The void fraction is a basic parameter to characterize the two-phase flow. Accurate measurement of its value is of great significance for industrial production. In this study, an optical correction system combined with software and hardware is developed. Stereoscopic particle image velocimetry (SPIV) technology is used to measure the void fraction in gas/liquid two-phase stratified flow. The measurement results are compared with those of 2D particle image velocimetry under the same conditions, and the correction effect appears to be reasonably good. The void fraction value is obtained by the image processing of the corrected image. The accuracy of the measurement results is verified by comparing the experimental liquid-phase volumetric flow rate with that measured through a standard flowmeter. Twenty types of void fraction prediction models are evaluated using the mean absolute relative deviation (MARD), root mean square error (RMSE), and gray correlation analysis methods. The results suggest that the Armand (1946) model has the smallest MARD, RMSE value, the largest correlation degree (Ri) value, and the best prediction performance. On this basis, an improved model is proposed by fitting the relationship between void fraction and gas volume holdup.

You can access this article if you purchase or spend a download.