"Kent, Russell" wrote:
>
> Paul Lamar wrote:
> > Way back in the early 30's when NACA was just getting started
> > they commissioned the National Bureau of Standards to study
> > radiator design for aircraft use. NACA Report No. 86
> > Properties of Special Types of Radiators by S.R. Parsons.
> > They coined a term they called "Head Resistance" and assumed
> > there were no ducts or diffusers. In other words the radiator
> > was stuck below or above the engine out in the air stream.
> > Air flowing through the rad was the same as aircraft forward
> > speed.
> Shame on you, Paul, you know better. There will be external diffusion
> ahead of the radiator, causing slower, higher pressure air to flow
> through the radiator.
>
> Russell Kent
That is factored in by the way they ran the test in the wind tunnel.
They noticed the effect by varying the spacing between the hollow
plates from 1/4 inch to 1/2 inch. 1/2 inch gave better results. No doubt
the study was flawed in several respects. It was also a strange mixture of theory
and experiment. I am just saying the concept might have potential and should be
revisited. What you are seeing here are probably Reynolds number effects.
Also by installing such a heat exchanger in a modern diffuser the effect of
air speed can be minimized. There is also not much chance of anybody coming
up with the wherewithal to optimize this design unless NASA gets involved.
However they seem too high and mighty.
Paul Lamar
My guess is that with the use of cowled rad with a diffuser, the same benefit of thicker rad can be had with the radiator fin construction now in common use. The key to high efficiency
is the high Delta T air temp through the rad.
Sooner or later, you'll fly a thick rad : )
Tracy Crook
I don't think so Tracy. Read Kays and London. This NACA work was done way before external
diffusion or the concept of a compact heat exchangers or the complete understanding of
what happens in a thick rad.
First of all the delta T in question is that between the rad metal and the cooling
air. You seem to be confused on that point. I think what you are referring to
is the increase in water temp as it flows through the block. I clarified this
on Bill's graph.
High increase in water temp actually hurts in a thick rad rather than helps. Assume
a 50 degree increase in cooling temp through the block. The water temp out of
the block is 200 F and the air temp is 100 F. The cooling water temp drop through
the rad must equal 50 F or the engine will eventually over heat. Since one degree
increase in cooling air temp is a one percent loss in cooling effectiveness.
Assume for the moment that temp increase is also achieved by the air flowing
through the rad. That means half way through the thick rad the air temp is up
to 125 degrees F leaving a delta T of only 75 degrees. The last half of the
rad is now only 75% as effective as the first half. This is a crude approximation
of what happens. The rule of thumb is the first half does 3/4 of the cooling
and the last half does 1/4. The ideal heat exchanger is zero thickness with
infinite frontal area.
You don't need to take my word for it. Here is what Stewart Warner has to
say about it.
Paul Lamar
Here is what Kays and London has to say about it.
Let me translate some of it [].
Paul Lamar
Heat Exchanger Thermal and Pressure-Drop Design 41
FLOW DISTRIBUTION AND HEADER DESIGNt
One disadvantage in the use of highly compact surface geometries is that the resulting core shapes are characterized by large flow frontal area and short flow lengths [read thin] for the gas-flow
path. The common automobile radiator is a close-at hand example-a frontal area of approximately 3.3 ft2(0.31 m2) and an air flow length of 1.2 in (0.0305 m) for a 220 hp (164 kW) automobile engine! In
this application, the large frontal area can be accommodated by a front mounting on the vehicle. However, heat exchangers in other installations must be located in ducting, and the flow header
[diffuser] configurations have a definitive influence on the engine system envelope geometry.
Figure 2-42a describes a folded-core Concept used to reduce header [diffuser] volume.
However, if the pressure drop across the core is not uniform, as is the case for the Fig. 2-42a configuration, the flow distribution over the heat transfer surfaces will not be uniform, and a penalty
in heat exchanger performance is the result. From this viewpoint, the design objective is to provide for acceptably uniform flow through the core with an acceptable header [diffuser] geometry and
!low-stream mechanical energy losses. Uniformity of flow distributioft over the core face is the dominating function of the headers [diffusers]. The importance of header [diffuser] design is
emphasized in the schematic representation of the methodology of heat exchanger design (Fig. 2-12), where it is listed in the mechanical design box. A poor header [diffuser] configuration may actually
result in a lower overall pressure drop, but the penalty in heat transfer performance can be very large.
London et al. [17] considered three header [diffuser] configurations, as shown in Fig.
2-42b to d. They provided the theory needed to specify the exit and inlet header [diffuser] shapes and sizes which yield the desired uniformity of the pressure drop, ~Pmatri" over the core.
Experimental results support the predictions quite well (:t5 percent nonuniformity).
As illustrated in the numerical example of Ref. 17, the parallel-flow configuration has a header [diffuser] pressure loss (total-to-total), exclusive of core pressure drop, equal to 2.4 7 times the
inlet velocity head based on the velocity Ui shown in Fig. 2-42b
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/
Copyright 1998-2004 All world wide rights reserved.