Predicting thermal instability in a closed loop pulsating heat pipe system

Mathematical models for a closed loop pulsating heat pipe (CLPHP) with multiple liquid slugs and vapor plugs are presented in this study. The model considers the effect of thermal instability in different sections of a CLPHP at different operational conditions. Based on a neural network, an approach of nonlinear autoregressive moving average model with exogenous inputs (NARMAX) can be applied to the thermal instability of CLPHP. This study approximates the nonlinear behavior of CLPHP with a linear approximation method that can establish the relationship among the response temperature differences between evaporator, adiabatic, and condenser sections. A multi-input single-output (MISO) strategy is adopted in this study to approximate nonlinear behavior of CLPHP. The predicted results show that the effect of the three sections to vapor condensation could be precisely distinguished; meanwhile, thermal performance of CLPHP would be predicted. The development of nonlinear identification technique will be helpful to optimize and understand the heat transfer performance of thermal instability in the different designs of CLPHP.