HEAT TRANSFER AND FLUID FLOW CHARACTERISTICS OF DISCRETE RIB ROUGHENED SOLAR AIR HEATERS

Abstract

The thermal performance of conventional solar air heater is generally poor because of low convective heat transfer coefficient between air and the absorber plate. The heat transfer coefficient of the absorber plate can be substantially increased by enhancing turbulence in the duct. The use of artificial roughness in a solar air heater duct has been proposed to be an excellent option to enhance the heat transfer from absorber plate to the air. However, the artificial roughness results in higher frictional losses leading to excessive power requirement for the fluid to flow through the duct. It is therefore desirable that turbulence must be created only in a region very close to the heat-transferring surface to break the viscous sub-layer for augmenting the heat transfer and the core flow should not be unduly disturbed to limit the increase in friction losses. The literature review shows that considerable amount of experimental work has been carried out to study the effect of artificial roughness on heat transfer and friction characteristics of various roughness geometries. It has been observed that the ribs of various shapes and orientations were used to enhance heat transfer in circular tubes, annuli, and ducts of rectangular and square cross sections having all or one of the walls roughened and subjected to constant heat flux. It was reported that the inclined ribs enhance the heat transfer more than that for the straight transverse ribs due to the generation of secondary flow. The concept of V-shaped ribs was floated to obtain still higher heat transfer rate due to development of two vertices cells compared to only one cell developed in case of inclined ribs. Further it is reported that inclined discrete ribs and V-shaped discrete ribs show higher heat transfer performance compared to that of the continuous inclined and continuous V-shaped ribs. Attempts have also been made to ii visualize the flow structure over the roughened surface by using different optical techniques including PIV system. In a recent investigation it has been reported that a small gap in inclined rib arrangement enhances the heat transfer with no significant increase in the pressure drop and it is observed that the heat transfer performance of such type of roughness is higher as compared to that of the V-shaped rib arrangement. It is thought that a gap in the inclined rib accelerates the flow and enhances the local turbulence which will result in an increase in the heat transfer performance. However detailed studies of heat transfer and flow characteristics of surfaces with such geometry of artificial roughness covering wide range of system and operating parameters has not been carried out so far. Further, most of the investigations carried out so far have applied the artificial roughness on two opposite walls with all four walls being heated. It is noted that for the application of this concept of enhancement of heat transfer in the case of solar air heaters, roughness elements have to be considered only on one wall, which is the only heated wall in the form of the absorber plate. These conditions make the fluid flow and heat transfer characteristics distinctly different from those found in the case of two roughened walls and four-heated wall duct. In the case of solar air heaters, only one wall of the rectangular air passage is subjected to uniform heat flux (insolation) while the remaining three walls are insulated. In the light of the foregoing discussion it appears that, there is need for a study of flow structure and heat transfer in a rectangular duct with one wall roughened with inclined ribs having a gap (i.e. inclined discrete rib) under a wide range of system and operating parameters suitable for solar air heater. The present investigation was, therefore, taken up to determine the optimum location and width of the gap in an inclined rib to form iii discrete rib. This study constitutes a basic building block for designing a discrete rib system with descetizing parameters of gap width and location. This study will help in determining the gap size and position while descretizing the inclined (non-transverse) ribs for enhancing the performance as compared to non-descretized ribs. Accordingly, the present research work has been taken up with the following objectives: 1. Experimental investigation of heat transfer and fluid flow characteristics of artificially roughened duct having inclined discrete type of roughness. 2. Development of heat transfer coefficient (Nusselt number) and friction factor correlations/relationships as a function of roughness geometry and flow parameters. 3. Investigation of flow structure inside the roughened duct using Particle Image Velocimetry (PIV) system. 4. Thermo-hydraulic performance optimization of solar collector system to obtain the optimal roughness geometry. Experimental set-ups have been designed and fabricated, one for investigating the heat transfer and friction characteristics of the ducts and the other for flow visualization using PIV system. These experimental set-ups have been used to conduct extensive experimentation on a rectangular duct with one broad wall roughened with inclined rib having a gap (inclined discrete rib). In case of investigation on heat transfer and friction characteristics, the artificially roughened wall was heated from the top by means of an electrical heater whereas visualization of the flow field of the roughened duct was carried out under ambient conditions. Experimental data have been collected for temperature rise iv of air, mass flow rate, heat flux, pressure drop across the duct and velocity vectors for various geometries of inclined discrete rib roughness. The data have also been collected for all these aspects of a conventional smooth duct under similar operating conditions with ment of heat transfer coefficient and the objective friction factor for the rough( system and operating conditions of the solar air heaters. The i in this investigation is given below: of parameters the ered S. No. Roughness and flow parameters 1 Reynolds number (Re) 2 Relative roughness pitch (P/e) 3 Relative gap position (d/ W) 4 Relative gap width (g/ e) 5 Angle of attack (a) 6 Relative roughness height (e/D) Range of parameters 3000-18000 (6 values) 4.0 to 10.0 (4 values) 0.16 to 0.67 (4 values) 0.5 to 2.0 (4 values) 30° to 90° (4 values) 0.018 to 0.037 (4 values) The values of Nusselt number and friction factor determined from experimental data in a smooth duct have been compared with the values obtained from the standard correlations i.e. Dittus Boelter equation for the Nusselt number and modified Blasius equation for the friction factor. The average deviation of experimental values of Nusselt number and friction factor from those predicted by standard correlations have been found to be ± 2.8% and± 2.3% respectively. This shows a reasonably good agreement which ensures the accuracy of the data being collected with the present experimental