Testing aircraft engines, whether reciprocating or jet, tries the mettle of wind tunnel designers. Simply blowing ambient air around a closed circuit is completely inadequate. Aircraft engines operate at pressures all the way from 1 atmosphere to 0.1 atmosphere and less. Large air-exhauster pumps must be added to the conventional tunnel to simulate high altitudes. Furthermore, high altitudes also mean very low temperatures: - 50°F and lower. In consequence, a huge refrigeration system must also be installed. A unique requirement appears when the engine is operated in the tunnel-as it usually is. Something has to be done with the hot exhaust. It cannot be recirculated because it is contaminated with combustion products. It must be captured and removed from the circulating airstream a real engineering challenge.
Lewis Flight Propulsion Laboratory, near Cleveland, Ohio, tackled the engine wind tunnel design task during World War II. In fact, the first big facility built at Lewis was the Altitude Wind Tunnel. It was rushed to completion in early 1944 and was immediately applied to testing B-29, P-47, and XTB2D-1 engine-propeller combinations. In less than a year, however, a full-scale YP-80A Shooting Star (with clipped wings) was mounted in the Lewis altitude tunnel to explore the operation of that revolutionary development, the turbojet engine.
 The Lewis tunnel met the high-altitude and low temperature requirements by incorporating four reciprocating-type exhauster units (7000 horsepower total) and a 21 OOO-ton refrigeration unit that cooled tunnel air passing by its coils down to -50° F. The singular feature was the exhaust air scoop immediately downstream of the engine under test. It collected the hot combustion products and removed them from the tunnel. To replace the lost air, engineers injected clean air just ahead of the engine being tested. The resulting wind tunnel, while belonging to the same species as those at Langley and Ames, had three unique components: exhauster, refrigerator, and exhaust scoop.
Engines, inherently more complex than airfoils, were adorned with much more instrumentation in wind tunnel tests. Tests requiring 1000 simultaneous readings were common. For example, to check out engine performance, one measured power level, engine RPM, fuel flow, supercharger setting, cowl flap opening, and hundreds of separate temperatures, pressures, and forces. All these data had to be collected, recorded, and analyzed. While this seemed an instrumentation nightmare in 1944, it was really a harbinger of things to come in the approaching missile and space ages.