Inverse parameters estimation an‎d optimal geometric design of adsorption bed for solar cooling applications

2/19/2027 12:00:00 AM

This doctoral research introduces a thorough modeling an‎d optimization framework for adsorption refrigeration systems commonly employed in solar cooling apparatus, concentrating on two primary elements: the estimation of adsorption parameters an‎d diffusion coefficients via a simplified two-dimensional model, an‎d the thermal-geometric optimization of a complete three-dimensional finned heat exchanger functioning as an adsorption bed. Therefore, efforts were devoted to address the majority of the challenges related with heat transfer an‎d geometric design within the adsorption bed, as well as the contradictions that have surfaced in recent research, in order to obtain the optimal performance for adsorption refrigeration applications. In the first part, a two-dimensional symmetric adsorption bed model is developed in the initial phase to derive the thermophysical properties of the activated carbon/methanol pair, utilizing experimental data from differential variance an‎d molecular weight analysis of the activated carbon sample in a private laboratory. Through an inverse parameter estimation method using numerical simulation coupled with a particle swarm optimization algorithm, the key transport parameters of the linear driving force model, including the effective thermal conductivity, diffusion coefficients, an‎d Dubinin–Astakhov adsorption equation constants, as well as other parameters of the governing equations such as Darcy’s law for vapor flow, transient energy balance, etc., were estimated under local thermal non-equilibrium conditions with minimization of the mean square temperature error of temperature between the simulation model an‎d the experiment. Next, simulations were done using the new estimated parameters an‎d resulted in a deviation of less than 1.2% for the results of temperature, pressure, an‎d uptake changes. The total mean square error for the entire model experiment was 0.87%. Using the previously estimated parameters, an accurate three-dimensional numerical simulation was performed to optimize the thermo-geometric design of the adsorption bed. Hence, the second part focused on developing a transient computational fluid dynamics model in COMSOL Multiphysics to simulate heat an‎d mass transfer in two heat exchanger configurations using longitudinal an‎d annular finned tubes. An experimental setup prepared an‎d the model validity confirmed based on experimental results under identical desorption conditions T_HTF = 353 [K], V_HTF = 0.2 [m/s], with a deviation of less than 3% in both temperature an‎d pressure. The model was then used to explore the effects of geometric variables, including the number of fins, fin height an‎d thickness, diameter an‎d thickness tube, on four key performance indicators: methanol mass generated, coefficient of performance, specific cooling power, an‎d volumetric cooling power. An initial parametric study identified the sensitive variables, an‎d an optimization routine based on the Nelder-Mead algorithm was used to fine-tune the design. Cross-sectional an‎d axial data on temperature, pressure, an‎d uptake profiles were extracted from the simulation to eva‎luate system performance. The optimized geometric design, which focused on fin number, fin height, an‎d tube diameter, resulted in improvements of 33%, 31%, 30%, an‎d 29%, respectively, for the four performance indicators compared to the reference design. The optimal longitudinal finned design provided the best trade-off between methanol desorption rate an‎d system compactness, achieving improvements 21% compared to annular finned design. This study establishes a 3D model simulation approach with experiment, combining geometric optimization with numerical parameters, enabling the design of high performance adsorption beds.

كربن فعال گرانولي , خواص ترموفيزيكي , تخمين پارامتر معكوس , پارامترهاي هندسي , بهينه‌سازي , جذب , واجذب

Granular activated carbon , Thermophysical properties , Inverse parameter estimation , Geometric parameters , Optimization , Adsorption , Desorption

Maaed Ossman

Dr. Majid Siavashi