[Tesla]

Just whent back through some previous discussions here trying to get an answer on how to calculate side taper,
seems to confirm that Half vehicle width is the basis to calculate tail length, Quotes below, note comments in bold:

Originally Posted by aerohead

The plan contour would mimic the contour in elevation.Everything depends on where the max roof camber point is with respect to the point of max body width,as this is the zero-point for beginning the plan taper.And the pan-taper would never exceed 22-degrees as well.
On the 'Template',technically,the zero points are at the same location,as we're working with a body of revolution.
In the real world though,the body needs to be more narrow,so we have to morph the top and side contours together best we can.
As Frank Lee has mentioned a number of times ,if the body is 'tall' let width be the dominant factor for the taper.If the car is wider,use the height.
If they're the same,do each equally.
-------------------------------------------------------------------------
If you want to see some minimums,GOOGLE:
Fairchild C-82 Packet
Fairchild C-119 Flying Boxcar
Fairchild C-119J Beaver tail
Bristol 170 Wayfarer
Gotha Go 345
Armstrong Whitworth Argossy

Originally Posted by aerohead

In the images posted earlier for the railroad boat-tailed tailcar,all the boat-tailing was done in plan-view.
The length of the tail is 1.74X the body width for the full stinger. It's within 2.5% of the template,so it's going to just slide in between Mair's 22-degree,and Emmelmann's 23-degree max slope angle protocol for attached flow.The base drag is completely eliminated.Which would be the lion's share of road load for a light truck on the highway.
The locomotives,self-powered railcars,and buses all show higher drag without the 'trailer.'
At 9" ground clearance,the ideal self-powered railcar comes in at Cd 0.08.NUNA territory.
We may not be able to go that low with ground clearance without active suspension,but nobody has said that we can't streamline the wheels.
The caveat is,that we can't have all the excressences like awnings and rooftop AC units,'n such scraping the sky as we go.These sorts of things must be better integrated into the body,as orbywan is doing on his rig.

Originally Posted by drmiller100

length of the tail would then be 3.5 times the half a body width (two sides).
the arctan(3.5) is about 74, leaving the angle 16 degrees to the direction of travel.

My theory is there is theory, and what works in the real world. I think for the sides you have to run a shallower angle because direction of wind in relation to the car is a big deal, where when looking at the top of the car, you can run a steeper angle as teh top of the car doesn't care what angle the wind is.