results from wind tunnel tests of eleven ( 11 ) different rear end configurations for the Golf/Rabbit-Jetta,at Volkswagens climatic wind tunnel.
----------------------------------------------------------------------------
In every case,the drag was asymptotic as the roofline architecture met the 'Template'.( please see 'Aerodynamic Streamlining Template Part-C').
At 30% aft-body length ( Golf/Rabbit),as derived from the 'Template,' the following coefficients of aerodynamic drag were observed:
Backlight Angle-------------------------- Cd------------------------------
28-degrees----------------------------- 0.44
26-degrees----------------------------- 0.43
24-degrees----------------------------- 0.42 ( production angle )
23.5-degrees--------------------------- 0.40
23-degrees----------------------------- 0.39
22-degrees----------------------------- 0.38
21.5-degrees--------------------------- 0.37
20-degrees----------------------------- 0.36
18-degrees----------------------------- 0.35
12-degrees----------------------------- 0.34 ( 'Template')( 19.04% reduc.)
10-degrees----------------------------- 0.34 ( Kamm recommends c.1955 )
-------------------------------------------------------------------------
The Jetta,at 42% aft-body could go no lower than Cd 0.365 even when the trunklid ( boot ) was raised 450mm ( 18-inches ) @ 'Template' height,netting only a 9.2% drag reduction.
The 'notch' costs us and cannot approach the low drag of the K-form 'Template' roof.
-------------------------------------------------------------------------
The Scirroco,@ 31 % aft-body reached a drag minimum of Cd 0.377 with the best rear spoiler configuration.From backlight header-to-trailing edge,the spoiler relaxed the down-slope angle from 24.5-degrees,to 18-degrees,far below the 'Template.'
-------------------------------------------------------------------------
With the 'Template' roofline,the shorter,frumpy Golf returns lower drag than the 'sporty',longer Scirroco.
--------------------------------------------------------------------------
Had the Golf had the 'Template" roof,and windshield of,and plan body taper of either the Renault Vesta II or Daimler M-B Boxfish,it would have seen a drop to Cd 0.19.
-------------------------------------------------------------------------
A few cars which have pushed their spoilers to the 'Template':
1988 Lotus Esprit Turbo
1994 Saleen S-351 Mustang
1995 Ferrari F355 Berlinetta
-------------------------------------------------------------------------
The original CAR and DRIVER mods to the 1974 Ford Pinto altered the rear angle from 26-degrees,to 23-degrees( Hucho's maximum angle recommendation[22-degrees by W.A.Mair]),@ 37% aft-body,using the 30-degree up-swept rear spoiler,for a 7% mpg improvement ( 11.6% drag reduction )
Same spoiler,straight back = 1.36 % mpg improvement
-------------------------------------------------------------------------
It appears that while the spoiler can help,it does not have the drag reduction potential offered by the fastback,K-form roof,which is what Hucho claims in his book.
And the 'Template' appears to hide in plain view,embedded within the data scatter-plot.
--------------------------------------------------------------------------
If you must do a spoiler,try and get close to the 'Template',even if it takes a bi-wing or tri-wing arrangement.

[brucepick]

I appreciate your effort to inform us here. But once again, I've discovered how little I know on this topic.

Are you saying that even the best optimized spoiler can't achieve much improvement in drag? Do you think that also holds true for cars like the Civic coupes that have a fairly gentle rear glass slope?

Beyond that concern, I have to admit two other confusions, right from the start.

I'm guessing that backlight angle is compared with a horizontal line continuing rearward from top of rear glass? So a 10º angle is a very gentle downslope, and a 0º angle would be a rear glass that is horizontal?

And pretty much the same question for the % aft-body length item. I'm guessing that refers to the dimension from base of rear glass (or from top glass edge?? to the rear bumper edge, straight back to a point even with the rear bumper. Am I close? That dimension as a percentage of what??

I'm sorry. A little knowledge is a dangerous thing. I thought I was going to build a rear spoiler, but maybe I'd be better off addressing other possibilities.

[California98Civic]

Originally Posted by aerohead

The original CAR and DRIVER mods to the 1974 Ford Pinto altered the rear angle from 26-degrees,to 23-degrees( Hucho's maximum angle recommendation[22-degrees by W.A.Mair]),@ 37% aft-body,using the 30-degree up-swept rear spoiler,for a 7% mpg improvement ( 11.6% drag reduction )
Same spoiler,straight back = 1.36 % mpg improvement

Thanks for a very useful post. I estimated the angle of my 98 Civic's rear window at about 23*. And I just did a tuft test that has me wondering: http://ecomodder.com/forum/showthrea...tml#post244751

[cfg83]

aerohead -

I took the liberty of adding a link to 'Aerodynamic Streamlining Template: Part-C'. I hope that's ok.

brucepick -

Aerodynamics is a black art. There are great gains to be made, but they need to be adapted and tested to work for *your* car. I think the only "easy" ones are the wheel covers.

