2011 Prius to get full boat tail w/ faired trailer

[aerohead]

Originally Posted by NeilBlanchard

Is there air separation on the rear window of the Prius?

Neil,I don't have a definitive answer on that one.I suspect that if it isn't,it's close.
There is an image of the low drag version of Audi's A2 ( Cd 0.25 ) @ Google Images.
That car is hypo-'Template' but the smoke flow looks really good over the backlight of that car,and none of the other streamlines above it seem perturbed.
Toyota's designer,who did the Prius,said that the shape is 'close' to ideal.We might infer from the comment that the roof is a little fast.
The 1st-gen Insight has separation.The 2nd gen Insight will have separation.It's about the same profile as Prius.
I can say that rear vision out the backlight of the new Prius is non-existant,so they may be doing what GM does,take the aero hit with the steeper glass,but ensure good outward vision.
Is that muddy enough for you?

[aerohead]

Spoke with Gary over the phone last night.
He has returned one of the trailers and decided to go with one straight-sided trailer as a compromise between having the pure aero trailer and a second utlity trailer.
He believes this configuration will allow him more utility,at a concession to better economy.
He has purchased a wire-feed welder and hood and is anxious to begin.Life has been intruding and he struggles like so many of us to 'find time.'
I have some wind tunnel data from Kamm that should allow a prediction of performance between the original plan and the new.

[jime57]

Originally Posted by aerohead

The 1st-gen Insight has separation.

Are you sure of this Phil. I have not seen wind tunnel smoke or tell tail data which would indicate separation. From the work that 3 Wheeler did, I though the rear met the template, if indeed that is an indicator. The rear most portion is only about 18 deg by my measurement. Could it have a Cd=.25 if it had separation?

[aerohead]

Originally Posted by jimepting

Are you sure of this Phil. I have not seen wind tunnel smoke or tell tail data which would indicate separation. From the work that 3 Wheeler did, I though the rear met the template, if indeed that is an indicator. The rear most portion is only about 18 deg by my measurement. Could it have a Cd=.25 if it had separation?

I'm not scientifically sure,but I'll explain my comment.
*Honda's wind tunnel photo of the smoke trace over the Insight is shot from about the cowl position and at about door handle height.
* From Honda's technical drawing,some of the roofline is obscured in the photo.
*The smoke trace is more 'flattering',than if compared to a shot taken at 100-meters,closer to true-length.
----------------------------------------------------------------------------
*From Honda's drawing,the departure angle at the roof terminus is 14-degrees.
*Jaray's pumpkin seed of 1921 had this angle at 32% 'Template' aft-body,for Cd 0.13 for his car.
* VW's 2009 1-liter car also uses it @ 32% for a Cd 0.195 car.
--------------------------------------------------------------------------
* At 32%,Mair limits the boat tail to 13-degrees.
* Lay was at 12.5-degrees @ 32% and fetched Cd 0.12 for his car.
* Kamm used 13.5-degrees " ".
* The Kamm-Jaray model was 12-degrees.
* GM Sunraycer,10-degrees
* "template",12-degrees
--------------------------------------------------------------------------
* Insight gets Cd 0.25 @ 34% aft-body
* 1999-2005 AUDI A-2 3-Liter gets Cd 0.25 @ 29% aft-body
* 1987 Renault Vesta-II gets Cd 0.187 @ 28.5% aft-body ( Vesta is a 'Template' aft-body form.
* New Beetle is Cd 0.38 @ 28% aft-body
--------------------------------------------------------------------------
It appears that the shallower 'template' tangent angles are providing lower Cd's per body length which would be in accord with Hucho,who defined the 'template',and it suggests that separation is the culprit.
The Insight I&II are close.Same for Prius.
I don't think Toyota or Honda considered that folks like us would be picking up where they left off,so they might leave a little on the table,as Toyota's styling chief has eluded to.
-------------------------------------------------------------------------
I am convinced that boat tail performance is absolutely dependent upon perfect onset flow and these tangent angles are of the upmost importance.
I'm not comfortable recommending anything less than the 'Template'.And I do believe that anything 'below' it is risking separation which is antithetical to streamlining.Again,I'm leaning on Hucho's words.
Sorry for the long answer,but the question merits the distance.

