طراحي، ساخت و ارزيابي سامانه شبيه‏ ساز جمع‏ كننده‏ تخت خورشيدي با استفاده از ضريب ترموهيدروليكي

Design, construction and evaluation of solar flat-plate collector simulator based on the thermohydraulic coefficient

استفاده از ناهمواری در سطح زیرین صفحه‌ی جاذب حرارت خورشید از جمله شیوه‏ های افزایش بازده ترموهیدرولیكی جمع كننده‏ های تخت خورشیدی می‏ باشد. سامانه‌ای با عنوان سامانه ‏ی شبیه ‏ساز جمع‏ كننده‏ ی صفحه ‏تخت خورشیدی، با مهیا ساختن شرایط تابشی خورشید، امكان بررسی و آزمایش هر نوع جمع‏ كننده‏ ی صفحه تخت خوشیدی را جهت به‏ دست آوردن مقدار عدد ناسلت، ضریب اصطكاك و در نهایت ضریب ترموهیدرولیكی ممكن ساخته است. در این آزمایش، سامانه‏ ی شبیه‏ سازی جمع‏ كننده‏ های تخت خورشیدی برحسب توان مورد نیاز مكنده‏ ی گریز از مركز، المنت‏های حرارتی و تجهیزات مورد نیاز جهت اندازه‏ گیری پارامترهای مورد نظر، طراحی و ساخته شد. ناهمواری مورب منقطع بر اساس تحقیقات پیشین، به‌عنوان ناهمواری بهینه، انتخاب و در مجرایی متشكل از سه سطح صاف كه كناره ‏های صفحه ‏ی جاذب حرارت صاف و ناهمواری‏ دار را تشكیل داده بودند، مورد آزمایش قرار گرفت. ضریب تبیین برای مقایسه‏ ی عدد ناسلت حاصل از آزمایش صفحه‏ ی جاذب حرارت صاف و رابطه‏ ی تجربی دیتوس- بولتر برابر با 69/0 و برای مقایسه‏ ی ضریب اصطكاك حاصل از آزمایش و رابطه‏ ی بلاسیوس برابر با 68/0 به ‏دست آمد. تغییرات ضریب ترموهیدرولیكی، برای ناهمواری مورد نظر نسبت به تغییرات عدد رینولدز بررسی گردیده و با تحقیقات پیشین مورد مقایسه قرار گرفت. افزایش ضریب اصطكاك نسبت به افزایش عدد استنتون ارجحیت داشته و در نتیجه ضریب ترموهیدرولیكی به ‏دست آمده برای ناهمواری آزمایش شده در مجرای مورد نظر، با افزایش عدد رینولدز روند كاهشی داشت. بالاترین مقدار ضریب ترموهیدرولیكی برای عدد رینولدز 3149، برابر با 08/1 حاصل گردید.

<strong >Introduction </strong > Increasing the area of absorber plate between the flowed air through the duct can be accomplished by corrugating the absorber plate or by using the artificial roughness underside of the absorber plate as the commercial methods for enhancing the thermohydraulic performance of the flat plate solar air heaters. Evaluation of this requires the construction of separated solar air heater which is costly and time consuming. The constructed solar flat-plate collector simulator can be a sufficient solution for obtaining the heat transfer and thermodynamic parameters for evaluating the absorber plate. The inclined broken roughness was chosen as the optimum roughness which is surrounded by three aluminum smooth walls. <strong >Materials and Methods </strong > The duct for both smooth and roughened plate have been constructed based on the ASHRAE 93-2010 standard. In order to achieve a fully thermal and hydraulic developed flow, the plenum is constructed. The centrifugal fan is considered by applying the required air volume at the pressure drop obtained by the duct, plenum and the orifice meter. The TSI velocity-meter 8355 is used to measure the velocity of air crossing through the pipe connected to the centrifugal fan. The micro manometer Kimo CPE310-s with the resolution of 0.1 Pa is used to measure the pressure drop across the test section of the smooth and roughened duct. The LM35 sensors are used to measure the absorber plate and air temperature through the test section. Obtained parameters are used to calculate the Nusselt number and friction factor across the test section for smooth and roughened absorber plate. The Nusselt number and friction factor parameters which is obtained for smooth absorber plate based on experimental set-up, is compared with Dittus-Bolter and Blasius equations, respectively, for validating the simulator. By calculating the Nusselt number and friction factor, Stanton number is obtained based on the equation (6), and thermohydraulic coefficient is calculated by the equation (5) for the desired roughness. <strong >Results and Discussion </strong > Pressure drop for smooth duct is obtained to be 20 Pa. Maximum velocity crossed through the plenum is calculated by the equation (8). Thereafter, pressure drop for plenum by considering the maximum velocity in equation (7), is obtained to be 1.16 Pa. The same procedure for maximum velocity which is crossed through the orifice meter is obtained by the equation (10) and then the pressure drop for orifice meter is calculated equal to 243 Pa by considering the velocity in equation (9). Total pressure is given by the equation (11) to be 246.16 Pa. The required power for centrifugal fan is obtained equal to 105 W from equations (12), (13) and (14), respectively. Both aforementioned Nusselt number variations with Reynolds number were monotonously increased by increasing the Reynolds number. The gained RMSE and coefficient of determination between the Nusselt numbers are 0.0566 and 0.6944, respectively. The obtained RMSE and coefficient of determination between the friction factors are 0.0004 and 0.6814, respectively. The low value of the RSME and high value of the R2 analysis for both Nusselt number and friction factor shows that there is a good agreement between the experimental data and empirical correlations. Fig. 8 demonstrates that the thermohydraulic coefficient is decreasing as the Reynolds number increased. The effect of friction factor related to the Stanton number is shown up more effective by increasing the Reynolds number. It should be noted that the same procedure is conducted for Han 's experiment where the thermohydraulic performance is decreased as the Reynolds number increased. The maximum magnitude of the thermohydraulic performance was achieved at minimum 3149 Reynolds number. <strong >Conclusions </strong > The flat-plate solar collector simulator was designed based on the ASHRAE 93-2010 standard which consists of the centrifugal fan, chosen based on the required air volume by considering the pressure drop in the duct, plenum and orifice meter. The experiment was conducted between 3149 to 19247 Reynolds numbers. The good agreement between the comparison of the Nusselt number and friction factor obtained by the experiment for smooth duct was achieved by the Dittus-Bolter and Blasius equations, respectively, to validate the simulator. The obtained thermohydraulic coefficient for optimized roughness surrounded by three smooth walls was lower than the former investigated roughnesses at each Reynolds number