X-Mozilla-Keys:
Al Gietzen wrote:
Paul;
Sincere thanks to you and others for the valuable information given in this
e-mail exchange. I've only been on the list for about 10 days and have
gotten answers that questions I've had for months; particularly in regard to
cooling system design. I had read all I could find and never had found the
most fundamental information needed to even begin; and that is 'how much
heat do we need to reject?'
The plot of the heat balance for a rotary tells me very quickly that about
28% of the fuel burn energy goes to turning the prop (disappointingly low
thermal efficiency, but there is a tradeoff); about 18% goes into the
cooling system; and 7 % into the oil cooler. Pretty much the rest goes out
the exhaust pipe.
I can now compute very quickly that for my intended normally aspirated 20B
installation at a max power of, say, 260 hp (11,000 Btu/min); that the heat
to the coolant is very close to 7000 Btu/min; and the heat to the oil cooler
is 2800 Btu/min.
For good effectiveness of the radiator area I'd like to have no more than
about 20 F coolant temp drop across the radiator (I agree with Jeff). Then,
using the constants you so kindly saved me the trouble of looking up, I can
compute the flow rate needed for pure water is 42 gpm. For 50/50 water/EG
it is about 58 gpm. These flow numbers are inversely proportional to the
delta T; i.e., as the T drop across the rad is increased the flow rate
decreases, so if you're willing to accept a 30 F drop, and increase your rad
area a bit because of the lower average temperature; you can get by with 28
gpm for water and about 38 gpm for water/EG.
For you 2-rotor folks, multiply these numbers by 2/3 (.6667). It appears
that for a normally aspirated 2-rotor the stock pumps should give quite
adequate coolant flow rate; provided that you don't have a unduly
restrictive flow path. It also indicates to me that for a 3-rotor needs
more flow if I want to achieve the 20 F max delta T. Anybody know if the
20B has a pump that is different than the 12 or 13B?
The flow from the pump is a function of its speed and the pressure drop that
it is pumping against. Does anyone know where one might get a map (flow
rate as a function of rpm and pressure drop) for the Mazda pumps?
The air flow required can be similarly computed; and as it turns out, for a
50 F temp increase in the air (a seemingly reasonable number that Paul
chose), the air flow in cubic feet/min (cfm) is very close to being equal to
the heat rejected in Btu/min. So for a 20B I'd need about 7000 cfm for a 50
F delta T. That's a lot of flow (and drag). Maybe the direct air-cooled
engines can get by with less flow because their average fin temp is probably
considerably higher than we can accept with a water coolant loop. I had
always heard the liquid cooled was better from the standpoint of cooling air
required. Maybe we need to design for a higher air temp increase.
I assume in all of this that the design operating point is a sustained max
power climb. Is this reasonable? Climbing at 1000 fpm means that the power
will be dropping after a few minutes no matter what you do.
So there you have a few answers and some more questions. Any input is
welcome.
Al Gietzen (planned 20B in a Velocity RG)
Yes Al the climb part seems to be worst case. Taxi can be a problem
but Tracy solves that by dribbiling water over the rads.
Electric fans would also work but watch out for driving the fans
with air flow at high air speeds. The fan bearings seems to be an issue
and I know of at least one person that has had to replace the fan
motor with a motor that had higher quality bearings.
Paul Lamar
The Aircraft Rotary Engine Newsletter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org