A Fuzzy Logic Framework for Rating Post-fire Structural Performance of Existing Structures via Visual Inspection

Visual inspection is the primary method for preliminary fire damage assessment; however, its reliability is questionable. Not only are inspection results highly subjective to the inspector’s judgment, but also, the results do not directly indicate the reduction in structural performance (e.g., gravity load capacity, lateral load capacity, deflection, etc.). Thus, this study proposes a new post-fire performance rating (PF-PR) system using fuzzy logic principles. The PF-PR system has five modules, including (1) the “Performance criteria” used as the basis for assessment, (2) the “Material damage models” which predict the reduction in material properties with fire damage (e.g., reduction in concrete strength and steel yield strength), (3) the “Fuzzy logic model” which maps subjective visual observations of fire damage into quantitative ranges of heating temperatures and material properties reductions, (4) the “Fuzzy-numerical analysis” which uses the material properties reductions as an input to a numerical model that evaluates the performance, and (5) the “Performance charts” which are the end-product of the framework and visual representations of the analysis results. These performance charts can support inspectors in mapping the findings of visual inspection (e.g., linguistic descriptions of fire damage) into quantitative structural performance measures. Not only would the proposed PF-PR system increase the reliability of visual inspections, but it would also allow stakeholders to reliably compare between different post-fire actions (e.g., repair or replace) from an economic perspective. The development of the PF-PR system is demonstrated for a case study of a reinforced concrete beam.