In this contribution we conjecture that systems of linearized Partial Differential Equations, viewed as consistent models for physically realizable systems, are diagonalizable. While this property is interesting for its own sake, our discussion will focus on technicalities and implications for advanced accelerated computing. We will demonstrate that diagonalization with respect to a chosen spatial coordinate systematically "reshuffles" the original PDEs and creates equivalent differential forms. It turns out that diagonalization automatically distinguishes the variables in the interface- and boundary conditions defined on surfaces normal to the diagonalization direction. © Springer-Verlag Berlin Heidelberg 2003.
CITATION STYLE
Baghai-Wadji, A. R. (2003). A symbolic procedure for the diagonalization of linear PDEs in accelerated computational engineering. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2630, 347–360. https://doi.org/10.1007/3-540-45084-x_17
Mendeley helps you to discover research relevant for your work.