REGISTER HEREBio:
Jinming Wan is a Ph. D student in the department of Systems Science and Industrial Engineering at State University of New York at Binghamton. His research interests include modeling and analysis, uncertainty quantification, machine learning, and optimal control techniques to provide compelling guidelines on coordinated policy design to enhance the preparedness for epidemics.
Presentation Description:
As described in the abstract below, Wan will present the following:
1. The application is to achieve the robust optimal design for ultra-precision machining processes (such as chip fabrication), which needs high material removal rate (MRR) and avoids regenerative chatter that leads to an unacceptable surface finish.
2. The proper selection of process parameters (e.g., spindle speed and depth of cut) can avoid regenerative chatter
Identification of the boundary that separates the stable from unstable design configurations in machining processes is a profound quest.
Craftsman knowledge and trial-and-error tests are traditionally used in machining process design, which, however, incurs tedious efforts and enormous costs
Numerical analytical approaches (physical simulation) have been utilized to predict the stability to avoid trial-and-error tests.
High depth of cut usually induces a large cutting force, which, in turn, magnifies the chance of chatter but enforces a high material removal rate/ high productivity
3. The boundary changes due to the inherent uncertainty in machining operations (e.g., material and process calibration)
quantifying the uncertainty and providing a robust design is inevitable to achieve a robust optimal design
the traditional method is using Monte Carlo analysis – it is time consuming and computationally intensive
We propose an active learning multi-fidelity approach to estimate the stable boundary, in which 95% the computational overhead is reduced
Abstract:
In this study, we carry out robust optimal design for the machining operations, one key process in wafer polishing in chip manufacturing, aiming to avoid the peculiar regenerative chatter and maximize the material removal rate (MRR) considering the inherent material and process uncertainty. More specifically, we characterize the cutting tool dynamics using a delay differential equation (DDE) and enlist temporal finite element method (TFEM) [1] to derive its approximate solution and stability index given process settings or design variables. To further quantify the inherent uncertainty, replications of TFEM under different realizations of random uncontrollable variables are performed, which however incurs extra computational burden. To eschew the deployment of such a crude Monte Carlo (MC) approach at each design setting, we integrate the stochastic TFEM with a stochastic surrogate model, stochastic kriging [2], in an active learning framework to sequentially approximate the stability boundary. The numerical
result suggests that the nominal stability boundary attained from this method is on par with that from the crude MC, but only demands a fraction of the computational overhead. To further ensure the robustness of process stability, we adopt another surrogate, Gaussian process, to predict the variance of the stability index at unexplored design points and identify the robust stability boundary per the conditional value at risk (CVaR) criterion [3]. Therefrom, an optimal design in the robust stable region that maximizes the MRR can be identified. References [1] P. V. Bayly, J. E. Halley, B. P. Mann, and M. A. Davies, “Stability of interrupted cutting by temporal finite element analysis,” Journal of Manufacturing Science and Engineering, vol. 125, no. 2, pp. 220–225, May 2003, doi: 10.1115/1.1556860. [2] B. Ankenman, B. L. Nelson, and J. Staum, “Stochastic kriging for simulation metamodeling,” Operations Research, vol. 58, no. 2, pp. 371–382, Apr. 2010, doi: 10.1287/opre.1090.0754. [3] A. Chaudhuri, B. Kramer, M. Norton, J. O. Royset, and K. Willcox, “Certifiable risk-based engineering design optimization,” AIAA Journal, pp. 1–15, Dec. 2021, doi: 10.2514/1.J060539