《PROGRAMMING THE FINITE ELEMENT METHOD 》（Fifth Edition）PDF高清原版下载

prologue Preface to Fifth Edition
    This edition maintains the successful theme of previous editions, namely a modular pro-gramming style which leads to concise, easy to read computer programs for the solution of a wide range of problems in engineering and science governed by partial differential equations.
The programming style has remained essentially the same despite huge advances in computer hardware. Readers will include beginners, making acquaintance with the finite element method for the first time, and specialists solving very large problems using the latest generation of parallel supercomputers.
     In this edition special attention is paid to interfacing with other open access software, for example ParaView for results visualisation, ABAQUS user subroutines for a range of material constitutive models, ARPACK for large eigenvalue analyses, and METIS for mesh partitioning.
Chapter 1 has been extensively rewritten to take account of rapid developments in com-puter hardware, for example the availability of GPUs and cloud computing environments. In Chapters 2 to 11 numerous additions have been made to enhance analytical options, for example new return algorithms for elastoplastic analyses, more general boundary condition specification and a complex response option for dynamic analyses.
      Chapter 12 has been updated to illustrate the rapidly advancing possibilities for finite element analyses in parallel computing environments. In the fourth edition the maxi-mum number of parallel ‘processes’ used was 64 whereas in this edition the number has increased to 64,000. The use of GPUs to accelerate computations is illustrated.

百度翻译文供参考：
第五版序言
       本版保持了先前版本的成功主题，即模块化编程风格，该风格导致简洁、易于阅读的计算机程序，用于解决偏微分方程控制的工程和科学中的广泛问题。尽管计算机硬件有了巨大的进步，但编程风格基本保持不变。读者将包括初学者，首次熟悉有限元法，以及使用最新一代并行超级计算机解决大型问题的专家。

        在本版中，特别注意与其他开放访问软件的接口，例如用于结果可视化的ParaView、用于一系列材料本构模型的ABAQUS用户子程序、用于大特征值分析的ARPACK以及用于网格划分的METIS。考虑到计算机硬件的快速发展，例如GPU和云计算环境的可用性，第1章被广泛重写。在第2章至第11章中，增加了许多分析选项，例如弹塑性分析的新返回算法、更一般的边界条件规范和动态分析的复杂响应选项。

          第12章已更新，以说明并行计算环境中有限元分析的快速发展可能性。在第四版中，使用的并行“进程”的最大数量为64个，而在本版中，数量增加到64000个。说明了使用GPU加速计算。