Fire behavior of sandwich panel core materials in the pre-flashover phase

This research is commissioned by the Dutch fire department and the Eindhoven University of Technology (TU/e). The Dutch fire department is increasing their scientific knowledge of events that are happening during a fire. They suspect sandwich panels with a combustible core to be a potential hazard, after several serious events in buildings constructed with these panels. This research contains a summary of the literature study and an indicative research. The literature study done together with L.L de Kluiver gives a short overview of the available literature about fire hazards regarding sandwich panels in an objective way. The research made use of different journals in the field of fire safety engineering and building materials, case study reports made by different companies, as well as information from manufacturers and Euro codes. Topics involved in this research are: sandwich panel, thin metal sheets, combustible insulation, polyurethane, polyisocyanurate, pyrolysis, and steel facings. The indicative research mainly contains experiments and simulations in order to create insight in the total amount of pyrolysis products in a smoke layer. Both parts of the research are to check the validity of the concern from the Dutch fire department. The results from the literature study show that the concern of the Dutch fire department might be correct although most of the literatures focuses on fully developed fires. The indicative research is designed to create more insight in the actual behavior in the pre-flashover phase. In order to determine if the sandwich panels, which are mainly applied on Dutch buildings pose, threats for fire fighters The literature study has focused on the hazards of the sandwich panel as a building product. The fire hazards of combustible cored sandwich panels are not clear yet. Fire fighters see things happening which are not supposed to happen according to official fire tests. Literature describes delamination and falling down of the metal facings, which are a potential hazards for fire fighters. Official fire tests do not describe this event. Besides that, cores can be exposed earlier in real fires than fire test show, due to a greater influence of the buckling effect. The mixture of smoke gases and pyrolysis gases can become a potential mix for smoke-gas explosions. The results of the indicative research show: that delamination does occur in the pre-flashover phase, all tested samples show signs of delamination or loss of structural strength at the upper range of the 6 Ing. A.W. Giunta d’Albani pre-flashover phase. Panels are normally mounted on purlins, so the chance of panels falling down is minimal. Some deformation of panels has been seen, but this test has been to small to make any statement about the buckling effect. Surprisingly the actual massloss by pyrolysis of synthetic cores and mineral wool based cores does not differ much up to 300 degrees Celsius. The mass-loss of PUR panels is exponential and PUR starts losing a significant amount of mass around 300 degrees Celcius, whereas the PIR and stone wool panels have a lower pyrolysis temperature and show a more linear trend. The radiation flux of the smokelayer is an important factor by which the maximum temperature can be calculated. The fire fighting tenability limits has been suggested as a maximum radiation flux of 4,5 Kw/m2 at 1.5 m above the floor and a minimum height to the bottom of the smoke layer of 2 m” [19]. A temperature of 350 °C would give a heat flux of 8 Kw/m2, this radiation level marks the upper limit for a fire fighters to apply an offensive fire repression within the fire compartment. Since the radiation level can be reduced by a factor 0.5 when working close to the sides of the compartments. When results of the experimental test are used in the Spreadsheet calculations. The simulations show that even in the most extreme situations the limit of the 39% mass of flammable gases of the total smoke layer will not be reached. The poultry farms (long and low buildings) show with 22.6% the highest amount of flammable gasses of the total smoke layer. Most buildings do not generate more than 5% pyrolysis gases of the total smoke layer at temperatures up to 270 degrees Celsius ( 4,5 Kw/m2), and 11% at temperatures up to 350 degrees Celcius ( 8 Kw/m2) During these simulations completely closed buildings where calculated. The outlet of pyrolysis gases has not been taken in account. All results generated in these calculations are project specific, and dot not contain compartment larger than 1600m2. To conclude, sandwich panels with synthetic and mineral wool cores both emit pyrolysis gases when exposed to heat. The potential hazard of a flammable mixture occurring in the smoke layer lies above the temperature range to which fire fighter can safely be exposed. Sandwich panels in any form are not dangerous as a building material, and do not pose any threat during normal use of the building. However, during a fire the presence of combustible additives in sandwich panels may in some cases, and mainly after pre-flashover increase the intensity of the fire.