Verification and Validation Best Practices for Integrated Computational Materials Engineering

In This Section

Synopsis

Due to the exponential increase in the availability of computing power over the last several decades, modeling and simulation have become an important part of virtually all of aspects of science, engineering, and manufacturing. For modeling and simulation to achieve the stated goal of Integrated Computational Materials Engineering (ICME) of delivering tools that aid the science and engineering decision-making process, thorough and unbiased assessments of the accuracy and credibility of model results are critical. This course gives an overview of the topics of verification and validation and uncertainty analysis, demonstrated by contextualized examples, with a focus on quantifying the confidence that can be attributed to results of computational materials science models. While these topics have been investigated for decades in the computational fluid dynamics and structural analysis communities, the ICME community is just starting to grapple with implementing these approaches, and will face some unique challenges due to the range of physical phenomena that are of interest to the materials science community.

Who Should Attend

Those who will benefit from this course include a broad cross section of ICME stakeholders, such as materials researchers, educators, design and manufacturing engineers, and program managers who seek to understand how to assess the accuracy of computational materials science and engineering simulations.

Why Attend

  • Learn the concepts of uncertainty quantification for ICME model verification and validation (V&V)
  • Understand the recommended process for assessing V&V needs
  • Learn the commercial and open-source tools and resources available to aid with V&V implementation 
  • Master the concepts of state-of-the-art probabilistic and uncertainty quantification methods
  • Understand what the results from a probabilistic/reliability analysis mean
  • Verification and Validation Best Practices for Integrated Computational Materials Engineering
Outline

Course Outline

  • V&V Framework
    • Definitions
      • ICME
      • Credibility
      • Verification
      • Calibration
      • Validation
       
    • Verification
      • Code Verification
      • Solution Verification
      • Error vs. Uncertainty
       
    • Validation
      • Tactical and Strategic Validation Goals
      • Validation Experiments
      • Model Accuracy Assessment and Validation Metrics
       
    • Uncertainty Quantification
      • Sources of Uncertainty
      • Sensitivity Analysis
      • Uncertainty Propagation
       
     
  • Process and Tools
    • V&V Process Planning
      • Guidelines for Application of ICME V&V to ICME Projects 
      • V&V System Level and Model Checklists 
      • Tool Maturity Level (TML) Assessment p\Process 
      • Recommended Risk vs. Consequences Assessment Process
       
    • V&V Computational Tools
      • Computational Tools Overview
      • Demonstration of V&V/UQ Analysis
       
     

 

Materials

Course Materials

Two weeks prior to the course, the instructors’ course notes will be available for registered students to download. Since course notes will not be distributed on site during this course, it is highly recommend that you download the course notes to your computer prior to the course and bring your computer with you.

 

Instructors

Course Instructors

Mark Benedict
Mark Benedict is currently a program manager and materials engineer working for the Air Force Research Laboratory (AFRL) Manufacturing Technology Division supporting programs that focus on Additive Manufacturing and Integrated Computation Materials Science and Engineering. Recently he has been working across the AFRL Materials and Manufacturing Directorate to modernize the IT infrastructure and encourage adoption of verification and validation best practices.


David Riha
David Riha’s technical experience includes predicting the probabilistic response and reliability of engineered systems using advanced probabilistic and uncertainty methodologies. Since 1991, he has coordinated the development of the NESSUS® probabilistic analysis computer program to support engineering analysis, design, and model verification and validation.