Probability analysis of the fire structural resistance of aluminium plate. (July 2016)

• Record Type: Journal Article
• Authors: Afaghi Khatibi, A.; Chevali, V.; Feih, S.; Mouritz, A.P.
• Is Part Of: Fire safety journal. Volume 83(2016:Jul.)
• Page Start: 15
• Page End: 24
• Publication Date: 2016-07
• Subjects: Fire -- Structures -- Aluminium -- Failure -- Model; Fire prevention -- Periodicals; Incendies -- Prévention -- Recherche -- Périodiques; Fire prevention -- Research; Periodicals; 628.92205
• DOI: 10.1016/j.firesaf.2016.04.004
• Languages: English
• ISSNs: 0379-7112

Abstract

Abstract: An experimental and numerical study into the intrinsic scatter in the fire structural resistance of aluminium plate supporting tension loads is presented. Small-scale simulated fire structural tests performed on two aluminium alloys (AA5083 and AA6061) show for the first time a large amount of scatter in the tensile deformation rate and rupture stress. Multiple simulated fire tests conducted under identical heat flux exposure and tensile load conditions reveal scatter in the softening behaviour of the two aluminium alloys; there is large variability in the deformation rate, rupture stress and time-to-rupture, particularly at low stresses when creep dominates the softening process. Finite element analysis (incorporating elastic, plastic and creep softening effects) and elevated temperature material testing reveals that the scatter is caused mainly by variability in creep properties such as creep activation energy. A probability density function is used to quantify the scatter in the creep activation energy, and the finite element model using Monte Carlo simulations computes the scatter in the fire structural resistance of aluminium, which is not possible with existing deterministic models. Graphical abstract: Highlights: We model fire structural properties of two aluminium alloys. We develop new statistical model for fire structural response. Experimental tests of scatter in fire structural resistance. Validation of statistical model using experimental data.