Multi-City Solar Chimney Dataset for Smart-City Applications

Résumé fourni par la source

This dataset contains the complete simulation output of a parametric study of a single-channel solar chimney (SC-SoCh) operating under real meteorological conditions across six representative cities in Mexico. It was generated using a validated transient Global Energy Balance (GEB) model that solves the coupled heat transfer interactions among the glass cover, air channel, absorber wall, and thermal insulation layer at each time step. The dataset is intended to support research in passive ventilation, building energy efficiency, surrogate modelling, and machine-learning applications for solar chimney systems. Coverage and parametric space The dataset spans the following parameter combinations: Cities (6): Mexico City (CDMX), Hermosillo, Mérida, Monterrey, Toluca, and Villahermosa — representing tropical, arid, semi-arid, and temperate climates according to the Köppen–García classification. Absorber wall materials (9): Aluminum, Concrete, Brick, RT25HC, RT28HC, RT35HC, RT42, Paraffin 46–50, and Mg29 (the last six are phase change materials, PCMs). Chimney orientations (4): North, South, East, West. Representative climatic days (2 per month × 12 months = 24 per city): warmest day (Calido) and coldest day (Frio) of each calendar month, derived from historical meteorological records. Total files: 6 cities × 9 materials × 4 orientations × 2 day types × 12 months = 5,184 CSV files.Total uncompressed size: approximately 133.64 GB. Simulation methodology Each simulation runs for 72 hours using real hourly meteorological data (solar irradiance, ambient temperature, relative humidity, wind speed, and atmospheric pressure) replicated across three consecutive 24-hour cycles. This initialization procedure eliminates dependence on the initial thermal condition. Only the output of the third 24-hour cycle is relevant for analysis; it corresponds to the last 17,280 rows of each file (time step Δt = 5 s). The GEB model was validated against experimental measurements obtained from a physical SC-SoCh prototype located at CENIDET (Cuernavaca, Mexico), reporting root-mean-square deviation (RMSD) values within acceptable ranges for temperatures and mass flow rate. The absorber wall is discretized into 21 layers in the thickness direction (x) and 21 nodes along the channel height (y). Phase change materials are modelled via an effective specific heat capacity method (CP,eff) that embeds the latent heat of fusion. Buoyancy-driven airflow is estimated using the Andersen–Bansal correlation with a discharge coefficient Cd = 0.57, and moist air density follows the Giacomo–Davis formulation to account for altitude effects. File structure The dataset is organized as follows: {CITY}/ {N}-{MonthName}/ Calido/ ← warmest day of the month Frio/ ← coldest day of the month Each folder at the leaf level contains 36 CSV files (9 materials × 4 orientations). File names follow this convention: Resultados-Chimney-OC-{CITY}-{orientation}-Hy-200-d-15-glass-1-paint-1-wall-{N}.csv where: {CITY} ∈ {CDMX, HERMOSILLO, MERIDA, MONTERREY, TOLUCA, VILLAHERMOSA} {orientation} ∈ {sur, norte, este, oeste} (South, North, East, West) {N} = material index: 1 = Aluminum, 2 = Concrete, 3 = Brick, 5 = RT25HC, 6 = RT28HC, 7 = RT35HC, 8 = RT42, 9 = Paraffin 46–50, 11 = Mg29 Variables recorded per file Each CSV file contains 51,841 rows (72 h at Δt = 5 s) and the following 17 columns: Column Description Units Tiempo Simulation time s y[m] Vertical position along the channel m Tg Glass cover temperature °C Tf Air channel (fluid) temperature °C Tw Absorber wall temperature (representative layer) °C Tais Thermal insulation temperature °C Tfout Channel outlet air temperature °C Fmas Mass flow rate kg s⁻¹ Fvol Volumetric flow rate m³ s⁻¹ Efi Instantaneous thermal efficiency – Residual Gauss–Seidel convergence residual – Gsolar Incident solar irradiance W m⁻² Tamb Ambient (outdoor) temperature °C Vwind Wind speed m s⁻¹ HR Relative humidity % Presion Atmospheric pressure MPa Qremovido Heat removed by the air stream W Intended uses This dataset is suitable for, but not limited to: Comparative analysis of absorber wall materials (massive vs. PCM) across climatic zones. Annual and seasonal assessment of air changes per hour (ACH) and passive ventilation potential. Training and validation of machine learning or artificial neural network (ANN) surrogate models for solar chimney performance prediction. Benchmarking of reduced-order thermal models against high-resolution transient simulations. Urban heat island and smart-city ventilation studies for Mexican and climatically analogous cities. Archive contents The dataset is distributed across three ZIP archives due to size constraints: Part_1.zip — cities CDMX and HERMOSILLO Part_2.zip — cities MERIDA and MONTERREY Part_3.zip — cities TOLUCA and VILLAHERMOSA

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.