Teaching Gallery

NOTE: This content is one of the topics in the UQ educational module series. UQ educational modules provide single-topic-focused lectures that can be plugged into an existing engineering course. Either SimCenter personnel can give lectures as guest lecturers or assist the instructors in developing tailored course materials. Other available topics include Bayesian updating, adaptive design of experiments in global surrogate modeling, and Bayesian model selection.

This course module provides an introduction to Global Sensitivity Analysis (GSA). Unlike local sensitivity analysis, where the variability of output is explored around a specific reference parameter value, GSA aims to evaluate the overall importance of each parameter in an engineering system by taking the full input space into account. Identification of the important parameters as well as the screening of trivial parameters is a fundamental task to any engineering analysis. Further, GSA can serve as a useful pre-analysis step to reduce the computational efforts in the main analysis, such as system optimization, system identification, model updating, and uncertainty analysis.

 

Target Audience: Graduate-level classes

  • Target class 1: Traditional engineering courses for which GSA can be useful. For example, classes that deal with model updating (GSA can assist in identifying important parameters to update) or classes that use computer simulations with multiple inputs of varying importance (such as performance-based earthquake engineering classes using nonlinear time history analysis).
  • Target class 2: Reliability, computational statistics courses in the engineering department. This audience will benefit from exploring uncertainty reduction using GSA.

NOTE: The contents and examples will be tailored to fit the level of each class. The materials for Target class 1 are designed to convey the concepts without UQ backgrounds. Therefore, the students do not require an understanding of random variables, probability density functions, or other UQ theories. Target class 2 will focus more on the UQ point of view both in theory (algorithms) and applications.

Learning Objectives:

  • Understand the concept of the global sensitivity index and perform efficient GSA
  • Interpret the GSA results with an understanding of their value and limitation
  • Discover the usage of GSA by investigating examples that use GSA to assist in solving various engineering problems, including optimization, parameter selection, and importance sampling

Tools Involved:

Example class materials for Target Class 1:

  • To be added

Example class materials for Target Class 2:

Instructions for Instructors:

  • Please contact us if you would like to bring this module in your courses or collaborate with us to develop new course materials.

Contact:

Sang-ri Yi, UC Berkeley, yisangri@berkeley.edu
Last Updated: November, 2022

This assignment provides a high-level introduction to FEMA P-58 to structural engineering students. The starter package includes the performance model and demand data for the analysis of a 12-story reinforced concrete shear wall building at three intensities of ground shaking.

Students can use the SimCenter PBE Application to calculate the expected damages in the building and the consequences of those damages. The assignment asks students to process the results andguides them through the kinds of questions they can answer with such a high-resolution analysis.

Target Audience:

Graduate students in Structural Engineering with basic knowledge of probabilities, earthquake engineering, and performance-based engineering.

 

 

Learning Objectives:

  • Describe the inputs required and the results of a FEMA P58 analysis.
  • Perform a FEMA P58 analysis in the SimCenter PBE Application using a pre-defined performancemodel and demand data.
  • Process the results of a FEMA P58 analysis and review the key performance indicators.

Tools Involved:

Introduction to FEMA P-58 Seismic Performance Assessment Course Materials:

Instructions for Instructors or TAs:

  • Introduction to SimCenter PBE
    Describes: i) how to download the files for the assignment; ii) how to load the inputs; iii) how to run the analysis iv) how to find the results in the working directory; and v) how to interpret the results. (Powerpoint)

Contact:

Adam Zsarnóczay, Stanford University, adamzs@stanford.edu
Last updated: May, 2022