EXPERIMENTAL AND NUMERICAL INVESTIGATIONS ON EXTERNAL AND INTERNAL FIRE SPREAD IN HIGH-RISE BUILDINGS

Abstract

From past few decades, need for high-rise buildings has increased to meet the modern urbanization demands. However, the complexities of construction and increased fuel loading have resulted in frequent re incidents in them resulting in considerable amount of loss in terms of life and property. This outburst has urged engineers and researchers worldwide to devise their safety measures. Therefore, it is necessary to understand the involved physics in re and smoke spread in high-rise buildings both internally and externally. Undoubtedly, such res cannot be reconstructed physically due to the large-scale involved. Therefore, the only way to investigate these res is by performing experiments at lab-scale with boundary conditions similar to the real scenarios or performing numerical simulations, using Computational Fluid Dynamics (CFD) codes when experiments are not possible or to save cost/time. This is the overall objective of the present work in this PhD thesis to explore and employ small-scale experimental methods and numerical tools for studying external and internal re spread in high-rise buildings. The whole work has been divided into three major parts - Experimental, Numerical and Performance-Based Design (PBD). In the rst part of the thesis on experimental investigations, a new test setup and method including the realistic physics (`chimney-e ect') for exterior wall assembly in high-rise structures has been developed. For the past few years, the externally spreading res (or fa cade res) on modern high-rise structures have occurred due to the mixed contribution of ammable materials in the chimney like construction. At present, only large-scale tests are available, which are expensive to perform while small-scale tests do not incorporate the correct physics of fa cade assembly. Therefore, an experimental setup is developed and demonstrated in the present work, which incorporates the physics for investigating the re response of fa cade materials and `chimney e ect' together at lab-scale. Di erent combinations of exterior cladding (Al-45, Al-45 Class O, Al-45 Class B) and continuous insulation (Expanded Polystyrene EPS, Polyisocyanurate PIR and Mineral wool MW) with and without chimney are selected for studying their re behaviour. The pressure di erential combining with re-radiation (between the samples) led to 3-6 times higher MBR (Mass Burning Rate), ame height and temperature in tests with `chimney e ect' to that of without for the same combination of products. Thermal analysis tests (TGA) for continuous insulation revealed that complete thermal decomposition of PIR occurs over a wide range of temperature, while for EPS, it occurs quite sharply. The low thermal decomposition temperature of insulation materials compared to exterior cladding suggests their vital role in deciding the ammability of fa cade assembly. With chimney, the secondary re sources are generated due to the dripping of products which further enhanced the MBR. A critical width of the chimney (13-50 mm) is established, at which maximum vertical re spread was recorded. Furthermore, the visual observations indicated that the products having thin outer Al sheets are prone to 30 % early structural failure. To correlate the results of the present study with large-scale tests, failure criteria is proposed for assemblies based on critical values of average visual ame height (H=L greater than 1.3, i.e. ames are coming outside the cavity) and the average temperature outside the cavity (higher than 500 °C). Comparison of results with large-scale tests validated the usefulness of the present setup as an intermediate test to study the burning behaviour of fa cade materials and their assembly combinations in the laboratory itself.