Electromagnetic field computations for multi-physics phenomena: new techniques for emerging applications

点击翻译按钮获取中文摘要

Many of the current advances in natural sciences and engineering evolve on and across rather than within conventional disciplinary boundaries. Electromagnetic field computations are now being employed in areas such as wireless bio-electronics, neuro-modulation, biomedical imaging and hyperthermia, in parallel with computations of associated thermal, mechanical and even biological interactions. Yet, the current approach to multiphysics modeling largely relies on interfacing independently developed solvers. Fundamental questions of numerical accuracy, stability and efficiency, thoroughly but separately addressed over the years for computational methods in electromagnetics, acoustics, thermodynamics, mechanics, quantum mechanics and chemistry, have received little attention in the process of integrating them in multiphysics packages. While the state-of-the-art in multiphysics modeling remains almost stationary and inherently fragmented, new and exciting multiphysics phenomena involving electromagnetic fields and light are actively explored, such as the use of opto-mechanical coupling for the design of non-reciprocal devices without magnetic materials (Miri et al, PRL, June 2017) and meta-surface based wireless links to biomedical implants (Ho et al, Phys. Rev. B, 2015). The applicant's group has also explored the multiphysics optimization of microwave ablation antennas (Sharma and Sarris, J. Appl. Phys., Feb. 2017) and wireless power transfer inspired hyperthermia applicators (Lang and Sarris, J. Appl. Phys., Aug. 2017), whereby electromagnetic devices are directly designed and optimized with respect to the desired temperature distribution in and around a biological tissue. The proposed program is a comprehensive effort to address this evident lack of synchronization between current numerical modeling techniques and the envisioned or experimentally explored multiphysics applications of electromagnetic fields. Joint optimization of numerical electromagnetics techniques, of the finite-difference and the discontinuous Galerkin time-domain type, and their counterparts for thermal, mechanical and quantum effects will be pursued to maximize their accuracy per computational resources used. Machine intelligence driven models of multiphysics interactions will be explored, along with efficient methods for space-time multiscale modeling, multiphysics uncertainty quantification and optimization under uncertainty. These advances will enable the systematic exploration of new and emerging communication, sensing, bio-engineering and computational technologies that are flourishing in Canada, directly benefiting the Canadian economy. Exciting opportunities for well-rounded, interdisciplinary training of highly qualified personnel will thus be created.