Development of a virtual EXV flow sensor for applications with two-phase flow inlet conditions

• A virtual EXV flow sensor for applications with two-phase inlet conditions is developed.
• The sensor is based on the nondimensional PI-group approach.
• The sensor correlations can be applied under both, single- as well as two-phase inlet conditions.
• The correlations were trained with half of the available data.
• The paper reports both, the coefficients and the performance of the correlations for the entire dataset.

In conventional vapor compression systems, electronic expansion valves (EXVs) are used for refrigerant flow control. In addition, the EXV control signal can be used along with a model and measurements of boundary conditions to provide a virtual sensor for refrigerant mass flow rate as outlined by Kim and Braun (2012). For typical single-phase inlet conditions, models for characterizing the relationship between the inlet and outlet conditions, the opening, and the mass flow rate have been extensively studied, e.g. by Park et al. (2007), and semi-empirical as well as empirical correlations have been developed. However, two-phase inlet conditions can exist at low refrigerant charge, or for certain applications such as when balancing valves are used in a hybrid control scheme as proposed by Kim et al. (2008). This paper presents a continuous correlation that can be used for both two-phase and subcooled valve inlet conditions.