Abstract
dc:description.abstractThe goal of this research is to enable accelerated qualification of metal parts produced using a specific additive manufacturing process called laser powder bed fusion. Realizing the goal will lead to a Born Qualified Additive Manufacturing paradigm for part quality assurance. Born Qualified Additive Manufacturing entails predicting part quality as it is manufactured and implementing corrective actions in real time to cure defects, thus enabling rapid, in-situ, online, and non-destructive assurance of part integrity. Despite its demonstrated ability to elevate performance while simultaneously reducing manufacturing cost, complexity, and time, the adoption of additive manufacturing processes in safety-critical applications remains limited. The main barrier is quality – the prevalence of defects and inconsistent (heterogeneous) microstructures lead to large uncertainty in safety and performance. High costs (~50% of manufacturing cost) and time are spent on post-build quality assurance through destructive characterization and testing. Hence, rapid, online, non-destructive, in-situ, and real-time quality assurance is vital for the sustainability of additive manufacturing. In pursuit of Born Qualified Additive Manufacturing, the overarching objective of this work is to establish a novel physics- and data-informed hierarchical digital twin framework to understand (model), observe (monitor), predict, and control the causal process phenomena that govern the critical-to-quality aspects of additively manufactured parts. The central hypothesis of the digital twin framework is that combining rapid physics-based computational modeling, in-situ sensor-based monitoring, and machine learning enables accurate prediction and control of the solidified microstructure and part properties. The hierarchical digital twin framework is stratified into the following three parts. Part 1: Modeling the causal process phenomena and understanding their effect (correlation) on solidified microstructure and mechanical properties. Part 2: Prediction of solidified microstructure using physics and data-integrated machine learning. Part 3: Control of part thermal history through computational modeling and real-time in-process sensing for improved part quality. Collectively, this work provides new theoretical and practical foundations in modeling, monitoring, machine learning, and machine control toward Born Qualified additive manufacturing.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Mechanical Engineering
- Department dc:contributor.department
- Mechanical Engineering
- Grantor dc:publisher
- Virginia Tech
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Deshmukh, Kaustubh Vilas
- Chairs dc:contributor.committeechair
-
- Williams, Christopher Bryant
- Rao, Prahalada Krishna
- Committee members dc:contributor.committeemember
-
- Raeymaekers, Bart
- Acar, Pinar
- Yu, Hang
Subjects
dc:subject × 7Rights
dc:rights- Statement dc:rights
-
- Creative Commons Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Dc Identifier Other
- vt_gsexam:46460