Using collaborative computing technologies to enable the sharing and integration of simulation services for product design

Hongwei Wang, Heming Zhang
2012 Simulation modelling practice and theory  
Collaborative Simulation (MCS) is often used to integrate multiple computational models and simulation tools, as well as to support the collaboration between members of a development team. MCS is a systematic approach to developing and performing accurate, effective, and efficient simulation, which emphasizes that system design should be optimized as a whole by taking into account the objectives and constraints from multiple disciplinary areas. For instance, simulations for mechanical
more » ... design and control system design are generally run together to identify the issues involved in the overall design [3] . Early work started in the 1990s and was focused on using advanced virtual prototyping simulation for design of mechanical systems [4] . In the context of mechatronic product design, MCS essentially involves constructing the mathematical models for a complex system and solving them by using one or more numerical solvers. Commercial CAE tools and bespoke simulation packages are widely applied in industry and as such pave the way for MCS implementation. There are two main ways for implementing MCS, namely the modeling using common language and the modular approach. The former focuses on describing the subsystems in a MCS problem using a common language from which system equations can be generated and solved using a single solver, see for example [3, 5] . The latter, on the other hand, offers much flexibility by utilizing multiple tools to create models and integrating these models at simulation runtime, see for example [1, [6] [7] [8] [9] . The advantages of the former mainly include no need for system integration, high accuracy achieved by using a single solver. The latter has advantages of modularity, flexibility, and better support for collaboration. Although system integration needs to be addressed for the modular approach, it is still a good choice for the MCS for product design. First, modular modeling has been identified as one of the requirements for modern simulation environments [10]. Second, it well supports distributed collaboration and integration of models created using different languages or tools. Last but not least, designers working on different disciplines tend to have diverse preferences (e.g. 3D modeling, diagrams, and human-in-the-loop) on the simulation tools and this diversity is difficult to be addressed by a single tool. It is therefore useful to develop an effective and efficient collaborative environment in which MCS can be undertaken by integrating models created for different disciplines and distributed on the Internet. The development of network and communication technologies opens the opportunity for supporting system integration and group collaboration for engineering applications [11] . For instance, many middleware technologies such as the Web, Web Services, the High Level Architecture (HLA), and semantic Web have been used for developing Web-based simulation environments [12] . These standards and technologies supply good support for the development of MCS environments. In this work, we focus on developing a MCS environment for product design and as such there are two main requirements raised in the first place. Firstly, this environment is aimed at supporting the collaborative development of distributed teams. This kind of collaboration can be either between members of a same organization or between teams from different organizations. In the latter case, source codes and model files may need to be kept confidential and only simulation results are allowed to be shared. Secondly, this environment should support the solving of equations in parallel and thus the runtime interaction between computational models also holds the key to system design and implementation. This requirement is overlooked in previous research [1, 6, 13, 14] as the focus was on enabling the transfer of simulation data. This research is motivated by these issues and aims to go a step further beyond traditional Web-based simulation by developing a service-oriented paradigm in which simulation services for product design can be shared, found, and integrated. The remainder of this paper is organized as follows. In Section 2, relevant research work is reviewed. In Section 3, MCS is analyzed in detail to achieve an understanding of the problem, as well as to provide insights into the development of simulation environments. In Section 4, the development of MCS system, in particular how collaborative computing technology is employed, is discussed. In Section 5, different ways for the runtime interaction between simulation models are discussed and an effective approach for interaction control is described. In Section 6, an engineering example is discussed to undertake preliminary evaluation on the proposed solution and the prototype system implemented. Finally the conclusions of this work are given in Section 7.
doi:10.1016/j.simpat.2012.05.002 fatcat:vfxzyfwavnhsxdsgjjadpvihii