Chapter 3: Transonic Wind Tunnel Development (1940 -1950)
[85] Early agency literature refers to these techniques as "Bomb-Drop" and "Missile-Test," revealing their wartime origins. It is not certain whether NACA supervisors Crowley and Thompson, who are said to have "considered" the body-drop approach in 1942 (ref. 54) had any [86] knowledge of the prior German use of the technique in 1941 (ref. 105). It is apparent, in any case, that the American development was a great improvement over the German (ref. 106). The technique supplied primarily zero-lift drag data at speeds up to about Mach 1.3. A question regarding possible errors due to acceleration effects was raised early in the program by von Karman, but it was proved later, by drops of identical models of varying weight, that the effect was negligible. The reliable drag data from the body drops were used, for example, to estimate the drag and power requirements for the transonic research airplanes (ref. 107). As the rocket models came into use in the latter forties the body drops diminished but continued to be used occasionally for special purposes. They have provided important comparative data for evaluation of slotted tunnels (ref. 108), but otherwise there was little interaction of this technique with the transonic wind tunnel developments.
The rocket-model approach started as a missile test and development program, but it rather quickly started to change character. Reflecting both the interests of its NACA operators and the growing demand for transonic aerodynamic data, it evolved into a program of general aerodynamic tests covering the entire transonic region and beyond into the supersonic regime. The flight data became increasingly more accurate and more comprehensive as time went on as a result of the impressive ingenuity brought to bear on the many challenging aspects of this technique. Inevitably, the practitioners of the technique tended to become as much interested in making the rocket models do more things more accurately as they were in the research problems. To a large degree, therefore, one finds that the Pilotless Aircraft Research Division (PARD) reports tended to be data reports for specific test objects rather than general or analytical treatments of research problems.
One aspect of the technique caused major interference with the wind tunnel programs: each firing required the sacrifice of the test model, including in many cases complex and costly internal instrumentation. For example, in the years 1947, 1948, and 1949 no less than 386 models were expended (ref. 109). This is roughly equivalent to the requirements of perhaps 10 major wind tunnels such as the 16-foot. Furthermore, the wind tunnel models generally carry only pressure taps, or [87] house a balance which can be used repeatedly in many models. There was a major slowdown in both wind-tunnel model and instrumentation production as a consequence of PARD's voracious appetite.
By mid-1946 transonic data reports had been published from all of the transonic methods then in use and a serious problem had arisen. Large discrepancies were apparent, and, understandably, queries had been received from industry users. The matter was discussed at the July 12, 1946, meeting of the Langley General Aerodynamics Committee, and a special ad hoc group was set up to study the problem. The group made its first report on September 17, 1946. W. H. Phillips showed that in one category, significant differences in the transonic drag of straight wings were believed explainable on the basis of large test Reynolds number differences. In the case of complete wing body configurations, very large and unexplainable differences existed; the gross trends, however, were similar. The group recommended that a specific wing and the X-2 aircraft configuration be tested by all of the techniques, including the 9-inch supersonic tunnel, for comparative study. Nine months later, on June 13, 1947, the group reported "no new conclusions." By that time the Langley bump was in full operation and the supply of discrepant data was growing rapidly. Cases were found where not only the data values disagreed, but also the trends were at variance. The low opinion of the bump data shared by a majority of Langley aerodynamicists found expression in a memorandum submitted to Langley's Chief of Research F. L. Thompson by E. C. Draley of the 8-foot tunnel, and discussed at the June 13, 1947, meeting of the Langley General Aerodynamics Committee. Draley was particularly concerned about the validity of stability and control data from the bump.
Thompson considered this problem of sufficient importance to take the unprecedented step of personally presenting an introductory paper on the subject at the NACA Conference on the Aerodynamic Problems of Transonic Airplane Design on September 27, 1949. He gave a brief objective assessment of all the different transonic techniques and focused special attention on the problems of the wing-flow and bump. He said in effect that the bump in most cases provided useful trends or comparisons, but bump data should not be used quantitatively.