Customer Reviews

Computational Methods for Fluid Dynamics

5 star:		(3)
4 star:		(1)
3 star:		(2)
2 star:		(1)
1 star:		(1)

 

 

This is a very good book for learning the details of implementing the Finite Volume method in Computational Fluid Dynamics. I view it as very complimentary to S. V. Patankar's book, which is more general in nature. Both books are geared toward people who want to write computer programs to solve fluid transport problems (heat transfer, Navier-Stokes, etc.) Both tend to focus on numerical issues (as opposed to the physics of transport phenomena). But while Pantankar's book provides a very easy, approachable introduction that is thin on the details, Ferziger & Peric have written a book rich in details. For instance, they devote an entire chapter to solving linear systems of equations. They compare all of the commonly used methods and contrast their rates of convergence for various test cases. This level of detail is great if you already understand the big picture. I think it may be somewhat difficult for someone new to CFD to really understand the finite volume method from this book alone. I would recommend Patankar's book for a good intro and this book for the implementation details.

I agree with the reviewer that said this book is complimentary to an intro text such as Patankar. This is not a cook book for a first time coder. But it is a really great reference for the Finite Volume dilettante.

I really appreciate that all numerical results presented are thoroughly documented. That counts.

Really, really nice chapter on iterative solvers....

Good overall description on many other topics such as multi-grid methods, turbulence, grid geometry and PV coupling.

This book really helped me speed up my homegrown quite a bit.

They also have all kinds of stuff available by ftp....

This is the best book on CFD I' ve read. It is more useful to those who want to develop their own codes rather than CFD users: It mostly describes in deep detail (although in a concise manner) a single method used by the authors and developed by themselves and their colleagues. It is not intended to be a list of the vast number of CFD techniques developed so far.

Their method is state-of-the-art and they provide plenty of results to support it's quality. It is mostly directed towards incompressible flows. They provide a chapter that extends their method to compressible flows but they do not describe any special convection schemes for flows with shock waves. It can be applied to both structured and arbitrarily unstructured grids, although their approach to the discretization of the convection and diffusion terms is particularly useful in the case of arbitrarily unstrucured grids. State of the art subjects such as multigrid and error-driven grid refinement are also covered and integrated into their method.

I agree with a previous reviewer that they provide a very good coverage of solution methods for linear equation systems which arise in CFD. Most other books on CFD (all the ones that I have read) have a poor coverage of the subject and describe only old and inneficient methods. However even this book does not sufficiently describe conjugate-gradient type methods or Krylov subspace methods in general, but references are provided.

In conclusion, this book is not for beginners but for code developers who have some basic knowlwdge of CFD and a relatively good mathematical background.

I recommend this book for the casual scientist who is interested in CFD. It covers the major issues, including finite volume discretizations, unsteady flows, inviscid and viscous flows, methods for solving systems of linear and nonlinear equations, moving grids, etc. But I do not recommend this book if you intend to use it as a reference for implementing a CFD code. It does not provide background on the mathematical nature of the governing equations. Also, it does not provide sufficient description of boundary conditions (subsonic/supersonic inflow/outflow, etc.) as they relate to the propagation of information. It also does not cover in sufficient detail the concepts of upwinding. Without this missing information, an implementation could *easily* produce invalid results. As a casual reference, though, it often presents gold nuggets of information that briefly answer the question you were asking. Look to Hirsch for a more comprehensive implementation reference.

I used this text as a part of an intro to CFD class I am taking. I wasn't very impressed by this book. There are a few good parts, but for the most part, the book is very much wanting. You have to trudge through some confusing text in places because of the author's desire to be general. Several figures add nothing to the discussion. For all the algorithms discussed, there is not a single flow chart or chunk of pseduocode - leaving very long-winded text-heavy explanations of what should be simple concepts.

