{"id":{"repo_id":"houston","oai_identifier":"oai:uh-ir.tdl.org:10657/20959"},"canonical_url":"https://search.dev.ndltd.org/etd/houston/oai:uh-ir.tdl.org:10657/20959","repository":{"repo_id":"houston","name":"University of Houston","base_url":"https://uh-ir.tdl.org/server/oai/request"},"display":{"title":"Real-Time Thermography for In Operando Additive Manufacturing Defect Mitigation using a Machine Learning Approach","abstract":"Directed Energy Deposition (DED), a subset of additive manufacturing, produces materials such as metal–matrix composites and functionally graded coatings by utilizing a concentrated high-power laser as the primary energy source. In spite of its benefits, DED continues to be confronted by process-induced defects, such as surface irregularities, geometric distortions, and porosity caused by multilayer effects and unoptimized parameters. The structural integrity and reliability of the printed components are compromised by these imperfections. Defect identification, especially porosity, has conventionally depended on non-destructive testing methods like X-ray and computed tomography (CT) imaging. While process improvement and post-processing can reduce flaws, total eradication is seldom accomplished, and comprehensive validation typically necessitates expensive testing protocols. This study presents in-situ thermal image analysis as an effective and scalable technique for fault identification during the DED process. The research concentrates on detecting porosity in inter-layer and intra-layer areas, cold spots resulting from the balling effect, and geometric deformations on the surfaces of maraging steel (M300) and Nickel-based Superalloy (In718) constructs and its correlation with process parameters. Post processed thermal image datasets were utilized to train a supervised machine learning model (Multi-Layer Perceptron) for defect prediction, while benchmarking against two ensemble learning models—Random Forest and XGBoost—for comparative accuracy evaluation. Additional experiments on stainless steel (316L) investigated the impact of variable hatch spacing and its correlation with lack-of-fusion defects, illustrating the role of insufficient thermal overlap. Overall, this study introduces a low-cost, in-situ thermal monitoring and predictive framework that establishes a connection between process parameters and the occurrence of defects and temperature evolution. The results improve comprehension of the relationship between process and defect in DED and lay the groundwork for intelligent process optimization through real-time monitoring and machine learning-driven prediction.","abstract_html":"Directed Energy Deposition (DED), a subset of additive manufacturing, produces materials such as metal–matrix composites and functionally graded coatings by utilizing a concentrated high-power laser as the primary energy source. In spite of its benefits, DED continues to be confronted by process-induced defects, such as surface irregularities, geometric distortions, and porosity caused by multilayer effects and unoptimized parameters. The structural integrity and reliability of the printed components are compromised by these imperfections. Defect identification, especially porosity, has conventionally depended on non-destructive testing methods like X-ray and computed tomography (CT) imaging. While process improvement and post-processing can reduce flaws, total eradication is seldom accomplished, and comprehensive validation typically necessitates expensive testing protocols. This study presents in-situ thermal image analysis as an effective and scalable technique for fault identification during the DED process. The research concentrates on detecting porosity in inter-layer and intra-layer areas, cold spots resulting from the balling effect, and geometric deformations on the surfaces of maraging steel (M300) and Nickel-based Superalloy (In718) constructs and its correlation with process parameters. Post processed thermal image datasets were utilized to train a supervised machine learning model (Multi-Layer Perceptron) for defect prediction, while benchmarking against two ensemble learning models—Random Forest and XGBoost—for comparative accuracy evaluation. Additional experiments on stainless steel (316L) investigated the impact of variable hatch spacing and its correlation with lack-of-fusion defects, illustrating the role of insufficient thermal overlap. Overall, this study introduces a low-cost, in-situ thermal monitoring and predictive framework that establishes a connection between process parameters and the occurrence of defects and temperature evolution. The results improve comprehension of the relationship between process and defect in DED and lay the groundwork for intelligent process optimization through real-time monitoring and machine learning-driven prediction.","abstract_has_math":false,"creators":["Lim, Aniqa Ibnat 1995-"],"institution":"University of Houston","degree_name":"Doctor of Philosophy","degree_level":null,"degree_discipline":"Electrical and Computer Engineering","degree_department":null,"school":null,"contributors":[],"advisors":["Hernandez, Francisco C Robles","Mayerich, David"],"committee_chairs":[],"committee_members":["Xu, Ben","Contreras-Vidal, Jose Luis","Litvinov, Dmitri"],"year":2025,"date_issued":"2025-12","date_published":"2025-12","updated_at":"2026-07-24T02:32:24Z","subjects":["Additive Manufacturing","Steel Alloy","Porosity","DED","Machine Learning","LENS","Super Aplloy"],"languages":["English"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/10657/20959","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Hernandez, Francisco C Robles","Mayerich, David"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Xu, Ben","Contreras-Vidal, Jose Luis","Litvinov, Dmitri"]},{"key":"dc:creator","label":"Author","values":["Lim, Aniqa Ibnat 1995-"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-02-17T17:40:30Z"]},{"key":"dc:date.issued","label":"Date","values":["2025-12"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical