Abstract
This paper describes a series of room fire tests using upholstered furniture items for comparison with their open burning rates, previously determined in a furniture calorimeter. For the four tests conducted good agreement was seen in all periods of the room fires, including post-flashover, noting that only fuel- controlled room fires were considered. Difficulties in making accurate mass and heat flow measurements in the room's window opening were found, and it is sug gested that with present day instrumentation only exhaust stack measurements are reliable. Finally, a number of simplified rules or theories for predicting room flashover based on room physical properties and open-burning heat release values were examined and compared. Broad agreement was generally found, with recommended ones selected on the basis of well-controlled asymptotic behavior.