The loss of Professor Ferziger was definitely to the detriment of the CFD field, but a lot of this book is the authors referencing their own work (this bugs the crap out of me). No offense, but their contributions thus far are not as important as they would have you believe by reading this book; they are not the only people to have papers in subjects that run the gamut of CFD topics. Speaking of referencing papers, the authors treatment of basic topics (numerical analysis, FDM/FVM, FEM, time integration methods) is very much lacking, while more "high brow" topics like turbulence and multiphase flows have a reference every other sentence. I found the treatment of linear equation solvers and general NS-equations solutions to be very good. The turbulence chapter was also very clear. It was good to see efficiency and error discussed in the context of actually solving problems, but alas, their treatment did not really do this very important area much justice.

I don't know what book is good for learning CFD. If you were completely new to the field, I suppose this would be a reasonable place to start, but once again, I don't think the text is THAT great.

Book chapters include: 1. Basic Concepts of Fluid Flow 2. Introduction to Numerical Methods 3. Finite Difference Methods 4. Finite Volume Methods 5. Solution of Linear Equation Systems 6. Methods for Unsteady Problems 7. Solution of the Navier-Stokes Equations 8. Complex Geometries 9. Turbulent Flows 10. Compressible Flow 11. Efficiency and Accuracy Improvement 12. Special Topics

Ferziger's book does a poor job of covering concrete details. It's revealing to see that many sections (and chapters) are simply long blocks of text, with no equations. The text gives a qualitative and fuzzy introduction to computational fluid dynamics, and leaves the reader wanting much more detail.

With respect to applications: I used the book in a graduate-level computational fluid dynamics course, and when I was actually writing CFD code, I found the book to be of no help at all. I used Tannehill, Anderson, and Pletcher's "Computational Fluid Mechanics and Heat Transfer" to be very helpful with finite difference, and Ptankar's "Numerical Heat Transfer and Fluid Flow" and Versteeg's "Introduction to Computational Fluid Dynamics" to be far superior with respect to the finite volume approach (although finite volume is the main focus of the book). With respect to the "special topics", nothing is covered with any substance, everything is glossed over, and nearly half of the book is, in my opinion, a waste. Chung's "Computational Fluid Dynamics" is a much better all-in-one reference book that covers special topics FAR better - with enough detail that one could actually attempt to implement them in a code.

With respect to the fundamental approach: I think Anderson's "Computational Fluid Dynamics" book gives a much better, more physically intuitive description of the governing equations and of finite volume discretization. Ferziger does a weak job of covering nearly everything in his book, and the finite volume introduction is no exception.

Overall, this is a watered-down introduction to CFD. It does not to justice to any of the subjects it covers, particularly the special topics. It is a terrible textbook for a CFD class, and although it has been sitting on my bookshelf for 3 years, I always pass it over in favor of other references (see above for list).

I found this book an excellent support in teaching and invaluable as an occasional reference in my practical work in industry. It is particularly gratifying to see that free surface flows are covered well. Best book I have seen so far in the field.

Peric's PhD thesis is one of my most precious possessions. I consider Ferziger to be one of THE greats in Large Eddy Simulation. So I had great hopes for this book. I was warned before I saw it that it contained nothing that I wouldn't already know, but even so it was a big disappointment. I could have written a better book myself and I don't have half the brain of either of the authors. I've looked at it three times and each time turned away in disgust.

This book will point a novice in the correct direction, unlike Roach's book, and unlike Patankar's in a post Rhie-Chow age, but it avoids like the plague any subject with even the slightest whiff of controversy. As any new technique is always controversial, the book was some 5 to 10 years out of date even before it hit the press. Further, it discusses each subject in a very superficial way. More depth is needed.

I want a book that will tell me what the advantages are and how to program the conjugate gradient method, algebraic multigrid, superbee, wall reflection terms for Reynolds Stress, multiphase flow, viscoelasticity, coalescence, the axis in cylindrical coordinates, grid generation, surface tension, adaptive grids, spectral elements etc. This book is no help with any of these.

I frequently get asked "What's a good book for Computational Fluid Dynamics?". I have to admit that there isn't one. The best that I know are the user manuals for the various commercial CFD packages.

To summarise, if programming CFD is a 100 step journey then this book will only take you the first 3 steps, but at least they're 3 steps in the right direction. After you've been programming CFD for a year you can probably toss the book out without any great loss.