Report Documentation Page Fire Suppression Efficiency Screening Using a Counterflow Cylindrical Burner N/a
Jiann-Cherng Yang,Michelle K. Donnelly,M. Prive,W. Grosshandler
Abstract:The design and validation of a counterflow cylindrical burner for fire suppression efficiency screening are described. The stability limits of the burner were mapped using various 0 o I Propane 0.03 L/s fuel (propane) and oxidizer (air) flows. The stability envelopes 4 compared favorably with those reported in the literature. The . No water application apparatus was characterized using inert gases (argon, helium, 0.3 0 With 0.005 g/s of water and nitrogen), and the relative fire suppression efficiency ranking of these three gases was found to be commensurate 0.2 with that from cup-burner tests. For liquid suppression 0) experiments, a piezoelectric droplet generator was used to form 0.1 * droplets ( < 100 gim). Water was used as a representative liquid 8 * suppressant to study the feasibility of using such a burner for 0.0 2. . .400 I00O 0 300 00 500 600 700 screening liquid agents. Extinction was facilitated with the addition of water droplets, and the effect of water became more (2 Vo /R)blow.off pronounced when its application rate was increased. Suppression experiments using water with and without nitrogen dilution in the oxidizer stream were also performed. Flame Figure A-I Mass fraction of nitrogen added in air as a extinction due to the combined effect of water and nitrogen function of stagnation velocity gradient at blow-off dilution was demonstrated. This paper was produced in the course of work by the U.S. Government and hence is in the public domain. INTRODUCTION 1971) or non-intrusive (Dreier el al., 1986; Sick et al., 1990) The recent ban on halon 1301 (CF3Br) production (as a probing techniques, thus enabling detailed studies of flame result of its ozone depleting potential) has resulted in extensive structure. search for its replacements and alternatives. Potential uses of liquid agents as replacements have been recently proposed in EXPERIMENTAL APPARATUS several applications (e.g., shipboard machinery spaces, engine Burner compartments in armored vehicles). The applications of fire The design of the burner is based on several important suppression efficiency screening methods constitute an criteria. The burner has to be robust, easily built, installed, and important aspect of this search process because good screening operated, and able to generate reliable screening test data. methods can facilitate the identification, comparison, and The burner is a replaceable porous (20 pm) sintered selection of potential candidates for halon replacement. Most stainless steel standard '/2" UJNF threaded cup filter with a of the current screening methods (e.g., cup burners) are length of 3.18 cm. (1.25"), an i.d. of 1.12 cm (0.44"), and an designed for evaluating fire suppressing agents that are o.d. (D) of 1.58 cmn (0.625"). The advantage of this burner delivered in the form of vapor although they have been design over those used in the past is that burner replacement occasionally used for evaluating condensed-phase agents (e.g., can be easily performed if partial or complete clogging of the sodium bicarbonate powder, Grosshandler et al., 1994). porous burner surface occurs. The burner is screwed onto an Currently there is no apparatus designed specifically for extended insert through which fuel is injected into the interior screening liquid agents; therefore, there is a need for the of the porous filter and double-pass cooling water runs. The development of a reliable screening method for liquid agents cooling water is used to cool the burner (to prevent damage to that can be delivered in droplet form. The objective of our work is to design, construct, and demonstrate a laboratory-scale Sintered stainless steel Insert Fe apparatus that can perform the screening of liquid agents in a well-controlled experimental setting. The design of the apparatus is based on a well-characterized flame, a means to I facilitate the introduction of small amount of liquid agents, and M =;. a way to generate liquid droplets that can be entrained into theVT TT flame. These individual components will be described in detailColn in the following sections.wae Since a porous cylindrical burner in a counterflow configuration has been extensively used to study flame Figure I Cross sectional view of burner structure (Tsuji and Yaxnaoka, 1967; 1969; 1971, Tsuji, 1982) and flame extinction using inert gases (Ishizuka and Tsuji, the porous surface structure) and .the fuel (to prevent fuel 1981) and halons (Milne et al, 1970), it is logical to extend the pyrolysis prior to its ejection through the porous surface). A application of such a burner configuration to liquid fire cross sectional schematic of the burner interior is shown in suppressants. hin our liquid agent screening apparatus, we make Figure 1. use of such a burner wherein a difrhsion flame is formed in the .The burner, together with the insert, does not forward stagnation region of the