Numerical methods in bifurcation problems. Lectures delivered at the Indian Institute of Science, Bangalore, under the T.I.F.R.-I.I.Sc. Programme in Applications of Mathematics. Notes by A. K. Nandakumaran and Mythily Ramaswamy (Q1110275)

Based on lecture notes prepared while the author visited the Tata Institute of Fundamental Research, this book provides an introductory survey of the issues associated with the numerical treatment of bifurcation problems. In Chapter 1, the author carefully and thoroughly illustrates the concepts of bifurcations, folds, perturbed bifurcation, stability, exchange of stability, pitchfork and Hopf bifurcations on three simple but meaningful examples. This chapter can probably be understood by a reader with a minimum of background. In Chapter 2, the author reviews continuation methods for numerically tracking the solution sets of parametrized nonlinear systems \(G(u,\lambda)=0\). To this end, he reviews contraction mapping theorems and implicit function theorems in Banach spaces. He then uses these and Lipschitz constants of G and \(G_ u\) to obtain step bounds for the continuation method. Additional background review includes Newton's method and the Kantorovitch theorem. The remainder of Chapter 2 is devoted to predictor-corrector methods for continuation. Catalogues of possible predictors and possible correctors, as well as a quasi-Newton method appear. Various theoretical notions appear in Chapter 3. These include Sard's theorem and various regularity theorems. The author then introduces the topological degree, and applies it to the Brouwer fixed point theorem, existence of periodic solutions to systems of ordinary differential equations, and existence of bifurcation points. The chapter concludes with a treatment of global Newton methods. Chapter 4 begins with a discussion of regularity of paths and limit points, and continues with pseudo-arclength continuation. Details are then given for practical techniques for solving the resulting linear systems; this includes the bordering algorithm and an asymptotic analysis when singular systems such as at limit points are solved approximately. The chapter concludes with a discussion of how to compute tangent vectors to arcs. Chapter 5 is the heart of the book. It begins with a derivation of the Keller/Langford equations (based on a quadratic model of the paths) for characterizing tangents to paths at a bifurcation point in a finite- dimensional setting. Properties of a simple bifurcating point, including exchange of stability, are considered in detail. Five methods, including the Keller/Langford equations, the Lyapunov-Schmidt method, use of the Crandall/Rabinowitz theory, and perturbation of the system are discussed for switching branches at a bifurcation point. We then see how to track paths of bifurcation points in a two-parameter (perturbed bifurcation) system with the bordering algorithm. The chapter ends with a treatment of methods for finding periodic solutions of Hopf bifurcation points. Chapter 6 contains numerical examples, which are second-order two-point boundary value problems with a semilinear forcing term and central difference discretization. Care is taken in presentation of details, and a flow diagram appears. Various tables and plots appear. The theorems are accompanied by proofs which illustrate underlying techniques and additional facts. The author states in the prefix that he is preparing an extended version to appear elsewhere. This book is recommended for scientists and mathematicians wishing to begin a study of numerical analysis of parametrized problems and bifurcation. It also is an excellent reference for the work of the author and associates.