| Popis: |
Empirical correlations are extensively utilized in applications such as the design of solar stills to estimate the evaporation rate from a pool of water. These correlations are based on specific experimental measurements carried out for a particular design and operating conditions. In a semiempirical approach, theoretical modeling is incorporated in deriving the expression for the evaporation rate. These models utilize assumptions such as zero temperature gradient in air and water and constant relative humidity in the surrounding air so as to circumvent whole field formulations. One such empirical correlation, called Dunkle's correlation, is improved in the present work by using a comprehensive ab initio approach in the background for estimating the evaporation rate, which incorporates two-layer natural convection with moisture transport for estimating the evaporative mass flux. The empirical and the ab initio approaches are compared to improve Dunkle's correlation in the inverse heat transfer framework. Here, the objective function is minimized using the combined artificial neural network-genetic algorithm (ANN-GA) tool. The uncertainties in estimating the parameters of interest are evaluated by sprinkling noise with a Gaussian distribution over the solutions of the comprehensive model. The improved Dunkle's correlation reduces the error in the predicted evaporation rate between 12% and 25%, depending on the water depth. |
