A Fast and Effective Sensitivity and Uncertainty Quantification Method for Additively Manufacturing Metals
Abstract
Metal-based additive manufacturing (AM) is considered a promising technology with many potential applications due to the unparalleled design flexibility of the process. The implications of this technology are far-reaching and represent a new paradigm in materials and systems design where engineered materials can be envisioned as an assembly of individually programmed material elements. Recent economic reports have shown that the AM industry has experienced rapid annual growth on the order of approximately 30% from 2010 to 2015, and it is projected to increase in sales totaling approximately $8 billion by 2025. AM brings unique opportunities for the defense sector by allowing the services to implement new weapon systems and to print spares and other vital components at the point of need; meeting the needs of the warfighter faster, more reliably, and with greater precision. However, despite the transformative potential of AM, the full utility of this material fabrication technology remains unrealized due to the lack of reproducibility and reliability in the process, and the uncertainty in mechanical properties of the fabricated parts. To overcome these challenges, it is essential to ascertain the parameters of the manufacturing process that drive the variability in the printed parts. This critical information is the foundation on which to modify, design, and control the AM process to improve quality and reliability. The purpose of this research is to develop, implement, verify, and validate a new simulation model for powder bed fusion additively manufactured metals that contains two new breakthrough technologies: i) a hypercomplex finite element method for arbitrary order sensitivity analysis, and ii) a fast uncertainty quantification (UQ) method that predicts the distribution of the material responses without sampling. This new methodology has the potential to revolutionize the AM procedure by providing near-real time feedback as to the primary causes of variation, and, hence, information how to improve the process. This capability can obviate much of the current practice of expensive design of experiment studies to determine important factors and process improvement. The anticipated outcomes of this research are clear: to provide DoD users with a new approach to analyzing and understanding the variability of AM manufactured parts such that the manufacturing process can be optimized to reduce variability and provide more reliable, higher quality parts suitable for deployment in DoD systems. The effort builds upon the recent DoD-funded development of the hypercomplex finite element method and its successful application to thermoelasticity, plasticity, fracture mechanics, creep, heat transfer, and other phenomenon. Hence, it stands ready to be extended and applied to model and provide insight into the physics of the additive manufacturing process. The new methodology will be easily transitioned to DoD laboratories since it will be implemented within the Abaqus commercial finite element program that is in widespread use across DoD. The program will be led by Dr. Harry Millwater, Professor, Mechanical Engineering, the University of Texas at San Antonio (UTSA) supported Co-PIs Dr. Arturo Montoya, Associate Professor, and Dr. David Restrepo, Assistant Professor, also from UTSA.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 31, 2020
- Source ID
- W911NF2010315
Entities
People
- Harry Millwater
Organizations
- Army Contracting Command
- Office of the Secretary of Defense
- University of Texas at San Antonio