Clemson University
Deriving Generalized Temperature-Dependent Material Models for Masonry Through Fire Tests and Machine Learning
Abstract
dc:description.abstractMasonry is one of the oldest and commonly used building materials in the construction industry. Among a variety of benefits, masonry provides low-cost construction, fire, and weather protection as well as thermal and sound insulation. In addition, masonry has superior material properties at elevated temperatures which is reflected by its slow degradation of its mechanical and thermal properties. Literature shows that we do not have a uniform material model that describes the mechanical degradation of masonry under fire conditions. As such, this limits the use of masonry in fire-based performance design of masonry structures. To bridge this knowledge gap, this thesis reviews regionally adopted fire testing methods on masonry and then presents findings from a fire experimental program aimed to explore the influence of elevated temperatures on the mechanical performance of concrete masonry units (CMUs). Our tests include heating and post heating evaluation of the compressive strength of CMUs exposed to realistic fire conditions. Then, this thesis delivers a methodology to derive generalized temperature-dependent material models for CMUs using statistical and Bayesian methods, as well as machine learning (by means of artificial neural networks). Finally, this work articulates limitations and research needs to be tackled in the near future.
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
-
- Daware, Aditya Avinash
- Contributors dc:contributor
-
- M Z Naser
- Laura Redmond
- Brandon Ross
Subjects
dc:subject × 1Identifiers
dc:identifier.*- Repository record dc:identifier
- https://open.clemson.edu/all_theses/3709