Thermal performance evaluation of a parabolic trough solar collector with helical and trapezoidal turbulators: experimental, CFD, and AI-based analysis

Abstract

Parabolic trough solar collectors (PTC) offer notable potential for solar thermal applications, but performance is hindered by limited convective heat transfer within the receiver tube. This research proposes an integrated experimental, three-dimensional computational fluid dynamics (CFD), and explainable AI framework to evaluate PTCs with helical and trapezoidal turbulators. The central contribution is the experimentally validated combination of CFD analysis and AI-based temperature prediction for PTCs with internal turbulator geometries. Outdoor experiments under Tokat climatic conditions validated the CFD model using measured outlettemperature data during a stable period with solar radiation near 900 W/m2 and outlet temperature of 140-150 degrees C. The CFD and experimental results are closely aligned, with mean absolute error (MAE) and root mean square error (RMSE) both around 0.85. Turbulators enhanced fluid mixing and outlet temperature compared to a reference tube; the helical turbulator increased outlet temperature by 7-9%, and the trapezoidal by 3-5%. Correspondingly, annual energy output rose from 546.9 kWh/year (reference) to 694.6 kWh/year (helical) and 661.8 kWh/year (trapezoidal), with COQ emissions reductions reaching 0.283 ton/year for the helical design. The M5P and Elastic.Net regression models achieved R2 scores of 0.95 and 0.90, respectively, for outlet temperature prediction using environmental parameters. This study demonstrates that the helical turbulator is the most effective configuration in terms of outlet-temperature enhancement under the investigated operating conditions, and that the combined experimental-CFD-AI approach provides an actionable, interpretable tool for improving air-based PTC systems used in thermal applications such as process-air heating and potential drying-related systems.