NUCLEATE POOL BOILING OF LIQUIDS pHN AND THEIR MIXTURES AT SUBATMOSPHERIC PRESSURES

Abstract

The present investigation pertains to the experimental research work related to the nucleate boiling heat transfer from a horizontal 410 ASIS stainless steel cylinder to the pool of saturated liquids, and to their binary liquid mixtures both at atmospheric and Bubatmospheric pressures. The pure liquids used for the investigation are distilled water, ethanol, methanol and isopropanol, and the binary liquid mixtures having varying concentrations of ethanol-water, methanol-water and isopropanol-water mixtures. The heat flux ranges from 9,618 W/m to 31,354 W/m2 and the system pressure from 25.33 kN/m to 98.63 kN/m2. Since this investigation aims to obtain experimental data for the pool boiling of pure liquids and their binary mixtures, an experimental facility was carefully designed and raised. The experimental set-up includes provisions for the measurement of concentration of the binary liquid mixtures, electrical energy input to the heating surface, pressure over the liquid pool and temperatures of the heating surface and the boiling liquid. The copper-constantan thermocouples measure the temperatures of the heating surface and the boiling liquid. The heating surface temperature is 11 measured circumferentially at the top-, the side- , and the bottom- positions at a given plane. The specially home-made travelling thermocouple probes measure the liquid bulk temperature at the three locations corresponding to the surface thermocouple positions. The surface temperature is corrected by subtracting the temperature drop across the wall thickness. From the readings of the corrected surface and the corresponding liquid tompcratum*,local values of At are calculated for the top-, the side-, and the bottom- positions of the heating surface. Using the 'mechanical quadrature' technique,the average values of aT are obtained to calculate average heat transfer coefficient, h over the circumference. The concentration of the boiling binary liquid mixture, X is determined by drawing the liquid sample from the liquid sampling unit and then comparing its refractive index with the calibration curve. The refractrometer used was supplied by li/s Carl Zeiss Jena Co., West Germany. The liquid concentration is checked at several intervals of time during a given test run for a given mixture composition. The concentration in the vapour phase, Y in equilibrium with the liquid phase concentration, X is obtained from the literature. The experimental data for the pool boiling of pure liquids at atmospheric as well as at subatmospheric pressures corroborate the validity of the well-established iii relationship between the heat transfer coefficient and 0 7 the heat flux for high pressures^.e.,h a q . However, the relationship between the boiling heat transfer coefficient and the pressure for the subatmospheric pressures differs from that at high pressures. In fact, the boiling heat transfer coefficient varies with the pressure raised to the power of 0.32 for the data conducted at subatmospheric pressures, i.e. h a P0-'2. The heat transfer data for the boiling of ethanol, methanol and isopropanol do not deviate amongst themselves, whereas they differ considerably from those of distilled water. The experimental data for the pool boiling of pure liquids as used in this investigation and those of earlier investigators conducted on widely differing heating surfaces for the liquids possessing differing physico-thermal properties for subatmospheric pressures are correlated by the following equation within + 15 per cent deviation : h* _ ( P. n0.32 h1 1 where E* ^(h/q0,7)Represents a ratio of average heat transfer coefficient to heat flux raised to the power of 0.7, and P is the system pressure. The subscript, 1 corresponds to 'reference1 pressure for which the value iv of h-, is known for a given liquid and heating surface. However, in the present investigation the 'reference' pressure chosen is one atmosphere. With the knowledge of h? and P, , the above correlation readily determines the value of E at any subatmospheric pressure for the same boiling liquid and the heating surface. Further, the above correlation is useful to check the consistency of boiling heat transfer data for a given liquid and heating surface at subatmospheric as well as atmospheric pressures. Since this correlation is for the data conducted for different liquids on the heating surfaces possessing differing surface characteristics at subatmospheric pressures, an implication of this is that the effect of the surface-liquid combination is the same for all the pressures, P <<C 1 atmosphere. It is important to note that the data for the pool boiling of liquids at high pressures could not be correlated by a correlation of the aforesaid type. This is due to the fact that the effect of surface-liquid combination is not the same for all the pressures, P > 1 atmosphere. The experimental data of binary liquid mixtures for subatmospheric pressures on a given heating surface 0.7 are also correlated by the relationships : h a q and h a P° which are applicable for the boiling of pure liquids. The data analysis of binary liquid mixtures shows that they are satisfied by the following correlation within + 15 per cent like for pure liquids: E* /JL%0.32 hl 1 where the terms have their same meaning as described for the correlation for the pure liquids. The addition of more volatile component to the water shows that the boiling heat transfer coefficient of the binary liquid mixture decreases upto a certain concentration, beyond which it increases. The concentration at which the heat transfer coefficient is minimum corresponds to a maximum value of [Y-X]. It is 31.10 wt. per cent ethanol, 30.80 wt. per cent methanol, and 22.5 wt. per cent isopropanol for ethanolwater, methanol-water and isopropanol-water mixtures respectively. This behaviour is shown at all the subatmospheric pressures studied. It may be noted that the actual heat transfer coefficient for any concentration of the binary liquid mixtures studied is less than the weighted heat transfer coefficient calculated from the heat transfer coefficients of the mixture in their pure states and the concentration of the mixture. This is a consistent behaviour for all the pressures investigated. The experimental data of all the binary liquid mixtures studied lead to correlations within + 15 per cent as follows : vi (a) For the values of x' ; 0 < x' <C 22.0 S*(Pfnl)0.32 = 5>70xl0-2u'r0.60 NU (b) For the values of X ; 30.0 ^ X < 78.0 m£)0M- 2.51xlO-4(X,)°'90 In the above equations Nu represents the average value of the normalised Nusselt number given by the quantity e* N where a is the surface tension; k, (,°jf ",°v)g the thermal conductivity of the boiling mixture; py , the liquid density and p , the vapour density. P represents the system pressure; P-,,the 'reference' pressure and X , the wt. per cent of more volatile component in the liquid phase. These correlations provide a procedure for calculating the boiling heat transfer coefficient of a binary liquid mixture for the aforesaid concentrations, X at subatmospheric and atmospheric pressures on a given heating surface.