SharePoint
Aide
Faire avancer la sûreté nucléaire

La Recherchev2

Publications

Smoke propagation between rooms in forced ventilated and confined multi-enclosures fire scenario – Preliminary investigations on reduced scale models


Fermer

Authentification

Email :

Mot de passe :

"Workshop on Fires in enclosures" – University of Ulster – Northern Ireland, 30 et 31 mai 2006

O. Vauquelin(1), P. Bournot(1), C. Lucchesi , H. Pretrel, JM. Such.

Type de document > *Congrès/colloque

Mots clés > sûreté, incendie, propagation

Unité de recherche > IRSN/DPAM/SEREA

Auteurs > LUCCHESI Caroline, PRETREL Hugues, SUCH Jean-Marc

Date de publication > 27/03/2006

Résumé

This paper deals with fire-induced smoke propagation in multi-mechanically ventilated enclosures that remains a key scenario for the safety assessment in nuclear industry. In this framework, the French "Institut de Radioprotection et de Sûreté Nucléaire" (IRSN) in collaboration with the University of Mediterranée has undertaken a research program to investigate smoke propagation mechanism between several confined and mechanically ventilated rooms. Carried on researches performed in semi-open configuration, the current objective is to investigate the effect of the complete confinement and the role of the mechanical ventilation on smoke movements with using experimental reduced scale models. Two models representing two connected rooms equipped with a ventilation system are used. The fire is initiated in one room and the smoke propagates in the other room through a door. Objectives of these preliminary experiments are to validate experimental approaches and also to get first results. The first model is a thermal one. Scale reduction is 1:5 and hydrocarbur pool fire is used. For several heat release rates (i.e. several pool diameters), local temperatures, burning rate and local conductive heat flux are then evaluated. Flow visualisations on the flow pattern. The second model is an isothermal model in which a buoyant fluid (a mixing of air and helium) is continuously released into the model. Scale reduction is also 1:5. Flow rate and density of the mixing are chosen in order to reproduce correctly the convictive part of the fire heat release rate. Similarity rules are based on Froude number conservation. Although this model does not reproduce combustion and the whole thermal effects, it is noticed (from visualisations and velocity measurements) a good qualitative agreement with the reduced scale thermal model. Conclusions are focused on the ability of reduced scale models to simulate smoke propagation in confined and forced ventilated enclosures and on future investigations.

(1) : IM2/UNIMECA -Marseille