Sustainability through flexibility: Building complex simulation programs for distributed computing systems

• We propose a methodology for building complex simulation programs for distributed computing systems.
• A software library specifically designed to support a sustainable development of simulation program components is presented.
• An application example in mechanical engineering is used to demonstrate the approach.

Complex simulation programs in science and engineering are often built up from a diverse set of existing applications. The large variety of application codes and their high computational demands lead to an increasing utilization of distributed computing systems. Furthermore, the need for developing sustainable simulation programs, especially with regard to ever increasing data sizes, requires a profound flexibility such that application codes and hardware resources can be easily replaced or extended. In this article, we propose a methodology for building complex simulation programs for distributed computing systems. A software library specifically designed to support a client–server-based development of simulation program components is presented. An application example for the simulation and optimization of lightweight structures in mechanical engineering is used to demonstrate the approach.