High-powered rifle bullets travel faster than sound. Given a large enough propulsive charge, bullets can penetrate the hypersonic range. Why not fashion bullets resembling hypersonic models and fire them out of guns and carefully watch them with instruments as they streak by? Although the idea sounds a bit radical, it is quite sound-provided that a properly instrumented firing range is available.
NACA owned several ideal firing ranges: the supersonic wind tunnels at Ames, Langley, and...
....Lewis. Not only did these wind tunnels have much of the appropriate instrumentation, they could provide a supersonic airstream for the bullets to fly into, thus extending their ranges into the hypersonic realm. In other words, the gun-launched models could be fired upstream to attain very high relative velocities. To illustrate: a gun firing a model at 4000 feet per second upstream in a tunnel operating at Mach 2 produces a relative Mach number of 7; the combination of 8000 feet per second and Mach 3 results in a relative Mach number of 15. Such high Mach numbers in a conventional wind tunnel would normally call for fantastically large expansion-ratio nozzles with the attendant danger of air liquefaction. However, with the supersonic wind tunnel actually operating at only Mach 2 or Mach 3, air liquefaction was no concern at all. The relative Mach number was what counted experimentally. Furthermore, the test Reynolds numbers would be realistically high, as would the air temperatures at critical points on the model. There was little doubt that a "counterflow" facility would solve some of the design problems of hypersonic wind tunnels, but what new problems would arise?
The concept of a counterflow tunnel was proposed by H. Julian Allen at Ames in 1946, but his so- called supersonic free-flight tunnel did not become a reality until late 1949. First, the problem of firing winged models from a gun with a cylindrical barrel had to be solved. Special sabots or projectile carriers were developed that enclosed the noncylindrical models while they were inside the gun. The sabots peeled off as the  model left the muzzle heading upstream into the wind tunnel test section. Inside the test section, the flight of the model was "stopped" by a light flash about 1/3 microsecond long. Photos of the bulletlike models streaking up the tunnel provided more insight into hypersonic airflow than the experimenters had dared hope. The high-temperature turbulent gases in the shock waves and boundary layers cast clear-cut shadows on the screens. (These shadowgraphs are not schlieren photos. They have the same origin as thermal shadows cast by hot air rising from a home radiator on a bright winter day or from a hot roadway in the summer.) In the Ames supersonic free-flight tunnel the shadowgraphs revealed all the intricacies of hypersonic airflow around various projecting portions of the model-a sort of aerodynamic X-ray of tunnel airflow.