     
|
STAF: Smoldering and Transition to Flaming in Microgravity Introduction and Overview The transition from smoldering to flaming is of interest both as a fundamental combustion problem and as a serious fire risk. It encompasses phenomena related to the ignition of a homogeneous gas phase reaction (flaming) that is induced by a heterogeneous surface reaction (smolder) that acts both as the source of gaseous fuel (pyrolyzate, CO, etc.) and the source of heat to initiate the homogeneous reaction. That it is a fire hazard is confirmed by the fact that more than 30% of US fire deaths can be attributed to smoldering. The transition from smoldering to flaming is also of concern in the space flight program; to date there have been a six incidents of overheated and charred cables and electrical components reported on Space Shuttle flights and significant smolder-related incidents aboard the Russian space station Mir. With the ongoing establishment of the International Space Station and the planning of other long-term space missions, there is an added need to study smoldering and its transition to flaming in microgravity in order to prevent and minimize the effects of a smolder-initiated fire. While considerable work has been conducted to understand the smoldering combustion of porous fuels, there has been considerably less research conducted on the transition from smoldering to flaming. A common observation of these works is that for the transition from smoldering to flaming to occur the fuel samples have to be fairly large, or the process must be assisted by increasing the oxygen concentration of the oxidizer flow and/or external heating. The smolder and transition to flaming project (a NASA funded project) has been on going since the late 1980’s.Experimental work was performed in both microgravity (experimental and normal gravity During this time period, experiments were focused on examining forward and opposed smolder in both microgravity and gravity environments. The experimental apparatus held a cylindrical sample, and was considered to be nominally one-dimensional. As the experimental efforts progressed, transition to flaming was investigated in both cylindrical and rectangular samples. Efforts grew to examine piloted transition to flaming as well as the behavior of various fire retardant materials. In a parallel fashion, modeling efforts were made to model smolder combustion. One dimensional, time dependent formulations were employed by Kallmann and Rein. Rein’s modeling efforts focused on decomposition kinetics to correctly model mass loss and heat absorption/generation during smoldering. ThermoGravametric Analysis data in both nitrogen and air environments for polyurethane foam was utilized to develop a kinetic mechanism. Data from microgravity experiments was utilized to fit the heats of reaction for each reaction, both forward and opposed smolder were modeled accurately with the same kinetic parameters. Current modeling efforts employ a new computational code that is capable of modeling one and two-dimensional smolder. The current focus is modeling the transition to flaming process. In order to do this appropriately, both appropriate condensed phase and gas phase reactions must be included in the reaction mechanism to capture the behavior up to the transition. Based on experimental results, the transition occurs upstream of the smolder front, which suggests an oxidative reaction releasing a substantial amount of heat is present in this area. The condensed phase present in this area is referred to as char. This reaction process has been called “secondary char oxidation”. A secondary char oxidation has been added to the kinetic scheme developed by Rein and has been fit to both TGA and smolder experimental data. Experiments The experimental effort ended in 2007, but the below is of interest because it relates to the modeling effort. The experimental work was part of a NASA-funded project. Because the microgravity experiments were planned for the International Space Station, the polyurethane foam samples had to be limited in size for safety and launch mass reasons. The maximum sample size permitted for the experiments is too small for smolder to self propagate due to heat losses to the surroundings. Thus, the smolder propagation had to be assisted by reducing the heat losses and by increasing the oxidizer oxygen concentration. The work has demonstrated that both can occur with relatively small fuel samples if the external ambient conditions are appropriate. Experimental Apparatus The experiments are conducted in a small vertically oriented flow duct. The fuel sample (flexible polyurethane foam of 50 mm x 50 mm cross section and 125 mm long) is oriented such that its front face is flush with the wall of the flow duct. The back and side walls of the sample holder have guard heaters. During the tests, there is a forced flow of oxidizer through the igniter, and into the foam. An infrared heater is mounted facing the fuel-sample free surface and the bottom of the sample is in contact with an igniter. Smolder ignition is induced at the bottom and smolder propagates upward in the same direction as the oxidizer (buoyant plus forced flows), i.e., forward smolder. Six thermocouples are located along the sample to monitor the foam temperature, and an infrared camera monitors the free-surface temperature of the foam. An ultrasound probing technique is used to measure changes in permeability of the sample. In addition, a high-speed camera, operating at 1000 frames/s, observes the free surface of the sample. A schlieren imaging system is used to observe density gradients in the duct flow along the free surface of the fuel sample.
