Date of Award

8-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Civil Engineering

Committee Chair/Advisor

M.Z naser

Committee Member

Dr. Laura Redmond

Committee Member

Dr. Brandon Ross

Committee Member

Dr. Prasad Rao Rangaraju

Abstract

When concrete is exposed to fire, it can suddenly break apart in a dangerous process called spalling. This can severely weaken buildings and even trigger a collapse. While researchers have studied this problem for years, we still don’t fully understand what causes it. My research uses a combination of data, and advanced AI techniques to identify and quantify the key factors that increase the risk of spalling. I collected data from over 1,000 fire tests and used AI models to predict when spalling might happen. I also applied explainable AI tools to understand why it happens.

The results showed that six key factors play a major role. These include how hot the fire gets, how strong the concrete is, how fast it heats up, the amount of moisture in the concrete, and whether it contains special fibers or additives. But prediction isn’t enough, we also need to understand cause and effect. To do this, I used causal AI techniques, which go beyond identifying patterns to uncovering which factors truly make spalling more or less likely. For example, I found that adding polypropylene fibers can reduce the risk of spalling by up to 31%, while too much moisture can increase the risk by more than 20%. These insights could help engineers design safer, more fire- resistant buildings and may eventually lead to clearer building code guidelines to prevent concrete failure in fires.

Author ORCID Identifier

0000-0002-3889-2564

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.