Solid Freeform Fabrication Conference 2019
August 12th through 14th, Austin, TX.
Topic Area: Thermal Modeling in Metal Additive Manufacturing
(1) Part-level Thermal Modeling in Additive Manufacturing Using Graph Theory
Day, Date, Time, and Location: Monday 8/13/2019, 1450 hrs (2:50 pm CT), Salon B.
Authors: Reza Yavari; Kevin Cole; and Prahalad Rao.
Abstract: The objective of this work is to develop and apply the mathematical concept of graph theory to approximate the instantaneous thermal distribution in metal additive manufactured parts. Compared to finite element analysis techniques, the proposed mesh-free graph theory-based approach facilitates layer-by-layer simulation of the heat flux within a few minutes on a desktop computer. To test this claim we compare the thermal distribution predictions obtained from the proposed graph theoretic approach with Goldak’s moving heat source finite element simulation, and a commercial solution. Results show that the graph theoretic approach agrees with the temperature distribution trends predicted using finite element analysis with error less than 10% and within 1/10th of the time of the finite element solution; the computational time for obtaining the thermal distribution for the three part geometries studied was less than 10 minutes using graph theory compared to over 3 hours with finite element analysis.
(2) Experimental Validation of the Thermal Distribution Predicted by the Graph Theory Approach: Application to Laser Powder Bed Fusion
Day, Date, Time, and Location: Tuesday 8/14/2019, 1400 hrs (2:00 pm CT), Salon A.
Authors: Paul Hooper; Reza Yavari; Prahalad Rao; and Kevin Cole.
Abstract: The objective of this paper is to experimentally validate the graph-based approach, advanced in our companion abstract, for predicting the thermal distribution in parts made using the laser powder bed fusion. The graph approach was used to predict the temperature on the top surface of two geometries, namely, an inverted cone and cylinder (stainless steel). Experimental data in the form thermal images were acquired using a calibrated staring-configuration shortwave infrared camera. The results show that the graph theory approach predicts the temperature at the top surface of the build within the time it takes to build the part with error within 10%. For instance, in the case of the inverted cone geometry (top diameter 20 mm, height 11.25 mm, 225 layers), the simulation time with graph theory was less than 20 minutes versus 50 minute build time.
Topic Area: In-situ Sensing and Analytics
(1) Heterogeneous Sensing-based In-process Quality Monitoring of Single-tracks Built using Laser Powder Bed Fusion Additive Manufacturing Process
Day, Date, Time, and Location: Tuesday 8/14/2019, 0915 hrs, 412.
Authors: Aniruddha Gaikwad; Brian Giera; Prahalad Rao
Abstract :The goal of this work is to monitor the quality of single-tracks made using laser powder bed fusion (LPBF) additive manufacturing (AM) process with the help of a dual sensing array comprising of a high speed camera, and a pyrometer. To realize this goal, multiple single tracks were printed at varying laser scan speeds and laser power. In-process data collected by the sensors was analyzed to classify the single-tracks into different Andrew’s numbers. Further, the senor data was analyzed to predict the quality metrics, viz., mean track width, standard deviation of the track width and track continuity.
(2) Digital Twin in Metal Additive Manufacturing – A Paradigm Integrating Modeling, Sensing, and Machine Learning for Defect Prediction
Day, Date, Time, and Location: Monday 8/13/2019, 1430 hrs (2:30 pm CT), 412.
Authors: Reza Yavari; Aniruddha Gaikwad; Mohammad Montazeri; Prahalad Rao; Kevin Cole; and Linkan Bian.
