Session: Government Agency Student Posters

Paper Number: 173513

Nonintrusive Identification of Flow Patterns in Parallel Microchannels via Statistical Analysis of Acoustic Emissions and Pressure Drop 

Flow boiling in parallel microchannels has emerged as a highly effective thermal management strategy for high heat flux applications, including advanced electronics cooling, microreactors, and aerospace systems. The small hydraulic diameters of microchannels enable enhanced heat transfer due to high surface-area-to-volume ratios and short thermal diffusion lengths. When boiling occurs within these confined geometries, it offers the potential for significant heat removal with relatively low coolant mass flow rates. However, the complex, multiphase nature of flow boiling in parallel channels introduces several challenges, particularly related to flow instabilities and non-uniform phase distribution. Understanding and monitoring flow boiling behavior in parallel microchannels are essential to improving the reliability and efficiency of microchannel heat exchangers. Traditional visualization methods for flow pattern identification, however, are often limited by optical access, scalability, and applicability in opaque or high-pressure systems. As a result, there is growing interest in developing nonintrusive diagnostic tools and data-driven modeling approaches to enable robust characterization of boiling phenomena in microscale devices.

This study investigates the statistical behavior of acoustic emissions (AE) and pressure drop as a diagnostic tool for detecting phase change and flow pattern transitions in parallel microchannels. Experiments were conducted across varying flow rates (0.4 – 0.6 LPM) and heater power levels (28V – 48V) at inlet temperatures of 60°C (single-phase) and 95°C (single- and two-phase). The tests leverage a 30 kHz AE sensor, a 10 kHz accelerometer, and a differential pressure transducer for acoustic and pressure sensing and data collection, respectively. The results show that the AE hits per unit time distinguish single and two-phase flow regimes, and AE counts indicate the intensity of the phase change process. Moreover, the pressure drop is also shown to significantly increase at the onset of boiling, making it a reliable indicator of phase transitions. Statistical metrics derived from time-domain analysis, such as the standard deviation, the margin metric, and the pulse metric, effectively captured pressure fluctuations associated with boiling, while frequency-domain analysis using Fast Fourier Transform (FFT) and Short-Time Fourier Transform (STFT) revealed dominant low-frequency oscillations (2 – 10 Hz) linked to flow instabilities. The findings highlight that lower flow rates exhibit greater pressure drop variability due to reduced flow inertia and larger void fractions and compressibility. In contrast, higher flow rates stabilize the boiling dynamics and enhance heat removal. This study demonstrates that AE and pressure drop, combined with spectral analysis, provide a robust framework for real-time monitoring of flow regime transitions in microchannel flow boiling.

Presenting Author: Mohammad Ishraq Hossain University of Arkansas

Presenting Author Biography: Mohammad Ishraq Hossain is a graduate student pursuing his MS in Mechanical Engineering from Fall 2024. He is from Bangladesh and graduated from the Islamic University of Technology in 2023. His current research centers on liquid-vapor phase-change processes for electronics cooling, specifically, pool boiling at subatmospheric pressures and acoustic sensing of flow boiling.

Authors:

Mohammad Ishraq Hossain University of Arkansas
Stephen Pierson University of Arkansas
Ying Sun University of Cincinnati
Todd Kingston Iowa State University
Han Hu University of Arkansas