Excessive Memory Usage of the ELLPACK Sparse Matrix Storage Scheme throughout the Finite Element Computations

2014-12-01
Akinci, Gokay
YILMAZ, ASIM EGEMEN
Kuzuoğlu, Mustafa
Sparse matrices are occasionally encountered during solution of various problems by means of numerical methods, particularly the finite element method ELLPACK sparse matrix storage scheme, one of the most widely used methods due to its implementation ease, is investigated in this study. The scheme uses excessive memory due to its definition. For the conventional finite element method, where the node elements are used, the excessive memory caused by redundant entries in the ELLPACK sparse matrix storage scheme becomes negligible for large scale problems. On the other hand, our analyses show that the redundancy is still considerable for the occasions where facet or edge elements have to be used
RADIOENGINEERING

Suggestions

How Generous is the ELLPACK Sparse Matrix Storage Scheme for Finite Element Computations
Akinci, Gokay; YILMAZ, ASIM EGEMEN; Kuzuoğlu, Mustafa (2012-11-03)
Sparse matrices are occasionally encountered during solutions of various problems by means of numerical methods, such as the finite element method. ELLPACK sparse matrix storage scheme, one of the most widely used methods due to its implementation ease, is investigated in this study. The scheme uses excessive memory due to its definition. For the conventional finite element method, where the node elements are used, the excessive memory caused by redundant entries in the ELLPACK sparse matrix storage scheme ...
Questioning Degree of Accuracy Offered by the Spectral Element Method in Computational Electromagnetics
Mahariq, I.; KURT, HAMZA; Kuzuoğlu, Mustafa (2015-07-01)
In this paper, a comparison amongst the spectral element method (SEM), the finite difference method (FDM), and the first-order finite element method (FEM) is presented. For the sake of consistency, the comparison is carried out on one-dimensional and two-dimensional boundary value problems based on the same measure of error in order to emphasize on the high accuracy gained by the SEM. Then, the deterioration in the accuracy of the SEM due to the elemental deformation is demonstrated. Following this, we try ...
Multiscale Modeling of Thin-Wire Coupling Problems Using Hybridization of Finite Element and Dipole Moment Methods and GPU Acceleration
ÖZGÜN, ÖZLEM; Mittra, Raj; Kuzuoğlu, Mustafa (2020-01-01)
In this article, a hybrid numerical method, called finite element method (FEM) + dipole moment (DM), is presented for efficient solution of multiscale electromagnetic radiation and scattering problems that involve structures with fine features, such as thin-wire antennas or objects. In this method, the FEM is hybridized with the DM approach to help ease certain computational burdens, such as mesh refinement, ill-conditioning, memory overload, and long computation times, when solving multiscale problems with...
MUTUAL COUPLING EFFECTS OF FINITE RECTANGULAR PHASED-ARRAYS
YAVUZ, H; BUYUKDURA, OM (1994-04-14)
A rigorous integral equation formulation for the analysis of a phased array of flangemounted waveguide apertures is given for a finite number of elements and nonuniform spacings. The resulting set of ihtegrd equations is reduced to a matrix equation called the coupling matrix which relates the coefficients of all the modes in all the waveguides to one another. The solution then yields the dominant mode reflection coefficient, coefficients of scattered modes and hence the field in each waveguide. The blockTo...
Fixed-frequency slice computation of discrete Cohen's bilinear class of time-frequency representations
Ozgen, MT (2000-02-01)
This communication derives DFT-sample-based discrete formulas directly in the spectral-correlation domain for computing fixed-frequency slices of discrete Cohen's class members with reduced computational cost, both for one-dimensional and multidimensional (specifically two-dimensional (2-D)) finite-extent sequence cases. Frequency domain integral expressions that define discrete representations are discretized to obtain these discrete implementation formulas. 2-D ambiguity function domain kernels are chosen...
Citation Formats
G. Akinci, A. E. YILMAZ, and M. Kuzuoğlu, “Excessive Memory Usage of the ELLPACK Sparse Matrix Storage Scheme throughout the Finite Element Computations,” RADIOENGINEERING, pp. 997–1004, 2014, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/53118.