Results The experiments show strong evidence that a transition from smolder to flaming occurs in the char region upstream of the smolder reaction, agreeing with previous observations of the process. A combination of infrared and video imaging with in-depth thermocouples adequately tracked the progress of the smolder reaction, and captured the transition to flaming event in sufficient detail to determine the approximate location of the transition and the time delay to the transition event. It is shown that the transition to flaming is sensitive to the external heat losses, and the heat generated by the secondary char reactions occurring in the char upstream of the smolder reaction. The data show that increasing the oxygen concentration of the oxidizer flow, reducing the velocity of the external flow, and/or increasing the external radiant flux increase the likelihood of a transition to flaming. These observations support the concept that the transition from smolder to flaming is basically a spontaneous gas-phase ignition reaction that is supported by the smolder reaction, which acts both as the source of gaseous fuel (pyrolyzate) and of heat to support the reaction.
The transition is more likely to occur when the heat released by both the heterogeneous smolder reaction and the homogeneous gas-phase reaction is larger than the heat losses to the surrounding environment. These findings are supported by a simplifled energy balance analysis that describes the transition as an autoignition process. The analysis is able to predict the boundaries between the transition/no transition regions.
Modeling
The current computational focus is to model smoldering in two-dimensions and the transition to flaming process. The computational model formulation includes two-dimensional conservation equations for a thermal and oxidative reacting porous material. The equations are solved numerically using a finite difference formulation that was also utilized in previous work by Lautenberger [1]. The reaction mechanism that is being utilized to model the condensed phase decomposition is an extension of that developed by Rein [2-3]. First steps in expanding the reaction mechanism were taken to add additional pyrolysis and oxidative pyrolysis decomposition steps. Two additional reaction steps corresponding to char degradation (reactions 3 and 7) are included. The mechanism consists of the following reactions: two foam pyrolysis steps (reactions 1 and 2), a char pyrolysis step (reaction 3), two foam oxidation steps (reaction 4 and 5), and two char oxidation steps (reactions 6 and 7).
The reaction mechanism was fit to TGA data using genetic algorithms. The two additional steps allowed for a better fit to the data at higher temperatures. The experiments simulated here [4] involve the forward propagation of a smolder wave through a polyurethane foam cylinder 12 cm in diameter and 14 cm in length. The experiments were conducted in microgravity on the NASA Space Shuttle (mission STS–108). Temperatures were measured with centerline thermocouples installed at eight axial locations. The heats of reaction were fit using a one-dimensional formulation of the model and data from a microgravity forward smolder experiment (STS-108). For the two-dimensional simulation, the computational domain is presented in Fig. 5. Convective boundary conditions are imposed on the condensed-phase energy equation at the top and sides of the domain. The temperature at the igniter (inlet) is specified to match that recorded experimentally. The inlet velocity and species composition is specified. The sides of the domain are impermeable to mass transfer. Figure 6 shows a snapshot of temperature, reaction rate of the first reaction rate, concentrations of foam and βfoam at an intermediate time during the simulation (t=635s) when the smolder front is self propagating. Figure 7 shows a qualitative comparison of experimental results and model data. It is evident that two dimensional effects in the experiment area captured by the model. Figure 8 compares centerline temperature contours of model and experimental data. The model compares fairly well to the experimental data both qualitatively and quantitatively. Several videos (temperature contours, the first reaction step, and concentrations of foam, βfoam, char, and thermal char) are shown below.
Videos
References for above writeup:
[1] C. Lautenberger, A Generalized Pyrolysis Model for Combustible Solids, PhD thesis, , University of California at Berkeley, Berkeley, CA, 2007.
Recent Scientific Publications on STAF:
|
Combustion Processes Laboratories
PI: Prof. A. Carlos Fernandez-Pello
Department of Mechanical Engineering
60 Hesse Hall
University of California
Berkeley, CA 94720-1740
Phone: +1 510 642-6554
Fax: +1 510 642-1850
Email: clauten (at) me.berkeley.edu