Ceramic materials have found various applications, especially under harsh conditions, thanks to their superior mechanical, electrical, optical, chemical, thermal, and biocompatible properties. However, since ceramics shatter upon impact rather than deform, manufacturing ceramic components with complex structures of high-quality surfaces is a challenge. Ultra-precision machining of ceramics has found a way to overcome this challenge by cutting or removing very tiny amounts of material. However, its productivity is not satisfactory and an understanding of the material behavior under cutting, especially at atomic scale, remains elusive. This award is to find optimized machining conditions for ceramic materials based on an improved understanding of material failure. This understanding is obtained by a combined strategy of state-of-the-art experiment and atomistic simulation approaches coupled with machine learning algorithms. This approach facilitates the machining of advanced ceramics without the need for extra post-processing, which is expensive and time consuming and, thus, achieves industry-required productivity. Moreover, by improving the fabrication process and damage control of ceramic materials, high quality ceramic components such as engine blocks, camera lenses, high energy lasers, and biomedical implants are possible, which benefits U.S. industry and economy. This research engages students from historically underrepresented groups in research experiences, leveraging programs such as Graduate Engineering Research Scholar and Women in Science and Engineering.
This collaborative research combines experiment and atomistic simulations to understand how residual stress and subsurface damage form during ultra-precision machining of ceramics by considering three representative ceramic materials; two hard ceramics, sapphire and zirconia, and one soft ceramic, potassium dihydrogen phosphate. Ultra-precision machining of ceramics depends on the anisotropy in their crystal structure and its influence on the critical depth-of-cut where the ductile-to-brittle transition occurs. The cutting experiments are designed to quantify changes in residual stress and subsurface damage under various cutting conditions while the atomistic simulations provide a detailed understanding of the ductile and brittle behaviors of ceramics at the atomic scale during machining. Molecular dynamics methodology is employed for atomistic simulations. In particular, the multiscale approach, based on the atomistic-continuum coupling, enables performing simulations in more realistic and near-experimental conditions. Moreover, experiments and simulations provide sampling conditions for the machine learning algorithm based on K-nearest neighbor calculations, which determine the optimal cutting conditions necessary to minimize residual stress and subsurface damage and cracking. The machine learning predictions are, in turn, verified by machining experiments and simulations. With this knowledge, aggressive rough cutting is applied to meet scalable material removal rate while controlling residual stress and subsurface damage, followed by finish ductile-mode cutting to remove cracks and smooth out the surface.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
