Search results

This paper aims to present a part of a coupled numerical model for prediction the fire resistance of elements in a horizontal furnace. Temperatures calculated inside the timber beam are compared to measured values from the fire test.

The paper presents a part of a coupled numerical model for prediction the fire resistance of elements in a horizontal furnace. The presented part lies in a virtual furnace which simulates temperature environment around tested elements in the furnace. Comparison of results show good agreement in the case when burning of timber is included in the numerical model.

The virtual furnace presented in this paper allows to calculate temperature environment around three timber beams. After validation of the fire dynamics simulator (FDS) model, the temperature conditions are passed to the FE model which solves heat transfer to the tested element. Temperatures inside the timber beam which are solved in software Atena Science are compared to measured temperatures from the fire test. The comparison of temperatures in three control points shows good accuracy of the calculation in the point closer to the heated edge. An inaccuracy is shown in points located deeper in the beam cross-section – below the char layer.

In conclusion, the virtual furnace has a great potential for investigating the thermal behaviour of fire-resistance tests. A huge advantage inheres in the evaluation of the thermal effect throughout the volume of the furnace, which allows an accurate prediction of fire-resistance tests and evaluation of large number of technical alternatives and boundary conditions. However, passing the temperature field from the FDS model into FE model may decrease the level of accuracy. The solution lies in a coupled CFD-FE model. A weakly coupled model including fluid dynamics, heat transfer and mechanical behaviour is under development at Faculty of Civil Engineering, Czech Technical University in Prague. The fluid dynamics part which is presented in this paper is solved by FDS and the thermo-mechanical part is computed by object-oriented finite element model (OOFEM). The interconnection of both software is made owing to MuPIF python library.

The virtual furnace takes advantage of great possibilities of computational fluid dynamics code FDS. The model is based on an accurate representation of a real fire furnace of fire laboratory PAVUS a.s. located in the Czech Republic. It includes geometry of the real furnace, material properties of the furnace linings, burners, ventilation conditions and tested elements. Gas temperature calculated in the virtual furnace is validated to temperatures measured during a fire test.

The virtual furnace has a great potential for investigating the thermal behaviour of fire-resistance tests. A huge advantage inheres in the evaluation of the thermal effect throughout the volume of the furnace, which allows an accurate prediction of fire-resistance tests and evaluation of large number of technical alternatives and boundary conditions.

The virtual furnace has a great potential for investigating the thermal behaviour of fire-resistance tests. A huge advantage inheres in the evaluation of the thermal effect throughout the volume of the furnace, which allows an accurate prediction of fire-resistance tests and evaluation of large number of technical alternatives and boundary conditions. However, passing the temperature field from the FDS model into FE model may decrease the level of accuracy. The solution lies in a coupled CFD-FE model. A weakly coupled model including fluid dynamics, heat transfer and mechanical behaviour is under development at Faculty of Civil Engineering, Czech Technical University in Prague. The fluid dynamics part which is presented in this paper is solved by FDS and the thermo-mechanical part is computed by OOFEM. The interconnection of both software is made thanks to MuPIF python library.

This paper aims to present the assumptions and the issues that arise when developing an integrated modelling methodology between a computational fluid dynamics (CFD) software applied to compartment fires and a finite element (FE) software applied to structural systems.

Particular emphasis is given to the weak coupling approach developed between the CFD code fire dynamics simulator (FDS) and the FE software SAFIR. Then, to show the potential benefits of such a methodology, a multi-storey steel-concrete composite open car park was considered.

Results show that the FDS–SAFIR coupling allows overcoming shortcomings of simplified models by performing the thermal analysis in the structural elements based on a more advanced modelling of the fire development, whereas it appears that the Hasemi model is more conservative in terms of thermal action.

A typical design approach using the Hasemi model is compared with a more advanced analysis that relies on the proposed FDS–SAFIR coupling.

Standard fire resistance curves such as ASTM E119 have been used for so long in structural fire practice. The issue with use of these curves that they do not represent real fire scenarios. As a result, the alternatives have been to either conduct experiments or find other tools to represent a real fire scenario. Therefore, the purpose of this paper is to understand the temperature effects resulted from a designed fire on steel beams and whether the standard fire curves represent a designed fire scenario.

Computational fluid dynamics (CFD) models were developed to simulate a designed fire scenario and to understand the structural responses on the beams under elevated temperatures. Consequently, the results obtained from the CFD models were compared with the results of three-dimensional (3D) non-linear finite element (FE) models developed by other researchers. The developed FE models were executed using a standard fire curve (ASTM E119). A parametric study including two case studies was conducted.

Results obtained from performing this study showed the importance of considering fire parameters such as fuel type and flame height during the thermal analysis compared to the standard fire curves, and this might lead to a non-conservative design as compared to the designed fire scenario. The studied cases showed that the steel beams experienced more degradation in their fire resistance at higher load levels under designed fires. Additionally, the models used the standard fire curves underestimated the temperatures at the early stages.

This paper shows results obtained by performing a comparison study of models used ASTM E119 curve and a designed fire scenario. The value of this study is to show the variability of using different fire scenarios; thus, more studies are required to see how temperature history curves can be used to represent real fire scenarios.