Additive Manufacturing: Building the Pathway towards Process and Material Qualification: Process Qualification Part I
Sponsored by: TMS Extraction and Processing Division, TMS Materials Processing and Manufacturing Division, TMS Structural Materials Division, TMS: Mechanical Behavior of Materials Committee, TMS: Powder Materials Committee, TMS: Process Technology and Modeling Committee, TMS: Additive Manufacturing Bridge Committee
Program Organizers: John Carpenter, Los Alamos National Laboratory; David Bourell, University of Texas - Austin; Allison Beese, Pennsylvania State University; James Sears, GE Global Research Center; Reginald Hamilton, Pennsylvania State University; Rajiv Mishra, University of North Texas; Edward Herderick, GE Corporate
Tuesday 2:00 PM
February 28, 2017
Location: San Diego Convention Ctr
Session Chair: Dan Thoma, University of Wisconsin; Deepankar Pal, 3DSIM
2:00 PM Invited
Challenges and Opportunities for Metal Additive Manufacturing: Dan Thoma1; 1University of Wisconsin-Madison
Metal additive manufacturing techniques offer exciting functionality for engineering components. For instance, topological optimization, novel design strategies, and functional grading are just a few examples that can be demonstrated. However, despite the overwhelming potential, the qualification of parts produced via metal additive manufacturing remains to be the key barrier for wide-spread utilization of the available technologies. Depending upon the final intent of the component, the performance will be dictated by the properties, and the properties can be tailored through the microstructural control during processing. More specifically, control of the solidification behavior can permit optimization of the material. Qualification for the performance requires new methodologies, models, and diagnostics. The use of in situ imaging can permit the validation of models for microstructural control as well as define the cause of defects emanating from the process methodology. Examples of the opportunities and challenges will be presented.
Microstructure Variation and Process Model Developments For LENS: Josh Sugar1; Lauren Beghini1; Michael Stender1; Michael Veilleux1; David Keicher2; Daryl Dagel2; Michael Maguire2; Chris San Marchi1; 1Sandia National Labs, Livermore, CA; 2Sandia National Labs, Albuquerque, NM
The optimized design and qualification of additively manufactured components requires that the fundamental physics underlying the heat transport during the rapid solidification and subsequent cyclic heating of parts be understood. The thermal gradients present during these AM processes, such as LENS, will impact the microstructure (e.g. grain size/texture, ferrite distribution, dislocation structure) and properties (e.g. tensile strength, residual stress) of these materials. Here, we will show how varying process parameters such as laser raster pattern, laser speed, and laser power affects the microstructure of resultant LENS-built parts. In addition, developments to a LENS process model that incorporates the physics of heat transport during LENS processing will be presented. The predictions of the process model are validated by in situ thermal measurements. The importance of robust process models to optimize the design and performance of parts built with LENS is also discussed.
Machine Learning Applications for Microstructure and Process Qualification in Additive Manufacturing: Brian DeCost1; Barnabas Poczos1; Elizabeth Holm1; 1Carnegie Mellon University
We explore novel applications of machine learning methods to metal powderbed fusion additive manufacturing process and materials characterization data. We focus on developing microstructure representations for both feedstock materials and as-built material that, in conjunction with process monitoring data and simple process models, can support a data-driven approach to gaining insight into processing-microstructure and microstructure performance relationships. For example, we demonstrate how computer vision systems might be applied to SEM micrographs of powder feedstock materials to monitor precursor consistency, and to search for and quantify rates of common defects in recycled powders, such as clustering of partially fused satellite particles. We also investigate how processing conditions such as the spatio-temporal distribution of laser energy input can be correlated to porosity and other microstructure features. We conclude by identifying promising future research directions in which machine learning can inform processing strategies to reliably achieve desired microstructural results.
Development of an Integrated Laser-aided Metal Additive Manufacturing System with Real-time Process, Dimensions, and Property Monitoring, Measurements and Control: Navin Sakthivel1; Joseph Fiordilino2; Deedee Banh3; Subrata Sanyal3; Hitesh Vora1; 1Oklahoma State University; 2University of Pittsburgh; 3Naval Surface Warfare Center
Even with the vast popularity of metal additive manufacturing (AM), the various technological challenges/limitations related to part quality and reproducibility are still major barriers for its widespread adoption. Also based on the various road-map studies, some of the common challenges of AM are related to: (i) real-time measurement and monitoring techniques; (ii) modeling systems that couple design and manufacturing; and (iii) fast closed loop control systems for AM. In light of this, the collaborative efforts undertaken here are directed towards developing a unique metal AM system integrated with robust in-situ process sensors for measuring and monitoring the part dimensions, materials properties, and process health parameters (power, speed, powder feed rate), which is synchronously coupled with modeling and simulations efforts that concurrently assist controlling the process in real-time with close-loop feedback arrangements to aid in producing the components of higher complexity and dimensional accuracy with superior mechanical, chemical, and functional properties. Note: DISTRIBUTION STATEMENT A. Approved for public release; distribution is unlimited.(NSWC Corona Public Release Control Number: 16-013)
3:30 PM Break
3:50 PM Invited
Identifying Critical Variables for Laser Powder Bed Fusion: Li Ma1; Brandon Lane1; Shawn Moylan1; Jeffrey Fong1; James Filliben1; Carelyn Campbell1; Lyle Levine1; 1National Institute of Standards and Technology
In laser-based powder bed fusion (L-PBF), the highly localized laser power input leads to extremely high local temperature gradients. These gradients produce complex thermal histories that vary from place to place, effecting the distribution of phases and localized residual stress in the as-built microstructure. In this research, 3D thermal-mechanical finite element analysis (FEA) models of L-PBF were developed. The temperature history and gradient predictions from the FEA thermal modeling are used for microstructure evolution simulations. Temperature distributions from in-situ thermography, melt pool geometry from microscope measurement, and residual stress from the x-ray measurements are used to validate the FEA models. The measurement and control of all possible material properties and processing parameters is challenging and resource consuming. We applied a computational design of experiments approach to investigate the sensitivity of the build process to numerous material and processing parameters.