and Computer Engineering"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Houston"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Additive Manufacturing","Steel Alloy","Porosity","DED","Machine Learning","LENS","Super Aplloy"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["English"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/10657/20959"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Directed Energy Deposition (DED), a subset of additive manufacturing, produces materials such as metal–matrix composites and functionally graded coatings by utilizing a concentrated high-power laser as the primary energy source. In spite of its benefits, DED continues to be confronted by process-induced defects, such as surface irregularities, geometric distortions, and porosity caused by multilayer effects and unoptimized parameters. The structural integrity and reliability of the printed components are compromised by these imperfections. Defect identification, especially porosity, has conventionally depended on non-destructive testing methods like X-ray and computed tomography (CT) imaging. While process improvement and post-processing can reduce flaws, total eradication is seldom accomplished, and comprehensive validation typically necessitates expensive testing protocols. This study presents in-situ thermal image analysis as an effective and scalable technique for fault identification during the DED process. The research concentrates on detecting porosity in inter-layer and intra-layer areas, cold spots resulting from the balling effect, and geometric deformations on the surfaces of maraging steel (M300) and Nickel-based Superalloy (In718) constructs and its correlation with process parameters. Post processed thermal image datasets were utilized to train a supervised machine learning model (Multi-Layer Perceptron) for defect prediction, while benchmarking against two ensemble learning models—Random Forest and XGBoost—for comparative accuracy evaluation. Additional experiments on stainless steel (316L) investigated the impact of variable hatch spacing and its correlation with lack-of-fusion defects, illustrating the role of insufficient thermal overlap. Overall, this study introduces a low-cost, in-situ thermal monitoring and predictive framework that establishes a connection between process parameters and the occurrence of defects and temperature evolution. The results improve comprehension of the relationship between process and defect in DED and lay the groundwork for intelligent process optimization through real-time monitoring and machine learning-driven prediction."]},{"key":"dc:format.mimetype","label":"Dc Format Mimetype","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Real-Time Thermography for In Operando Additive Manufacturing Defect Mitigation using a Machine Learning Approach"]}]}],"canonical_facts":{"dc:contributor.advisor":["Hernandez, Francisco C Robles","Mayerich, David"],"dc:contributor.committeemember":["Xu, Ben","Contreras-Vidal, Jose Luis","Litvinov, Dmitri"],"dc:creator":["Lim, Aniqa Ibnat 1995-"],"dc:date.accessioned":["2026-02-17T17:40:30Z"],"dc:date.issued":["2025-12"],"dc:description.abstract":["Directed Energy Deposition (DED), a subset of additive manufacturing, produces materials such as metal–matrix composites and functionally graded coatings by utilizing a concentrated high-power laser as the primary energy source. In spite of its benefits, DED continues to be confronted by process-induced defects, such as surface irregularities, geometric distortions, and porosity caused by multilayer effects and unoptimized parameters. The structural integrity and reliability of the printed components are compromised by these imperfections. Defect identification, especially porosity, has conventionally depended on non-destructive testing methods like X-ray and computed tomography (CT) imaging. While process improvement and post-processing can reduce flaws, total eradication is seldom accomplished, and comprehensive validation typically necessitates expensive testing protocols. This study presents in-situ thermal image analysis as an effective and scalable technique for fault identification during the DED process. The research concentrates on detecting porosity in inter-layer and intra-layer areas, cold spots resulting from the balling effect, and geometric deformations on the surfaces of maraging steel (M300) and Nickel-based Superalloy (In718) constructs and its correlation with process parameters. Post processed thermal image datasets were utilized to train a supervised machine learning model (Multi-Layer Perceptron) for defect prediction, while benchmarking against two ensemble learning models—Random Forest and XGBoost—for comparative accuracy evaluation. Additional experiments on stainless steel (316L) investigated the impact of variable hatch spacing and its correlation with lack-of-fusion defects, illustrating the role of insufficient thermal overlap. Overall, this study introduces a low-cost, in-situ thermal monitoring and predictive framework that establishes a connection between process parameters and the occurrence of defects and temperature evolution. The results improve comprehension of the relationship between process and defect in DED and lay the groundwork for intelligent process optimization through real-time monitoring and machine learning-driven prediction."],"dc:format.mimetype":["application/pdf"],"dc:identifier.uri":["https://hdl.handle.net/10657/20959"],"dc:language.iso":["English"],"dc:subject":["Additive Manufacturing","Steel Alloy","Porosity","DED","Machine Learning","LENS","Super Aplloy"],"dc:title":["Real-Time Thermography for In Operando Additive Manufacturing Defect Mitigation using a Machine Learning Approach"],"dc:type":["Thesis"],"thesis:degree_discipline":["Electrical and Computer Engineering"],"thesis:degree_name":["Doctor of Philosophy"],"thesis:institution_name":["University of Houston"]},"updated_at":"2026-07-24T02:32:24Z"}