Georgia Institute of Technology
In-Situ Monitoring and Estimation of Thermal Histories in Powder-Blown Directed energy Deposition
Abstract
dc:description.abstractPowder-blown directed energy deposition (DED-LP) is an emerging additive manufacturing process for producing near-net-shape metallic components. Materials commonly used in DED-LP, such as 17-4 precipitation-hardening (PH) stainless steel, exhibit strong sensitivity to thermal history due to repeated melting and solidification cycles. These transient thermal effects lead to heat accumulation and significantly influence microstructural evolution and mechanical properties, particularly given the temperature-dependent hardening behavior of 17-4 PH. Predicting these material outcomes typically relies on finite element analysis (FEA); however, such models are difficult to calibrate and often suffer from significant uncertainty due to process variability and poorly constrained boundary conditions. This motivates the need for in-situ monitoring techniques capable of improving model fidelity and enabling real-time process insight. In practice, thermal measurements are sparse and noisy due to high spatial gradients and occlusions from the deposition head, making accurate temperature estimation both computationally intensive and highly qualitative. This work presents a novel melt-pool tracking framework that establishes a direct correspondence between video data and simulation nodes. A state estimation approach is then used to model the time-varying error between FEA predictions and observed thermal behavior, reducing sensor noise and reconstructing incomplete temperature fields. The proposed method enables efficient, parallelizable temperature estimation and improves predictive capability for subsequent layer deposition. Experimental results demonstrate melt-pool tracking accuracy within 1.5 mm without camera recalibration across layers. Additionally, the proposed thermal estimation framework reduces variance while preserving critical thermal cycling characteristics. All methods are computationally lightweight and suitable for deployment on edge devices operating at frame rates of 30 Hz or higher
Degree
thesis:*- Name thesis:degree_name
- Mechanical Engineering, MS
- Department thesis:degree_department
- Mechanical Engineering
- Grantor
- Georgia Institute of Technology
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Bourgikos, George Spiros
- Advisor dc:contributor.advisor
-
- Kurfess, Thomas R.
- Committee members dc:contributor.committeemember
-
- Saleeby, Kyle
- Melkote, Shreyes
Rights
- Language dc:language.iso
- English
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/1853/81619