Field monitoring and numerical analysis of thermal behavior of the National Stadium under solar radiation

Abstract

Due to high static indeterminacy, large-span spatial structures are sensitive to temperature changes. But dynamic boundary conditions like solar radiation, wind, and atmospheric factors cause spatiotemporal thermal non-uniformity, deviating from the uniformity assumption. Thus, non-uniform thermal behavior of the National Stadium was studied by long-term monitoring and numerical analysis. Sensor node was developed with vibrating-wire stress/temperature sensor and wireless communication module comprising ATmega64, CC1101, TPS7333, AT45DB641E, and ADS1115. A robust tree-type wireless sensor network was deployed, featuring time-aligned data acquisition and cloud computing-based remote platform. The National Stadium's system adopts 160 sensors to monitor temperature and stress in 40 members, including 9 top chords, 22 bottom chords, 6 webs, and 3 columns. Long-term monitoring revealed significant spatiotemporal non-uniformity in the structural temperature field, with solar radiation causing >10 °C differences and daytime temperatures well above ambient. The structure showed marked thermal-sensitivity, with clear stress gradients, strong stress-temperature correlation, and annual stress variations of many members exceeding 20 MPa. Based on ray-tracing and computer graphics, an algorithm was established to identify dynamic shading effects among structural members and applied to thermal/structural analysis. A numerical method was developed that incorporates heat conduction, convection, and radiation, while accounting for realistic thermal boundary conditions induced by solar radiation, wind, and ambient temperature. In the simulation, temperature distribution was modeled using link elements, and thermal stresses were modeled using beam elements. Measured and simulated data closely replicate each other, with average error rates for temperature and stress in monitored members below 8 % and 16 %, respectively, and average deviations within 2.9 °C and 1.1 MPa. These results highlight the reliability of the numerical approach and support the validity of the simulation method in representing structural thermal behavior. Methodologies and conclusions of this study provides practical insights for thermal design, monitoring, and control of spatial structures.