[MetroMPG]

Originally Posted by aerohead

I don't think Toyota or Honda considered that folks like us would be picking up where they left off,so they might leave a little on the table,as Toyota's styling chief has eluded to.
...

I'm not comfortable recommending anything less than the 'Template'.And I do believe that anything 'below' it is risking separation

Phil, excuse the basic question, but does separation imply reversed flow, or might we only see turbulence if the angle is not too far departed from the ideal line?

Reason I ask: I'd be willing to tuft test (& make video) of the flow over the rear of my 1st generation Insight if you thought it would be instructive.

Of course, reversed flow would be easier to spot in a test.

Darin

[jime57]

Phil, I thank you for the obvious investment of time you have made on this question. As you say, both Honda and Toyota may have "left a bit on the table," but as modders, we are pretty much stuck with their rear body angles. We aren't likely to be able to make changes there. Still, in the interest of science, it entirely appropriate that we try to figure this out.

I want to do more reading, but I'm somewhat suspicious of some of the older Cd data like that of Lay, Jaray, Kamm and Mair. (Please remember that skepticism is one of the cornerstones of science. It motivates us to challenge the conventional wisdom.) The numbers they provide, .15-.17 or so for bodies WITH WHEELS has just never seemed likely, in light of the insight's .25 Cd. Remember that these folks were in the business of selling "streamlining" and their incomes depended greatly on low Cd numbers. In addition, instrumentation wasn't nearly as good as current, and even Hucho has commented on this issue. Some of the "reasonable" data has been verified by Hucho, in the modern VW windtunnel, but other data such as the Tatra(if memory serves) did not stand up.

Here's and interesting example of the "marketing" problem. I have build and raced small sports cars. Of late, dyno testing has become all the rage. On the Dyno Dynamics chassis dyno, there is a software variable, this allows the operator to "correct" the power output for various "errors" which the operator perceives to be in play. The operator who adjusts the "correction" to the high side gets more business than the operator to sets the variable to a lower level. The car owner is much happier if his result says 110 hp vs. 106 hp, though nothing has actually changed!

[jime57]

OK, my above post was a bit shakey on factual data, so I've gone back to Hucho to provide some real quotes.

On the subject of trusting "old" data, refer to page 25 and the description of the Tatra 87. Lange measure a model of this car at Cd=.244. Quoting directly from Hucho, "The drag figures published for this car caused a lot of confusion, and are a good example for how "old" figures must be handled......But the correct figure turned out to be Cd=.36 as measured in 1979 in the large wind tunnel of Volkswagen AG with an original car."

Page 26 of Hucho provides a very interesting comparison between the theoretical and the real. Midpage Hucho describes the development of a model, including wheels, by Lange which was "long and lean" and measured at Cd=.16 in the VW windtunnel. When this shape was translated into a real world car, the Porsche 911 Carrera, the Cd came in at Cd=.33. Huco comments that the Lange model must be considered in the light that it "was completely smooth, with no recesses for windows, no real wheel houses, and no undercarriage." I would add that it had no radiator opening and no mirrors. That hardly makes for a practical Prius or Insight.

To me the bottom line is that I personally need to have a practial goal of aeromodding in mind. The GM Aero, Huco p.57, perhaps shows what a idealized prototype can achieve, Cd=.14(by their own measurements) without mirrors, but no modder is ever going to achieve that shape with an existing car. I can see lowering the Insight Cd to the order of Cd=.20-.22, but I don't see modders getting any better than that. We don't have the option of doing away with window openings, wheel openings, undercarriage, and perhaps mirrors. JMHO

But, perhaps I degress. Your original point was that the Insight and Prius were too steep, to early on. I really don't have the basis to object to this assessment. but I thought it had been decided when 3 Wheeler was designing his boattail that his overall profile met the template. That would imply that the Insight angles were within spec. I guess I'll have to go back and read that thread again. I look forward to seeing some tuft testing on the actual car, after he finishes. I would only point out that the determination of the "high point" of the car has a rather profound impact on departure angles and that the criticality of the high point location is not really know for real world mods. Don't mean to offend and I think the conversation is really worth having. Perhaps Metro will be able to do some tuft testing and that would certainly help in answering this question. I would do it myself, but currently I'm very busy with some more critical things like repairing a bent racecar for sale and FINALLY getting my complicated taxes finished.

