In May 2016, a fire occurred unexpectedly in a building of the industrial (meat processing) plant in Sevce (Laško, Slovenia). The most likely cause of the fire was the ignition of flammable polyurethane insulation (PUR) of a wall panel, from where the fire slowly spread onwards across its interior and gradually embraced the entire ceiling (also from the same PUR panels) of the fire compartment. The unfortunate event did not require serious consequences; nevertheless, after the fire the safety of the steel roof truss (this extended directly above the burnt-out ceiling panels and had no fire protection) was brought to question. Since this event was thematically related to topics discussed in the postdoctoral project Z7-7677, collecting the data from this fire and using them also for the needs of this project was considered as an excellent opportunity. Using the data that one manages to collect directly from a fire site and by talking to eyewitnesses is, namely, often invaluable in terms of the development and validation of computational models of fires, because the alternative way (i.e. searching for the information about past fires in the available literature) is often too superficial and unreliable. Following a conversation with the owner of the plant in Sevce, the leader of this postdoctoral project obtained a permission for the inspection of the burnt site of the fire and, accordingly, the site was carefully photographed and the statements of eyewitnesses were thoroughly documented. In addition, samples of PUR were taken from the unburned wall panels and then analysed with a cone calorimeter in the ZAG Fire laboratory (in accordance with the specifics of the fire, i.e. slow burning of PUR underneath metal sheets of the panels, smouldering combustion and combustion at low oxygen concentrations were inspected as a priority). Finally, the data obtained in this way were used for the preparation of a simplified computational simulation of the fire and its consequences using the model FIRESIM-I and, thereby, for a (partial) validation of this model and its supplementary tool GeneticMat (the first validation case as mentioned in the penultimate paragraph of Section 3 of this report). The results of the simulation showed a good agreement with the conclusions of other investigators of the fire scene who concluded (after taking samples of steel from the exposed steel roof truss) that the maximum temperature of the steel did not exceed 300°C in the fire and that the steel roof structural system was not damaged seriously. Correspondingly, it is still able to bear the load for which it was originally designed. Publication of the results of this test case in professional journal POŽAR (eng. 'FIRE') also took care of the promotion of the project, which in this way certainly reached the widest possible group of experts working in the field of fire safety in Slovenia. This paper is available at the project webpage: http://arrs-firesim.zag.si/ .
F.06 Development of a new product
COBISS.SI-ID: 2270823As in the case of the fire of the industrial building of the meat processing plant in Sevce (Laško), described above, the project leader also had the opportunity to be involved in the after-fire inspection of the ASP building (vehicle store) in Jesenice which was damaged by a fire at the end of 2016. The fire started by the ignition of ceiling lamps in the basement of the building, where more than 3000 automotive tires were stored. Since the fire of this building was thematically related to the topics discussed in this postdoctoral project, this was a great opportunity for the collected data to be also used for a simplified simulation of the fire and its consequences and, thereby, for a partial validation of the GeneticMat tool and the model FIRESIM-I. Samples of unburnt tires were taken from the fire-site for the purpose of this study and then the combustion of the samples was carefully studied in the cone calorimeter of ZAG Fire laboratory. Publication of the results of this test case in professional journal POŽAR (eng. 'FIRE') also took care of the promotion of the project, which in this way certainly reached the widest possible group of experts working in the field of fire safety in Slovenia. This paper is available at the project webpage: http://arrs-firesim.zag.si/ .
