Think Flight: Using Simple Aerodynamics to Beat High Gas Prices
 

Published Aug 31st. Delete if a repost.

35.64mpg to 40.52mpg

https://www.youtube.com/watch?v=O-FY...el=ThinkFlight

I put it in "Interesting Aeordynamic..." but don't delete. It's just lost there.

What can one see from the thumbnail, that he took tuft testing to find out?

what see
 

The preliminary conclusion from the tuft study infers that both the plan-taper and diffuser angle are too extreme to protect the boundary layer from premature separation, counter-flow, eddies, and turbulence.
Also, from other reportings, the addition of rear wheel skirts and a full belly pan with diffuser would have amplified the 13.7% mpg result.
So, anyone wishing to improve on the tail would select smaller angles on the sides and bottom, for the 'second-approximation' design.
When the tufts behave as up top, you're essentially 'there.'
Additionally, if the coefficient of pressure profile were to be determined, this would be 'what you're looking for', by default.

Admitedly the plan taper doesn't show in the thumbnail.

Compare the Breer high-tailed stinger vs Harry Westergard's taildraggers.

Westergard
 

They all look too aggressive to me. Pseudo-Jaray 'fastback'. I'd bet a cup a Joe and a sugar-dum-dum that they'd suffer premature separation.
'Like the look though.:)

I didn't watch, but I don't think aviation is a simpler frame for aerodynamics. The entire subject is extremely complex, perhaps with a few easy to remember "rules of thumb".

As with any subject, the basics are low hanging fruit, with mastery involving lifetime dedication and proclivity.

Same four forces, same Reynolds numbers; what better 'frame'?

He's a pilot, it's a build story with tuft testing.

Meh. Laminar flow better than turbulent? Is that what thinking about flight is supposed to reveal? I still refuse to watch based on the dumb title.

Perhaps we should use simple golf ball dimples to beat high gas prices?

The aerodynamicist's tool?

https://www.matcotools.com/productas...yImage_400.jpg
https://www.matcotools.com/productas...yImage_400.jpg

Close but not quite. That would be for peening the edges of metal shafts to retain them.

You need a pick hammer to make proper dimples.

I came across the video two days ago.
I hate to admit it, but I was unaware that he is a big time YouTuber.
I was amazed at how many views he had on the video, and left a reply back to him with some 'advice' on how he could get better flow with different side angles. ( I should have just not said anything, since after seeing some of his other videos, I see the guy really knows his stuff. I feel like a fool now.)
It's great to see his video get so many views.
I just wish that he would have given some credit to Darin's 'Flea' design.
If you are into aero and like quality videos, have a look at his channel and subscribe !

aviation
 

The thing about 'aviation' which sets it apart from road vehicle aerodynamics is that, the fineness ratios of fuselage, wings, and tail surfaces are typically so 'large' as to make flow separation impossible.
And in addition, virtually all fuselage have 'compound surfaces'.
--------------------------------------------------------------------------------------
The Subaru remains a 'bluff' body, and pressure drag, which dominates total drag, is due the the intentional absence of the aft-body, which is responsible for the large wake's low base pressure.
The Subaru electric sibling ought to have significantly lower drag, due to a smaller radiator opening and a 'ideal' underbody.
Adding the same boat-tail to the BEV variant might boost battery range closer to 20%.
And optimizing the boat-tails' sides and bottom, push the improvement to over 20%.
AeroStealth and wife are leaving San Diego this morning, having spent the weekend at the 'Fully-Charged' BEV convention. We talked last night and he said no one presenting at the fair had any clue about aerodynamics and range.
This video could help stimulate a conversation.

I'm actually a little disappointed that he made the tail with such an aggressive diffuser angle. I figured he would know better. But I'm sure he'll get it figured out.

angle
 

Seems like the tufts were enough to convince him.
If he'd been able to see the elevation above ground for the 'source' flow traveling under the Subie, he would have never expected it to rise 'above' where it came from at the 'sink'.;)

I replied to as many comments as I could.
Does anyone use Discord? :) I want him to see my diagram! :D

https://discord.gg/b7BfHq2QJM

https://ecomodder.com/forum/attachme...1&d=1663040585

I haven't done one of those in years! :D

The roof looks pretty close to the template. I moved the bottom forward to line up with the bumper.

He said that he tried to match the body curves, but those probably detach at the front of the bumper.

I am not sure that the length is wrong, just too much taper in the front, and then not enough in the rear.

Hey, do you guys remember fanfold foam boards?
Do you think that he could use individual layers of that to build up the front and then either use a longboard to sand it down, or fill it in with 2-part foam?

I am sure that he would want to remove the foil first.

How hard would that be?

It was a pain to remove the perforated film from my fanfold foam.