As always, thank you very much for your insightful thinking and your long term interest in aero topics. We couldn't do without you

[jime57]

Originally Posted by aerohead

I am convinced that boat tail performance is absolutely dependent upon perfect onset flow and these tangent angles are of the upmost importance.
I'm not comfortable recommending anything less than the 'Template'.And I do believe that anything 'below' it is risking separation which is antithetical to streamlining.Again,I'm leaning on Hucho's words.
Sorry for the long answer,but the question merits the distance.

Phil, would you please give me a reference to the appropriate section in Hucho's book. I haven't been able to find it, though I have little doubt that it is somewhere in that vast book

[aerohead]

Originally Posted by MetroMPG

Phil, excuse the basic question, but does separation imply reversed flow, or might we only see turbulence if the angle is not too far departed from the ideal line?

Reason I ask: I'd be willing to tuft test (& make video) of the flow over the rear of my 1st generation Insight if you thought it would be instructive.

Of course, reversed flow would be easier to spot in a test.

Darin

At the separation point the velocity of the turbulent boundary layer is already zero.
As the roofline or body sides continue to reduce in cross-section,Bernouli's Theorem dictates that the air must decelerate further,but that's impossible,as it is already at rest against the boundary of the body.
In the presence of this adverse pressure gradient the boundary layer lifts off the surface of the body and the air behind rushes forward into the void beginning the back-flow which will grow into eddies.
If the air can re-attach onto,say,a spoiler,then full-blown turbulence is avoided until after the spoiler.
Without a spoiler it will grow into turbulence with its pressure the same as at the point of separation.
On the Insight,even though the slope is not ideal,the body provides a wake-stuffing function,forbidding air to occupy its 'space,' and in effect,tailors the wake to some extent.
When Carl Breer hung the 3-foot 'stinger' off the back of the 1934 DeSoto Airflow,it was far from ideal in shape,but it altered the wake enough for the car to pull off Cd 0.244.
I think that if the camera angle was better for the photos of cars like the surviving Kamm car,AUDI's A2 3-liter,and Insight,we'd have a more accurate look at the streamline filament over the roof.
------------------------------------------------------------------------
The last point I'd like to make,and then Ill shut up,is that if Hucho is correct about the 'Template',if we streamline short of it,we shoot ourselves in the foot.There won't be any plug-and-play additions,adding length at a later time for even lower drag.The air will be trash,not the robust 'onset' flow which Hucho describes as essential to" moving the point of separation back",the whole premise of streamlining.
Okay,I'm shuttin' up now!

[aerohead]

Originally Posted by jimepting

OK, my above post was a bit shakey on factual data, so I've gone back to Hucho to provide some real quotes.

On the subject of trusting "old" data, refer to page 25 and the description of the Tatra 87. Lange measure a model of this car at Cd=.244. Quoting directly from Hucho, "The drag figures published for this car caused a lot of confusion, and are a good example for how "old" figures must be handled......But the correct figure turned out to be Cd=.36 as measured in 1979 in the large wind tunnel of Volkswagen AG with an original car."

Page 26 of Hucho provides a very interesting comparison between the theoretical and the real. Midpage Hucho describes the development of a model, including wheels, by Lange which was "long and lean" and measured at Cd=.16 in the VW windtunnel. When this shape was translated into a real world car, the Porsche 911 Carrera, the Cd came in at Cd=.33. Huco comments that the Lange model must be considered in the light that it "was completely smooth, with no recesses for windows, no real wheel houses, and no undercarriage." I would add that it had no radiator opening and no mirrors. That hardly makes for a practical Prius or Insight.

