Statistical and Dynamic Model of Surface Morphology Evolution during Polishing in Additive Manufacturing

AbstractMany industrial components, especially those realized through 3D printing undergo surface finishing processes, predominantly, in the form of mechanical polishing. The polishing processes for custom components remains manual and iterative. Determination of the polishing endpoints, i.e., when to stop the process to achieve a desired surface finish, remains a major obstacle to process automation and in the cost-effective custom/3D printing process chains. With the motivation to automate the polishing process of 3D printed materials to a desired level of surface smoothness, we propose a dynamic model of surface morphology evolution of 3D printed materials during a polishing process. The dynamic model can account for both material removal and redistribution during the polishing process. In addition, the model accounts for increased material flow due to heat generated during the polishing process. We also provide an initial random surface model that matches the initial surface statistics. We propose an optimization problem for model parameter estimation based on empirical data using KL-divergence and surface roughness as two metrics of the objective. We validate the proposed model using data from polishing of a 3D printed sample. The procedures developed makes the model applicable to other 3D printed materials and polishing processes. We obtain a network formation model as a representation of the surface evolution from the heights and radii of asperities. We use the network connectivity (Fiedler number) as a metric for surface smoothness that can be used to determine whether a desired level of smoothness is reached or not.Keywords: Additive ManufacturingPolishingSurface Morphology EvolutionNetwork FormationDynamic modelsMaterial Removal and Material RedistributionDisclaimerAs a service to authors and researchers we are providing this version of an accepted manuscript (AM). Copyediting, typesetting, and review of the resulting proofs will be undertaken on this manuscript before final publication of the Version of Record (VoR). During production and pre-press, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal relate to these versions also. Additional informationNotes on contributorsAdithyaa KarthikeyanAdithyaa Karthikeyan is currently a PhD student specializing in Advanced Manufacturing within the Department of Industrial and Systems Engineering, Texas A&M University, USA. He received his B.Tech (Hons) degree in Mechanical Engineering from National Institute of Technology, Tiruchirappalli, India in 2017 and MS degree in Interdisciplinary Engineering from Texas A&M University in 2020. He was a summer intern with Micron Technology Inc. in 2023 with their CMP (Chemical Mechanical Planarization) process development team for semiconductor manufacturing. His research interests include mathematical modelling and data analytics for manufacturing processes and systems.Soham DasSoham Das is a PhD student in the Department of Industrial and Systems Engineering at Texas A&M University, USA. He received his B.Tech degree in Mechanical Engineering from NIT Durgapur, India in 2017. His research lies at the intersection of game theory and combinatorial optimization, such as in the control of learning in network games.Satish T.S. BukkapatnamSatish T.S. Bukkapatnam is the Rockwell International Professor of Industrial and Systems Engineering at Texas A&M University, and the Director of Texas A&M Engineering Experiment Station Institute of Manufacturing Systems. He received his PhD degree in Industrial and Manufacturing Engineering from Pennsylvania State University (1997). His research interests are broadly in smart manufacturing systems, and ultraprecision manufacturing. Dr. Bukkapatnam is a Fellow of IISE and SME, Associate Member of CIRP, and was a Fulbright-Tocqueville Distinguished Chair.Ceyhun EksinCeyhun Eksin received his B.S. degree in control engineering from Istanbul Technical University, Istanbul, Turkey, in 2005, the M.S. degree in industrial engineering from Boğaziçi University, Istanbul, Turkey in 2008, the M.A. degree in statistics from Wharton Statistics Department, University of Pennsylvania, Philadelphia, PA, 19103, USA in 2015, and the Ph.D. degree in electrical and systems engineering from the Department of Electrical & Systems Engineering, University of Pennsylvania, Philadelphia, PA, USA in 2015. He was a postdoctoral researcher jointly affiliated with Scholls of Biological Sciences, and Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA. He is currently an assistant professor in the Department of Industrial and Systems Engineering at Texas A&M University, College Station, TX, USA. He is a recipient of NSF CAREER award in 2023. His research interests are in the areas of networks, game theory, control theory, and distributed optimization. His current research focuses on game theoretic learning and decentralized optimization with applications to autonomous teams, epidemics and energy systems.