Modern aerodynamics formed by its fathers Kutta, Zhukovsky and Prandtl, resulted from a crisis of classical mechanics with the prediction of zero drag and lift in slightly viscous flow falsified by observation as expressed in d’Alembert’s paradox. To resolve the paradox two ad hoc assumptions were introduced:
- An airfoil with sharp trailing edge will generate around itself a circulation of air which gives lift.
- The thin boundary layer of slightly viscous flow gives rise to drag.
These two assumptions have become the basis of modern aerodynamics, which in the spirit of Popper have been viewed to be correct as long as they have not been falsified. The two assumptions are thus not directly justified, but instead motivated as causes of observed effects of lift and drag. The two assumptions are further formulated so as to make falsification difficult. For example, if 2. is falsified for a certain viscosity, 2. can still be upheld by arguing that the viscosity in the falsification test was not small enough for 2. to hold.
The New Theory of Flight shows that both 1. and 2. are incorrect, or rather scientifically irrelevant, by showing that both lift and drag originate from a specific modification of potential flow in the form of 3d rotational separation, and thus not from circulation and boundary layer.
We learn from this example that in science one should not be allowed to introduce assumptions with the only motivation that they lead to desired effects. It should not be allowed to view assumptions to be correct as long as they have not been falsified. Instead the assumptions themselves must be justified or at least be made likely by observation or argument.
This means that Popper’s falsifiability test is to weak to serve the development of correct useful science. The classical test by justification, rather than absence of falsification, should still be the norm.
The reason the classical requirement of justification was given up in modern physics, was the introduction of statistics which was beyond direct test and justification.
I have advanced the idea that statistical mechanics can be replaced by a form of finite precision mechanics, which can be directly tested.
See also the video: Popper on the Old Theory of Flight .