The cargo compartment tests were to further evaluate SF₆ and to study HFC 125 in this type of scenario. The main objective of this evaluation program was to determine the decay characteristics of these simulants while exposing them to different air leakage rates and to compare the results with Halon 1301. SF₆ was further evaluated by running two series of tests utilizing a fire suppression metering system after the activation of the main total-flooding system. In this scenario, the comparison was based on the similarity of volumetric concentrations after activating the metering system.

The evaluation tests were conducted in the Fire Safety Section facilities at the FAA William J. Hughes Technical Center; the test articles used in this study were two different size aircraft, a narrow-body aircraft and a wide-body aircraft. The narrow-body cargo compartment had a total cargo volume of approximately 625 cubic feet while the wide-body aircraft had a cargo volume of 2000 cubic feet. Both compartments had a ventilation system to control the leakage rate.

For economic reasons, the narrow-body compartment was divided into two sections; the section selected for this study had a volume of 300 cubic feet. This small cargo bay was equipped with a leak system, a modular total-flooding fire suppression system, an open-loop fire suppression metering system, pressure valves, blowout panels, gas sampling probes, and measuring instruments (Refer to figure 1). A variable-speed fan and a perforated duct controlled the leakage rate of the cargo compartment. The leak system was capable of providing a maximum flow of 21 CFM. The leakage rate of the cargo bay was set and measured using a digital anemometer (Omega Model HH-30). The L-shaped perforated duct had a height of 28", a width of 39", and a 3" diameter orifice. This duct had a total of thirteen 1/2" diameter holes. The holes on the horizontal leg were spaced 5" apart from each other while the holes on the vertical leg of the duct were spaced 2.5" apart. A Powerstat Variable Autotransformer (PVA) controlled the fan rotational speed. It was planned that three fan flow settings were going to be used during the testing program: 2 CFM, 11 CFM, and 21 CFM. The leakage rate of 2 CFM was obtained by opening two holes, one on each end of the duct legs; the remaining holes remained closed. The PVA was set to 40%. The 11- and 21-CFM leakage rates were obtained by opening all of the holes and removing the duct cap found at the top of the vertical leg. The only difference between their settings was how the PVA was dialed: 60% for the 11-CFM leakage rate and 100% for the 21-CFM leakage rate.

The small compartment’s fire suppression system consisted of two modular systems: a total-flooding system and a metering system. The total-flooding system consisted of a Fenwal fire bottle connected to a 7/8" manifold with a single nozzle at the end. The manifold was approximately 10' long. The fire bottle was activated using an electronic control head. The open-loop metering system comprised of a pressure-regulated cylinder, regulated to 40 psig, connected to a 0.25" copper tube that had a single nozzle with a 0.35" orifice at the other end; the metering system setting and size were arbitrarily selected. This system was activated manually.

There were two types of gas analyzers used in this test program: the Heat Technology Laboratory (HTL) Halonyzer and the Rosemount 880A analyzer. The Halonyzer, calibrated to detect the volumetric concentrations of HFC 125 and Halon 1301, offered a gas collection capability of 12 probes. This system was used in the small compartment and with HFC 125 and Halon 1301 only due to its temporary availability. Five Rosemount 880A analyzers, calibrated for HFC 125, Halon 1301, and SF₆, were available for both compartments; only one of the five analyzers was calibrated for SF₆. This equipment provided a gas collection capability of four probes that could be used for Halon 1301 or HFC 125 and one probe for SF₆. Even though there were five analyzers on hand, only one was used at a time during some of the testing in order to maintain the same flow conditions and gas dynamics when testing SF₆. When the Halonyzer was used at the beginning of this study, all 12 probes were used and where located in two different trees. Each tree had six levels spaced 7.5" apart vertically starting from the floor. The sampling probes, 0.25" copper pipes and equal in length, were designed to convey samples at a uniform flow rate from the sampling area to the analyzer units.

During the wide-body aircraft test series the entire 2000 ft³ compartment was used. The ventilation system of this aircraft provided a leakage rate of 50 CFM. This bay was equipped with four main gas collection probes and eight backup probes. The fire suppression system consisted of an unbalanced banked system with five nozzles. A closed-loop fire suppression metering system was present but was not used during these tests.

Discharge tests were conducted for all of the identified candidates as well as the baseline agent using the following procedures.

The installed fire bottle was charged with the desired agent to a weight that produced a 5% volumetric concentration in the cargo compartment; there is one exception to this initial concentration in the case of HFC 125. During the first series of testing with HFC 125 in the small compartment, the fire bottle was filled based on the equal liquid fill approach rather than the equal volumetric concentration approach; this resulted in an initial volumetric concentration of 4.7%. This liquid fill approach maximizes agent density at the point of discharge, provides a consistent energy source during discharge, conserves the amount of gaseous nitrogen, and results in equivalent nozzle discharge pressure (Grosshandler et al., 1995). This approach was later changed during the large compartment testing to produce an initial volumetric concentration of 5%. The amount of Halon 1301 and SF₆ used to achieve 5% in the small compartment was 5.8 pounds and 5.7 pounds, respectively. In the case of HFC 125, 4.4 pounds were transferred into the fire bottle to achieve 4.7% volumetric concentration in this compartment. The amount of agent for the large compartment testing was 38.5 pounds of Halon 1301, 31.0 pounds of HFC 125, and 37.7 pounds of SF₆. After charging the bottle with the agent, the bottle was pressurized with nitrogen gas to 375 pounds per square inch absolute.

After charging and plumbing the bottle, the leak fan was set to the appropriate leak rate. Once this was set, and the analyzer(s) was (were) ready, the fire bottle was discharged simultaneously with the activation of the data acquisition system. The data acquisition then collected the compartment's concentration history, temperature, ambient humidity, and leak rate. During the tests that were designed to use the fire suppression metering system, the system was activated when the volumetric concentration dropped below 3%. Tests without the metering system ran for 30 minutes, and the tests with the metering system ran for 45 minutes.