This is not exactly what we are looking for but there are
some clues here.

This is from SAE paper 821149 "The Effects of Turbocharger Design
and Installation on Gasoline Vehicles Transient Response" I provided
some of the instrumentation on the R&D reported in this paper.
Note the 20 HP loss in the overall steady state engine net HP out going
from a .63 to a .38 A/R. This is a T3 turbo charger. R, the radius
of the turbine, is fixed and A, the area of the nozzles, is changing.
Basically what is going on here is more back pressure and net HP loss
is being being generated by the smaller nozzle.

Just like electric motors you can get the same HP from a
turbine if you make a small one running at high RPM or a large
one running at low RPM. In this case the weight penalty is less
severe in the over all scheme of things. The larger turbine
will have the same HP loss with the same nozzle area but the
RPM will be less and the turbine torque higher. Consequently turbine
power will be nearly the same. Turbine power is the bottom line
for turbo compounding.

Paul Lamar

BTW Doug I made contact with my old friend at Garret, David Elpern (He is still
there. Now Honeywell). See SAE paper 821149 The "Effects of Turbocharger Design
and Installation on Gasoline Vehicles Transient Response". He is working on gathering
HP data on the turbo chargers but don't hold your breath as that is not what they
are really interested in. You might have to build a turbo dyno. (Not as hard as
it sounds.)

Here is a chart from Corky Bell's book "Maximum Boost".
For a turbo compound it looks like we need the largest  A/R available for 
the  TO4. Checking the 1999 Turbonetics catalog for the 13B that is 1.52 Garrett
part number 407263-0039 Tang Divided (One pipe from each exhaust port).
P-trim with a 2.544 Exducer.  

For the RX8 engine with the confused side exhaust system an undivided
A/R = 1.71 should work as well. Garrett part number 407264-0039. P-trim 
with a 2.544 Exducer. Big shaft in both cases.

Here too is a clarification on the A/R number. As you move around the scroll
the A/R ratio remains the same.

BTW the A/R as near as I can calculate it for the 88 89 Turbo II 13B
appears to be about .82. 

Paul Lamar
 
The Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/            
Copyright 1998-2004 All world wide rights reserved.