X-Mozilla-Keys:
Keith Moore wrote:
This concept [oil-to-coolant heat exchanger] still sounds worth further review, > > and also raises the question; why doesn't the "industry" do it?
Russell Kent wrote:
Paul,
I'm struggling to reach a conclusion about that "triple
flow" radiator offered by US Radiator. What they have
effectively done is taken a close-to-square radiator and turned
it into a long-skinny-folded radiator. For example, a Ford 1965
Mustang "triple flow" radiator is listed as 16 3/8 x 16 7/8 x 1
1/4 (276 sq. in. area).
This is really more like a 5 3/8 x 50 5/8
x 1 1/4 (272 sq. in. area) radiator that has been folded.
Compare this to two back-to-back 16 3/8 x 16 7/8 x 5/8 (276 sq.
in. or 552 sq. in. ?) radiators that flow coolant from
back-to-front. The The back-to-back radiators should have the
same drag (same frontal area and volume) as the "triple flow,"
but the back-to-back arrangement manages the delta Temp to be
more uniform from coolant entry to coolant exit. The big
questions are:
1. Which arrangement should yield a lower coolant exit
temperature?
To simplify lets look at everything with equal core volumes
and equal frontal areas. The back to front flow would have
lower exit temps.
2. Which arrangement has the lower pumping losses?
The pumping losses would be the same with a three pass
back to front assuming there were three rows of tubes
to achive this.
3. Do both arrangements actually have equal drag?
If the frontal area is the same and the over all thickness the
same.. probably.
Russell Kent wrote:
I assume (perhaps incorrectly :-) that US Radiator wouldn't offer
the "triple flow" for the hell of it, and so it must actually
give lower coolant exit temperatures (greater entry/exit delta
Temp) than a conventional radiator would in their automotive
applications. Why would that be so?
Russell
Paul Wrote
As Fred mentioned it is a question of too much flow. If all tubes
were in parallel like they are in stock rads the water would get
through the rad too fast and not give up enough heat. Don't
forget the car manufactures do a very good job of optimizing
the coolant system for climbing 7% hills at legal highway speeds
which may not be the same amount of HP and speed used in racing
cars and certainly not the HP and speed required in an airplane.
So at the expense of higher coolant flow restriction they put
one third of the tubes in series with each other. That means
you need a better water pump or the orginal pump was too good
to begin with.
Then of course you have the complicating factor of the speed
of the flow through the cooling jacket inside the engine. This
should be the first thing to optimise as we can't change that
by much. After that is determined a water pump can be sized to
achieve that flow as a first cut. Then the rad should be optimised
for that flow rate and the water pump re-optimzed for the additional
restriction of the rad to keep water jacket flow on target.
The main point about the rad is heat rejection to air drag
ratio. It should be as high as possible. The optimum flow
rate through the rad can be adjusted by the number of passes
used and the total area of the tubes. Back to front passes
being more efficient from a heat rejection point of view
than side to side passes.
It is not that these guys are dis-honest. They are just
ignorant and that is not hard to be when it comes to heat
exchangers and cooling systems :-)
BTW I have no doubt that these rads sold are optimum
from a flow restriction point of view for a small block
Chevy in the kind of racing they are selling the rads
for. They would be better if they used back
to front techniques and they could have the same
restriction to flow.
They may not be the lowest drag possible however
and they are certainly not optimized for a Mazda rotary
engine in aircraft use.
This is how I see it at the present time. I reserve the
right to change my opinion in the future :-)
Paul Lamar
The Aircraft Rotary Engine Newsletter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org