F.06 Development of a new product
COBISS.SI-ID: 2340967The key objective of the paper is to present modern, more reliable approaches to computer modelling of the development of fires. The presentation is carried out on the example of the model for simulation of a fire in a factory for production of XPS panels (aplication of model FIRESIM-I, which was developed within the post-doctoral project; see Section 3 of this report, DP III). This model is primarily used for analyses of the progressive ignition of combustible XPS panels within the analysed hall and the release of heat and gaseous products during their combustion. The basis for these analyses are pyrolysis and combustion sub-models of the analysed XPS which were developed by the project leader and then embedded into one of the well-known and well-established fire analysis software, i.e. FDS (Fire Dynamics Simulator). Finally, an analysis of the transport of released heat and gases through the interior of the analysed hall was carried out, where the system of transport equations, better known as numerical models of fluid dynamics (CFD models) and thermal radiation, was used as the basis. The developed pyrolysis and combustion material sub-models represent the greatest novelty of the content presented in this paper and the greatest contribution to the development of science and profession. Thus, the paper is devoted primarily to the description of the characteristics of basic features of these sub-models and to how they were developed. With the combustion sub-model, the heat of combustion of the analysed XPS (measured with a bomb calorimeter) and the equation of the chemical reaction in the gaseous phase (stoichiometric equation) were first defined. Because XPS is a simple thermoplast and styrene polymer with a molecular formula (H5C6-CH = CH2)n, the equation was taken of the form CxHyOzNv + nO2 O2 → nCO2 CO2 + nH2O H2O + nCO CO + nS soot + nN2 N2, where the coefficients nCO2, nH2O, nCO, nS, nN2 were determined by testing XPS samples in a cone calorimeter with FTIR analysers. Furthermore, the pyrolysis sub-model was prepared defining heat decomposition of the material ("break-down" of bonds between molecules, which are of different types and decompose at different temperatures, during material heating and consequently loss of material mass). For the basis the following equations were used: (i) a system of Arrhenius equations of thermogravimetric material decomposition, and (ii) a group of Fourier's equations of heat transfer through a solid, where the internal heat source due to heat of the above-mentioned internal reactions was also considered. The parameters of these equations (set of the so-called thermal and kinetic parameters) were determined by numerical fitting of the pyrolysis sub-model to results of high-temperature material experiments (TGA / DSC and cone calorimeter tests). In order to capture all the possible situations that the material could be exposed to during a fire, the experiments were repeated at different heating rates and different concentrations of available oxygen. The final part of the paper also presents the results of validation of the FIRESIM-I model. This was done by comparing the results of the computerized simulation of the fire that happened in 2013 in the factory for production of XPS panels Fibran Nord d.o.o. (Sodražica, Slovenia) and the actual data collected from this fire (data from eyewitnesses, firefighters, forensics and other inspectors). The agreement between the data in terms of the speed of the fire development, the duration of the fire, the released amount of smoke (visibility in the hall during the fire) and the evolved maximum air temperature was very good. This paper tries to present new scientific methods in a simple way, so that they can be understood also by engineers and other fire safety experts who are regular readers of the Požar journal. This paper is available at the project webpage: http://arrs-firesim.zag.si/ .
F.18 Transfer of new know-how to direct users (seminars, fora, conferences)
COBISS.SI-ID: 2340711This is a paper presented at scientific conference ISAMA’18: International Symposium on Advanced Materials and Application, January 19-21, 2018 (Seoul, Korea). The paper presented the possible deviations between “realistic” (performance-based) calculations of fire resistance of a selected steel structure and corresponding calculations made by one of the often used simplified (prescriptive) procedures of EN 1993-1-2, i.e. the method of critical temperature (MCT). The comparison was done for a case of an assembly consisting of a steel beam and a steel girder connected to each other by a bolted fin-plate connection. For such structure the MCT method suggested that the structural fire resistance was 50 minutes. However, the realistic fire resistance calculated by an advanced performance-based procedure was evaluated to 44 minutes. Although the discrepancy between the results of both methods was not significant in the presented case, this finding implied that MCT can be on the unsafe side for some cases. More future debates and clarifications were therefore ecouraged with this presentetion regarding the actual limits of the applicability of the simplified procedures. Model FIRESIM-II, developed under workpackage DP III, was implemented for analyses of this paper. As it is customary in practice, the considered steel structure was assumed to be fire-protected. A protection with a fire-protective intumescent coating was assumed. For this purpose, the project leader also proposed a new (modification of an existing) material model for calculation of insulating efficiency of the coating, which she developed using experimental data, accessible through the archives of ZAG Fire laboratory. Poster with the summary of this paper is available at the project webpage: http://arrs-firesim.zag.si/ .
F.18 Transfer of new know-how to direct users (seminars, fora, conferences)
COBISS.SI-ID: 2350183This is a paper presented by the project leader at scientific conference ISAMA’18: International Symposium on Advanced Materials and Application, January 19-21, 2018 (Seoul, Korea). The paper presented possible deviations between advanced (performance-based or ‘realistic’) and simplified calculations of initiation and spread of a fire in a factory for the production of insulation panels made of extruded polystyrene (XPS). Simplified procedures, discussed within the paper, are founded on several assumptions regarding the evolution of the fire. They are normally based on a great deal of practical experience of the fire designer and/or reports of eyewitnesses of similar fires. Provided that the latter are easily accessible, such procedures are fast and can provide a reasonably good first impression regarding the expected fire as the paper shows. However, they also almost inevitably lead to temperatures which are somewhat underestimated (up to 30%) in the most intense fire phase but are overestimated in the decaying phase. Hence, for a more reliable calculation, advanced procedures are recommended. Model FIRESIM-I, developed under workpackage DP III, was implemented for analyses of this paper. Poster with the summary of this paper is available at the project webpage: http://arrs-firesim.zag.si/ .
F.18 Transfer of new know-how to direct users (seminars, fora, conferences)
COBISS.SI-ID: 2349927