HPCL: integrated software for parameter characterization and mechanical properties of rubber nanocomposite

Abstract

This paper presents the integrated scientific software called High Performance Computing Library (HPCL) as a breakthrough strategy for predicting, visualizing and observing some parameter characteristics and mechanical properties of rubber nanocomposite for advance polymer processing. The function and use of HPCL excessively to manage a big data set and organized the computational library related to mechanical parameters of composite material searchable through a user-friendly web-based system. This software will involve some mathematical modeling, numerical simulations and high performance computing technology to improve the accuracy of prediction, quality of resolution and performance measurements. The parameter characterization of rubber nanocomposite will focus on heat flow, rubber elasticity, stress and stiffness, density and pressure distribution, stress-displacement relation, phase change condition, cooling and heating processes, solid-liquid interface, time step and dimensional of the targeted area. The implementations of parallel algorithms for investigating the parameter changes involving Partial Differential Equation (PDE) are based on large sparse parabolic and elliptic-Poisson types. PDE models are efficiently used to govern the complex dynamic systems for rubber nanocomposite processing. The discretization technique to obtain a large sparse linear system of equations of PDE is based on Finite Difference Method (FDM). A set of mathematical modeling are the computational engine for HPCL visualization. The HPCL software supports high performance computing of numerical simulation library and repository on distributed parallel computing system (DPCS) using multiple computers communicating over the local area network. The contribution of modeling and simulation involve the algorithm provider, organizing process by scientific librarian and application by user. The sequences and parallel programs are supported by Parallel Virtual Machine (PVM) and C language on DPCS based on Linux operating system. The performance measurement will be investigated by run time, accuracy, convergence, errors, speedup, efficiency and effectiveness. As a conclusion, the integrated of HPCL software will be an alternative software system to speed up the large sparse computational complexity and optimize the visualization quality for predicting, visualizing and observing some parameter characteristics of the unique combination of rubber composite to save resources, time consuming and space allocation during the injection molding process, melt-molding process, assemble process of rubber nanocomposite material and manufacturing the assembled product. The validation of mathematical modeling and simulation of HPCL is compared to the real data set from Malaysia Rubber Board (LGM).