Search for: All records

Additive manufacturing (AM) enables the fabrication of complex, highly customized geometries. However, the design and fabrication of structures with advanced functionalities, such as multistability and fail-safe mechanism, remain challenging due to the significant time and costs required for high-fidelity simulations and iterative prototyping. In this study, we investigate the application of Bayesian Optimization (BO), an advanced machine learning framework, to accelerate the discovery of optimal AM compatible designs with such advanced properties. BO uses a probabilistic surrogate to strategically balances the exploration of design space with few test designs and the exploitation of design space near current best performing designs, thereby reducing the number of design simulations needed. While existing studies have demonstrated the potential of BO in AM, most have focused on static or simple designs. Here, we target multistable structures that can reconfigure among multiple stable states in response to external conditions. Since mechanical performance (e.g., strength) is configuration-dependent, our goal is to identify high performing designs while ensuring that strength in all stable configurations exceeds a prescribed threshold for structural robustness. 

The present work concerns the transferability of coarse-grained (CG) modeling in reproducing the dynamic properties of the reference atomistic systems across a range of parameters. In particular, we focus on implicit-solvent CG modeling of polymer solutions. The CG model is based on the generalized Langevin equation, where the memory kernel plays the critical role in determining the dynamics in all time scales. Thus, we propose methods for transfer learning of memory kernels. The key ingredient of our methods is Gaussian process regression. By integration with the model order reduction via proper orthogonal decomposition and the active learning technique, the transfer learning can be practically efficient and requires minimum training data. Through two example polymer solution systems, we demonstrate the accuracy and efficiency of the proposed transfer learning methods in the construction of transferable memory kernels. The transferability allows for out-of-sample predictions, even in the extrapolated domain of parameters. Built on the transferable memory kernels, the CG models can reproduce the dynamic properties of polymers in all time scales at different thermodynamic conditions (such as temperature and solvent viscosity) and for different systems with varying concentrations and lengths of polymers.