CFD is three subjects wearing one name. There is the fluid physics, there is the numerical analysis of partial differential equations, and there is the software engineering of getting a solver to converge on a real mesh. Books that assume you have the first two are useless if you do not, and books that teach the first two do not tell you why your simulation diverged at iteration forty.
The single most common failure is starting with a CFD textbook without the fluid mechanics underneath it. A discretisation scheme is meaningless until you know what a boundary layer is and why the convective term is the difficult one. Read the physics first even if you are impatient to run something.
Expect real prerequisites: vector calculus, ordinary and partial differential equations, linear algebra, and enough programming to write a solver for a one-dimensional problem yourself. No book on this list substitutes for actually running cases — the judgement that matters in CFD is developed at a terminal, not in a chair.
The physics first
Fluid mechanics by Frank M. White is the standard undergraduate text and the right foundation: viscosity, dimensional analysis, boundary layers, pipe flow, all with worked problems. If you have not seen the Navier-Stokes equations derived, start here and do not skip the exercises.
Fundamentals of aerodynamics is John Anderson covering compressible flow, shocks and the aerodynamic applications that motivate most of the numerics later on. Anderson writes better than almost anyone in engineering, and the historical asides are not padding — they explain why particular methods exist.
The first CFD book
Computational fluid dynamics edited by John F. Wendt collects the von Karman Institute introductory lecture series, and it is the gentlest bridge from the physics into the discretised world.
Then choose one of the two standard introductions rather than working through both. An introduction to computational fluid dynamics by Versteeg and Malalasekera builds the finite volume method from scratch with unusual patience and is the friendlier of the two. Computational methods for fluid dynamics by Ferziger and Peric covers the same territory more rigorously, with better treatment of grids, pressure-velocity coupling and error estimation. Read one now and the other later; reading them back to back is mostly repetition.
Numerical heat transfer and fluid flow is Suhas V. Patankar's 1980 book, and it is on this list for a specific reason: it is where the SIMPLE algorithm is derived, and that derivation is still the clearest available even though the book predates unstructured meshes entirely. Read it for that chapter.
The Finite Volume Method in Computational Fluid Dynamics by Moukalled and colleagues is the modern, exhaustive treatment, written alongside working code and directly useful if you are reading an open-source solver's source.
The numerics tier
Finite Volume Methods for Hyperbolic Problems is Randall J. LeVeque, and it is where the subject becomes mathematics. Conservation laws, characteristics, shock capturing, why naive schemes produce oscillations at discontinuities. It is demanding and it is the book that separates people who use CFD from people who understand it.
Riemann Solvers and Numerical Methods for Fluid Dynamics by Eleuterio F. Toro is the reference for compressible flow — approximate Riemann solvers, Godunov methods, high-resolution schemes. Keep it on the desk rather than reading it front to back.
Turbulence, last and least settled
Turbulence modeling for CFD is David C. Wilcox's practitioner reference on the RANS models you will actually select from a menu in commercial software, with the derivations and the calibration constants that menu hides.
Turbulent Flows by Stephen B. Pope is the graduate theory text and the more permanent book. Read it knowing that turbulence modelling is the least settled part of the field: the standard models are calibrated approximations that work well on flows resembling the ones they were tuned against and fail quietly on flows that do not. A solver that runs is not a solver that is right.
More engineering paths sit under fluid dynamics.
Follow the full ordered path here: How to Learn Computational Fluid Dynamics from Books, in Order.
