Grid-Supported Simulation of Vapour-Liquid Equilibria with GridSFEA I.L. Muntean, E. Elts, M. Buchholz, and H.-J. Bungartz Technische Universit¨at M¨ unchen, Dept. of Informatics, Scientific Computing in Computer Science, Boltzmannstr. 3, 85748 Garching, Germany {muntean,elts,buchholm,bungartz} @in.tum.de http://www5.in.tum.de Abstract. In order to benefit from grid computing, software applica- tions in CSE often need to be substantially modified or rewritten to a large extent. To reduce the required grid know-how and effort the compu- tational scientist (end user and software developer) needs for this task, we developed a framework for engineering simulations in grid environments (GridSFEA). This paper presents two novel features of GridSFEA: the integrated support for parameter investigations and the controlled exe- cution of long-running simulations in grids. They allow the grid enabling of CSE applications with minimal or even without changes of their source code. Furthermore, the overhead for working in grid environments intro- duced by our approach, compared to working on classical HPC platforms, is very low. We provide two examples of using GridSFEA for performing vapour-liquid equilibria (VLE) simulations using Molecular Dynamics and Monte Carlo methods. To develop VLE models, parameter investi- gations are carried out. Large VLE scenarios are computed over a long time, to create test cases for the development of HPC software. Keywords: grid application, grid services, HPC molecular simulation, vapour-liquid equilibria, two-centre Lennard-Jones, polar fluid. 1 Introduction Recently, grid computing environments [1] evolved from research and experimen- tal status towards production systems, providing scientists with access to large aggregated computing and storage resources. Although the scope of such envi- ronments is limited – scenarios of capability computing typically requiring high performance resources at one place – esp. capacity computing applications (e.g., Monte Carlo (MC) simulations, parameter studies) offer a huge potential for the grid. Despite this attractiveness (due to mechanisms for security or access to dis- tributed resources e.g.) of computing grids, they still remain underutilised and underexploited by the computational science and engineering (CSE) community. One reason for this is esp. the tedious development of grid applications and the grid middleware know-how necessary for the scientist (developer) to master. M. Bubak et al. (Eds.): ICCS 2008, Part I, LNCS 5101, pp. 45–55, 2008. c  Springer-Verlag Berlin Heidelberg 2008