Non-intrusive flow regime detection in narrow rectangular channels using gauge pressure based pseudo acoustic analysis

• A gauge-pressure-based pseudo-acoustic approach (P-SPL) was developed without hydrophones. • P-SPL quantitatively distinguishes boiling-induced pressure fluctuations from conventional SPL. • High-speed imaging and image-based void fraction validation confirmed reliable regime detection. • The method successfully identified slug-flow transitions in narrow rectangular channels. • The non-intrusive and real-time capable technique is applicable to SMR and passive system monitoring. When boiling occurs in nuclear fuel rods, identifying the flow regime becomes critical, as excessive bubble formation can deteriorate heat transfer performance and induce flow instability. Therefore, non-intrusive monitoring methods are essential for ensuring the safety of nuclear fuels and reactor systems. The Pseudo-Sound Pressure Level (P-SPL), although not a direct measurement of acoustic SPL, was employed to classify flow regimes based on SPL principles, using inlet gauge pressure fluctuations. In the bubbly flow regime, the onset of flow instability (OFI) caused the P-SPL values to fluctuate beyond the average error range, regardless of inlet temperature. When the void fraction exceeded the 95 % Confidence Interval (CI), P-SPL increased proportionally—and vice versa. In the slug flow regime, backflow was observed. As the void fraction increased, the mass flow rate and the pressure dropped across the test section sharply decreased, and the P-SPL value exceeded the upper bound of the 95 % CI. These results indicate that P-SPL can be used to monitor transient variation in void fraction. Flow regimes across all power levels were identified using the total P-SPL. At the OFI, the total P-SPL increased from the previous power level to 69.4 dB (T in : 30 ℃) and 75.4 dB (T in : 60 ℃). The total P-SPL sharply rose to 89.0 dB (T in : 30 ℃) and 92.6 dB (T in : 60 ℃), attributed to the presence of large vapor structures, intermittent liquid slugs, and intense flow disturbances that generate broadband pressure fluctuations. Therefore, distinguishing flow regimes based solely on pressure signals—without the use of acoustic sensors—was found to be a sufficiently effective approach.