Dataset of On-tube local condensation heat transfer coefficient of R134a, R513A and R450A on smooth stainless-steel tube at saturation temperature of 35 and 40℃

Published: 22 November 2024| Version 1 | DOI: 10.17632/8zfbfj4t7m.1
Contributors:
Michael Akinseloyin,

Description

These data contain local condensation heat transfer coefficients of three refrigerants (i.e., R134a, R4513A and R450A) on a single, horizontal, smooth stainless tube at saturation temperatures of 35 and 40℃. A custom experimental apparatus was constructed to measure on-tube condensation heat transfer coefficients which consisted of two main parts, the chamber and the condensation tube. Refrigerant was vaporized in the experimental chamber and then condensed on the condensation tube. Two recirculating chillers were connected to the experimental chamber. One supplied cooling water through the condensation tube and the other supplied water used to boil the liquid refrigerant in order to achieve saturation temperatures of 35 or 40℃ . The refrigerant boiled and subsequently condensed on the smooth condensation tube. Thermocouples were installed in the flow to measure temperature at the inlet and outlet of the tube. The condensation tube was machined using electron discharge machining and 1.016-mm-diameter thermocouples(T1-T8) were installed in the tube wall in order to measure heat flux through the wall at two radial positions, β=11o from the vertical and β=109o from the vertical. Four thermocouples (T1-T4) were used to compute heat flux at β=11o, and an additional four thermocouples (T5-T8) were used to compute heat flux at β=109o. Saturation pressures equivalent to 35 ± 0.5℃ and 40±0.5℃ were monitored in LabVIEW. Steady state was attained when the vapor pressure and temperature variation across the wall were in the range of ±0.34 kPa and ±0.05℃. After steady state was reached, experimental data were recorded for five minutes. Each experimental datapoint is the average of data collected within five minutes, and they include the eight thermocouple wall temperatures, mass flow rate of cooling water, temperature inlet and outlet of cooling water, saturation temperature and pressure of refrigerants. The outer wall temperature varied around the tube, and the measured wall temperatures were used to compute the outer wall temperatures and temperature gradients at radial positions around the tube using a natural logarithm curve fit with the measured wall temperatures as input. These values were used to compute heat fluxes and local heat transfer coefficients were computed using thermal conductivity, heat flux, saturation temperature (determined from the measured saturation pressure) and wall temperatures. Uncertainties of local heat transfer coefficients for each experimental trial were calculated using the propagation of uncertainty method and uncertainties of the instrumentation.

Files

Steps to reproduce

The main data collection program used for this experiment was LabVIEW, integrated with a data acquisition hardware (cDAQ-9185), including a chassis that contained the voltage input (NI-9205) and thermocouple input (NI-9213) modules for collecting measurements from the pressure transducer and thermocouples, respectively. Wall temperatures of the condensation tube, temperature of cooling water inlet and outlet and temperature and pressure in the chamber were measured and displayed in LabVIEW. The inlet water mass flow rate was measured using a Coriolis mass flow meter connected to the pump outlet of the recirculating chiller(condenser). The Coriolis mass meter was integrated with ProlinkIII software to display the mass flow rate, volume flow rate, density and other flow properties of the inlet water through the condensation tube. After attaining steady state(i.e., when the variation in pressure was within ±0.34 kPa and thermocouple readings were steady), data from LabVIEW and the ProlinkIII were recorded simultaneously for five minutes, then the insulation to the two sight glasses were opened to visualize the condensate while the system was running at steady state conditions. A point light source was placed at one sight glass and an iPhone 14 pro max camera at the other to take visualization images and videos. To prevent condensation on the sight glass, a heat gun was pointed at it for about three minutes before taking the images. For data reduction, the data from LabVIEW and Prolink III were averaged using Excel and exported to Engineering Equation Solver (EES). The equations, such as Nusselt on-tube condensation heat transfer correlation, propagation of uncertainties, used for computing the local heat transfer coefficients, heat flux, subcooling, etc. were solved in EES. The experiments were validated via the Nusselt correlation for on-tube condensation and energy balances. To get the next datapoint, the insulation was replaced, cooling water mass flow rate, evaporator and condenser temperatures were varied and the chamber conditions were adjusted to attain the saturation pressure equivalent to the saturation temperature of 35 ± 0.5℃ and 40±0.5℃ for each refrigerant.

Institutions

Kansas State University

Categories

Heat Transfer, Thermal Engineering, Condensation, Thermofluids, Alternative Refrigerant, Global Warming Potential

Funding

Office of Naval Research

N00014-22-1-2328

Licence