The NACA Lewis wind tunnel group had already wrestled with the special problems of testing operating subsonic engines, with their attendant floods of combustion products and ear-splitting noise levels. With the advent of the Unitary Plan, Lewis was assigned the task of providing a supersonic tunnel large enough to encompass full-scale jet and rocket engines. The central problem was not building a supersonic tunnel per se, but rather adding to the basic tunnel the auxiliary equipment that would properly simulate the air density, temperature, humidity, and purity at anticipated operating conditions. This was compounded because supersonic tunnels achieve their high velocities by expanding air through a nozzle; as this air expands, it cools rapidly and there may be condensation of contained water vapor. Consequently, control of the engine environment and the facility environment becomes much more difficult.
Under the leadership of Abe Silverstein and Eugene Wasliewski, the basic test section was sized at 10 x 10 feet and incorporated a symmetrical, flexible-wall supersonic nozzle. The design of the flexible wall was a great engineering challenge. Carefully polished 10-foot-wide stainless steel plates, 1-3/8 inches thick and 76 feet long, had to be bent to conform to a range of nozzle shapes covering the Mach range from 2.0 to 3.5. A series of hydraulic-operated jack screws positioned the flexible plates to an accuracy of 0.001 inch. Twenty-five years of successful operation have demonstrated the reliability of the system.
Because the earlier Lewis 8 x 6-foot propulsion tunnel had proven much more useful after it had been converted into an optional open- or closedcircuit facility, Lewis engineers decided to design the supersonic tunnel for dual operation from the start.
 The Lewis 10 x 10-foot supersonic wind tunnel built under the Unitary Plan.
The new tunnel operated as a closed system in its "aerodynamic mode" and as an open system in the "propulsion mode." In the former mode, it operates like a conventional supersonic tunnel. An 8-stage compressor driven by four electric motors (150000 horsepower total) feeds up to 4.6 million cubic feet of air per minute into a cooler, which is followed by a 10 stage compressor (100 000 horsepower). In this, its simplest and more conventional mode the Lewis tunnel is most impressive in terms of size and power. The propulsion mode is much more demanding. Since the tunnel runs as an open cycle, incoming air must be dried and heated while exhaust air must be muffled by a huge baffle to avoid deafening the local populace. The intake air dryer employs 1900 tons of activated alumina that dehumidifies 1 ton of air per second to a dewpoint of -40° during a 2-hour run. A 4-hour reactivation cycle must follow. In addition to the air dryer, an air-heater was added upstream to keep the air expanding in the nozzle from reaching temperatures far below anticipated flight conditions. Even with these efforts to control input air conditions the tunnel nozzle expands the air a bit too far at the higher Mach numbers, and it is impossible to simulate altitudes below 55 000 feet where the air is more dense.
The usefulness of the Lewis 10 x 10-foot supersonic propulsion tunnel has not been severely compromised by this limitation. The jet engines powering such famous aircraft as the Navy F-14 and USAF F-111 were tested at Lewis. In addition, the Space Shuttle liquidrocket engines were checked for heating effects on the space vehicle's structures. Looking to the future, the Lewis supersonic propulsion tunnel was designed with ample capability for testing the engines of the next generations of aircraft, air-breathing missiles, and manned spacecraft, whatever they may be.