Experimental investigation of the surface temperature and water retention effects on the frosting performance of a compact microchannel heat exchanger for heat pump systems

Abstract

Frost formation on a louvered fin microchannel heat exchanger was experimentally investigated in this paper with the aim of determining the dominant factors affecting the time of frosting and frost growth rate. A novel methodology was developed to measure frost thickness and frost weight at intervals during the frosting period. Frost mass and thickness growth rates, corresponding coil heat transfer, capacity degradations and air pressure drop are measured and discussed. The experimental data showed that at a given air dry bulb temperature, the fin surface temperature and air humidity are the primary parameters that influence the frost growth rates. Water retention and air velocity had a secondary impact on the frosting performance. From digital images of the frost growth it was observed that frost does not nucleate from the water droplets retained in between fins but it developed from the leading edges of the fins.

Introduction

In modern heat pump systems, heat exchangers use enhanced heat transfer surfaces for both air and refrigerant sides. Conventional fin and tube coils are slowly being replaced by microchannel heat exchangers, which use flat multi port tubes and louvered fin design. These heat exchangers are usually made of aluminum and because of the low conductive thermal resistance of the microchannel tubes, the fin base temperature is closer to the local saturation temperature of the refrigerant in comparison to conventional fin and tube type heat exchangers. Moreover, they reduce the volume and weight of condensers and evaporators and their reduced refrigerant charges could potentially lower the direct contribution to global warming due to refrigerant leakage (Garimella, 2003, Kim and Groll, 2003, Kim and Bullard, 2002, Park and Hrnjak, 2007). In heat pump applications, microchannel heat exchangers are usually mounted with their tubes oriented vertically to promote drainage of water condensate in the corrugated fin bends. While in cooling mode the increase in energy efficiency is in the range of 6–10% compared to spine fin or plate fin and tube coils with similar face area, during heating mode the energy performance of heat pump systems with microchannel outdoor coils are generally low due to a higher frequency of defrost cycles (Garimella, 2003, Kim and Groll, 2003, Padhmanabhan et al., 2008, Subramaniam and Garimella, 2005). Defrost cycles are periodically executed in between the heating times to melt the ice, drain the water from the outdoor coil, and free its surface from accumulated frost before the heating service could start again. Because frequent defrost cycles penalize the heating seasonal energy efficiency, it is crucial to understand the characteristics of frost growth on outdoor coils and develop heat exchangers that would minimize, if not eliminate, defrost cycles.

There are several parameters that affect frost growth on outdoor coils, such as air velocity, air humidity, air temperature, surface temperature, surface energy (including coatings or roughness), fin geometry and water retention (Kondepudi and O’Neal, 1989, Lee et al., 1997, Na and Webb, 2003, Shin et al., 2003, Xia et al., 2006). Frost thickness influences the heat transfer resistance and reduces the free flow area, ultimately causing a capacity degradation of the outdoor coil (Kondepudi and O’Neal, 1989, Padhmanabhan et al., 2010).

Studies of frost growth on microchannel heat exchangers with surface temperatures just below freezing point are rather sporadic and sometimes inconsistent in the literature. The complexity of the phenomena makes the theoretical analysis problematic and most of the results are based on a limited range of experimental tests in which the effects from the operating parameters and geometry are difficult to isolate and quantify (Kim and Groll, 2003, Xia et al., 2006, Yang et al., 2006). Other studies focused on simplified geometries, such as flat plates or channel flows between parallel flat plates (Lenic et al., 2009, Lüer and Beer, 2000, Na and Webb, 2004b).

Various techniques have been used in the literature to measure frost thickness such as mechanical methods with micrometers by Lee et al., 2003, Na and Webb, 2004a and Fossa and Tanda (2010); optical or laser methods by Kennedy and Goodman (1974) and Chen et al. (1999); and image analysis by Xia et al. (2005). With mechanical measurements methods, accurate measurements of frost build up might not be possible due to the fact that it is necessary to stop the experiment to obtain the measurement or the micrometer might interrupt the local air stream and hinder the onset of frost growth. Researchers reported laser beam methods to be impractical in most cases due to high roughness of the frost surface (Na and Webb, 2004a). Image analysis proved to be successful in measuring the frost thickness with good accuracy but the data reduction is a very lengthy process.

