Transportation and deposition of particles in metal foam heat exchanger

Abstract

Open-cell metal foams have been investigated extensively as a potential heat exchanger in many applications due to their unique microstructures, made of interconnected pore-ligament constructions. Many studies have proved their excellent heat transfer performances. However, excessive pressure drop becomes one of its major drawbacks. In addition, fluid flow structures induced by the metal foam are always associated with various instabilities. For established thermohydraulics properties of the metal foam heat exchangers in clean and fouled conditions, more experimental works are called for. Until recently, the fouled condition has received relatively little attention among researchers. It is understandable as there are still many unsolved issues in the clean condition due to various geometrical considerations such as annulus pipe, duct, baffle, porous sandwich, including the trade-off analyses between pressure drop and thermal performances for both fully and partially filled configurations at different operating conditions.This present study conducted a series of experiments to investigate the thermohydraulic properties of (1) a metal foam-filled annulus and (2) a partially filled channel with metal foam. The latter configuration was also subjected to a fouled condition to gain insight of underlying mechanisms of particle transport and deposition processes in the metal foam structure. The experiments were conducted using Particle Image Velocimetry (PIV), Laser Doppler Anemometry (LDA), Hot-Wire Anemometry (HWA), Phantom high-speed camera and infrared radiation (IR) camera, depending on the experimental purposes. The PIV measurement provides instantaneous whole flow-field velocity, and the LDA and HWA offer local flow velocity in non-porous regions. The investigation of flow behaviours inside the porous structure of the metal foam was conducted based on the temperature distributions from the IR images. Meanwhile, in-situ observation of particles movement and deposition was performed using the high-speed camera, where the particle trajectories in a region of interest (ROI) could be determined.The result shows that the pressure drop of the metal foam-filled annulus is increasing with the airflow rates (20, 85, and 150 SLPM) and foam length. However, a significant amount of heat transfer takes place only in the first half of the foam, suggesting a need for an effective length for this kind of metal foam heat exchanger. For a certain case, for example, 20 PPI at 20 SLPM, a longer metal foam than 0.05 to 0.06 m would be unnecessary since it will only increase the pressure drop. The temperature distributions inside the porous structure are non-linear with a developed flow could be only seen at the lower half of the annulus of 20 PPI at 150 SLPM.For the partially filled channel, a part of the incoming airflows would enter the porous structure of the metal foam block, while the rest is going into the free stream region (a non-porous region that located on the top of the foam block). The free stream velocity is increasing with the blockage ratio (a ratio of foam height to channel height) and a formation of recirculation zone could be seen at the downstream region of a high enough blockage ratio. Interestingly, this phenomenon did not occur with the low pore density, e.g., 10 PPI foam but the 30 PPI foam and solid block. The flow inside the foam block also tends to leave into the free stream region through the fluid/foam interface for a certain foam length, pore density and blockage ratio. While the pressure drop is expected to be increased with the pore density in a fully filled channel, it is not necessarily always true for the partially filled channel. The fluctuating velocity and changes of the flow direction at the interface, as well as the formation of a recirculation zone would affect the total pressure drop and resistance coefficient.Furthermore, the metal foam causes a higher pressure drop than the solid block of the same size, indicating the substantial effects of porous structure on the total pressure drop. Our proposed friction factor model for the partially filled channel agrees with the experimental data with ±16% deviation. The result from the fouling experiment shows that the particle size has a significant influence on the particle transport and deposition processes. The large particles tend to keep their inertia, while the fine particles would follow the fluid streamlines. For the high blockage ratio, more particles tend to deposit on the frontal area of the foam block and the deposition is building up in the upstream region. When the blockage ratio is low, more deposited particles could be observed on the top surface of the foam block. As expected, the deposition causes an increase in the pressure drop performance. However, fluctuated pressure drop values could be seen at the beginning of the fouling process before reaching a similar value over time, regardless of pore density.