Abstract of the paper: ---------------------------- External Thermal Insulation Composite System (ETICS) facades with expanded polystyrene (EPS) insulation and thin rendering are applied frequently in buildings. Considering high combustibility of EPS, with these facades concerns also arise regarding spread of a possible fire between neighbouring compartments of high-rise buildings. Fire tests of two large-scale facades were performed to study two parameters presumably influencing significantly the fires of such facades in real-life settings, i.e. incident heat flux upon the facade%s surface (IHFFS) and damage of the facade%s render (the latter being a consequence of poor or unfinished construction work, ageing or fire-induced thermal strain). The first facade was rendered fully and was exposed to moderately-fast increasing IHFFS. In the second (partially unrendered) facade case the IHFFS progressed faster. The facade flame body (temperatures and shape) was monitored by thermocouples, photo and video cameras. For detection of melting of EPS and internal burning, thermocameras were used within the facades areas outside the visible plume. In the plume zone, however, a group of thermocouples was embedded inside EPS and the shapes (plateaus and slopes) of the collected time-temperature graphs were observed for these purposes. The IHFFS imposed on the facades during fire testing were estimated by numerical calculations. In both cases the first pronounced render crack was evolved at the estimated average between-windows IHFFS of around 30 kW/m2 and was followed by internal burning of EPS. While the latter did not seem to spread across the facade for the fully-rendered facade, a fast fire spread was detected for the second specimen. Meaning of the paper for the project: ----------------------------------------------- This paper is highlighted as one of the most important scientific result of the postdoctoral project because it not only presents experimental data from a fire test of two EPS ETICS facades (Section 2.2.2) but also explains results from the corresponding computational simulations (see Section 2.3 of the paper). These were carried out as two of the four test cases mentioned in the penultimate paragraph of point 3 of this report. The simplified computational models of both façades were prepared in FDS and mainly served for comparison (validation) of the measured and calculated façade flame temperatures. A good agreement between both results was discovered. Combustion of ignitable materials (wooden cribs at the foot of the facades and wooden cladding of the second façade) was for purposes of the facade fire models described by simplified material submodels prepared in the developed tool GeneticMat.
COBISS.SI-ID: 2240871
The article compares two approaches to modeling creep of steel in numerical analyses of the mechanical response of a steel structure during a fire; the simpler approach where creep strains are considered implicitly in the plasticity model and the more complex approach where creep strains are considered separately (in an explicit manner). For conventional engineering applications, the first approach is most commonly used, however the limits of its validity are often neglected. The severity of the error that can happen as a result, has only been shown for individual structural members in the literature so far. This paper presents such comparison on structural assemblies, where the influence of creep is most pronounced in the between-member connections. The results show that failure times predicted by the implicit-creep plasticity model are significantly longer (up to 40%) than failure times predicted by the explicit creep model, if the implicit model is used outside the limits of applicability. Within the limits, the discrepancies of about 20% were found for one of the explored steels. The possible consequences of improper use of the simplified methods of modeling the fire response of steel structures, as described in this article, were investigated using the model FIRESIM-II which was developed within workpackage DP III of the project. Presentation of the results at the conference, which is organized every year by the Slovenian Association of Construction Constructors and is visited by a large number of structural designers and scientists from Slovenia and elsewhere, ensured proper promotion and dissemination of the collected findings. This paper is available at the project webpage: http://arrs-firesim.zag.si/ .
COBISS.SI-ID: 8234849