Synchronous Displacement Acquisition Approach for Early Warning of Fire-induced Collapse of Steel Portal Frames

Unexpected collapse accidents of burning buildings have greatly threatened firefighters and trapped people. Hence, accurate real-time early warning of fire-induced collapse of burning buildings is urgently needed to avoid unnecessary casualties. Existing research evaluated the collapse state of the burning building and issued an early warning before the collapse by analyzing the variation laws of displacements at key positions. However, some displacements are hard to acquire rapidly and accurately at fire rescue scenes, which restricts the application of early-warning methods in real fires. This paper proposes a synchronous real-time displacement acquisition approach for early warning of fire-induced collapse of steel portal frames. The three-order polynomial function is adopted to represent the deformation curve of members in frames under fire conditions. The displacements measured by on-site arranged microwave radars coupled with rotations measured by embedded inclinometers are adopted for determining the polynomial function and synchronously predicting displacements of the burning frame at any location. The research findings show that placing five inclinometers on each roof and three inclinometers on each internal column with equal spacing can accurately predict displacements of the overall frame. Slope-based weights and reset conditions with thresholds are proposed to consider the effects of heating conditions and structural parameters in order to improve the accuracy and reliability of the acquisition approach. Parametric analyses and case studies are presented to illustrate the application and robustness of the proposed approach in early warning of fire-induced collapse of steel portal frames.