Polymeric materials are easily molded into complex shapes, are relatively lightweight, and may be tailored to a specific task by modification of properties through chemical composition and the use of additives. These highly attractive properties have made polymers as ubiquitous in space vehicles as they are in houses, office buildings, and transportation vehicles. The flammability of these materials poses challenge for the fire safety of personnel and spacecraft. To make wise choices about spacecraft design and their fire suppression systems, we must improve our understanding of how polymeric materials burn in microgravity environment.
Combustion behavior in microgravity differs from that in normal-gravity environments. The primary difference is in the elimination of buoyancy-driven flow, which considerably reduces the convection of heat and combustion products away from the flame and of oxygen into the flame. In early microgravity experiments on paraffin, neoprene, foam rubber, and other combusting solids in quiescent environment [1], the flames did not reach steady-state but gradually darkened and shrank in size. Some samples self-extinguished. These experiments supported the conclusion that combustion is less hazardous at low gravity. Testing of fire hazards in normal gravity was therefore assumed to adequately assess the behavior of burning materials in low gravity.
The behavior of polymeric materials during pyrolysis and combustion is highly complex. As the temperature rises, the molecules in thermoplastic solid become increasingly mobile, giving the material the viscous properties of fluid. For some polymers, phase change from solid to liquid takes place; for others, the viscosity changes more slowly with temperature. Chemical bond breaking reduces the average molecular weight of the polymer, further reducing the viscosity. The breaking of bonds eventually results in polymer fragments that are small enough to constitute gas molecules. Conduction of heat from the surface often results in gasification deep within the polymeric melt, causing formation of internal bubbles, swelling of the outer surface, and sputtering of gases and small droplets as the bubbles burst [3]. The bubbles act as mechanism for the mass transport of volatile gases. They also affect the transport of heat by decreasing the local thermal conductivity, causing convection of the surrounding fluid, and changing the internal absorption of radiation. As they burst, the bubbles may expose larger region near the surface to chemical attack from the surrounding gases (such as oxygen). This effect is due to entrainment and, for highly viscous melts, to the distortion of the surface geometry.
The bursting of bubbles results in fire hazard unique to microgravity. In combustion experiments on Velcro fasteners made of nylon [4], the ignition of flammable gases from bursting bubble resulted in flamelets spurting from the flame zone. Occasionally, burning liquid droplets were expelled from the sample with velocities higher than 30 cm/s. These droplets burned robustly until all fuel was consumed, demonstrating potential to contribute to the spread of fire by serving as hot ignition sources. Bursting fuel vapor bubbles were also observed in the WIF experiment [5], which studied the burning behavior of polyethylene insulation covering nichrome wire in both quiescent and flowing air. Bubble bursting was inferred by pulsations of the flame and ejection of small particles of molten polyethylene.
The purpose of the numerical model described in this report is to improve our understanding of the development of bubbles in burning thermoplastic object and their effects on heat and mass transport under microgravity conditions. The model was designed to help understand experimental data from combusting spheres of polypropylene (PP) and polymethylmethacrylate (PMMA) collected under a NASA grant headed by Dr. Jiann Yang [6], [7].
The bubble model combines one-dimensional (radial) finite element model of sphere with the behavior of individual bubbles located in three-dimensional space. The thermoplastic sphere is initialized as volume of highly viscous material divided into specified number of elements of equal volume in the shape of concentric spherical shells. Each element is "material volume" of polymeric material; the element travels with the same chunk of polymer throughout the calculation, although the amount of polymer decreases with time as it gasifies.
At the beginning of the calculation, bubble nucleation site is placed at random location within each element. Each bubble is initialized with zero volume. A heat flux is applied to the outer surface of the thermoplastic sphere, and the energy equation is solved to determine the temperature field at the next timestep. Radiative and convective losses at the surface are included. As the sphere heats, the polymer contained in each element gasifies according to first order Arrhenius expression. This gas collects in the bubble associated with the element, which grows and migrates toward the surface of the sphere due to the viscosity and surface tension gradients caused by the temperature gradient within the element. If the midpoint of the bubble has exceeded specified distance from the surface of the sphere and satisfied delay time representing the time required to drain the thin film between surface and bubble, the bubble is presumed to burst. The mass of its gaseous contents is then subtracted from the spherical sample.
At the end of the timestep, the size, location, and content (polymer and gas) of each element are recomputed and the material properties are determined by the properties of polymer and gas weighted by their respective volume fractions. The energy equation is solved again, and the process continues.