Simulation of partial differential equations models in Java. Issue 3 (2nd May 2017)

• Record Type: Journal Article
• Authors: Cano, María José; Chacon-Vera, Eliseo; Esquembre, Francisco
• Is Part Of: Engineering computations. Volume 34:Issue 3(2017)
• Page Start: 800
• Page End: 813
• Publication Date: 2017-05-02
• Subjects: Engineering -- Java -- Simulations -- Partial differential equations -- Physical sciences; Computer-aided engineering -- Periodicals; Computer graphics -- Periodicals; 620.00285
• DOI: 10.1108/EC-05-2015-0111
• Languages: English
• ISSNs: 0264-4401

Abstract

Abstract : Purpose: Computer simulations improve the knowledge of physical models and are widely used in teaching and research. Key aspects are to understand their solutions and to make interactive changes to the models, observing their effects in real-time. The drawback of creating interactive simulations of physical models is the high level of programming expertise required. The purpose of this study is to facilitate this task. Design/methodology/approach: Java is the perfect language for this task; it yields high-quality graphics and is widely spread in the scientific community. Because many important physical models are described by means of partial differential equations (PDEs), the combination of Java with FreeFem++, a C++ PDE solver based on the finite element method, is considered. Findings: In this study, a Java library is introduced to numerically solve PDE equations via a run-time connection with FreeFem++. The solution is encapsulated into Java objects that are ready to be used in different programming tasks. The library also includes new Java visualization elements for solutions and meshes in the context of the Open Source Physics project library. Together, the connection features and the visualization elements facilitate the creation of Java simulations by programming researchers. For those with less programming capabilities, this work has been included into Easy Java Simulations, a tool to further ease the creation of interactive simulations. Originality/value: The present study approach allows simulating models given PDEs. The equations are solved either in local or in remote mode (e.g. by a network accessible to a high-performance computer) and visualized locally, providing a high degree of interactivity to the end user.