Benchmarking of two flexible multibody dynamic simulation software in engine simulations
In this paper, two different commercial multibody dynamic (MBD) simulation software cases are studied. Due to the restrictions determined in the conditions of contract, the names of the software are not revealed, instead being called Software S and Software E. The central purpose of this research was to investigate the abilities of Software S in the simulation of a large engine, as a part of the strength analysis process. The abilities were studied by comparing the program with another, here
... th another, here called Software E, which is designed primarily for engine simulations. The capabilities of Software E have been proven after years of usage at Wärtsilä, resulting in its essential role in the strength analysis process today. The aim was to find the shortcomings and restrictions of Software S but also advantages it could bring to the strength analysis process for Wärtsilä. Similar simulation models were also built using both programs during this research. A 16-cylinder V-engine was selected as the subject because of its size in order to obtain further information about the behavior of the program when working with extensive model files. The components of the engine were flexible and were reduced FE models, also called super elements. The forces and contact situations that occur inside the engine were modeled using elements provided by the MBD programs. Different levels of detail of the modeling elements were used to obtain information about the flexibility of the program. The results obtained from time integrations were compared to ensure the similarity of both modeling elements used. Also, this paper reports the calculation times. In addition, a small-scale study was performed for Software S to clarify the effect of the modes used in time integrations towards results accuracy and calculation times. Simulation models were built successfully in both programs, and the results obtained correlated with each other on an adequate level. Significant differences appeared in the features and usability of the programs in general. The GUI of Software S is advanced and user-friendly, whereas Software E is not focused on these features. On the other hand, the modeling element library of Software E covers all of the required features related to large engine simulations, some of which Software S is lacking. This work can be used in assistance when considering buying new software for a company as well as when investigating new development areas that could be improved with new software.