PHYSICAL REVIEW APPLIED 14, 054050 (2020) Subcooled Flow Boiling of Carbon Dioxide Near the Critical Point Inside a Microchannel Anatoly Parahovnik , * Mostafa Asadzadeh, Subith S. Vasu , and Yoav Peles Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, Florida 32816, USA (Received 15 March 2020; revised 13 October 2020; accepted 14 October 2020; published 19 November 2020) Subcooled flow boiling heat transfer of carbon dioxide (CO 2 ) at the microscale near the critical pres- sure is experimentally studied. The onset of nucleate boiling and the local heat-transfer coefficient are obtained under 26 experimental conditions, consisting of 17 sample points each, yielding 442 data points. Excess temperatures of about 1 K at the onset of nucleate boiling and heat-transfer coefficients of up to 200 kW/m 2 K are recorded. The heat-transfer coefficient and the onset of nucleate boiling models and correlations developed for low reduced pressures are compared to current experimental results. Modifica- tions to these models are introduced to account for conditions pertinent to reduced pressures around unity. Overall, the correlations do not fully capture current measurements, suggesting that the physics of flow boiling heat transfer near the critical conditions exhibits some unique characteristics. Furthermore, it is observed that the experimentally obtained heat-transfer coefficient is better predicted by a model devel- oped for flow boiling of CO 2 . However, with increasing pressure, deviations between experiments and predictions increase. DOI: 10.1103/PhysRevApplied.14.054050 I. INTRODUCTION Carbon dioxide (CO 2 ) has long been used in various industries, such as the food and beverage industry, the oil and gas industry, and the manufacturing and construc- tion industry. In its supercritical state, it has been recently studied as a potential fluid for Brayton power cycles. Con- currently, a century-long effort revealed many of the char- acteristics of flow boiling heat transfer under a range of flow and thermal conditions. Since early 2000, it has been researched at the microscale. However, thermal engineers have been reluctant to integrate microchannel flow boiling in their systems because of flow instabilities, the critical heat flux (CHF) condition, and inconsistencies between available correlations. Recent results [1] suggest that these issues can be resolved with CO 2 , especially at reduced pressures close to unity. Nevertheless, besides a couple of studies at reduced pressures of up to 0.87 [2–6], not much is known about flow boiling heat-transfer characteristics under such conditions. Flow boiling scientists seek to obtain data and reveal the underlining physics controlling the onset of nucleate boiling (ONB), two-phase heat-transfer coefficient (HTC), and CHF condition. Flow morphologies and a range of other effects, such as flow instabilities, have been meticu- lously documented and modeled. Frequently used models to predict ONB include Hsu’s model [7], the Davis and * tolik@knights.ucf.edu Anderson model [8], and the Lienhard correlation [9]. Numerous correlations and models have been developed to predict the two-phase HTC. Several of the more notable ones include the Shah correlation [10], Kandlikar’s cor- relation [11], and the Cheng correlation [6]. They were developed based on experimental results at lower reduced pressures and applied to boiling numbers that were, for the most part, below those pertinent to CO 2 near the critical condition. Despite many years of research, flow boiling near the critical condition at the microscale has rarely been studied, and data pertinent to heat transfer are missing. There- fore, knowledge about the mechanisms controlling the heat-transfer process under such important conditions is lacking. This study seeks to address this shortcoming by providing experimental data about the onset of nucleate boiling and the heat-transfer coefficient of microscale flow boiling of carbon dioxide near its critical condition. A comparison with the correlations mentioned above that are developed for lower reduced pressures, lower boiling num- bers, and larger channels is made in an attempt to assess their validity for CO 2 near the critical condition. II. EXPERIMENTAL METHODS A. Microfluidic device The current microfluidic device is constructed from two substrates [Fig. 1(a)]. The 0.1 mm high and 2 mm wide microchannel is housed in the top 5-mm-thick piece, and 2331-7019/20/14(5)/054050(11) 054050-1 © 2020 American Physical Society