VERIFICATION AND VALIDATION IN SCIENTIFIC COMPUTING

Transcription

1 VERIFICATION AND VALIDATION IN SCIENTIFIC COMPUTING WILLIAM L. OBERKAMPF CHRISTOPHER J. ROY CAMBRIDGE UNIVERSITY PRESS

2 Contents Preface page xi Acknowledgments xiii 1 Introduction Historical and modern role of modeling and simulation Credibility of scientific computing Outline and use of the book References 17 Part I Fundamental concepts 19 2 Fundamental concepts and terminology Development of concepts and terminology Primary terms and concepts Types and sources of uncertainties Error in a quantity Integration of verification, validation, and prediction References 75 3 Modeling and computational simulation Fundamentals of system specifications Fundamentals of models and simulations Risk and failure Phases of computational simulation Example problem: missile flight dynamics References 137 Part II Code verification Software engineering Software development Version control Software verification and validation Software quality and reliability Case study in reliability: the T experiments 161 Vll

3 Vlll Contents 4.6 Software engineering for large software projects References Code verification Code verification criteria Definitions Order of accuracy Systematic mesh refinement Order verification procedures Responsibility for code verification References Exact solutions Introduction to differential equations Traditional exact solutions Method of manufactured solutions (MMS) Physically realistic manufactured solutions Approximate solution methods References 244 Part III Solution verification Solution verification Elements of solution verification Round-off error Statistical sampling error Iterative error Numerical error versus numerical uncertainty References Discretization error Elements of the discretization process Approaches for estimating discretization error Richardson extrapolation Reliability of discretization error estimators Discretization error and uncertainty Roache's grid convergence index (GCI) Mesh refinement issues Open research issues References Solution adaptation Factors affecting the discretization error Adaptation criteria Adaptation approaches Comparison of methods for driving mesh adaptation References 366

4 Contents ix Part IV Model validation and prediction Model validation fundamentals Philosophy of validation experiments Validation experiment hierarchy Example problem: hypersonic cruise missile Conceptual, technical, and practical difficulties of validation References Design and execution of validation experiments Guidelines for validation experiments Validation experiment example: Joint Computational/ Experimental Aerodynamics Program (JCEAP) Example of estimation of experimental measurement uncertainties in JCEAP Example of further computational-experimental synergism in JCEAP References Model accuracy assessment Elements of model accuracy assessment Approaches to parameter estimation and validation metrics Recommended features for validation metrics Introduction to the approach for comparing means Comparison of means using interpolation of experimental data Comparison of means requiring linear regression of the experimental data Comparison of means requiring nonlinear regression of the experimental data Validation metric for comparing p-boxes References Predictive capability Step 1: identify all relevant sources of uncertainty Step 2: characterize each source of uncertainty Step 3: estimate numerical solution error Step 4: estimate output uncertainty Step 5: conduct model updating Step 6: conduct sensitivity analysis Example problem: thermal heating of a safety component Bayesian approach as opposed to РВА References 665

5 x Contents PartV Planning, management, and implementation issues Planning and prioritization in modeling and simulation Methodology for planning and prioritization Phenomena identification and ranking table (PIRT) Gap analysis process Planning and prioritization with commercial codes Example problem: aircraft fire spread during crash landing References Maturity assessment of modeling and simulation Survey of maturity assessment procedures Predictive capability maturity model Additional uses of the PCMM References Development and responsibilities for verification, validation and uncertainty quantification Needed technical developments Staff responsibilities Management actions and responsibilities Development of databases Development of standards References 755 Appendix: Programming practices 757 Index 762 The color plates will be found between pages 370 and 371.