måndag 15 februari 2010

d'Alembertgate

Another example of the (mal)practice of peer-review in science revealed in Climategate, is my own experience with the reception of my resolution of d'Alembert's paradox with Johan Hoffman, first submitted to the leading journal in the field Journal of Fluid Mechanics and rejected with the following referee's reports:

Referee 1:
  • I strongly recommend that you reject this paper.
  • 1. The authors show no understanding of fluid mechanics, of how small viscosity leads to experimentally verified thin boundary layers, and of how experimentally verified these boundary layers detach when the external flow decelerates.
  • 2. The authors show no understanding of mathematics, of how in the Euler equations streamlines leaving a surface will be vortex sheets which divide the domain into regions in which there is separately potential flow.
  • 3. The authors show no understanding of numerics, of how finite element methods introduce an artificial viscosity (or hyper-viscosity in high-order methods) through the truncation error, and so their solutions are not solutions of the Euler equation but solutions of a Navier-Stokes equation.
  • I am alarmed that these authors have another paper accepted by JFM: I recommend that someone checks it for logic.
Referee 2:
  • THE AUTHORS PROPOSE RESOLVE D'ALEMBERT'S PARADOX BY SHOWING THAT THE ZERO DRAG POTENTIAL SOLUTION OF EULER'S EQUATIONS IS UNSTABLE AND INSTEAD A TURBULENT (APPROXIMATE) SOLUTION DEVELOPS WITH A NON ZERO DRAG, EVEN WITHOUT BOUNDARY LAYER EFFECTS.
  • THEY BASE THEIR RESULT ON NUMERICAL CALCULATION MOSTLY DONE IN 2006 WITH THEIR OWN NUMERICAL PACKAGE.
  • THE EQUATIONS THEY STUDY (2.1) HAVE NO PARAMETERS OTHER THAN THE VELOCITY SO THAT THEIR CONCLUSIONS MUST APPLY TO ALL REYNOLDS NUMBERS.
  • SINCE TURBULENCE DOES NOT DEVELOP AT LOW RE THEIR RESULT GOES TOO FAR. I DO NOT KNOW HOW THEY GET VORTICITY WHEN THE INTIAL FLOW IS IRROTATIONAL, BAROTROPIC AND WITHOUT SHEAR LAYERS.
  • I DO NOT NOT UNDERSTAND THEIR USE OF SEPARATION WHEN THERE ARE NO BOUNDARY LAYERS TO SEPARATE. I HAVE ATTACHED A CITATION FROM LIGHTHILL IN WHICH HE ADVOCATES REPLACING d'ALEMBERTS PARADOX WITH d'ALEMBERTS THEOREM.
  • THIS PAPER SHOULD BE REJECTED.
  • I DID NOT READ THEIR NUMERICAL PAPERS WHICH ARE VERY NEW AND APPARENTLY NOT VALIDATED AGAINST EXACT SOLUTIONS AND OTHER TURBULENT CODES.
It is clear that these referee's are out kill, and they do it! The paper was then submitted to Journal of Mathematical Fluid Mechanics and was accepted and published in Dec 2008.

The story is told in my previous post The Sciencegate of Fluid Mechanics: It shows how a small group of scientists can control a whole scientific discipline by suppressing new information and new discoveries showing classical "truths" to be empty/false. See my interview with the Editors of Journal of Fluid Mechanics JFM.

It shows how this small group of scientists also controls Wikipedia, and blocks any reference

Climategate has shown that some of climate science has been a dirty business, and I have experienced similar practice in fluid mechanics. The Climategate drama is now unfolding and its consequences must be far-reaching...carbon trading is losing momentum...Maybe the emails behind dAlembertgate will also be made public...maybe according to FOI...

If you look into the details of dAlembertgate you will discover that the new resolution of d'Alembert's paradox fundamentally changes the mathematical basis of fluid dynamics and that the attitude of JFM is untenable...

For more insight into the essentials of peer-review see an example from 1945.

1 kommentar: