信息工程学院

报告题目：A Spatially Resolved Uncertainty Quantification Framework in Aerosol Jet Printing via Hybrid Collocation Polynomial Chaos Expansion

报告人：梁楠楠（副教授）

讲座对象：教师

报告时间：2026年3月18日14:30

报告地点：工科楼B315

梁楠楠，女，副教授，安徽师范大学硕士生导师，安徽省计算机学会理事，安徽省教坛新秀。主要从事智能制造等相关领域的研究工作。

报告摘要：Aerosol jet printing (AJP) has emerged as a transformative additive manufacturing (AM) technology for high-resolution microelectronics. However, its industrial scalability is fundamentally hindered by inherent process instabilities induced by stochastic perturbations. While traditional deterministic models provide insights into the underlying aerosol transport mechanisms within the printhead, they lack the capacity to characterize the spatial evolution of stochastic uncertainties driven by gas flow fluctuations and particle polydispersity, which obscures the stochastic dependency of deposition fidelity on input fluctuations. In response, this research proposes a non-intrusive uncertainty quantification (UQ) framework to capture the intricate, spatially resolved stochasticity of the AJP process. Specifically, a hybrid collocation strategy based on Smolyak sparse grids is proposed, which facilitates systematic uncertainty propagation through the developed computational fluid dynamics (CFD) model by accommodating heterogeneous stochastic inputs. Subsequently, polynomial chaos expansion (PCE) method is employed to quantify spatially resolved statistical distributions of flow velocity and pressure, revealing the propagation mechanisms of aerodynamic perturbations across complex printhead geometries. Finally, a Lagrangian discrete phase model (DPM) is integrated into the analysis to evaluate the statistical impact of droplet polydispersity on trajectory divergence, providing a quantitative relationship between particle-size variability and printing outcomes. The results identify the combination chamber as the primary region for nonlinear uncertainty magnification driven by shear instabilities, whereas the size-dependent competition between Saffman lift and inertial forces constitutes the governing mechanism for trajectory dispersion in the unconstrained free-flight region. This mechanistic understanding enables targeted printhead optimization and uncertainty-informed process tuning, facilitating robust deposition precision in aerosol-based AM.