How To Line Up Template - Sports Car Design

[kach22i]

Originally Posted by winkosmosis

Trying to align a template arbitrarily with a whole front end that has zero relevance to a normal car is a waste of time and energy.

Sooooo, it's the front part of the template which bugs you the most?

I have to admit, that it's one of the stranger parts. Feels silly, looks silly, but it's the tail end which makes the big difference, right? Drag mostly comes from everything after the mid-point, at least from what I can tell.

[aerohead]

Originally Posted by winkosmosis

You pulled the "template" out of your butt. You can't just overlay a half teardrop shape on a car arbitrarily and claim that following that shape on part of the body is the optimal way to improve aerodynamics. It's just a crutch to avoid critical thinking.

The OP's images demonstrate perfectly how arbitrary this baloney is.


What matters here is whether airflow can stay attached to the rear window and depart smoothly when the body ends. Like I said... The air doesn't care about your "template".

If you'll actually read the thread,you'll discover the science which is the genesis for the 'Template.'
If the vehicle has attached flow up to the point of maximum roof camber,then the 'Template' describes a separation-free curvature,which if followed to its conclusion,along with plan-taper,results in the elimination of pressure drag.
The critical thinking was done by the phDs dating to 1921.
If you have Hucho's book,then you already know this.It's his Template.

[aerohead]

Originally Posted by winkosmosis

According to the almighty template, the 911 sucks. And yet it's one of the most aerodynamic sports cars at 0.28 Cd. You know why? Because the roof and rear window are sloped shallow enough that airflow remains attached and doesn't detach until the rear spoiler. Those angles are what matter here. If you made the 911 look like the template, there would be no advantage for flow attachment, and the cross section of the rear end would be much bigger.

This is a perfect example of the arbitrary and nonsensical nature of the template.

Forget templates and just make sure the curves of the body allow attached airflow all the way back at the desired speed! BTW, you notice there aren't different versions of the template for different speeds?

Can you explain why Boxfish,with 12% less roof has Cd 0.19 vs 0.28 for the 911.
Can you explain why Porsche has abandoned the Lange body with the new 918 and that it is more 'Template?'

[aerohead]

Originally Posted by winkosmosis

But the 911 is disadvantaged by huge wheel wells and big spokey wheels

Perhaps the 911 is disadvantaged by the flow separation off the pseudo-Jaray roofline.

[autogyro]

Oh, good grief! I suppose everybody on this thread has a hobby horse to ride, so I'll saddle up too.
At least in my copy of Hucho (chapter 4) he mentions that there are other basic body shapes that surpass the Klemperer/Gottingen shape form the 20's, specifically the Morelli shape from 1976. (I actually have the paper)
That's not my point. A big part of chapter 4 is "detail optimization". This is where the rubber meets the road, literally. Drag caused by wheel turbulence is a huge factor, and has only recently begun to be modeled and tested with any sophistication. Trailing wheel vortices from the rear wheels alone can consume 5kW. (Morelli again)
Likewise cooling system drag.
Hucho's answer, and mine, is to test, and test again, using reasonably sized (30%) models to get reasonable Reynolds numbers. My belief is that you don't need the MIRA wind tunnel to do this.
Automotive body engineering - Roof Rack "Wind Tunnel": A Naive Idea?
Sure, play with all the basic Shapes you want, but the devil is in the Detail Optimization.

[aerohead]

Originally Posted by autogyro

Oh, good grief! I suppose everybody on this thread has a hobby horse to ride, so I'll saddle up too.
At least in my copy of Hucho (chapter 4) he mentions that there are other basic body shapes that surpass the Klemperer/Gottingen shape form the 20's, specifically the Morelli shape from 1976. (I actually have the paper)
That's not my point. A big part of chapter 4 is "detail optimization". This is where the rubber meets the road, literally. Drag caused by wheel turbulence is a huge factor, and has only recently begun to be modeled and tested with any sophistication. Trailing wheel vortices from the rear wheels alone can consume 5kW. (Morelli again)
Likewise cooling system drag.
Hucho's answer, and mine, is to test, and test again, using reasonably sized (30%) models to get reasonable Reynolds numbers. My belief is that you don't need the MIRA wind tunnel to do this.
Automotive body engineering - Roof Rack "Wind Tunnel": A Naive Idea?
Sure, play with all the basic Shapes you want, but the devil is in the Detail Optimization.

*If you look at a transverse section slice of the Morrelli shape, the degree of diffuser angle necessary for that form to work,and necessary ground clearance,you may wonder if this 'template' would lend itself to modifications to a conventional passenger car.
*The 'Template' encompasses full wheel skirts at all four wheels and delivers Cd 0.13.
*When detail optimization of the wheels is performed drag potential rises to sub-Cd 0.1.
Walter Korff gave us the lowest drag form cooling system as of 1963.That technology has been available to all ever since.A no-brainer.
*Hucho says that detail optimization can only get you so far.
*Hucho also says that the only way to lower drag is the 'Template.'
*The basic shape for the 'Template' is Hucho's.Some of the references are listed in Part-C of the 'Template thread.
*My belief is that enough wind tunnel testing has already been done.
*Low-drag shapes are off-the-shelf technology.There are no mysteries regarding low drag.
*People appear to suffer from the tyranny of unity.It's impossible for them to believe that there can be a single solution to low drag.
* And it smacks of intellectual dishonesty when high-efficiency is presumed for sub-mariners going aboard an Akula,kayakers entering white-water,travelers boarding a commercial airliner,or a perrigrine falcon departing from a stoop at 280 mph,but then this same 'technology' couldn't possibly impact their road vehicle.

