The presence of a sudden expansion generates a variation of the static pressure commonly called Pressure Recovery (PR). In this paper, we made firstly an extensive literature survey to list existing gas-liquid two-phase flow pressure recovery models and to collect an experimental database. Thus, a total of 305 data was collected from 6 recent works and 18 predictive models was identified. An analysis of the different existing models was carried out firstly. Then, the predictive capability of nine existing models was assessed using the collected database. It was reported that none of the models can predict the experimental results for a large range of experimental conditions. This finding highlighted the necessity to propose a new model. The proposed predictive model was developed using the two-phase multiplier and mass quality. These two parameters were correlated using 157 data points from the collected database, while the other data was used to validate it. It was found that the proposed model gives better predictions compared to existing ones in the range of conditions and parameters of the experimental database used in this analysis.

The presence of a sudden expansion generates a variation of the static pressure commonly called Pressure Recovery (PR). In this paper, we made firstly an extensive literature survey to list existing gas-liquid two-phase flow pressure recovery models and to collect an experimental database. Thus, a total of 305 data was collected from 6 recent works and 18 predictive models was identified. An analysis of the different existing models was carried out firstly. Then, the predictive capability of nine existing models was assessed using the collected database. It was reported that none of the models can predict the experimental results for a large range of experimental conditions. This finding highlighted the necessity to propose a new model. The proposed predictive model was developed using the two-phase multiplier and mass quality. These two parameters were correlated using 157 data points from the collected database, while the other data was used to validate it. It was found that the proposed model gives better predictions compared to existing ones in the range of conditions and parameters of the experimental database used in this analysis.