onsdag 15 augusti 2012

Analysis of AIAA Referee Reports: Reviewer 2

Let us now somewhat more closely scrutinize the referee report AIAAreview2 of the New Theory of Flight submitted to AIAA Journal. The tone is set in the first sentence:
  • The classical theory of flight is one the most beautiful and subtle achievements of applied mathematics.
This is copied from Stoker's review of Birkhoff's critical book Hydromechanics, which elevates classical theory to a level beyond critique. Nevertheless the reviewer delivers a 6 page defense of classical theory:
  • All of the criticisms that comprise Section I of their paper can be answered, and I will try to do this below.
The defense contains many remarkable and incorrect statements with my comments in parenthesis:
  • There is a mathematical subtlety involved because the flow of a fluid at infinite Reynolds number (zero viscosity) is not always the same as the limit at very small viscosity (it is a singular perturbation problem), and so the  question is what light can be shed by the former on the latter? (This is so subtle that it is meaningless.)
  • There is no way to create vorticity within a viscous incompressible fluid. Vorticity can be created only at a solid boundary. (This is incorrect as shown on Kelvin's Theorem Unphysical.)
  • In the limit of vanishingly small viscosity, the boundary layer has no thickness, but is still present as an infinitesimal layer of infinite vorticity and hence making a finite contribution to circulation. (Meaningless statement.)
  • Vorticity can be created only at a solid boundary. It travels into the interior solely by diffusion....There is a tendency to suppose that the vorticity must be spread by viscosity, which does not seem plausible...(Contradiction.)
  • But what happens is that the circulation at infinity is set up by acoustic waves, and, if the flow really were incompressible, these travel infinitely fast. What acoustic waves cannot do is create vorticity. (Mind boggling.) 
  • The authors greatly underestimate the classical theory, most likely because the usual truncated exposition has not shown it to them in its proper light. If they take time to realize how its parts fit together, they will come to see that is a masterpiece of physical modeling. (The reviewer greatly overestimates classical theory; if it was such a masterpiece the reviewer's rescue operation would not be needed.)
  • The real flow (by which they mean their computed flow) always contains a boundary layer whose influence is not negligible at any Reynolds number. This is characteristic of singular perturbation problems, and is the reason why Prandtl’s insight was transformative to the theory of flight. (This is incorrect and is precisely the key element of our criticism of Prandtl which the reviewer does not address.)
  • As described earlier, the desirability of the sharp edge lies in forcing the boundary layer to negotiate an adverse pressure gradient before it could reach any other stagnation point. It is not necessary for the trailing edge to be absolutely sharp to achieve this aim. But the shaper the edge is, the more certain the effect, and the more likely to remain effective at high angles of attack. (This is incorrect. If it was correct that only a sharp trailing edge would a have a "certain effect", there would be no air transportation.) 
  • The “trick” (Kutta condition) if it deserves to be so called, lies in condensing this to a simple boundary condition, the effect of which is to force the zero-viscosity solution to obey the boundary condition for the small-viscosity solution. (This is incorrect as shown on The Kutta Trick is Illegal.)
  • It is well known that it is extremely hard, and probably impossible, to produce two-dimensional flow experimentally. It should be, and usually is, impossible to produce it in a three-dimensional computation. (Confusion about the non-physical 2d problem, which lacks all relevance.)
  • Calculations of this kind (Unicorn) are often referred to as Implicit Large Eddy Simulation, and are a recognized, but somewhat controversial, approach to modeling some aspects of turbulence. Is that what is being done? In any case, the mere fact of vorticity being observed means that the code did not simulate a potential flow. (Total confusion concerning the computational solution of the Navier-Stokes equations supporting the theory.) 
  • There follows a stability analysis of the linearized Euler equations. This is of doubtful validity (But is it valid?) because it assumes that a perturbation with non-zero curl can be introduced into an irrotational flow. Physically this cannot be done; I have already explained that there is no mechanism even within the Navier-Stokes equations for vorticity creation, merely the evolution of vorticity already present. Creation must take place at solid surfaces and involve viscosity, or must require externabody forces. There is nothing at all wrong with Kelvin’s Theorem. (Yes, it is, see Kelvin's Theorem Unphysical.)
  • Regrettably, it is my conclusion that publication of any of this material, in any form, would be highly retrogressive. (Not anything in any form?)
  • The authors have put their fingers accurately on many of the defects in the truncated 
  • versions of aerodynamic theory that are now current. (Compare previous statement)
  • However, they have not realized that all these difficult issues were struggled with years ago by the founding fathers of the subject, and resolved in completely satisfactory ways. (Incorrect, as shown by the reviewers attempt to rescue the fathers)
  • Sadly, the outcomes of those struggles have since been simplified or discarded in modern presentations to create a pragmatic treatment focusing on utility. (Yes, it is a sad state-of-the-art.)
  • This review is very much longer than I would normally write, because I believe that serious issues of substantial public interest are involved. (Yes the issues are important and require reviewers with deep insight into both mathematics, computation and fluid mechanics, and Reviewer 2 does not meet these requirements.)
Summary: The reviewer has not read and understood the article. The reviewer seeks to stop the article, because it questions the dogmas set by the fathers of aerodynamics 100 years ago. We share the criticism with many, but we are unique by offering a new correct understandable theory of flight backed by solid math, computation, physics and observation. It is not very clever by AIAA to dismiss our work on loose grounds. It will not disappear.

Inga kommentarer:

Skicka en kommentar