This work presents a novel approach to increasing the efficiency of photovoltaic (PV) panels by integrating them with a cooling tower (CT). An infusion of water cools the hot, dry ambient air at the top of the CT. Due to gravity, the cooled air drops toward the base of the CT, where it interacts with a turbine placed at the bottom of the CT to produce electricity. The air then exits the CT base, creating a cooled air jet stream. The PV panels were placed at the base of the CT, right at the stream’s exit. As the cooled air passes underneath the PV panels, it exchanges energy with the PV, reducing the panels’ temperature. The results showed that the maximum annual efficiency improvement (6.831%) was observed using two rows of PV panels. The efficiency declined incrementally from 6.831% to 4.652% when the number of rows of PV panels was increased from two to twelve. The results also showed a significant improvement in the temperature of the PV panels. The best results were obtained at noon (maximum ambient temperature), where the solar panel temperature was lowered to 25 °C from 55 °C. Furthermore, the annual electrical energy generated with two rows of panels was 39,207.4 kWh without the CT, compared to 41,768.2 kWh with the CT. In addition, the results showed that with a 10 m diameter and 200 m height CT, the maximum number of PV rows that can be effectively cooled is 24. Future work will investigate integrating additional techniques to improve the system’s efficiency further.

The significance of this work lies in the development of a novel code-based, detailed, and deterministic geometrical approach that couples the optimization of the Fresnel lens primary optical element (POE) and the dome-shaped secondary optical element (SOE). The objective was to maximize the concentration acceptance product (CAP), while using the minimum SOE and receiver geometry at a given f-number and incidence angle (also referred to as the tracking error angle). The laws of polychromatic light refraction along with trigonometry and spherical geometry were utilized to optimize the POE grooves, SOE radius, receiver size, and SOE–receiver spacing. Two literature case studies were analyzed to verify this work’s optimization, both with a spot Fresnel lens POE and a spherical dome SOE. Case 1 had a 625 cm² POE at an f-number of 1.5, and Case 2 had a 314.2 cm² POE at an f-number of 1.34. The equivalent POE designed by this work, with optimized SOE radiuses of 13.6 and 11.4 mm, respectively, enhanced the CAP value of Case 1 by 52% to 0.426 and that of Case 2 by 32.4% to 0.45. The SOE’s analytical optimization of Case 1 was checked by a simulated comparative analysis to ensure the validity of the results. Fine-tuning this design for thermal applications and concentrated photovoltaics is also discussed in this paper. The algorithm can be further improved for more optimization parameters and other SOE shapes.

Generative AI is rapidly reshaping higher education, yet its influence on student learning in engineering remains insufficiently understood. In courses that require conceptual understanding, design reasoning, and problem-solving, AI tools may support learning by providing explanations, solution pathways, and feedback. At the same time, they raise questions about trust, verification, and the quality of student thinking. This study investigates student perceptions of generative AI in engineering education, focusing on learning, trust, and information-seeking. This work extends a prior conceptual framework grounded in Ellis’s theory of information-seeking. The study applies a six-stage subset of Ellis’s model—starting, chaining, browsing, differentiating, monitoring, and verifying—to examine how students perceived generative AI during an engineering learning activity. It also examines themes related to perceived usefulness, learning support, trust, judgment, and AI use. Within this sample, respondents generally reported perceiving AI as a support tool for conceptual understanding, efficiency, and checking their reasoning. Among respondents who reported at least some AI use, higher agreement was observed for items concerning ongoing task support, perceived deeper learning, and the application of their own judgment, whereas lower agreement was observed for using AI to understand the problem requirements initially or to check and validate technical work.