View Single Post
Old 01-24-2011, 05:47 PM   #9 (permalink)
aerohead
Master EcoModder
 
aerohead's Avatar
 
Join Date: Jan 2008
Location: Sanger,Texas,U.S.A.
Posts: 15,861
Thanks: 23,922
Thanked 7,207 Times in 4,640 Posts
eddies

Quote:
Originally Posted by Piwoslaw View Post
I believe the answer is in the following diagrams:




It's all about eddies. In all cars there are eddies behind the rear wheels, but in a sloped-back car there also counter-rotating eddies from the rear pillars. The two (partially) cancel out - by increasing the diffuser angle (up to a point) the lower eddies are increased in size to be a better match for the upper eddies (allowing better cancellation). See figures Ba and E in the second diagram.

In a square-backed vehicle there are no upper eddies, only lower. By increasing the diffuser angle these are increased in size, but do not cancel with anything. Figure C in second diagram.

Looking at figure D makes me wonder about adding vertical fins to the sides of a square-back's diffuser. On the other hand, adding a Kammback would probably be a better idea.
I think the distinction needs to made between eddy flow ( turbulence ) and 'attached vorticity'.
Whenever the flow separates from an automobiles body,that region will be one of turbulence,or,random eddy flow.Once the air spins up into an eddy we've lost any chance of recovering the kinetic energy of rotation back into static pressure.The loss of pressure is by definition pressure drag,the major component of form drag,and the pressure of the entire wake will be that of the point of separation.This is why the tail is so important to us.The tail can completely prevent separation hence,pressure drag.
-------------------------------------------------------------------------
Attached-vorticity is a different animal.
The front of a vehicle defines how the atmosphere will be distributed as it passes the vehicles body.
In order to comply with conservation of mass,the windshield and roof cause the air to move at a greater velocity than down the sides of a vehicle,or underneath,due to the greater distance traveled by the air in the same time.
To comply with Daniel Bournouli's (sp?) Theorem,this accelerated flow must exist at a lower pressure than elsewhere around the vehicles body.
If the body designer is competent,the roof curvature,tumblehome, and upper radius of the greenhouse/C-pillar region will be such that when the flow off the roof rejoins the slower flow off the sides,these flows are at comparable velocity and pressure,the they simply continue on in a laminar flow fashion.
If the flows should meet beyond a critical velocity or pressure differential ( delta-V/delta-P) the slower,higher pressure flow will seek out the faster,lower pressure flow,and the two streams will spin up into a vortex as they attempt to blend together.
The vacuum created at the center of the vortex creates tremendous drag,and the kinetic energy required to feed the vortex robs precious kinetic energy from the flow,which,again,can never be converted to useful static pressure,hence the high drag and lower mpg.
If you've ever had a window seat on an airliner you may have seen wingtip vortices form during landing,caused when the higher pressure air below the wing is bleeding over the wingtip into the lower pressure air above,spinning it into a vortex.The pressure drop is extreme enough to create the refrigeration effect which brings the temperature of the water vapor in the airstream below dew-point,creating the clearly visible water fog condensing at the center.
--------------------------------------------------------------------------
The three different body styles presented in the book emphasize the aerodynamic peculiarities of each 'style.' The models are simplistic but do point to a large body of research in this area which have attempted to isolate these differences.
It's best to deal with every vehicle on a strictly case-specific basis but much is gained by recognizing and honoring the limitations illustrated in the book.
Up to recently,this is where the value of a good, large scale wind tunnel came in.Now we have CFD which can predict complex aft-body
'solutions' in 3-D flow.Short of both of those tools,the books are invaluable.
-------------------------------------------------------------------------
With respect to the diffuser and vertical plates,the diffuser won't work without a really good full bellypan ahead of it.So that's the first challenge.If that's behind us,then the sides of the diffuser should be out to the extremity of the body,following it up at it projects rearward.The departure angle will interfere with 'pure' aerodynamics,so there might be need of 'something' back there to help compensate.
Tuft-testing? Videos?
  Reply With Quote
The Following 2 Users Say Thank You to aerohead For This Useful Post:
Piwoslaw (01-25-2011), Sven7 (12-11-2011)