John A. Rockett

Fire Induced Gas Flow in an Enclosure

Abstract The gas flow induced by a small fire in a large room is considered. The fire plume acts as a pump and the window opening as a throttle. Generalizations of Kawagoes expressions for the window air flow and height of the neutral plane are developed and used to rationalize previously unexplained features of Gross and Robertsons enclosure fire data.

Protecting fire fighters exposed in room fires: Comparison of results of bench scale test for thermal protection and conditions during room flashover

Heat flux conditions measured in seven room fires are discussed. The conditions varied from just below flashover in a sparsely furnished bedroom to flashover and severe postflashover fire in a typically furnished recreation room. These heat flux conditions are compared with the protection level provided by fire fighter turnout coats conforming to NFPA 1971,Protective Clothing for Structural Fire Fighting. This standard requires that the turnout coat or pants assembly must protect the wearer against second degree burns when a heat flux of 84 kW/m2 (2 cal/cm2.s) is applied to its outside surface for a minimum of 17.5 seconds [“thermal protective performance (TPP) of 35”]. The results imply that fire fighters have only ten seconds or less to escape under most flashover conditions. However, the turnout coats provide good protection in many other fire situations. Practical definitions for flashover are given, and possible means for making the TPP test more relevant for research and development work are discussed.

Comparisons of NBS/Harvard VI simulations and data from all runs of a full-scale multi-room fire test program

The NBS/Harvard VI multi-room fire model computer code was used to simulate results of previously reported full-scale, multi-room fire experiments. The tests and simulations involved four different compartment configurations made up of two or three rooms connected by open doorways, four different fire types generated by a methane burner located in the room identified as the burn room, and up to four different doorway openings between the burn room and the adjacent space. A total of nineteen different tests were carried out and simulated. Comparisons between simulated and measured parameters of the fire-generated environments are reviewed. While the computer code is generally found to provide favorable simulations for the entire range of tests, several areas in modeling detail are identified as requiring clarification, research and further improvement. The improvements should be incorporated in future versions of the NBS/Harvard multi-room fire model.

Protecting fire fighters exposed in room fires, part 2: Performance of turnout coat materials under actual fire conditions

Seven experimental fires varying in fire load were conducted in a simulated townhouse. Specimens of various current fire fighters turnout coat materials were exposed in the room of fire origin. The time at which conditions would become untenable for the fire fighter due to pain, as well as the time to second degree burn, were calculated. These times ranked the coat specimens in roughly the same order as the “Thermal Protection Performance” measured according to NFPA 1971–1986, especially if the heat in the room developed rapidly.

Objectives and pitfalls in the simulation of building fires with a computer

The complex interactions between a building and a fire are being studied using the National Bureau of Standards computer facility. By highlighting the information needs for a successful calculation, the fire simulation study provides a guide to future research.

Data for Room Fire Models

Abstract Data needs for state-of-the-art single room fire models are discussed using several examples. Three types of data are needed: geometric, thermal and chemical. Needed geometric data generally present no problem and are not discussed. Under thermal data those quantities which determine the transient surface temperature of objects in the room are considered. For inert materials, these should present no problem but few materials are inert. Even materials used for non-combustible walls or ceilings may have thermal properties which charge significantly during the fire development. Chemical properties are the additional data needed to define the burning behavior of materials. These present a more serious problem. Currently, the Harvard simulation needs fifteen properties and pseudo properties (surrogates for complex combinations of more fundamental properties not currently resolved by the models). The importance (relative sensitivity of the simulation predictions) and the availability of useful data are...

Two approaches to the analysis of actual fires

Two calculations are described. One used only simple algebra to show the rate of development of a critical aspect of a fire. The other used one of our most elaborate computer based schemes to extend the results of full-scale fire tests to additional, important situations. Both provided useful results. The significance of this is that it is not the complexity of a calculation that is important but its relevance to the problem at hand.