A Dynamic p-Adaptive DGTD Algorithm for Electromagnetic and Multiphysics Simulations

Research output: Contribution to journalArticlepeer-review


In the time-domain simulation of electromagnetic and multiphysics problems, the distributions of physical quantities of interest vary in both space and time. To achieve a good spatial resolution, high-order basis functions can be used to expand the unknown quantities, which is known as the p-refinement. However, a global and static p-refinement will increase the computational cost significantly. In this paper, a dynamic p-adaptation algorithm is proposed based on the discontinuous Galerkin time-domain method, which is able to determine and adjust the basis order in a given discretization element in real time of the simulation. Based on the relation between the nodal and modal approximations defined on unstructured tetrahedral elements, an error estimator, which is very cheap to compute, is developed to determine the proper basis order to achieve a desired numerical accuracy. The dynamic p-adaptation algorithm proposed in this paper is able to capture the fast varying physics by changing the order of basis functions wherever and whenever needed. Several numerical examples adopted from multiple physical disciplines are presented to demonstrate the accuracy, efficiency, and flexibility of the proposed algorithm in the simulation of electromagnetic and multiphysics problems.

Original languageEnglish (US)
Article number7867868
Pages (from-to)2446-2459
Number of pages14
JournalIEEE Transactions on Antennas and Propagation
Issue number5
StatePublished - May 2017


  • Discontinuous Galerkin time-domain (DGTD) method
  • dynamic adaptation
  • electromagnetic simulation
  • error estimation
  • high-order method
  • multiphysics simulation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering


Dive into the research topics of 'A Dynamic p-Adaptive DGTD Algorithm for Electromagnetic and Multiphysics Simulations'. Together they form a unique fingerprint.

Cite this