I am sure that the foil is fine--until a rock or grocery cart hits it, but after he adds to it he would need to fiberglass it.

There were some questions about how much benefit different lengths of tail would provide.

Does anyone remember where those tables are?

Aren't there rules of thumbs for major mods, like 5% for an air dam, 10% for an undertray, and 20% for a boat tail?

Are those first two high? Maybe 3 and 5%, respectively?

I hope he isn't planning on leaving the license plate on the car. He talked about it and in one shot it was gone, but it sure seemed like he drove around with a boat tail and the license plate on the car.

I was trying to figure out what year his Imprezza is.

It seems like most comments were the same assumptions over and over--and some people really don't like making a small car longer.

How long do you think his goop will keep the wheel covers attached?

If he makes a proper boat tail, should he remove the hatch?

This is where they should be, but I didn't check. Too busy linkifying the quote.
Rule of Thumbs are merely first approximations. The various elements are interactive; for example adding an air dam without a spoiler.

Sometimes boundary-layer turbulators can save a marginal taper from separation. The Aero-Gare prototype flew OK, but they had to rough up the production ones to keep them flying straight, as they would develop separation on one side or the other.
I was surprised that the first bit of advice I got from a PhD who earned it was to avoid using a skin-on-frame nose cone like a glider, because gliders only get away with it by not having crosswinds like cars.

Think Flight Subaru = approx. Cd 0.156 potential
 

I froze the screen of the video, and Al captured it on a memory stick, then printed out 11" X 17" copy, at the scale of my x-ray overlays.
Using the dimensions of an Outback, the Think Flight came in at approx. 252" overall length, a 60" elongation.
Length / Height = 3.7672
Length / height = 4.2637
Length / width = 3.3753
Length / three-ratio average = 3.80206
Af estimated @ 85.5% of gross frontal area = 29.656 sq-ft.
Length / square-root of Af = 3.8562
forebody = 49.5% of L
Aft-body = 50.5% of L
The elongation effectively creates a 'double-ended' body
Verungungsverhatnisse ( V ), ( aft-body length / h ) = 2.1654
-----------------------------------------------------------------------------------
I found six cars with L / square-root of Af similar to Think Flight. Cds ranged from 0.20, to 0.137, with a 6-car average of Cd 0.1568.
--------------------------------------------------------------------------------------
I found six cars with V similar to Think Flight. Cds ranged from 0.19, to 0.137, with a 6-car average of Cd 0.1558.
-------------------------------------------------------------------------------------
The comparison cars are:
1983 GM Aero 2002 concept
1984 GM Chevy CITATION-IV concept
1985 Ford PROBE-V concept
1993 GM EV1 LSR
2006 LOREMO
2016 M-B IAA concept
2017 Eindhoven University Stella Vie
2019 GAC ENO. 146
-------------------------------------------------------------------------------------
With all- wheel- skirts, wheel covers, a full belly pan, diffuser, and 'corrected' boat-tail, there's a high probability that Think Flight would indicate near Cd 0.156 or less.
About a 50.9% drag reduction.
And 'back of the napkin' mpg improvement, at 65-mph, of 25.5%, or around 44.5-mpg.:)
If the same mods were applied to the 2023 Subaru Solterra BEV, we'd be looking at around 155-mpg-e.:):)
--------------------------------------------------------------------------------------
PS, before I forget, I measured a 19-degree downslope angle on top.
TF reported that he used 20-degrees for the sides.
I measured 20.5-degrees for the diffuser.
The upper profile is very similar to Mair's 1969 torpedo, Volkswagen's 1981 'Flow-Body', and the 2007 AST-I contours.
--------------------------------------------------------------------------------------
PPS, over the weekend I'll go over the Subaru Impreza Wagon specs now that I have them.

Think Flight- diffuser design
 

If this links, the point of interest is the 'elevation' of the lower smoke filament passing under the XL1, and the 'elevation' of the actual diffuser panel at it's trailing edge.
You'll notice that VW doesn't expect the lowest streamline at the rear to 'recover' to an 'elevation' which is 'higher' than where it originated from.
The 'sink' is the same as the 'source.'
https://www.vwvortex.com/threads/vol...48569&slide=35

different lengths different Cds
 

1) it's about wake area.
2) if you reduce wake area by half, you've reduced drag by half ( Variable trailer design drastically cuts aerodynamic drag,' SAE International Paper 2013-01-2414,' by Mario Hirz and Severin Stadler, Graz University of Technology.
3) before you boat-tail though, one wants to accomplish all the optimizations upstream of the tail, to provide fully-attached onset flow, or you risk compromising the tail's performance.
4) the tail itself must be 'streamlined.' At no point, top, sides, or bottom may the pressure increase exceed that seen in a 'streamlined' surface.
5) the 'top' is the easiest part, as the carmaker has already 'begun' to reduce the cross-section there, and the flow is typically attached until the rear edge.
6) the sides, while cambered, are typically not as 'boat-tailed' as the roof. And the flow on the sides of the vehicle is typically 'slower' than on top, is at a higher pressure already, is less 'displaced', has lower kinetic energy, and cannot survive as much cross-section reduction as on top without separation.
secondly, as you're creating a 'fastback' form, the fastback requires the highest degree of tumblehome, of the basic three body types.
7) The underbody has the most feeble flow, and asking a diffuser to position a separation edge at a 'higher' elevation than that from which the underbody flow originated at the front is a recipe for disaster.

