Weissinger and Maass 



REPLY TO THE DISCUSSION 



J. Weissinger and D. Maab 



Similar difficulties (e.g., crossing of streamlines) in using the boundary- 

 condition method have also been observed by Dickmann and Weissinger during 

 their computations. Therefore, in (102), the figures showing the streamlines in- 

 side the duct were calculated by means of the stream function which expresses 

 the law of continuity. K an iteration procedure is applied, both methods should 

 give the same results. An iterated-boundary-condition method has been used by 

 Wiedemer (113), and Fig. 16 above shows that the duct profile can be changed 

 considerably by iteration. 



It does seem to be a question if the changes are also as great in the direct 

 problem, i.e., if the flow field has to be determined for a given duct shape. From 

 Bollheimer's results (116), there is some indication that the location of the ring 

 vortices does not matter so much, at least for the over-all characteristics. 



As long as we do not know bounds for the errors caused by the approxima- 

 tions (and they are unknown even in two-dimensional airfoil theory), we must 

 rely on intuition and numerical experience in judging the accuracy of a method. 

 The experience brought forward in the discussion is highly appreciated by the 

 authors. 



1264 



