Semigroup Theory with Applications to Systems and Control (N. U. Ahmed)
J. Neuberger
DOI: https://doi.org/10.1137/1036039
IF: 10.2
1994-03-01
SIAM Review
Abstract:If nonlinear oscillations are the weak point of the book, matching is its strong point. Beginning with a simple linear example, the author computes the outer solution and the exact solution, and by comparing them, discovers the existence of a boundary layer. Nice motivations are given for the choice of an inner variable and for Prandtl’s matching condition for the first order approximation. Next, some examples having different behavior are given: a layer at the opposite end, a layer of thickness -g, an interior layer. Then higher-order matching is presented, using an intermediate variable; it is surprising that the author does not follow up his use of Prandtl’s matching condition by a development of Van Dyke’s rules, which are certainly easier to use than the intermediate variable method, although not as general in their applicability. A short section on nonlinear examples follows. Finally, there is a long (almost 40 pages) section dealing with the Navier-Stokes equation, Reynolds and Prandtl numbers, the Blasius solution, thermal boundary layers, and so forth, containing several computer programs and tables of output. Much of this material is unknown to me, but it looks quite readable and if wanted to learn it would certainly begin here. Unfortunately, this book commits the canonical mistake made, as far as am aware, by every book in this field that is intended for a strictly "applied" audience. For the sake of illustration, have made this mistake myself when said above that renormalizing takes a nonuniform expansion and renders it uniform. In fact, renormalization only takes a "disordered" expansion and renders it "uniformly ordered." All textbooks begin with the correct definition of uniform, which involves an estimate of the difference between the exact solution and an approximation; but they quickly lapse into the practice of "checking uniformity" by examining the sizes of successive terms of the expansion, making no reference to the (unknown) exact solution. This practice is partly responsible for the undeserved reputation that perturbation theory sometimes has among mathematicians, a reputation roughly comparable to that of tabloid newspapers. do not ask that a book at this level provide proofs of actual uniformity. But it is not unreasonable to ask two things: first, that it make clear that uniform ordering is only a necessary, not a sufficient condition for uniformity; and second, that it make students aware that for proving actual uniformity there exists a good deal of useful theory, as well as many interesting unsolved problems. These things could be done in the book underreview with the addition ofonly a few pages in Chapter 2 and the changing of a few words elsewhere. This much honesty would not, of course, change the reputation of the field overnight, but it might gradually elevate the level of awareness with which perturbation theory is used. would urge those adopting this textbook to make these points clear to their students.