CarloSW2

[aerohead]

Originally Posted by brucepick

I appreciate your effort to inform us here. But once again, I've discovered how little I know on this topic.

Are you saying that even the best optimized spoiler can't achieve much improvement in drag? Do you think that also holds true for cars like the Civic coupes that have a fairly gentle rear glass slope?

Beyond that concern, I have to admit two other confusions, right from the start.

I'm guessing that backlight angle is compared with a horizontal line continuing rearward from top of rear glass? So a 10º angle is a very gentle downslope, and a 0º angle would be a rear glass that is horizontal?

And pretty much the same question for the % aft-body length item. I'm guessing that refers to the dimension from base of rear glass (or from top glass edge?? to the rear bumper edge, straight back to a point even with the rear bumper. Am I close? That dimension as a percentage of what??

I'm sorry. A little knowledge is a dangerous thing. I thought I was going to build a rear spoiler, but maybe I'd be better off addressing other possibilities.

brucepick,I'll try and address your questions in order.
* the best 'idealized' spoiler could have significant potential for drag reduction,although due to the turbulence associated with the separated flow above the spoiler,it could never achieve the drag reduction of the K-form 'Template' roofline, which is devoid of such turbulence.
*Turbulence cannot be converted to useful pressure regain,only heat,during viscous attrition as the eddies are eventually worn down by calmer surrounding air.It's a second law of thermodynamics thing we strive to avoid.
* If the CIVIC backlight is steeper than the 'Template' it must have separation.The spoiler is a palliative to help with re-attachment and restoration of laminar flow of the outer flow field 'jumping' over the created locked-vortex of the spoiler.
* On a hatchback type car the backlight and its pillars define the departure angle/(curvature if any) for the aft-body roofline.This would be the case for the the Golf/Rabbit,where there exists no trunk( boot).And it is measured from the horizontal downwards.
* On a notchback coupe or sedan,the line described by a straight-edge laid across from the roof rear center-line to the trailing edge of the trunklid (boot) center-line, defines the aft-body departure angle.( an example would be the 1968 Dodge Daytona Charger.Its backlight is 22-degrees,but when you factor in the trunklid,the angle relaxes to only 18-degrees,an angle W.E.Lay used to achieve Cd 0.12 cars of 'impractical' length.
* The percentage of aft-body is taken from the 'Aerodynamic Streamlining Template Part-C' and is defined within that thread.The 'Template ' will show you the maximum aft-body 'angle' the boundary layer can tolerate without separation.
I recommend that members ballast their cars with 300-pounds( a couple of friends) in the front seat, when this determination is made so as to reflect the inclination of a 'loaded' car when actually moving.

[brucepick]

Thanks aerohead!

Originally Posted by aerohead

...
* If the CIVIC backlight is steeper than the 'Template' it must have separation.The spoiler is a palliative to help with re-attachment and restoration of laminar flow of the outer flow field 'jumping' over the created locked-vortex of the spoiler....

Is this similar to the Mythbusters pickup truck with the tailgate up, where a vortex is set up in the bed and the airflow skips over it almost as if there were a cover over the bed?

Assuming the spoiler's back edge is positioned at the template line, restoration of laminar flow would be brief at best. Would occur only at the spoiler edge. Is there value to that restoration? Would it somehow tend to restore the airflow to the ideal path and result in a smaller cross section of wake?

And - if the spoiler is positioned lower or a bit further forward so that it doesn't reach the template line, would it still "catch" the air flow and give a smaller wake? Or would it just be a mess of turbulence due to not being on the template?

Green lines show how mirror would see spoiler's edge. I drew two lines, not knowing precisely where the mirror is.

[aerohead]

Originally Posted by brucepick

Thanks aerohead!



Is this similar to the Mythbusters pickup truck with the tailgate up, where a vortex is set up in the bed and the airflow skips over it almost as if there were a cover over the bed?

Assuming the spoiler's back edge is positioned at the template line, restoration of laminar flow would be brief at best. Would occur only at the spoiler edge. Is there value to that restoration? Would it somehow tend to restore the airflow to the ideal path and result in a smaller cross section of wake?

And - if the spoiler is positioned lower or a bit further forward so that it doesn't reach the template line, would it still "catch" the air flow and give a smaller wake? Or would it just be a mess of turbulence due to not being on the template?

Green lines show how mirror would see spoiler's edge. I drew two lines, not knowing precisely where the mirror is.

brucepick,it does function exactly like the raised tailgate.
If the trailing edge is close to the template the spoiler captures the low-pressure vortex and isolates that low pressure in front of it,preventing it from communicating that pressure to the wake behind which would otherwise lower the base pressure,increasing the delta-P across the car,increasing drag.
--------------------------------------------------------------------------

When the outer flow does separate at the trailing edge of the spoiler, it has decelerated to a lower velocity/higher pressure ( that evil Daniel Bournoulli ) leaving the base pressure of the wake higher,the delta-P lower,an concomitant lower drag.
--------------------------------------------------------------------------

If you tilt the spoiler up at 30-degrees from the back of the trunklid( boot),it will hit the 'Template' sooner,and it can be 'shorter'.
--------------------------------------------------------------------------

If you were to also extend the C-pillars back along the 'Template' line as I did with the CRX,you can tailor the vortex even finer with this buttress,and in so doing,foster even better side flow towards the rear by,eliminating a potential breeding ground for longitudinal attached-vortices on each side.
You'll have a bit of new blind-spot,but you can compensate for that with your mirrors.
Anyone with a 1st-gen Insight would be familiar with this.
--------------------------------------------------------------------------
If you can't reconcile yourself to the 'Template' try to set 22-degrees as your maximum local tangent angle.Air cannot follow an angle steeper than this,and typically,you need to have close to a body height worth of length in progressive curvature before you can get to 22-degrees.

[California98Civic]

Originally Posted by aerohead

brucepick,it does function exactly like the raised tailgate.
If the trailing edge is close to the template the spoiler captures the low-pressure vortex and isolates that low pressure in front of it,preventing it from communicating that pressure to the wake behind which would otherwise lower the base pressure,increasing the delta-P across the car,increasing drag.
--------------------------------------------------------------------------

When the outer flow does separate at the trailing edge of the spoiler, it has decelerated to a lower velocity/higher pressure ( that evil Daniel Bournoulli ) leaving the base pressure of the wake higher,the delta-P lower,an concomitant lower drag.
--------------------------------------------------------------------------

This raises questions for me. The tuft testing I have done of my sixth gen Civic coupe shows a recirculation zone for about 4 or 5% of the lid/window area near the spine of the car. Air reattaches a few inches before detaching from the trailing edge of the trunk lid. Good. And better than I thought. Certainly a solid well-placed surface is better than a recirculation zone for allowing air to slip along, no? Certainly the longer, descending-angled raised lid and lip would slow the air more than the 4" where the air currently reattaches, therefore gaining the pressure benefits you describe here, no? It seems to me the key is making sure the contours of a raised lid do not add drag, and if they don't that a 2% drag reduction should be possible. Modest, but measurable in FE terms in a careful test, no?

[brucepick]

Aerohead -
All other things being equal - which they never are of course,
would a spoiler angled upward 30º generate more downforce than a level one ending at the same point? And would that downforce carry a drag penalty?

Measured angles:

I measured some of the angles on the '96-00 Civic coupe using a print of the drawing that's been posted in these threads, and an old fashioned protractor. Also used a somewhat sketchy straightedge to establish the base lines, so absolute precision is not guaranteed!

The rear glass angle is about 20-22º downward, vs. horizontal. So, its local angle vs. the immediate prior roof is going to be less than 22º because the roof has already begun to curve downward at that point. This may explain the pretty good attached flow that I believe 98CaliformiaCivic found on the rear glass and trunk lid.
- - - tuft video
- - - damp tufts still #1
- - - damp tufts still #2

I measured the angle from roof to trunk lid edge, based from a point on roof slightly rear of the small side window's forward edge line. This is not the roof peak but is (to my eye) where it departs visibly from the template curve. I got 14º.

I measured from top edge of rear glass to trunk lid edge. I found about 16º there.
---------------------------------
I've seen recommendations for rear down angles of 14º from roof edge to trunk rear edge, and aerohead's recent suggestion to keep local angles at less than 22º. Based on those, it looks like the Civic coupe's stock back end is already in pretty good shape. However the trunk edge isn't anywhere near the template line. A "false trunk lid" to meet the template curve would be some 7"-9" higher than the original trunk, even if the false trunk lid is extended back a few inches to be even with the rear bumper. If not extended rearward, the false lid would need to be higher if it is to meet the template.

Aerohead, what do you think? Would there be gain in a false trunk lid if its sides are formed by extending the C-pillars rearward on their own existing angle? Or have the designers already worked the angles to the point where they've minimized the wake cross section - and a raised lid would only increase that cross section?

[Pendragon]

aerohead,

I doubt that the airflow which would be affected by the spoiler you describe is laminar in flow. In fact, it should be rather turbulent. Isn't the spoiler simply moving the point of complete separation aft so that the area of the notch back is more stable, even if it is a "dead" area?

If one were to place a spoiler at the top of the rear window it should extend the separation some distance and then a second (lower) spoiler could extend the complete separation even further.