Back to search

Clemson University

Creation and Validation of a Novel Bayesian Calibration Method with Griddy Gibbs Sampling

Abstract

dc:description.abstract

Computer models of physical systems are widely used in lieu of, or in tandem with, experimental testing. It is critical to verify the accuracy of computer models through the process of calibration. Typical calibration methods are often computationally expensive and therefore cannot be performed in real time. This thesis presents a novel Bayesian calibration method using a Griddy Gibbs sampling algorithm to improve calibration speeds. This method was verified in two applications: a location-dependent dataset in the heat transfer analysis of an engine piston, and time-dependent tire forces in a drum test. The proposed method was directly compared to a traditional Bayesian calibration method in the engine piston application. It was found that the two methods were close in accuracy with large amounts of calibration data, and the Griddy Gibbs method was significantly less computationally expensive; it could calibrate in less than a minute, while the traditional method took several days.

Degree

thesis:*
Name thesis:degree_name
Master of Science (MS)
Level thesis:degree_level
Thesis
Discipline thesis:degree_discipline
Civil Engineering

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Stewart, Hannah
Contributors dc:contributor
  • Laura Redmond
  • Christopher McMahan
  • Weichiang Pang

Subjects

dc:subject × 4

Identifiers

dc:identifier.*
Repository record dc:identifier
https://open.clemson.edu/all_theses/4177

Chain of custody

source
Harvested from
Clemson University
Base URL
open.clemson.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Stewart, Hannah. Creation and Validation of a Novel Bayesian Calibration Method with Griddy Gibbs Sampling. Thesis thesis, https://open.clemson.edu/all_theses/4177