<VV> Rocket Scientists

[Rad Davis]

At 09:05 PM 8/5/2005 -0400, Padgett wrote:

...

>What this means is that the fan and shroud for this century would be of 
>composite material have a much different appearance, and would move enough 
>cooling air with an energy budget of around 2 hp. Further the vents behind 
>the rear window would be gone because that is a low pressure area (would 
>be interesting to instrument the engine compartment above the shroud at 70 
>mph - anyone ever do that ?) Instead the engine would probably be a bottom 
>breather and use baffles to create a high pressure area at the fan inlet 
>and a low pressure area under the engine.

Yes, I have done that.  Turns out that the air intake vents are just about 
perfectly placed on the late coupe.  Can't speak to the late sedan.  If you 
go zipping down the road at 60 MPH, you have a considerable positive 
pressure in the compartment above the fan.  If you then measure the delta-p 
against the exhaust grille area, you get even better.  The Kamm tail does a 
pretty good job of making a low pressure area back there.  Of course, 
ducting air from the nose of the car would work better, but I suspect that 
GM found it problematic to have a 10 " dia duct running from the nose, 
through the trunk and passenger compartment into the engine bay.

It's fairly well documented that GM put the late car body in the wind 
tunnel.  It's obvious that they looked at cooling flow as well as 
crosswinds and drag coefficient.

I also have some visual data from my own '65 coupe:  I drove it to work 
year-round in North Carolina for about five years.  My commute was mostly 
interstate and the car was parked outside.  Cruising speed was 60-70 
mph.  Since this was NC, there was usually a thick layer of dew on the car 
in the morning when I walked out to it.  I'd start the car, run the wipers 
one stroke to clear the windshield, then carefully back out of my spot and 
drive to work.  As you might expect, the warm air of the rush-hour 
interstate dried the car off nicely by the time I got to work.  Now for the 
observation:  The first parts of the backlight to clear were two 
longitudinal patches in line with the air intake grilles.  It was as 
reliable as the carburetor icing that I'd get on the long climb uphill to 
the interstate with a cold engine during the fall.  It would seem that the 
angle of the backlight is gentle enough on late coupes, at least, to 
prevent boundary layer separation.  So much for GM not knowing where to put 
the air intake grilles...

Similarly, ask any Corsa convertible owner what happens to head temperature 
when s/he puts the top down and hits the interstate...

>The current design tries to do this by positioning the air exit at the 
>rear where the underbody airflow will provide an assist. That the engine 
>cools better with the lower shroud removed says this thinking was faulty.

I disagree again.  The reason the engine cools better with the lower 
shrouds removed is because of the proximity of the red-hot exhaust manifold 
to the heads.  Black body radiation is a very real concern when the exhaust 
manifold is 1/2" from the head.  Anything you can do to cool the manifolds 
off will cool the heads, and exposure to 70 MPH ambient air certainly 
qualifies.  Further, the lower shrouds do reduce total airflow over the 
engine, which would obviously raise head temperature as well.  GM needed 
the bottom of the engine to stay fairly hot in order to provide heat and 
defrost once the gasoline heater became optional.  Likewise, it's pretty 
clear that they decided that normal operating head temps below 475 F were 
acceptable, so who cared if the heads were at 450 instead of 400?

>So the engineers in Detroit, Tonowanda, and Waterford in the 60's were 
>doing the very best they could given the constraints they were working 
>under. However even a decade later at GMI I was able to get all of the 
>computer time on the 360 mainframe I wanted because few others were using 
>it. Nothing like a mainframe for brute force calculations of all possible 
>ratios for a THM-400 or simulating over 1 gee excursions with a B/P 
>Corvette, somthing we can do on a PC in seconds now but took overnight 
>then (and one of the first things a programmer learned, after the proper 
>use of a magic marker, was how to increase priority and get unlimited 
>runtime). Have forgotten JCL three times now and hope it stays that way.

I'll agree that a lot of the Corvair was designed to minimum cost or rule 
of thumb.  The post-1961 fan design is obviously deficient.  There's 
cardboard all over the car.  the early direct-air heat is almost as bad as 
a VW beetle's.  GM wouldn't spring for Viton O-rings.  Late shfiter tube 
bushings were cheesy plastic that broke all too quickly.  Having said that, 
rule of thumb via iterative testing works fine if you have enough 
iterations.  Likewise, with a maximum design lifespan of 10 years or 100 K 
miles, who cares if the alternator bearings seize at 120 K?

The simple fact that they started off with 80 HP Corvairs and ended up 
building 180 HP Corvairs with a bolt-on third-party turbo and needed no 
major design changes other than materials upgrades and 17% more 
displacement to do so points out that the original design was not exactly 
on the hairy edge of practicality.  Was it optimal?  No.  I'm certain that 
a Corvair built to 1960s aerospace specifications under government contract 
would have been an amazing car.  Of course, it would have cost as much as a 
T-38...

Having said that, I'd happily pay $200 for a properly designed composite 
fan for my van.  I could certainly use more cooling and the ~10 HP I'd get 
at 3500 RPM, and I suspect that fuel economy would improve as well.  It 
would be a much lower priority for the 140 HP Corsa coupes.  They're 
overcooled as delivered, so the only benefit would be the horsepower gain.