W.A.Mair's Boat Tail (1969)
 

I've been referring to this work online for 7-years now and thought that I could save some breath if everyone had the image.
--------------------------------------------------------------------------
Mair's work appeared in the November,1969 issue of AERONAUTICAL QUARTERLY,pgs 307-320.His work is also illustrated in Wolf H.Hucho's textbooks.
Only the tail itself is depicted in the literature but I've gone ahead and drawn out the entire wind tunnel model Mair used in his research according to specifications.
Mair's model is essentially a 16.13% thickness circular fuselage with prolate-ellipsoid nose, with a free-air frontal area drag coefficient of 0.066.
Without the boat tail,the model is essentially a Gavre-type artillery projectile of frontal area-based Cd 0.204,as measured and reported by the U.S.ARMY's Aberdeen Proving Grounds.
Also,Mair's drag curve depicts only the pressure drag component,leaving off the skin friction.
I've constructed a drag table which includes the total profile drag,also superimposing ground proximity,wheel drag.
In lieu of the 'Template',Mair's boat tail could be used for aft-body streamlining.It would not be as strong,lacking the full compound curvature,however,it would be easier to construct once past the curvelinear transition into the constant angular cone.
If you used this architecture for a passenger car you might expect as low as Cd 0.166.
If you 'lost' the straight-walled tubular section,leaving only the nose and tail you might expect Cd 0.12.
http://i1271.photobucket.com/albums/...ad2/WAMair.jpg

That appears to me to approximate the frontal prolate-ellipsoid to about station 50, with a simple cone from there.

Arranging in order from roundest to pointed-est, I have Schlorwagen, Dymaxion, Mair and then The Template. Rounding off saves overall length. Is there any guidance here?

guidence
 

What I took away from Mair's work was:
*Be careful about the transition into the curve. As it is depicted,the gentle curvature will protect the fragile boundary layer,preventing separation.
*Once you're to 22-degrees,you can just project that line as far as you like.This construction would require internal bracing to prevent tin-canning as without the ovoid compound-curvature she'd be rubbery.
*Mair's drag curve suggests the decreasing returns you get from extending beyond about 52% of the full tail.
*At this length,a car would have Cd 0.284.
*If you ran an inflated tail section out to 90% you'd have Cd 0.166.
*If you lose the constant-velocity portion of the model,leaving only the nose and tail,you end up with a body of revolution 'template' of L/D= 2.52,very similar to the 'Template,' of Cd 0.04,and Cd 0.08 in ground effect,and around Cd 0.12 as a half-body with skinny wheels.
*This is where I took the 22-degree limit for the 'Template.'
*Hucho's partner,Rolf Buchheim (sp?) allowed 23-degrees as a maximum.

Why does the tubular section help? Do its benefits transfer to ground effect?

tube
 

The tubular section is just pure skin friction.
If you chopped the model in half for a limo body or 18-wheeler,without wheels,its minimum drag would be Cd 0.13.Cd 0.17 with wheels.

Oh, I need to work on reading comprehension... or perhaps sleep! Thanks. :)

Does Mair outline how the rear transition section is defined?

What I am taking away from this is the acute angle at the rear would be 2x(22 or 23°) or, on average, 45°. Which is what I use on the 3-foot tail in the Collapsible Boat-tail thread.

When you truncate, doesn't a recessed box section work like a half-tonneau on a pickup, and further reduce drag? Is there a rule-of-thumb for determining the depth of the box?

There was a study posted here somewhere in the past which indicated this to be the case. I've wonder if even more "innovative" ways to truncate would be productive, like perhaps a 1-2 degree recurve at the very end of the tail. It's just thinking out loud - no supporting data;)

defined
 

All I've seen is the broken line drawing depicting the aft portion of the body and tail,along with the drag table as depicted in Hucho,Chapter 4,Figure 4.41.