[aerohead]

Originally Posted by autogyro

Oops! My Bad. Heres the right one

I'm seeing curvature leading into the rear slope.

[aerohead]

Originally Posted by kach22i

Autogryo, I trust you don't mind my marking up some of your images so that I can explain the whole tangent line thing as I see it.

Your first one:
Automobile pictures by kach22i - Photobucket


Your second one:


Here is my car at 30 degrees - approx.


Some other P-Cars:


An old racing Porsche modified with roof line per race rule allowances.


I stand by my earlier statement that 10-15 degrees is best, you can get away with 20 degrees on a good day, and that 30 degrees has air lifting off of it quite easily at speed. I also think the template works and is an "Idea" we should strive for but maybe not marry.

W.A.Mair,who's work is shown in Hucho's book would argue that 'angles' must be taken within the context of 'where' they occur,with respect to the position of max. roof camber.
This is why the 'point' is a centerpiece of the 'Template.'
The aft-body can slope as steep as 22-degrees,but you cannot get to 22-degrees without separation until you are 1.1 free-air body diameters ( twice the body height )behind the point of max. roof camber.
-------------------------------------------------------------------------
*Looking at the Porsche 911 RS America,the backlight is already 21-degrees ( nearly Mair's limit of a maximum 22-degrees )by 20% of 'Template' aft-body.
*As the rear contour just comes onto the 'tea-tray' rear spoiler it is at 23-degrees,already exceeding Mair's maximum for attached flow.
* At 44.4% 'Template' aft-body the trailing edge of the tea-tray gives the 'apparent' backlight/angle 13-degrees,rather than 32-degrees without the spoiler.
*If the Porsche 'template is followed to the ground,the 911's exit tangent angle is 40-degrees.
-------------------------------------------------------------------------
The 911 has survived since 1963 when it bowed at the Frankfurt Auto Show as the 901.They already had to 'fix' the roofline a bit for 1964 production,which has become the iconic sillouette.
None of the other Porsches use the 911 roofline.
*With the 1996 911 GT1 they spent 33-days in the wind tunnel trying to kill the 'inherent' lift caused by the 911s steep roofline.They ended up lowering the car 5.6" and adding 16.6" to the rear,ending up with a 3.991:1 fineness ratio,excedding even Jaray's pumpkin seed of 1921.
------------------------------------------------------------------------
The early Porsche designs,dating to 1910 were all tulip or teardrop forms.
*The 1937 R-Type Auto-Union GP teardrop racer of to-be Porsche designer Erwin Komenda clinched the victory,setting Porsche in such a favorable light with Herr Hitler,that he would beat out GM and Ford for The People's Car program.
-------------------------------------------------------------------------
Porsche 'Template' aero-modded race cars include the 907,917,and 956 long-tails,which all ran 100%'template' rooflines and boat-tails.
*The short-tail 917 spyder,Cd 0.57,would see Cd 0.44 when the windshield and roof were installed.
* When the long-tail was added,the drag dropped to Cd 0.33.
* This is exactly what the 'Template' is structured to do.
* The 'EcoModded' LeMans long-tail 907,of 1968,saw Cd 0.27 when the 911 had the Cd of a VW Bus.

[aerohead]

Originally Posted by winkosmosis

But my point is that the template is meaningless in this context. People are overlaying it arbitrarily on their cars and then thinking that where the lines fall happen to describe some kind of ideal shape. The shape of the template is only ideal for the whole shape described by it.

It's completely backward reasoning.

You're right, it's not just attached flow that matters. It's attached flow that allows the car to leave as small and clean a hole as possible in the atmosphere once air is done flowing around it.

You can do better with common sense. Start with a horizontal surface (most roofs), increase the curve gradually, and don't exceed 20°, or 15°, or 10°, or whatever (depending on airspeed). Trying to align a template arbitrarily with a whole front end that has zero relevance to a normal car is a waste of time and energy.

*Hucho's point is that once the front end has attached flow,you go after the aft-body.
*Hucho lays out the rulebook for doing the aft-body,and the 'Template' is the embodiment of all the 'critical thinking' requirements to pull that off, with a minimum of material(mass,cost,inertia,momentum,tooling,invest ment,etc.).
*The 'Template' cannot precipitate separation,so for as long as you dare follow it,you're guaranteed attached flow.
*When you do decide to do the Fachsenfeld/Kamm/Breer 'chop',the wake will have the highest base pressure,i.e. the lowest pressure drag which is what streamlining is all about.
*The gradual curve that you speak of is the scientific premise of the 'Template'
* There is no 'arbitrary' alignment allowed with the 'Template.'
* As far as 'ideal',if that means the biggest bang for the buck,then yes it's ideal.
* if you're steeper you lose.
* if your less steep you lose.
* The 'front' of the 'Template' is ideal for sub-transonic flow.The emphasis for the 'Template' is on the aft-body.This is where all fluid mechanics text will emphasize that we focus our attention,including road vehicle aerodynamics which is only a branch of fluid mechanics.
*The 'small and clean hole' you mention is the wake area.If you use the 'Template' there is no effective wake at all.Pressure drag is completely elliminated.
*If you're going to debate with an anecdotal brief such as 'common sense' your going to lose the debate every time.There is no such thing as common sense.It doesn't stand as prima facie evidence.

[kach22i]

aerohead, thank you for all of the information.