It always seems like the rear corners of cars are optimized for not hitting things, independent of aerodynamics.

So, rear skirts before a boat tail?

Undertray, too?

skirts and belly
 

Yep! :)

TF's $ 15,000 hubcaps and tail
 

I ran the numbers for the Impreza dimensions.
The frontal area is approx. 23.815- sq-ft ( 2.212 meters-square ).
The L/ sq-rt of Af basically doesn't change.
'Kamm's' V-criteria does, increasing to 2.4643.
A comparison of seven cars with V's in the neighborhood, averaging 2.3372 yielded an average Cd 0.1475.
--------------------------------------------------------------------------------------
If the car were given the TF wheel covers and tail as a 'new' car, over it's statistical, 12-year lifespan, for the EPA-weighted HWY portion of it's total mileage, the Impreza would return:
* approx. 600-million less Btu waste heat to the atmosphere ( entropy )
* 5,274 fewer gallons of fuel consumed.
* @ $ 3.00/gallon, saving $ 15,800.
* And lowering carbon-dioxide emissions by 52-tons.
The fuel savings would cover about half the new car price, in constant dollars.

He has part 2 up now. ��
https://youtu.be/4ykw_8lpjco

I'm sure he does not need that step as a turbulator after two open wheelwells. It would be easy to put skirts on the rear as a blister running past the wheel openings. You could make a mold for one side, and then just pop a thin splash of fiberglass into the opposite curve for the other side before finishing it. Copies would probably fit many cars with a little trimming. Vacuum forming might be cheaper for production.

I was admiring just how flush that his wheels were with the body in stock form.
I cringe each time I see those huge 'elephant ear ' mirrors out destroying the drag though.
I notice he didn't add any plexi for windows in back, yet he used it to cover the lights after viewers suggested it.
If he added windows, he could have folded his mirrors ( or at least one ) back.
In this latest video, he added a low air dam, rather than creating a belly pan.
He is familiar with all sorts of materials for creating R/C planes, and I'm puzzled why he chose not to cover the underside in Coroplast ( with cutouts around hot bits )
The air dam added too much to the frontal area, and actually made things worse if i remember correctly.
I think maybe he chose the airdam due to it being the faster option.
He got really low on the departure angle of the tail too.
He's lucky to have not scraped it.
One thing I plan to do if I ever create a full tail, is to have this same angle, but have it flex upward and drop back down by using the floor mounted to a hinge and held up by wires.
He was asking for advice on how to get a more precise way of measuring the MPG over tank to tank fills at the pump.
I saw a few comments that mentioned the Scanguage, and I hope he reads them.
The guy is brilliant with R/C stuff and electronics, and I have been a fan of his videos for a while now without knowing who he was. ( Been into R/C since childhood, but just could never afford one )
One really clever thing he did, was to create a LIDAR controlled Ekronoplan ( one of my favorite machines ! )
He really does come up with some interesting stuff.
Looking at the shape of his latest tail, I'm reminded of others that have created them on this site.
He even used foamboard.
After looking at some at work today, I'm now interested in using it myself. If flexes to curves, unlike Coroplast.

I am reminded of Darin's ( MetroMPG ) work each time I see someone post a video like these.
It would be nice to see him mentioned, since he is one of the first to really put this stuff out there on Youtube in educational form along with A-B-A data.

Okay, but what do you do with it?This guy again?!https://ecomodder.com/forum/attachme...1&d=1665800225
I don't have any idea what he thinks he is saying.

Someone mentioned variance in accelerating and I remembered people here adding hand throttles, then wondered about adding a servo or something so that it accelerates up to speed the same way every time, then you can use cruise control.

You would just need an immediate way to deactivate it.Get our name out of your mouth! The last thing we need over here is activity!In reply:Excalibur John Deere pointed out that you get better fuel economy at slower speeds.This says that is for cargo, but I am sure that is open to interpretation.I finished reading the comments?! :D

Many comments recommended installing a separate gas tank and weighing it before and after.

How did they do that on the Centurion?

Youse yor words.

