This paper investigates the combined effect of the kiln or brick surface roughness and the brick lattice setting density on the fluid flow and heat transfer characteristics in tunnel kilns. The flow uniformity, pressure drop, convective heat transfer coefficient (CHTC), and pumping power are studied. A high-density setting (HDS), which comprises 768 bricks, and a low-density setting (LDS), which comprises 512 bricks, are tested for kiln boundaries and brick surface roughness levels of 0, 1, 2, 3, and 4 mm. The investigation is conducted using a 3D-CFD model with the k-ω turbulence model. The surface roughness changes from 0 to 4 mm for either kiln walls or bricks while fixing the other. The results show that increasing the tunnel kiln surface roughness from 0 to 4 mm increases the pressure drop of both the HDS and LDS by about 13.5%. It also increases the established CHTC value of the LDS more than the HDS by about 23% for all tested roughness levels. Changing the brick surface roughness from 0 to 4 mm increases the pressure drop and CHTC value for the LDS more than for the HDS by about 10% and 12%, respectively. Additionally, the total heat transfer rate-to-pumping power ratio for the LDS is larger than for the HDS by 17.4% for smooth bricks and 23.1% for the brick roughness of 2 mm, i.e., the brick roughness provides a greater advantage to the LDS. The results confirm that the LDS for rough and smooth bricks loaded in tunnel kilns attains a better brick quality, a higher heat transfer rate, and a lower pumping power than the HDS.

This paper presents the effect of brick and kiln wall roughness on the fluid flow, pressure drop, and convection and radiation heat transfer in tunnel kilns. The surface roughness of 0–4 mm is investigated for bricks and tunnel boundary. Another wall roughness of 10 mm is considered to explore the effect of significant defects in the tunnel boundary. The study is conducted using a three-dimensional computational fluid dynamics (CFD) model based on the finite volume method with the k – ω turbulence model. The convective heat transfer coefficients enhance by 45% and 97%, and the pressure drop increases by 25.1% and 80.4% as the brick roughness is increased from 0 to 1 mm and 0 to 4 mm, respectively. The ratio of heat transfer rate to pumping power reaches its maximum at a brick roughness of 2 mm. These results provide essential knowledge about the acceptable range of brick roughness for manufacturers. As the tunnel boundary roughness is increased from 0 to 1 and 0 to 10 mm, the heat transfer rates increase by 1.34% and 3.88%, while the pressure drops increase by 7.5% and 18.2%, respectively. These results are supportive of scheduling the maintenance of tunnel kilns’ interior structure. Moreover, the enhancement of the radiation heat transfer depends on the brick emissivity and the area ratio of rough to smooth surfaces.

This paper proposes correlation equations of heat transfer and pressure drop for the design and operation of tunnel kilns loaded with lattice brick settings of different geometrical parameters in the form of Nusselt number and friction factor. The developed correlation equations considered parameters that were not investigated in previous studies, such as the relative roughness of the bricks and the stack channels, and they also extended the Reynolds numbers to a practical range that was not covered before in a simple, practical form. The correlation equations are valid for Reynolds number between 125 and 10,200, Prandtl number between 0.68 and 0.73, brick’s relative roughness between 0.23 and 0.93, voidage fraction between 0.48 and 0.653, and the geometrical parameters of the tested lattice brick settings. The achieved correlations of the Nusselt number and the friction factor are well compared with the available correlations in the literature in their valid range of parameters. It is found that Nusselt numbers and the friction factors for low-density are higher than those of high-density settings for all considered parameters except the voidage fraction. The effect of the considered parameters confirms that Nusselt numbers increase and the friction factors decrease substantially with the Reynolds number and slightly with the Prandtl number. At a constant Reynolds number, both the Nusselt number and the friction factor increase as the brick’s relative roughness is increased. Moreover, as the stack channel spacing is increased, the Nusselt number decreases, and the friction factor increases. The voidage fraction of the setting has a monotonic effect on both Nusselt numbers and friction factors. Nusselt numbers for high-density are higher than those for low-density settings as the voidage fraction varies.

Estimating Global Solar Radiation (GSR) is crucial for various engineering applications, including agriculture, residential planning, and energy systems. Reliable ground-based GSR measurements are limited in desert climates due to sensor degradation, sparse coverage and harsh environmental conditions. This study evaluates four ensemble-based machine learning (ML) models—Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting (G-Boost), and Extreme Gradient Boosting (XG-Boost)—alongside two deep learning (DL) models, Multi-Layer Perceptron (MLP) and hybrid Long Short-Term Memory–MLP (LSTM-MLP)—for predicting daily GSR in Kuwait, one of the world’s hottest regions. The models incorporate 13 daily meteorological and time-related input parameters, including temperature (average, maximum, and minimum), potential sunshine duration, cloud cover, humidity, wind speed, dew point temperature, sea-level pressure, and time-based factors (day of month, day of year, month, and year). Data were collected from a ground weather station between March 2015 and December 2020. Performance was assessed using five metrics: Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and coefficient of determination (R²). Results showed strong predictive capabilities across all models, with G-Boost and XG-Boost demonstrating the best predictive performance (R² = 0.996 and 0.994, respectively), while SVM achieved the highest held-out test R² (0.9729) among the ensemble-based ML models. An additional hourly validation using the XG-Boost model demonstrated excellent agreement between observed and predicted hourly solar radiation, confirming that the proposed framework accurately captures short-term variability and is applicable for both daily solar resource assessment and short-term photovoltaic forecasting under extreme desert climate conditions. Among DL models, MLP outperformed LSTM-MLP. The models maintained consistently high accuracy across seasonal variations, including peak summer periods characterized by extremely high ambient temperatures. Feature importance analysis identified potential sunshine duration as the most influential factor, followed by maximum temperature. These findings demonstrate that the proposed framework provides accurate and reliable prediction of global solar radiation across multiple temporal resolutions, making it suitable for both long-term solar resource assessment and short-term photovoltaic forecasting in extreme desert climates.