Finding shape with lowest drag coefficient

[julien.decharentenay]

I totally agree with you. It is quite an evolving field, and it is difficult to keep up to date with research and best practice.

From a methodology point of view, there has been some interested development with the so-called meshless models (I think that Autodesk Falcon is one of them) - but I have not used any and find it difficult to make the difference between claims and reality.

This being considered, simple models using combination of empirical/physical test/models can be far more accurate than very advanced models... It is often about validation... validation... validation... and experience...

[julien.decharentenay]

I just tried a sphere:

- Cd: 0.0846
- Frontal area: 2.2 m2
- Reynolds number: 1.6e6

Results: https://aerodynamicondemand.appspot....5-74a3a2e8fc3a

One of the very interesting feature is that the flow seems attached at the downwind side apart from the lower part (close to the moving ground surface) of the downwind side.

[Awd180]

Shape Effects on Drag

Just saying, but really fluid dynamics of an object in a jet stream is simple. Mostly the object that yields the lowest cd is a tear drop or double ended bullet. Relative to the above mentioned site and any lecture about it begins with talking about a rain drop. Another relavent way to conceive its design is fire, either waving a stick around or more interesting ways of seeing how the shape ultimately forms a tear drop shape.

[aerohead]

Originally Posted by julien.decharentenay

I totally agree with you. It is quite an evolving field, and it is difficult to keep up to date with research and best practice.

From a methodology point of view, there has been some interested development with the so-called meshless models (I think that Autodesk Falcon is one of them) - but I have not used any and find it difficult to make the difference between claims and reality.

This being considered, simple models using combination of empirical/physical test/models can be far more accurate than very advanced models... It is often about validation... validation... validation... and experience...

Even in wind tunnel tests of 3-D models things get weird and contextual.
Professor Alberto Morelli developed a CFD tool from a 1947 NACA (NASA) numerical algorithm.
In 1976 a scale model of Morelli's sent an earthquake and aftershocks through the aero community,registering Cd 0.161 with mild tail truncation.
The thing that wasn't mentioned was that the body was not at a practical ground clearance.
In 1978 the Italian govt. funded a full-scale demonstrator and with wheels,and realistic ground clearance,the car registered Cd 0.35,much higher than what was anticipated.
Aptera attempted to capitalize on Morelli's form and to my knowledge attained no remarkable Cd when compared to more old school,retro,vintage,antique,or Pleistocenic forms.
PS If my memory serves me,a NOVA broadcast showed a computer graphic simulation of cosmic radiation deflection by Earth's magnetosphere,creating a classical teardrop form.

[8307c4]

I'll save you (but mostly myself) the time, the old Lamborghini Countach had the world's lowest drag-coefficient, you can take that car's shape and put it to the test anyday, wheels and all...

Here's another:
http://cr4.globalspec.com/PostImages...FE10BC3519.png

It's a simple thing, starts with a point, ends in a point.
No getting around it, that is the world's most aerodynamic shape.

Actually the newest Lamborghini Veneno is super close, too:
http://en.wikipedia.org/wiki/File:La...profile%29.jpg

Now my work is done.

[Xist]

Originally Posted by 8307c4

Actually the newest Lamborghini Veneno is super close, too:

Needs a boattail! :P

[Gasoline Fumes]

Originally Posted by 8307c4

I'll save you (but mostly myself) the time, the old Lamborghini Countach had the world's lowest drag-coefficient, you can take that car's shape and put it to the test anyday, wheels and all...

What's the number and where did you find it? I'm finding a 0.42 Cd on Wikipedia for the 1974 Lamborghini Countach.

http://en.wikipedia.org/wiki/Automob...ag_coefficient

[Xist]

Originally Posted by Gasoline Fumes

What's the number and where did you find it? I'm finding a 0.42 Cd on Wikipedia for the 1974 Lamborghini Countach.

Automobile drag coefficient - Wikipedia, the free encyclopedia

My Forester is only .45!

[julien.decharentenay]

Thanks for the idea. I refine the concept a little bit as follows:
- Design is to include a box of 1100mm W x 1200mm H x 1900mm L
- Drag is measured by drag area (Cd.A) in place of Cd.

1) Simulation a - Cd.A = 0.6501 m2
Results


2) Simulation b - Cd.A = 0.5737 m2
Results


3) Simulation c - Cd.A = 0.5318 m2
Results


4) Simulation d - Cd.A = 0.4159 m2
Results


5) Simulation e - Cd.A = 0.1661 m2
Results


6) Simuation f - Cd.A = 0.1431 m2
Results


7) Simulation g - Cd.A = 0.1176 m2
Results