The process of aerodynamic design of modern aircraft heavily relies on numerical simulations. Advances in high-performance computing hardware and algorithms have accelerated the design process and enabled better understanding of design trade-offs. The combination of computational fluid dynamics solvers using either inviscid or viscous compressible flow discretisation with numerical optimisation methods allows evaluation of many design solutions.
Despite considerable progress the ability to determine optimal solutions is still restricted due to the large computational cost of such simulations. The high cost results from numerical handling of underlying non-linear equations describing flow physics and a constant need to increase the number of design variables. Surrogate-based optimisation attempts to alleviate prohibitive simulations costs by utilising computationally cheaper surrogate prediction tools.
The authors of Simulation-Driven Aerodynamic Design Using Variable-Fidelity Models monograph demonstrate that the design process of aerodynamic shapes, such as aerofoils and wings, can be efficiently conducted using surrogate-based optimisation techniques, which could be used as an alternative to high-fidelity computational fluid dynamics models providing satisfactory aerodynamic design solutions at lower computation cost.
The first chapter of the book gives a general introduction to the aerodynamic shape optimisation. This is rather short description of the subject but gives references to important contributions in the subject as well as introduces a description of the representative aerodynamic shape optimisation problem with some explanation of major terms: an objective function (one-point and multi-point), constraints and the inverse design problem.
Chapter 2 gives a general overview of conventional optimisation techniques including adjoint sensitivity techniques. It also explains the idea of surrogate-based optimisation and its variants with approximation-based and physics-based surrogates. In this chapter, some typical problems are formulated together with advantages and disadvantages of various techniques. This is a good overview of major concepts and fundamental principles and an excellent starting point for further studies.
The next section of the book is focused on the aerodynamic design process starting with geometry parameterisation. The major steps involved in solving computational fluid dynamic problems are discussed. It also presents low-fidelity aerodynamics models that lead to the construction of physics-based surrogate models. The following major section presents examples of the shape optimisation applied to a transonic as well as subsonic aerofoil and wing.
The text is concise, clearly written and systematically organised. It contains the equations and formulae needed to solve practical problems. The text is amply illustrated. The supportive drawings facilitate understanding of the theory and are of good quality. The book also successfully presents selected applications of surrogate-based optimisation in other areas, such as the design optimisation of microwave filters. These were very carefully selected to demonstrate major challenges related to the performance of the optimisation algorithm. The practical implementation of some optimisation concepts is supported by MATLAB examples, which are included at the end of the book.
The book is well presented and should appeal to a wider audience interested in computational optimisation design. It explains the major concepts, gives enough depth and provides references for further studies.