burner placed in a uniform completely span the entire test section of the wind tunnel. A oxidizer flow, with fuel being ejected uniformly from the burner cylindrical brass rod (same diameter as the burner) with internal surface. Other counterfiow geometry has recently been used for water cooling is inserted from the opposite wall and is used as extinction studies with water aerosols (Fleming et al, 1998). an extension so that the burner assembly can be treated as a There are many advantages associated with the use of a single cylinder across the test section. counterfiow cylindrical burner. The fuel and the oxidizer flows The rearward 1800 of the burner surface is coated with can be independently adjusted, if required. The flame is a thin layer of high-temperature resistant black paint in order to laminar, two-dimensional, and very stable in the forward prevent fuel ejection into the wake region. The high pressure stagnation region. The geometry of the burner and the flow drop across the porous sintered surface assures a very uniform field allow for relatively simple analysis of the forward fuel flow over the burner surface, thus the fuel ejection velocity stagnation region (Dixon-Lewis et al., 1984; Dreier et al., 1986; (Vj) is calculated by dividing the fuiel volumetric flow rate by Peters and Kee, 1987; Olson and T'ien, 1987; Dixon-Lewis and the available fuel ejection area of the burner surface. Missaghi, 1988; Chen and Weng, 1990; Sick et al., 1990). Both Propane is used as fuel, and its flow rate is regulated wake and enveloped flames can be easily maintained over a by a mass-flow controller which is controlled by a personal wide range of fuel and oxidizer flows. The flame is easily computer. A bubble flow meter was used to calibrate the massobserved, and critical stages such as the blow-off limit (abrupt flow controller. transition from an enveloped flame to a wake flame) can be ascertained with ease and high reproducibility. The flame front can be easily accessed by intrusive (Tsuji and Yamnaoka, 1969; Flow Facility measured average velocities (V0) and the cross sectional area of A small-scale vertically upward open-circuit wind the test section. tunnel is used to provide oxidizer (air) flow to the burnier. The The diffuser is 30 cm long, with an inlet cross tunnel, except the test section, is made of clear polycarbonate sectional area of 10 cm x 10 cm and an expansion ratio (based for visual observation of droplet transport toward the burner, on areas) of 1:9. The flow straightener consists of a The tunnel consists of a blower, a diff-user, a flow straightener, polycarbonate honeycomb with cell diameter of 0.32 cm a contraction section, and a test section where the burner is (0.125") and a 50 mesh stainless steel screen with 30.3 % open placed in cross-flow. This configuration not only allows for the area. The contraction section with a contraction ratio (based on delivery of a uniform flow of oxidizer to the burner at a low areas) of 9:1 has an inlet cross sectional area of 30 cm x 30 cm turbulence intensity but also assists in the delivery of liquid and is 30 cm long. A square flange, which was machined to agent droplets to the flame. Figure 2 is a photograph of the have a smooth transition passage to minimize flow separation, flow facility, is placed between the outlet of the contraction section and the inlet of the test section. The test section has a cross sectional area of 10 cm x 10 cm and a length of 20 cm. It is made of anodized aluminum with three Pyrex®1 observation windows (12 cm x 7 cm x 0.64 cm) flushly mounted on the three walls of the test section using high-temperature silicone RTV®&. The burner is inserted through the fourth wall. The combustion products from the burnier are vented to an exhaust hood. Droplet Generator One of the most important steps in the liquid suppressant screening is to be able to deliver the liquid agent in droplet form to the flame zone to cause suppression or extinction. A droplet generator is used in the proposed screening in lieu of a spray because all the droplets, in principle, can be directed to the flame zone, thus minimizing droplet loss to the wall of the flow facility caused by the fan-out of a spray. In addition, the application of a spray may require more agent to perform a test than using a droplet generator, which may not be ideal or feasible for some potential new liquid agents that can only be synthesized in minute quantity. The droplet generator is based on the following wellestablished liquid jet break-up theory. According to Rayleigh's analysis of the instability of capillary jets, the frequency!f for maximum instability is given by the following equation (Bayvel and Orzechowski, 1993): Figure 2 Photograph of flow facility f Ui The air flow in the tunnel is provided by a variable4.508 (1) speed (frequency controlled) blower, whose outlet is connected to the main part of the wind tunnel via a flexible aluminum tube where ujis the jet velocity and dj is the jet diameter. When the and a coupling to convert a circular cross sectio