'part-2' and un-plucked fruit
 

I'm appreciative of the project.
Some impressions:
1) without the belly pan we might be looking at a Cd 0.035 penalty.
2) compromising the diffuser, due to the missing belly might add another Cd 0.02 penalty.
3) lack of rear skirts could mean another Cd 0.0.006 penalty.
4) and since a 'fastback' body is the most aerodynamically sensitive to tumblehome, there's a little extra fruit there for the taking.
------------------------------------------------------------------------------------
All of which adds to build complication and expense. And outside the scope of the project.
--------------------------------------------------------------------------------------
The 'skirted' configuration reflects an aggravated frontal area situation, so whatever the Cd was, it would be a little lower, if adjusted for the higher Af.
I give it all a big thumbs-up!:thumbup:
Thanks again for sharing it.

'Think Flight' V2 noodling
 

I ran an energy balance for all the Subaru configurations.
The conditions were:
65-mph ( 95.333 feet/second )
estimated Af= 23.8153 sq-ft OEM, estimated 24.4394 sq-ft with airdam
rho = approx. 0.0023256823
Cd 0.29 ( CarsDirect.com )
tires- 205mm width
REGULAR Unleaded, reformulated gasoline @ 111,836-Btu/gallon
6.138-pounds/ gallon
CVT transmission of 87.71% mechanical efficiency
40% thermal efficiency for engine
2546 Btu/ brake-horsepower
2.2% engine accessory losses
Rolling resistance power absorption is carried as a constant
Brake Specific Fuel Consumption ( BSFC) is carried as a constant
--------------------------------------------------------------------------------------
configuration-1: Cd 0.29, CdA= 6.9064 sq-ft, 35.644-mpg
configuration-2: Cd 0.214, CdA= 5.0981 sq-ft, 40.528-mpg
configuration-3: Cd 0.2671, CdA= 6.5289 sq-ft, 36.5637-mpg
configuration-4: Cd 0.1764, CdA= 4.3114 sq-ft, 43.0971-mpg
configuration-5: Cd 0.2291, CdA= 5.5467 sq-ft, 39.4562-mpg
configuration-'6'* Cd 0.1721, CdA= 4.2060 sq-ft, 46.5376-mpg
-------------------------------------------------------------------------------------
'6' represents a 17.3% radiator blockage to compensate for a now, 'oversized' radiator area.
-------------------------------------------------------------------------------------
The 'source' flow for the underbody and diffuser originates at approx. 16.11-inches above ground level before displacement. The separation edge for the slanted diffuser should not be any higher than 16.11-inches, as, at this elevation, the displacement kinetic energy nulls, and constitutes the 'reversal point.'
------------------------------------------------------------------------------------
Additional 'fruit' which lay outside the scope of the Impreza project, if added, could easily bring the Subaru into Cd 0.145 range

I wish he could build my Ioniq a rear boat tail like that. Wonder how stable it is at 100 mph with “ duct tape” though lol

Aircraft mechanics often refer to "hundred mile per hour tape," but it even gets used on jets. A case of tape was used to temporarily replace much of the fabric on the fuselage and tail of a bush plane that smelled of fish in bear country. Workmanship matters. When sealing a glider canopy for a record run along a windy ridge, someone didn't press hard and consistently, and a small section turned into a screaming reed for over ten hours.

Tape for aircraft comes in flavors, 100, 300, 500 mph mostly dependent on who you are talking to and the construction materials... safe to say 500 is really structural, not used much and super expensive. Doesn't come off either.

Not a duct tape, but 3M has a double-sided tape that is popular for replacing lines of rivets on box-truck bodies. I have a sample, and I'm a firm believer.

I was thinking of using 3m double sided tape to attach my extended box cavity versus a full boat tail versus drilling the box cavity in and hurting the value of my car more

Best would be a tape that can be removed with some solvent that doesn't affect the paint.

some Think Flight geometry vs performance
 

With the original modifications, the upper contour of the boat tail best matched the W.A.Mair, and FKFS (Kamm) 1/2-body silhouettes.
Maximum down-slope angle is 19-degrees. Diffuser upsweep is 20.5-degrees.
The 'highest' streamline filament traveling under the Impreza begins about 16.1-inches above ground level.
Along with the other mods, the tail gets the car to approximately Cd 0.214.
--------------------------------------------------------------------------------------
For the second-generation, #4 configuration modifications, the upper contour of the boat tail is almost an exact match to W.A.Mair's and Buchheim et al.'s 1981 VW Flow-body tails.
Maximum down-slope angle reaches 23-degrees. The diffuser angle is 5-degrees upsweep.
All mods contribute to an approximate Cd 0.1764, and achieves an almost 21% improvement in mpg.
--------------------------------------------------------------------------------------
Averaging for all configurations, the delta-drag/delta-mpg relationship works out at:
delta-10% drag = delta-5.15% mpg ( assuming constant BSFC )