Publications

Abstract not provided.

Abstract not provided.

Abstract not provided.

Metallic enclosures are commonly used to protect electronic circuits against unwanted electromagnetic (EM) interactions. However, these enclosures may be sealed with imperfect mechanical seams or joints. These joints form narrow slots that allow external EM energy to couple into the cavity and then to the internal circuits. This coupled EM energy can severely affect circuit operations, particularly at the cavity resonance frequencies when the cavity has a high Q factor. To model these slots and the corresponding EM coupling, a thin-slot sub-cell model [1] , developed for slots in infinite ground plane and extended to numerical modeling of cavity-backed apertures, was successfully implemented in Sandia's electromagnetic code EIGER [2] and its next-generation counterpart Gemma [3]. However, this thin-slot model only considers resonances along the length of the slot. At sufficiently high frequencies, the resonances due to the slot depth must also be considered. Currently, slots must be explicitly meshed to capture these depth resonances, which can lead to low-frequency instability (due to electrically small mesh elements). Therefore, a slot sub-cell model that considers resonances in both length and depth is needed to efficiently and accurately capture the slot coupling.

Abstract not provided.

Solving dense systems of linear equations is essential in applications encountered in physics, mathematics, and engineering. This paper describes our current efforts toward the development of the ADELUS package for current and next generation distributed, accelerator-based, high-performance computing platforms. The package solves dense linear systems using partial pivoting LU factorization on distributed-memory systems with CPUs/GPUs. The matrix is block-mapped onto distributed memory on CPUs/GPUs and is solved as if it was torus-wrapped for an optimal balance of computation and communication. A permutation operation is performed to restore the results so the torus-wrap distribution is transparent to the user. This package targets performance portability by leveraging the abstractions provided in the Kokkos and Kokkos Kernels libraries. Comparison of the performance gains versus the state-of-the-art SLATE and DPLASMA GESV functionalities on the Summit supercomputer are provided. Preliminary performance results from large-scale electromagnetic simulations using ADELUS are also presented. The solver achieves 7.7 Petaflops on 7600 GPUs of the Sierra supercomputer translating to 16.9% efficiency.

Abstract not provided.

The response of a high - voltage (HV) transformer to fast rise time voltages, such as that from a electromagnetic pulse (EMP) can result in interruption of power distribution and possibly system failure. To help identify these potential occurrences, it is necessary to develop a transformer model that not only captures the input/output response of the transformer but also the internal behavior. The model constructed should cover the frequency band of interest while capturing the internal physical and electrical characteristics. This broad-band, high-fidelity model would enable the prediction of unwanted effects through simulation. A proposed modeling scheme for a HV transformer is described in Part 1 of this report. Part 2 of this report details assessments of internal voltage and electrical field holdoff testing of transformer insulation dielectric breakdown, including comparison of low frequency (DC/60 Hz) holdoff to rise times characteristic of lightning (1 s) and EMP E1 transients (10 - 30 ns). This initial project is a path toward establishing electrical grid transformer failure criteria for EMP voltage transients. We developed modeling methods and measured breakdown electrical field statistical distributions for direct current, 60 Hz, lightning and EMP characteristic voltage rise times. Methods of nanosecond-scale capacitive discharge unit high voltage source development, suggestions for derating of 60 Hz insulation maximum electrical fields for EMP nanosecond pulse voltage withstand rating, and potential methods for increasing transformer resilience to such fast rise time pulses are described.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

This paper describes the implementation of the adaptive cross approximation(ACA) in the method of moments code EIGER. This purely algebraic method provides a mechanism to reduce memory usage and overall computation time. In addition, this work has been targeted for massively parallel platforms to extend the viable frequency range for electromagnetic compatibility and interference problems.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.