To me the bottom line is that I personally need to have a practial goal of aeromodding in mind. The GM Aero, Huco p.57, perhaps shows what a idealized prototype can achieve, Cd=.14(by their own measurements) without mirrors, but no modder is ever going to achieve that shape with an existing car. I can see lowering the Insight Cd to the order of Cd=.20-.22, but I don't see modders getting any better than that. We don't have the option of doing away with window openings, wheel openings, undercarriage, and perhaps mirrors. JMHO

But, perhaps I degress. Your original point was that the Insight and Prius were too steep, to early on. I really don't have the basis to object to this assessment. but I thought it had been decided when 3 Wheeler was designing his boattail that his overall profile met the template. That would imply that the Insight angles were within spec. I guess I'll have to go back and read that thread again. I look forward to seeing some tuft testing on the actual car, after he finishes. I would only point out that the determination of the "high point" of the car has a rather profound impact on departure angles and that the criticality of the high point location is not really know for real world mods. Don't mean to offend and I think the conversation is really worth having. Perhaps Metro will be able to do some tuft testing and that would certainly help in answering this question. I would do it myself, but currently I'm very busy with some more critical things like repairing a bent racecar for sale and FINALLY getting my complicated taxes finished.

As always, thank you very much for your insightful thinking and your long term interest in aero topics. We couldn't do without you

The funny thing about the early numbers is that some,when tested in modern tunnels yield numbers actually lower than originally published ( Jaray small car,Cd 0.235,1920s,Rumpler saloon car,Cd 0.28,1924,Klemperer 'minivan'1921,Cd 0.15.
* Jaray's Cd 0.13 pumpkin seed of 1921 goes off 'Template' @ 32%,and descends to an exit slope of 26-degrees,violating what Mair would discover later in 1968 to be a limit of 22-degrees maximum slope for attached flow.
* In 1933,Walter E. Lay stays on 'Template' for the whole enchilada and nets Cd 0.12,a limit for a non-wheel fairing car.
* In 1947 you'll find pumpkin seeds hiding below Boeing PB-1W aircraft.
* In 1949 you'll find 'em under Grumman AF-2 Guardians.
* In 1953 the US NAVY begins a second look at fish shapes with USS Albacore,begins to break every underwater record,and sets the stage for all modern nuclear submarines.
* In 1957 the MG EX 181 borrows the pumpkin seed to set land speed records at Bonneville.
* Same year,the 'Template' pumpkin seed hitches a ride atop Grumman's E-1(WF-2 Willy Fudd) Tracer.
* 1961,Mickey Thompson goes to Bonneville with "The Pumpkin Seed."
* 1968,W.A.Mair establishes the optimum boat-tail angle at no more than 22-degrees,achieving that angle at no sooner than one diameter body thickness behind the point of boat tail origin.
* 1981,VW uses 'Template' with 'Drop Shape' for Cd 0.16 @ 42% aft-body.
* 1981,VW uses 'Template' with 'Flow Body' for Cd 0.15 @ 46.5% aft-body.
* 1981,VW uses 'Template' with 'Blunt Body' for Cd 0.15@ 43% aft-body.
* 1987 Chevrolet Express Cd 0.195
* 1987 Vesta II Cd 0.187
* 1987 GM Sunraycer Cd 0.89-0.12.
* 1990,ITworks CRX, Cd 0.235
* 1991,GM HX3,Cd 0.258
* 1994,Honda Dream II,Cd 0.10
* 2004 Team Corvette C5-R
* 2008 M-B,Bionic car,'Boxfish',Cd 0.19
* 2009 VW 1-liter,Cd 0.195
* 2010 Lamborghini Murcielago
* 2010 Cadillac CTS Wagon
* 2010 Ford GT
* 2010 Ferrari 458 Italia
* 2010 Ferrari 599XX
* 2012 Porsche 918
* 2012 Lamborghini Esto Elemento
* 2011 Honda Type R
* 2011 Nissan Leaf
* 2011 Nissan Leaf NISMO race car
--------------------------------------------------------------------------
As to Hucho and the origin of the 'Template" it requires a full reading of the text.I can tell you that the 'Template' is hiding in full view.You just have to be looking for it.
I gave some details in the 'Template' thread.
-------------------------------------------------------------------------
I'll be trying the full tail in September.It's an unknown quantity until then.And it's obviously longer than need be.It's just that it's where I'll be sleeping,so I made it accordingly.