Comment on 'Afterbody configuration effects on model forebody and afterbody drag'

Abstract

I the above paper Thompson and Smith show an appreciable difference between pressure drag values obtained from AEDC and pressure measurements (see Fig. 1). As the basic data, i.e., the measurements of the pressure distributions, were obtained at ten years ago in a small wind tunnel and also with much smaller effort in instrumentation, the unprejudiced reader will certainly assume the data to be incorrect. However, what Thompson and Smith present as drag data is in fact a not sufficiently accurate integration of the original pressures. From their reported need to use for the integration also the stagnation pressure at the nose of the body and from the discrepancy between the two drag distribution curves it is clear that they have conducted a trapezoidal integration of Cp=f(R) instead of RCP=f(R), with R being the radius of the body varying with x. Particularly, if there are so few pressure orifices in the nose region, as was the case with our model, the first method used by Thompson and Smith will result in a too large forebody pressure drag. This is because the errors due to the trapezoidal approximation in the regions of overand underpressure, respectively, add to each other, while with the second integration method used by us they are partly compensated, as shown in Fig. 2. Note that these two methods of approximating one single set of tabulated data (CP and R) result in very different forebody pressure drag coefficients, the magnitude of which differs by more than a factor of 5 if only the actually measured pressures are used. The drag value according to the less accurate method No. 1 yields a forebody pressure drag coefficient of +0.0162, which agrees perfectly with the DFVLR forebody pressure drag given by Thompson and Smith (Fig. 1). The above two integrations were performed in November 1981, after we had seen the original version of the Thompson and Smith paper. The purpose was to understand their integration of the data. In our earlier data reduction in 1972 and 1974 we had further improved method No. 2 by obtaining additional CP values for the polynomial approximation through interpolation of the CP=f(x) values. Depending on the number of interpolated CP's, the above forebody pressure drag coefficient is raised from -0.003 to 0.001 or 0.004. Our 1972 data, which have been added in Fig. 1 to the results presented by Thompson and Smith, are found to agree very well with their data, both forebody and afterbody drag. Similar statements hold also for afterbody No. 5.