setup. Based on the analysis of the errors in the experimental measurements using different instruments, the maximum uncertainties in the calculated values of Reynolds number, Nusselt number, and friction factor have been estimated as ± 2.6 %, ± 2.03 % and ± 9.24 % respectively for minimum Reynolds number (3000) and ± 1.8%, ± 3.42% and ± 1.84 % respectively for maximum Reynolds number. It has been found that the value of Nusselt number increases monotonically with increase in Reynolds number whereas, the friction factor decreases initially and then becomes nearly asymptotic to a constant value with the increase in Reynolds number. The inclined discrete rib roughened duct has shown higher values of Nusselt number as compared to that produced by the continuous inclined ribbed duct for the same values of roughness parameters. It seems that introduction of a gap in the rib allows release of fluid partly belonging to secondary flow and partly to the main flow. As a result of the presence of the gap, the secondary flow along the rib joins the main flow to accelerate it which energizes the retarded main flow over the surface to increase the level of turbulence and hence the heat transfer coefficient as well as the friction factor increases. It has been observed that the values of Nusselt number and friction factor increase with increase in relative gap position, attaining maxima at the value of 0.25 and then decrease with further increase in the relative gap position. It is known that an inclined rib gives rise to a high heat transfer region at the leading edge due to cold fluid entering at this point and a low heat transfer region at the trailing edge because the fluid flowing along the rib gets heated continuously and leaves the rib at this point. If a gap is created vi towards the trailing edge region, it will help in increasing the heat transfer in the low heat transfer trailing edge region by introducing colder fluid near the gap and hence the overall heat transfer is enhanced. It is seen that this happens till the gap is positioned about one fourth of the duct width, possibly because placing a gap at a position too close to the duct wall will not produce similar effect of enhancement of heat transfer due to the presence of lateral boundary layer near the wall and hence further decrease of gap position i.e. placing the gap closer to the trailing edge (d/W < 0.25) results in a decrease in the overall heat transfer. It is also observed that the values of Nusselt number and friction factor increase with increase in the value of relative gap width (from 0.5 to 1.0) and then decrease with further increase in the relative gap width. The maximum values of these parameters are observed at the relative gap width of 1.0 and minimum at the relative gap width of 2.0. It seems that although providing a gap enhances the turbulence but a wider gap reduces the flow velocities substantially through the gap and hence the effect of the gap almost vanishes and the surface behaves almost like a smooth surface. At the same time too small a gap will also not allow sufficient amount of secondary flow fluid to pass through it and thus the main flow will not be energized sufficient enough. The turbulence intensity at different relative gap widths has been measured using 2-D PIV system and it is seen that the highest value of turbulence intensity is observed at the relative gap width of 1.0 and its lowest value is observed at the relative gap width of 0.5 which shows good agreement between the heat transfer studies and flow investigations. Similarly it is observed that the maximum values of Nusselt number and friction factor occur at the relative roughness pitch of 8.0 and minimum values are observed at the vii relative roughness pitch of 4.0. The value of Nusselt number is highest at an angle of attack of 60° and lowest at an angle of attack of 30° whereas the highest value of friction factor is observed at an angle of attack of 90°. The value of Nusselt number increases monotonically with increase in relative roughness height at all the Reynolds numbers. The inclined discrete rib roughened duct with relative gap position (d/W) of 0.25, relative gap width (g/e) of 1.0, relative roughness pitch (P/e) of 8.0, angle of attack (a) of 60° and with relative roughness height (e/D) of 0.037 has shown the maximum value of the Nusselt number which is of the order of 2.83 times of that of the smooth duct in the experimental range of investigation. Similarly for friction factor, the maximum value of the friction factor of inclined discrete rib roughened duct is of the order of 2.93 times of that of the smooth duct for the same range of roughness parameters. The following correlations have been developed for Nusselt number and friction factor in terms of roughness and flow parameters: Nu = 0.002 Re' °8 (a 1 60)0.006-exp( \ 2 \ I p 2 \ ( (P / 01.87 exp –0.45 In — –0.65 In a e 60 ) \ 1 \ 1 ( z\ ( \ 2 \ 0 5 exp –0.12 In — (g. e,)-0.03 exp –0.18 In (D) W ) ) W i e) D / i d \2\ f = 0.071Re-0 133 p( / ey 83 exp –0.44(1n 2i d °43 exp –0.14 In — e W W j j ( ( 2 \ 0 69 0.12 in-KJ (y \ e i ) D It has been found that 92% of the data points of Nusselt number lie within the deviation limits of ± 10% and the average absolute deviation of the predicted values from d \ -0 32 W ) (a / 60)067 (gle)-0.052 exp viii the experimental results has been found to be 5.85 % and the corresponding values of friction factor are 96% and 4.12 % respectively. This reveals a good agreement between the predicted and experimental values of Nusselt number and friction factor. It has been pointed out that the heat transfer coefficient and hence thermal performance can be enhanced by providing artificial roughness; however, this also increases the pumping power substantially. Thus it becomes important to select roughness geometry in such a way that it will improve heat transfer to the maximum possible extent while keeping the penalty of increase in the pumping power at the minimum possible level. The selection of optimum values of roughness parameters, therefore, involves the consideration of thermo-hydraulic performance of the collector. Accordingly, the values of roughness parameters that correspond to maximum thermo-hydraulic benefit for specified operating conditions relevant to the present investigation have been determined. The results obtained from the optimization process based on the following criteria have been studied and compared. (i) Thermal efficiency, 'nth (ii) Effective efficiency, gat- (iii) Thermo-hydraulic performance parameter, ri The values of optimizing parameters corresponding to each of the above criteria have been determined for all possible sets of roughness parameters in the range of specified operating parameters (temperature rise parameter, AT/I and average insolation, I). For each set of calculations with specified values of operating parameters, the values of optimizing parameters are compared and the optimum set of roughness parameters is identified as the one that results in maximum value of the optimizing parameter. ix