2026 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2026): Thermal Models
Program Organizers: David Leigh, University of Texas at Austin
Tuesday 10:00 AM
August 4, 2026
Room: Lavaca
Location: AT&T Center
10:00 AM
A Graph-Based Thermal History Simulation for Material Extrusion Additive Manufacturing: Joseph Bartolai1; Callie Zawaski1; Jacob Morgan1; Jan Petrich1; 1Pennsylvania State University
A method for simulating thermal history of Material Extrusion (MEX) Additive Manufacturing (AM) processes using a graph-based representation of the part and an energy-based thermal history simulation. This graph-based simulation generates a thermal history prediction significantly more quickly than finite element based methods. The graph is generated directly from gcode toolpath data, with nodes representing discrete volumes of deposited material and edges connecting adjacent material volumes. Energy is added to the simulated part at a node representing the current position of the deposition nozzle. Energy is removed at boundary nodes. Applying polymer weld theory to this simulated thermal history, part performance can be predicted. These performance predictions can be used to optimize part and process design, or when combined with in-situ assessment used as part of a quality control process. Simulated thermal histories are compared to in-situ data collected with embedded thermocouples, and the accuracy of the simulation is discussed.
10:20 AM
Benchmarking Thermal Models for Directed Energy Deposition Processes: Elizabeth Chang-Davidson1; 1Wesleyan University
Within fusion-based additive manufacturing (AM) processes, understanding the temperature across the part during the build is useful for predicting cooling rates, material properties, and process defects. A significant portion of the thermal models developed specifically for AM are based on laser power bed fusion (LPBF) data and parameters. However, directed energy deposition (DED) processes have no less need for specialized thermal models, since heat buildup and the consequent part deformation are significant issues. Benchmarking models for DED requires evaluating both model accuracy and computational cost. In many cases, these two performance metrics have opposing effects; a classic example would be mesh density in finite element models. This work, still in progress, focuses on developing real-world data to evaluate models and comparisons of computational cost on identical computers. The models evaluated include classic finite element models, semi-analytical models, and models originally developed for LPBF being explored for use in DED processes.
10:40 AM
Temperature Distributions of Traditional Deposition Trajectories Towards Thermal-Model-Informed Path Planning Strategies in WAAM: Hutchison Peter1; Bradley Jared1; 1University of Tennessee, Knoxville
Advancements in numerical and analytical modeling of heat transfer in wire arc additive manufacturing (WAAM) have led to improved agreement between simulated and experimental temperature results. Additionally, thermo-mechanical models of the WAAM process have been developed for improved prediction of residual stress, distortion, and resulting microstructure. However, implementation of thermal modeling into path planning for WAAM to influence resulting temperature distribution remains a topic for further investigation and development. Temperature distributions of part surfaces were measured during WAAM of four parts. Although designed from a constant part model, a unique and traditional deposition trajectory was used to manufacture each part. From the temperature distributions, metrics relevant to residual stress, grain size, and grain morphology were calculated and are discussed. The experimentally acquired temperature distribution measurements will be compared to temperature distributions from new deposition strategies in future work.
11:00 AM
Finite Element Thermal Simulation of Forced Convective Cooling in Wire-Arc Directed Energy Deposition: Brenna Betts1; Jeffery Betts1; David Johnson1; Reid Schaff1; Matthew Priddy1; 1Mississippi State University
Wire-arc directed energy deposition offers extensive advantages in deposition rate and material efficiency for large-scale metal components, yet thermal management remains a critical challenge. Heat accumulation, varying cooling rates, and lack of standardized interpass conditions contribute to microstructural evolution, residual stress accumulation, and geometric distortion. To mitigate impacts, practitioners have adopted forced convective cooling strategies, including end-effector mounted cold air guns applied during dwell periods. Despite growing adoption, a validated finite element (FE) modeling framework to accurately capture the thermal influence of forced convective cooling in wire-arc DED has not been established. This study presents development and validation of an FE thermal simulation approach for wire-arc DED with integrated forced cooling implemented in Abaqus, validated against experimental thermal histories. The model reproduces thermal cycling behavior, instantaneous cooling rates, and the 3-D internal cooling field, providing a foundation for integrating active cooling strategies into physics-based process models and digital twins.
11:20 AM
A Neural ODE Approach for Thermomechanical Field Prediction in Directed Energy Deposition: Dhruba Aryal1; Praveen Vulimiri1; Todd Sparks1; Albert To1; 1University of Pittsburgh
In metal directed energy deposition, accurate prediction of thermal and mechanical fields such as temperature, residual stress and displacement enables improved design and optimization of parts and process parameters without relying on expensive experimental trials. High fidelity thermo-mechanical simulations using the Finite Element Method can provide such predictions. However, these simulations remain computationally expensive. This work proposes a deep learning method by training a neural ordinary differential equation (ODE) model that leverages features derived from the governing coupled thermo-mechanical equations to predict both thermal and mechanical responses. The model demonstrates excellent agreement with unseen datasets for both fields, accurately capturing residual stresses and deformations while achieving significantly faster inference compared to full finite element simulations. Additionally, the model enables localized prediction of thermo-mechanical fields at regions of interest without requiring full-domain simulation, offering further computational savings for targeted analysis.
11:40 AM
Accelerating Phase Field Simulations of Microstructure Evolution in Additive Manufacturing by Tensor Decomposition: Ye Lu1; 1University of Maryland Baltimore County
Phase-field simulations have been recognized as a powerful tool for understanding the detailed microstructure evolution in metal additive manufacturing (AM). However, their prohibitive computational cost has limited the achievable simulation domain size, leading to their predominant use in two-dimensional (2D) studies. This limitation has also hindered direct validation against larger experimental specimens and constrained their practical application in AM process optimization. This work presents a tensor decomposition-based model reduction framework for accelerating these phase field simulations. The core idea is to perform a spatial decomposition of the solution field to overcome the difficulties related to the growing degrees of freedom in large volume and high-resolution phase field simulations. The performance of the method will be demonstrated using 3D large volume phase field simulations of grain structure in AM processes.
12:00 PM
Investigation of Reduced-Order Temperature Models for Warpage Prediction in Composites Additive Manufacturing: Ethan Kessel1; Eduardo Barocio1; 1Purdue University
Fiber-reinforced polymers printed via Material Extrusion additive manufacturing experience residual stresses and warpage due to thermal gradients and anisotropic phenomena of the composite material. Warpage can be effectively predicted using FEA, however the detailed thermal histories required are expensive to compute using full-3D transient heat transfer. We investigated the use of a reduced-order heat transfer model to cut the computation time needed for temperature predictions from hours to minutes. The temperature predictions from the reduced heat transfer model were mapped onto 3D geometry used in a thermomechanical analysis of the print process. Temperature predictions, residual stress, and warpage are compared for geometries prone to deformation for both computational methods. This presentation discusses the computation savings, accuracy, and limitations of the reduced-order method against a full-3D thermomechanical analysis.