Simulation and Experimental Validation of Thermal Cycling Motivated Distortion on Parts Produced Using Alloy IN 625 via Selective Laser Melting: Deepankar Pal1; Samuel Dilip Jangam2; Nachiket Patil1; Sally Xu1; Pradeep Chalavadi1; Kevin Briggs1; Brent Stucker1; 13DSIM LLC; 2University of Louisville
A variety of cantilever builds and their structural modifications have been fabricated using Inconel625 on an EOSM270 powder bed fusion machine at different laser power and speed combinations. It was found that slow scan speeds at high power leads to extraneous deformations of the magnitude of ‘a’ layer thickness in the +Z direction leading to blade crashes during layer spreading. In addition, the parts which survived blade crash events led to intermediate deformations with huge deviations with respect to their dimensional tolerances. To address these deformation anomalies, a microstructurally inclined thermal process model has been developed at 3DSIM where the cyclic thermal events were simulated and the shrinkage variations in and out of the build plane have been turned into 3 dimensional structural boundary conditions as inputs to the Structural solver. The simulated and experimentally measured deformation behavior was further compared and the goodness of match has been obtained.
Modeling the Effects of Microstructure on the Strength of Additively Manufactured Ti-6Al-4V: Jeffrey Florando1; Darren Pagan1; Jonathan Lind1; Rupalee Mulay1; Joseph McKeown1; John Moore1; Nathan Barton1; Mukul Kumar1; 1Lawrence Livermore National Laboratory
Current strength models are often constructed to be parameterized for one microstructure, and may not accurately capture effects due to changes in the microstructure. With additive manufacturing (AM), the microstructure can evolve due to rapid solidification and subsequent thermal treatments. Developing strength models that can capture these complex effects is still an on-going challenge. In this work, the microstructure of AM Ti-6Al-4V is modified to create material with different phase fractions and morphologies, and characterized using electron and x-ray microscopy techniques to understand the partitioning in the stress-strain response from the microstructural features. Using these phase specific properties, crystal plasticity and continuum strength models have been implemented to capture these effects, and will be presented. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract No. DE-AC52-07NA27344.
Effect of Process Time Interval on Mechanical Behavior of Metallic Parts Fabricated via Directed Energy Deposition: Aref Yadollahi1; MJ Mahtabi1; Shuai Shao1; Nima Shamsaei1; Scott Thompson1; 1Mississippi State University
The additive manufacture of various-sized components and/or multiple parts requires one to know the relationship between specific energy input and resultant thermal response. Since densification of material is directly coupled with part heat transfer, the process parameters found optimal for fabricating a single specimen are not necessarily optimal, or even usable, for fabricating larger lots or parts of different size and/or geometry (e.g. components). In this study, thermographic inspection is employed to characterize additively manufactured Ti-6Al-4V, Inconel 718 and stainless steel 316L. For a given material and a set of process parameters, various inter-layer time intervals, i.e. the time between successive layer deposits, are employed using a Laser Engineered Net Shaping (LENS) machine. The microstructure and mechanical behavior of fabricated parts are characterized and the results are explained based on measured thermal history during the build.
Cellular Automata based Microstructural Modeling for Additive Manufacturing Processes: Deepankar Pal1; Javed Akram1; Pradeep Chalavadi1; Brent Stucker1; 13DSIM
Evolution of microstructural features in additively manufactured parts is one of the most important factors in evaluation of the part properties and performance; therefore, prior knowledge of microstructure is essential to predict the part behavior during service conditions. In this study, simulation studies based on cellular automata method is adopted to predict the microstructure at different process conditions. The model combines the effect of thermal, solute, and curvature undercooling to track the velocity of solid-liquid interface. Solidification Rate (R) and Thermal Gradient (G) was computed from the thermal model to identify the types of grains such as the columnar or equiaxed grains. Different scan power and material compositions were varied to study the evolving microstructure. The simulated results were validated against the experimental results with measures of interest such as grain size, grain orientation, and micro-segregation.