<title>Concurrent simulation of xerographic imaging subsystems</title>

Abstract

This paper presents first principles simulations of xerographic imaging where photoreceptors are exposed and discharged to result in a spatial image charge distribution on the surface. The exposure stage involves the solution of the Maxwell equations for wave propagation through the lossy transport layer. The real component of the Poynting vector is then calculated to estimate the intensity of illumination. A direct boundary integral equation method (BIEM) is used to formulate and solve the coupled Helmholtz equations for the tangential component of electric field and its normal derivative. The discharge stage requires consideration of both charge conservation and current continuity. A hybrid BIEM-MOC (boundary integral equation method - method of characteristics) algorithm is used to solve for the electrostatic fields, and to track the space charge migration. Both steady- state and time-transient cases may be treated. Electron-hole pair generation is controlled by a field-dependent quantum generation efficiency. Computed results include: discharged surface voltage, contours of potential through the photoreceptor cross-section, and adjacency effects at the edges of exposed lines. This simulation is implemented within a distributed computing environment that supports interactive steering, interactive browsing and visualization, concurrent processing, iconic assembly of dataflow networks, and multi-level (process-device) simulation. Parallel Virtual Machine, a public domain parallel programming software toolkit from Oak Ridge National Laboratory, is used to attain desktop supercomputing level performance by harnessing the computational power of a heterogeneous cluster of Unix workstations into a single loosely coupled parallel computing resource.