Srinivasan K

This paper presents the experimental investigation of acoustic spectral variations in saturated pool boiling regimes of water on a heated wire. In the current study, two different wires of standard wire gauge (SWG) 36 and 42 are considered to investigate the acoustic spectral variations through the pool boiling curve. The amplitude and frequency changes are evaluated for each regime of pool boiling. In a single regime, amplitude rise is observed with respect to the heat flux without any significant change in dominant frequencies. On the other hand, frequency shifts are observed in regime transitions. A change in the diameter of the heater wire has no significant effect on the boiling acoustic spectra. However, the number of high-frequency components increased for the SWG – 42 than the SWG – 36 wire. A frequency peak near 2000 Hz is found to be crucial for boiling regime identification. The sound pressure level (SPL) for SWG – 36 is higher than the SWG – 42, and it is further noted that SPL follows an ‘N’ shaped pattern for both wires owing to the frequency shifts and variation of mean bubble departure diameter at that heat flux.

Ever-growing miniaturization of electronic devices and space-conserving endeavours of heat transfer systems pose a challenging task to the current cooling strategies. Boiling acoustics is one of the most potent and efficacious methodologies to reliably predict the boiling regime inside the cooling systems to assay the safety and address the emergency conditions. Boiling acoustics is gripping attention out of the few foreseeable technologies for the future cooling requirements. Though the potentiality of boiling acoustics was unravelled in the late 1990s, the research data present in this arena is however lacking. This paper presents a comprehensive review of the literature reported from the 1970s to the present date. Furthermore, this paper also details the evolution of boiling acoustics from the initial application of boiling incipience detection to regime identification. Much focus is given on salient features of boiling acoustic characterization dealing with the regime detection. Effects of various parameters such as thermos-physical properties of the heater surface and the boiling liquid that directly or indirectly influence the acoustic spectra are also presented. The prediction of the boiling regime constitutes the first necessary step in producing autonomous cooling systems. Hence, the detection and characterization of boiling noise under various conditions such as pressure, heat flux, and flow rate is essential.

The present study focuses on the performance of the machine learning methods in classifying the boiling regimes of water up to critical heat flux conditions based on the acoustic characteristics of boiling. The data set is generated by conducting a pool boiling experiment on a wire heater at various heat fluxes varying from 54.95 kW/m2 to 2898.67 kW/m2. A Kanthal D wire of standard wire gauge 36 is used. The data set is divided into three classes: no boiling, nucleate boiling, and critical heat flux to identify the boiling incipience and critical heat flux. Much focus is insisted on identifying critical heat flux as it carries more practical importance in the safety of the cooling systems. Data set size optimization is performed to find the lowest number of records required for each method. Three machine-learning methods are employed to predict the boiling regime, namely, binary decision tree method, decision tree ensemble method and naive Bayes method. Out of these, the decision tree ensemble outperformed the binary decision tree and naive Bayes classifiers. The decision tree ensemble classified the regimes in the given data with the lowest classification error and inference time. The accurate classification of boiling regimes based on boiling acoustics strengthens the safety measures in real-time monitoring of cooling systems.