Effect of Natural Ageing on the Reaction-to-Fire Performance of Fire-Retardant-Treated Timber Façade Systems

Sammanfattning

This study investigates the influence of natural ageing on the reaction-to-fire performance of fire-retardant-treated (FRT) timber façade systems. Timber samples were obtained from an existing building façade in İzmir, Türkiye, where they had been exposed to outdoor environmental conditions for approximately 14–18 months. Three sample categories were considered: unaged reference samples (UA), less severely exposed samples (A), and more severely exposed samples (B), including variations with and without surface coating. The experimental program combined medium-scale corner tests based on the EN 13823 configuration and cone calorimeter tests according to ISO 5660-1. The medium-scale tests were used to evaluate fire growth behavior through parameters such as heat release rate (HRR), total heat release (THR), and Fire Growth Rate Index (FIGRA), which is the primary indicator for Euroclass classification. Cone calorimeter tests were conducted at multiple heat flux levels to assess ignition characteristics and heat release behavior, and to derive additional fire performance parameters, including the Thermal Response Parameter (TRP) and Fire Propagation Index (FPI). The results indicate that natural ageing influences the fire behavior of FRT timber, although the effect depends on exposure conditions and surface characteristics. In general, aged samples exhibited changes in ignition behavior and heat release characteristics compared with unaged samples, consistent with previous findings on weathering-induced degradation of fire-retardant treatments. The analysis also shows that ageing is not the only governing factor; surface coating condition plays a significant role. Furthermore, correlation analysis between cone calorimeter and medium-scale corner test results demonstrated that cone-derived parameters can provide useful indications of fire growth behavior at larger scales, although uncertainties remain due to differences in test configurations and scale effects.