Chapter 5: High-speed Cowlings, Air Inlets and Outlets, and Internal-Flow Systems
[161] In 1936, F. W. Meredith pointed out that the waste heat of a piston engine which is transferred to the cooling-air flow in a radiator is not all lost; it produces a small thrust provided the pressure at the exhaust of the radiator tubes is higher than the free static pressure of flight (ref. 192). This phenomenon became known as the "Meredith effect." Its mechanism was something of a mystery to many engineers of that period. A common fallacious notion was that the radial engine, because its fins were hotter than usual radiator temperatures of liquid-cooled engines, would enjoy greater benefits. (This mistaken notion still existed as late as 1949 and is stated by Schlaifer to constitute an "inherent advantage of the radial engine" (ref. 41).) The Meredith effect was so small at 1936 airspeeds that it could conveniently be neglected in performance estimates both by those who did not understand it and by those who doubted that such an effect really existed.
In our engineering analysis of the effects of heat in internal flow systems, the conversion of heat to thrust power was clearly the most [162] intriguing aspect. Thinking in terms of flight speeds of 550 mph, we calculated ideal thermal efficiencies of as much as 10 percent, and by Mach 1.5 the heated duct would have a thermal efficiency comparable to an internal combustion engine. Clearly, the insignificant "Meredith effect" had the potential to become a primary jet-propulsion system. (The term "ramjet" was not then in general use, and we were unaware that there were several discussions of propulsive ducts in the literature starting with Lorin in 1913 and including later treatments by Carter, V Leduc, Roy, and others.)
Excited at these prospects, I arranged a meeting with Langley's leading propulsion analyst at our Power Plant Division, Ben Pinkel. I also talked briefly with D. T. Williams, a young physicist whom Pinkel had recently assigned to analyze propulsive ducts at high subsonic speeds, including the effect of an engine-driver blower typical of the Campini system under study by Jacobs. Neither man showed any real hope for these systems, and Pinkel, reflecting the general attitude of most of the propulsion community at that time, patiently explained "the great weakness of all forms of jet propulsion-excessive fuel consumption compared to piston engines". When Williams' work was published about a year later (ref. 193), its primary conclusion emphasized the same point, showing on overall propulsive efficiency at Mach 0.8 on the order of one-sixth that of a piston-engine driving a propeller. Both men felt that tests of a propulsive duct in the 8-foot high-speed tunnel would be of little value. The duct and heater losses would, they speculated, largely nullify any possibility of net thrust at Mach 0.75.
In fairness to Pinkel and Williams it should be recalled that in 1940 the aircraft industry generally saw no possibility for supersonic aircraft. Mach 0.8 was regarded as a rather optimistic upper limit for the future. The potential of the turbojet for large improvements over the Campini cycle was not recognized either, and it is not mentioned in Williams' paper.
In spite of my disappointing session with Pinkel and Williams I resolved to proceed with the propulsive duct test. At the very least it would establish the Meredith effect as a major design factor at high speeds. Our 8-foot, high-speed tunnel afforded a unique tool for such an experiment. Stack solidly supported the idea. In promoting the project....

photo of 'heat model'
[163] FIGURE 42.- "Heat model" used in the first NACA investigation of a propulsive-duct (ramjet) system in the 8-Foot High-Speed Tunnel in February and March 1941. Model incorporated a 160-kw heater. Nose B and cusped outlet from ref. 179.
....we decided not to mention the jet propulsion implications in order to avoid the negative reactions of the propulsion people.
The nacelle model chosen for the tests embodied our universal Nose B shape together with our most effective cusped tail outlet (fig. 42). The all-metal nacelle was supported on a new thin metal wing selected to avoid the local area of flow separation that existed in the wing/body juncture of my inlet-outlet model. (In reviewing my original work at the request of Mr. Miller, A. M. Kuethe, who was employed briefly by NACA during the war, had endorsed my findings generally but had raised questions about possible drag interactions involving the separated flow. These would now be answered. By comparing the inlet results from the new model with the original data, we found no measurable effect of the separated flow.)
How to add heat at a high rate was our primary design problem. Combustion of fuel in the 8-foot tunnel was quite out of the question for many reasons. A search of the electrical heater catalogs with help from G. T. Strailman, Langley's principal electrical engineer, turned up no [164] high-output heater capable of being fitted into our 11-inch diameter duct. Baals and I therefore became high-capacity heater designers and produced a 160-kw, three-phase, 15000 F heat exchanger with 32 square feet of surface area in the form of 1.5-inch-wide Nichrome ribbon woven on reinforced asbestos millboard supports. This heater produced air temperature rises of about 3000 F at high speeds with very small frictional losses. The rates of heat input were larger than those due to piston-engine cooling, but still only a small fraction of the heat of combustion of kerosene.
Testing of the "heat model" started in February 1941, the first NACA wind tunnel investigation of a propulsive duct producing thrust. At a Mach number of about 0.5, the propulsive effect had become equal to the internal drag, and beyond this speed substantial net thrust was developed by the internal flow. At the highest test speed, Mach 0.75, the heated duct developed the respectable thermal efficiency of some 9.5 percent, close to the ideal theoretical value. As expected, the phenomena depended on the ratio of duct pressure to stream pressure, and was independent of heater surface temperatures per se. In all other respects, the careful measurements of these tests confirmed the calculations made by our engineering relations for analysis of this kind of internal flow system (ref.187).
In 1941 during the period of our propulsive-duct investigations, Stewart Way, of Westinghouse, made an analysis of the subsonic propulsion possibilities of "open-duct jet propulsion," his name for what was later called the ramjet. He also apparently conducted some tests with an electrically-heated model at about the same time of our high-speed tests in February and March of 1941, although the experimental work was never published (ref. 194), and we knew nothing of Way's work until years later. In the first version of our internal-flow-system report which was issued in September 1942 as a confidential document (ref. 189), the propulsive duct data were included but there was no emphasis in the title or text that the first NACA tests of a potentially important jet-propulsion system had [165] been made. Our "heat-model" tests rather definitely settled once and for all the doubts and arguments about the Meredith effect. Whether they had any impact on ramjet development is questionable. The revelation of the British and German turbojets shortly after our paper was issued had such an enormous impact that all the scattered U.S. activities in jet propulsion were in effect rendered insignificant. Almost overnight the propulsion community reversed its attitudes. By war's end, the ramjet was under vigorous development for missile applications. Both the Langley and the Lewis Laboratories of NACA had organized ramjet projects, concentrating on the prime problems of combustion and burner design which we had not been able to deal with in our 1941 project.
NACA was now being severely criticized for its prior general neglect of jet propulsion and it was clearly desirable to highlight whatever had been done. Accordingly, our report was reorganized to emphasize the tests of the ramjet system, and the words "Ram-jet System" were added to the title. The revised version is included in the 28th Annual Report of the NACA, dated 1943 but actually issued after the war.