DOI: 10.5593/sgem2017/51/S20.007


J. Mullerova
Wednesday 13 September 2017 by Libadmin2017

References: 17th International Multidisciplinary Scientific GeoConference SGEM 2017, www.sgem.org, SGEM2017 Conference Proceedings, ISBN 978-619-7408-08-9 / ISSN 1314-2704, 29 June - 5 July, 2017, Vol. 17, Issue 51, 47-52 pp, DOI: 10.5593/sgem2017/51/S20.007


Generally, there are 3 basic approaches to fire modelling. The oldest one is full scale modelling, than the small-scale modelling and finally PC simulation modelling such as FDS. If we compare the basic properties of these three approaches, we can compare following criteria. The most important is accuracy, relative costs, professional competence, time of preparation and very important is also the number of possible iterations within a short time period. The highest accuracy still belongs to full-scale models. Full-scale modelling is very often used to get the input data for PC modelling. The realistic accuracy is the biggest advantage. The future belongs to the PC modelling known. But the accuracy is questionable. There are many simplifications and many approximations included in the software. The accuracy depends on the quality of input data. The output of the model can be accurate only with the high quality inputs. That is the reason why PC software need data of real full scale experiments. Many data and many realistic measurements are needed. But full scale modelling is very slow. Full-scale models take too much time for their realisation. Especially models of entire buildings are very rare. Therefore there is the place for small-scale modelling. It enables much more iterations at the same time comparing to the full-scale experiments. Its accuracy is comparable to full-scale modelling. The biggest advantage is the price for experiment reflecting much shorter preparation time. It enables more iterations in various conditions. Afterwards, the output data might be used in the PC modelling.

Keywords: fire modelling, small-scale experiments, full-scale model, software modelling, comparison, flashover