Abstract: The goal of this work is to prevent occurrence of defects in parts made using metal additive manufacturing (AM) though closed-loop control and in-situ process correction. As a step towards this goal, the objective of this work is to develop and apply a theoretical model-based approach to track the thermal profile of parts made using the directed energy deposition (DED) metal AM process, and subsequently, use the model-derived thermal predictions in conjunction with in-process temperature measurements to detect occurrence of lack-of-fusion flaws. In other words, instantiate the digital twin concept to identify and isolate defects in the DED metal AM process. The central hypothesis of this work is that, a deviation in the observed melt pool temperature from its model-derived counterpart signifies a process drift, indicative of an impending fault. This approach presented herein is demonstrated to predict the occurrence of lack-of-fusion flaws with statistical fidelity approaching 90% F-score.
(3) Using Heterogeneous In-process Sensor Data To Detect Lack-of-fusion Defects In Directed Energy Deposition of Titanium Alloy (Ti-6Al-4V) Parts
Day, Date, Time, and Location: Monday 8/14/2019, 0935 hrs, 412.
Authors: Mohammad Montazeri; Abdalla Nassar; Christopher Stutzman; and Prahalad Rao;
Abstract: The objective of this work is to detect in situ the occurrence of lack-of-fusion defects in titanium alloy (Ti-6Al-4V) parts made using directed energy deposition (DED) additive manufacturing (AM). We use data from two types of in-process sensors, namely, a spectrometer and a camera filtered around 430 nm which are integrated into an Optomec MR-7 DED machine. Both sensors are focused on capturing the dynamic phenomena around the melt pool region. The spectrometer measures optical emissions from the melt pool and vapor/plasma plume above the melt pool, whilst the filtered, optical camera captures images of the plume. To detect lack-of-fusion defects, we fuse (combine) the data from the in-process sensors invoking the concept of Kronecker product of graphs. We show that the features that manifest from the combined sensor data in the Kronecker graph product domain classify the pore severity with statistical fidelity (F-score) close to 85%.
Topic Area: Processing Science
(1) Material-structure-property Optimization and In-process Monitoring of 3D Printed Bone Tissue
Day, Date, Time, and Location: Tuesday 8/14/2019, 0815 hrs , 415AB.
Authors: Samuel Gerdes; Azadeh Mostafavi; Srikanthan Ramesh; Ali Tamayol; Iris Rivero; and Prahalad Rao.
Abstract: With developments in additive manufacturing processes, it has become possible to fabricate entirely new classes of materials; biomaterials. These biomaterials and composite materials feature favorable mechanical properties, cell adhesion, and biodegradability for the formation of biologically relevant additively manufactured tissues. With hydrogels and hard polymers targeting soft tissues, such as skin or muscle, and hard tissues such as bone, respectively, tissues of varying composition could be created. However, the process of biological additive manufacturing has yet to have a standardized approach for reproducible, optimized material printability. This study employs a bottom-up approach to optimizing printability using in-process sensor data for the linking of process parameters and print quality. Composites of thermoplastic polycaprolactone polymer and biologically relevant hydroxyapatite were utilized to target the formation of bone tissue. These composites were rheologically assessed and extrusion was first optimized one-dimensionally then through two-dimensional geometric analysis. Furthermore, three-dimensional scaffolds were evaluated for cell viability.
(2) Process-structure-property Relationships in the Coating of Stellite on Inconel 718 by Directed Energy Deposition Process
Day, Date, Time, and Location: Tuesday 8/14/2019, 1045 hrs , 416AB.
Authors: Ziyad Smoqi; Jordan Severson; Joshua Toddy; Harold Halliday; Tom Cobbs; and Prahalad Rao
Abstract: The objective of this work is to quantify the effect of process parameters on the microstructure and physical properties of Stellite 21 and Stellite 6 coatings on Inconel 718 by Directed Energy Deposition (DED) Process. Two process factors were varied: preheating and laser power. Preheating was achieved by scanning the surface of the Inconel 718 coupon with the laser. In-process thermal data was acquired via thermocouple probes. The major finding is that the combination of preheating the substrate and laser power for deposition minimizes cracking and results in micro-hardness of up to approximately 450 HV. To explain this observation, the surface temperature data acquired during the process was analyzed and correlated with the coating hardness and surface integrity.