Model Accuracy of Variable Fidelity Vapor Cycle System Simulations 2014-01-2140

As the cost and complexity of modern aircraft systems advance, emphasis has been placed on model-based design as a means for cost effective subsystem optimization. The success of the model-based design process is contingent on accurate prediction of the system response prior to hardware fabrication, but the level of fidelity necessary to achieve this objective is often called into question. Identifying the key benefits and limitations of model fidelity along with the key parameters that drive model accuracy will help improve the model-based design process enabling low cost, optimized solutions for current and future programs. In this effort, the accuracy and capability of a vapor cycle system (VCS) model were considered from a model fidelity and parameter accuracy standpoint. A range of model fidelity was evaluated in terms of accuracy, capability, simulation speed, and development time. Inability to implement control strategies and reliance on external sources for operating assumptions were seen as significant limitations of the lower fidelity models. Characteristics of the highest fidelity model were noted for their ability to accurately capture the dynamic nature of the hardware data through physics based modeling and control implementation, but at a cost in development time and simulation speed. In addition to characterization of model accuracy and capability, design of experiment results demonstrated that parameters, such as refrigerant charge, compressor efficiency, and heat exchanger heat transfer coefficients are more significant to model accuracy than parameters such as pipe length or heat exchanger dimensions.