Different methods for measuring the frost weight were also proposed in the literature. The direct weight measurements using precision scales in air flow wind tunnels or load cells in psychrometric chambers generally yield more accurate results. Scraping the frost from the coil and measuring its weight at the end of the test (Hosoda and Uzuhashi, 1967), or melting the frost and weighting the water condensate (Kondepudi and O’Neal, 1989, Song et al., 2002), or adopting sophisticated scale mechanism to balance the frost weight (Yonko and Sepsy, 1967) are examples used in past studies but are not suitable to determine the onset of frost nucleation and the instantaneous frost growth rate in compact microchannel coils. A more promising method to measure frost growth is described by Verma et al. (2002) and more recently by Xia et al. (2005) for compact heat exchangers. The authors proposed to weigh directly small coil samples placed on a digital balance and compare the final weight with the amount of condensate. An experimental investigation of the impact of frost growth on the system performance and reliability of microchannel heat exchangers in heat pump systems was presented by Kim and Groll (2003). Frost growth and the frequency of the defrost cycles were considered the major factors in determining system performance. The investigators considered microchannel evaporator coils with two different fin density of 15 and 20 fins per inch (590 and 787 fins per meter) in two different fin-tube orientations (vertical tubes and 15° slanted tubes). For heating mode, the vertical coil installation with lower fin density leads to slightly increased capacity and energy efficiency ratio (EER) and longer frost time. The authors concluded that water condensate drainage removal during defrost cycles needed to be improved and that the manifolds should be designed for more even refrigerant distribution. In their work, the investigators reported overall UA average coefficients but did not measure the amount of frost accumulated on the coil, amount of water retained in between the fins or effects of various surface temperatures.

The effect of water retention on louvered fin flat tube microchannel evaporators was experimentally investigated in a recent work of Xia et al. (2006). The authors tested five louvered fin flat tube microchannel coils under frosting, defrosting, and subsequent refrosting conditions. From images they noticed that droplets retained between fins from the defrost cycles provide sites for future frost growth. Xia et al. concluded that water retention has a significant effect on the air pressure drop in the next cycles and they also observed that the refrosting behavior became periodic after about four cycles. Their work is one of the first examples showing the direct impact of water retention on frosting and defrosting performance. However the effects of varying surface temperature and air humidity were not considered and Xia and co-workers suggested additional experiments to parametrically extend the geometric dimensions in the data set and understand the effects from flow-depth, water retention, and fin geometry.

A recent study by Padhmanabhan et al. (2008) focused on experimental comparisons of frost and defrost performance of fin and tube and microchannel coils on a heat pump unit. Their study showed that microchannel coils frost faster than conventional fin and tube coil working under similar conditions. The experimental data also showed that frost cycle time for the first dry-start cycle is quite different and in fact is 70% longer than other next wet-start frost cycles. The authors also found that removing the water residual at the end of the defrost cycle by flushing the coil with pressurized nitrogen improved the next frost cycle time by no more than 4%. In their work, direct control of entering refrigerant temperature to the evaporator was not possible and the fin surface temperature might have been affected by other system operating parameters such as fluctuations of the evaporator and condenser saturation pressures, adjusting of the refrigerant flow due to the thermal expansion valve settings during transient periods, refrigerant charge migration into the system accumulator and cycling temperatures from switch of the four-way valve and outdoor fan. A recent experimental study on frosting performance of parallel-flow parallel-fin (PF²) flat tube microchannel heat exchangers with horizontally installed tubes was presented by Zhang and Hrnjak (2010). They found that the time of the first frost cycle was 21% longer than the second frost cycle and considerably longer (92%) than the time of the 15th consecutive frost cycle for PF² heat exchanger. Results showed that 10% decrease in air relative humidity from 80 to 70% caused the frosting time to increase 42%. They observed an improvement in frosting performance over a conventional serpentine fin which they attributed to better drainage capability of the PF² heat exchanger.

From the literature review it is obvious that several root causes that promote frost growth on microchannel heat exchangers were identified, but quantifying the main parameters that affect frost growth on microchannel heat exchangers is still an open question. To our knowledge, in the above mentioned studies on microchannel heat exchangers, the geometries were changed to search for the best option for thermal performance but surface temperature was not investigated or independently controlled. As a result, the effect of geometry modification was coupled with the effect of surface temperature change. Some studies in the literature included the effect of surface temperature but the geometries were taken as fin and tube coil or flat plates. Their findings and experimental results could not be generalized directly to louvered fins microchannel heat exchangers. The objectives of this paper are to investigate the effects and possibly prioritize the impact of operating variables that influence the frost growth and the time of frosting for a microchannel heat exchanger. Assuming the air dry bulb temperature is set, we identified four main factors that could compete on promoting frost growth on a microchannel heat exchanger: fin temperature, water retention, air humidity, and air velocity. We aimed to quantify the influence on frost growth of each factor. We critically assessed their impact on the thermal hydraulic performance of the microchannel coils in frosting operating conditions. A new methodology to measure frost thickness and frost mass accumulated on the heat transfer surface was introduced. Findings from the experimental work are summarized and discussed in the following sections.