4.6 Article

Correlations in a Posteriori Error Trends for the Finite Element Method in the Presence of Changing Material Parameters

Journal

IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS
Volume 20, Issue 12, Pages 2516-2518

Publisher

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/LAWP.2021.3116167

Keywords

Finite element analysis; Permittivity; Scattering; Correlation; Correlation coefficient; Error analysis; Uncertainty; Adjoint methods; computational electromagnetics (CEM); error estimation; finite element method (FEM); material parameters

Funding

  1. US Air Force Research Laboratory
  2. RF Analysis and Validation Engineering Software (RAVEnS), Applied Research Associates [FA8650-20-C-1132]

Ask authors/readers for more resources

In this study, it was demonstrated that elementwise error contribution estimates remain highly correlated as material parameters change in scattering problems. The correlation between these estimates is shown to be stronger than that for quantities of interest or gradients of such quantities across the same range of dielectric parameters. Additionally, a correlation was found between the mean and standard deviation of elementwise error contribution estimate magnitudes.
In this letter, we demonstrate that adjoint-based a posteriori elementwise error contribution estimates can be highly correlated in the presence of changing material parameters for finite element method scattering problems. Using a simple lossy dielectric sphere scattering problem set, we explore trends in elementwise a posteriori error contribution estimates as the real permittivity of the spherical scatterer is varied. We show that, not only do elementwise error contribution estimates for this problem remain highly correlated as material parameters vary, but that this correlation is stronger than that apparent for quantities of interest or gradients of such quantities across the same range of dielectric parameters. We also show correlation between the mean and standard deviation of elementwise error contribution estimate magnitudes.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.6
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available