G'day
NutLoose. The figures I quoted, ironically, come from a speech made by Lee Atwood. Lee had made claim to something that he was not. He did not design the P-51, that came down to Edgar Schmued, Lee has a reputation for promoting himself as doing things he did not with respect to the P-51. It’s interesting that he claims he was the Chief Engineer, when he is listed by Edgar, and others, as Vice President and Assistant General Manager. The following is what Ed Schmued had to say re the radiator design,
“Then there were some other features of the airplane that are essential. For instance, the radiator of the airplane; if it is a ducted radiator (surrounded by cowling), it has a tendency to spill air. The air that is coming into the opening cannot all go through the core of the radiator itself and spills around the edges. Here it is very important to locate the radiator in a position where it cannot interfere with either wing or main-fuselage drag, because any spillage is turbulent air, producing drag. To keep that to a minimum, I moved the radiator as far aft as I could, and as far below the wing as I could. With the radiator located in this position, we really got an optimum.
”Since the radiator was far enough aft, it interfered as little as possible with fuselage drag and with wing characteristics. We had a world beater! We also found out, later on, that the heat from the engine actually produced thrust in the radiator by increasing the velocity of the air flowing through.
“That horsepower gained by the radiator was only discovered by wind-tunnel investigation. We were contractually required to wind-tunnel test the P-51, and long after the first airplane flew, we got around to test a model which had an electric motor to drive a three-bladed propeller. We found from wind-tunnel data that the P-51 should not be as fast as it was actually clocked. Our chief thermodynamicist, Joe Beerer, studied the problem and noticed the favorable effect of the radiator." This added thrust is called the Meredith effect, after the British engineer F. W. Meredith, who described it in a paper published in 1935.
The distribution of power is approximately as follows: 100 percent energy in fuel is put into the engine, but about 30 percent of this has to be radiated into the air to cool the engine. Another 30 percent of the energy is lost in the exhaust heat of the engine. Then 25 percent is usually used on the propeller. This amount of power drives the airplane, overcoming the drag, so 15 percent of the fuel energy is lost in mechanical friction. This indicates that an airplane, or any other heat machine, is not a particularly economical device.
“The British Air Ministry was extremely helpful. Among others, they sent us Dr. B. S. Shenstone [who arrived February 25, 1941], to assist us in some of the airflow problems into the radiator. The radiator, as we had it, consisted primarily of a fairing, which started at the bottom of the fuselage and enclosed the radiator. Dr. Shenstone advised us to provide an upper lip on the radiator housing, which was about 1.5 inches below the fuselage contour. By doing this, we got a much better pressure distribution in the air scoop.
“Previously, air would go in at the bottom of the scoop and spill out on top. By providing this lip and the gap between the fuselage and the radiator lip, we actually equalized the pressure distribution in the duct and got a much better cooling system. We reduced some of the loss due to spilling, which is always detrimental and will produce a certain amount of drag. lt is always a problem with any kind of a ducted radiator installation, so this all helped to reduce the spillage."
Aerodynamicist Ed Horkey fills in the details. “The back was the greatest place in the world to put [the radiator]. However, we started out with an opening to the radiator duct, the top line of which was the bottom line of the fuselage. The problem was that being so far back in the fuselage caused the boundary-layer build up and the airflow wasn’t doing the job of furnishing enough [air] to the radiator to give efficiency or enough cooling.
“Meredith had brought forth the theory or proposal to take in air at a high velocity and slow it down, which, of course, builds up pressure. As it goes through the radiator core, the pressure helps some, but primarily you have more dwell time. Then with the increased pressure and temperature, you squeeze the air down again as it goes out the back and you actually get some thrust from this, or what can be called negative radiator drag. It was great, but with the problems we had at the inlet, we weren't achieving cooling or a drag reduction.
“Ed certainly looks at it from a different viewpoint than we in aerodynamics did at that time. Actually, with the one-quarter scale model at Cal Tech, Irving Ashkenas, who had been working with me, was doing the night shift and I was doing the day shift. He was over one night and came up with the idea 0f why not put a boundary bleed in. ln other words, take that top line of the radiator duct and bring it down from the bottom line of the fuselage and let the turbulent boundary layer that built up under the fuselage go by the entrance and therefore you would get more efficient air into the duct. He went ahead one night and did this on the model and the results were great. Other people may have come up with that later on. I want to give full credit to Irving Ashkenas as really being the developer of the boundary-layer bleed. This is in full use yet today. For instance, one can look at the F-16 and see the tremendous boundary-layer gutter that they use.
Perfection of the cooling system was an ongoing process, requiring the input of Horkey’s department, wind-tunnel model work, and, above all, continuous flight testing. And most of it would have to be redone in 1943 and 1944 when the Mustang shifted to the more powerful Merlin engine.
Writes Horkey: “The boundary-layer bleed wasn't all that simple. If you drop the radiator inlet down far enough to get rid of all of the turbulent boundary layer, the drag would be too high; so it ended where it was a compromise of the bleed depth to get an acceptable cooling performance with minimum drag.
“It so happened that later on when we started the P-51B project, we got the boundary-layer bleed a little too small. What would happen was that it caused a duct rumble. Chilton described it to us as somebody pounding on a locker. The boundary layer would build up, and airflow would go around the duct inlet, and then it would all of a sudden go inside again, and this would create a large impact load. We did two things. We got the model out again and started checking the bleed. We also took an actual P-51B up to Ames Aeronautical Lab and cut the wing span down a little and mounted in the 16-foot high-speed wind tunnel.
“I took the first ride, and when we got up to 500 mph in the tunnel, we got the rumble. lt was quite a thrill. Smith J. DeFrance of NACA at Ames, Manley Hood, and Bill Harper all worked with us. We lowered the top inlet of the radiator duct a small amount and also went to the cutback, or slanted, inlet shape and solved the problem on the P-51B. Later on, on other airplanes like the P-51H, we had to make an eighth of an inch change, and again make sure we had this problem in hand.
“l am absolutely certain, having been there, that the boundary-layer-bleed solution credit should be given to Irving Ashkenas. He later went on to Northrop and participated in the development of the P-61 and their flying wings. He still has a company that does aerodynamic consulting on drag, stability, and control, etc."
What gets lost in the discussion is that, yes the radiator produces thrust, but not sufficient to overcome cooling drag. You can see evidence of this in one of the P-51 Reno racers that has dispensed with the radiator completely, and has a complete loss system, whereby the coolant boils off and is vented overboard. You can see the steam venting from the outlet in the fuselage immediately above and behind the wing trailing edge, they claimed removing the scoop and radiator substantially reduced drag.
