No abstract

It has been well established that the rate of heat transfer associated with boiling systems is strongly dependent on the nucleation site density. Over many years attempts have been made to predict nucleation site density in boiling systems using a variety of techniques. With the exception of specially prepared surfaces, these attempts have met with little success. This paper presents an experimental investigation of nucleation site density measured on roughly polished brass and stainless steel surfaces for gas nucleation and pool boiling over a large parameter space. A statistical model used to predict the nucleation site density in saturated pool boiling is also investigated. The fluids used for this study, distilled water and ethanol, are moderately wetting and highly wetting, respectively. Using distilled water it has been observed that the trends of nucleation site density versus the inverse of the critical radius are similar for pool boiling and gas nucleation. The nucleation site density is higher for gas nucleation than for pool boiling. An unexpected result has been observed with ethanol as the heat transfer fluid, which casts doubt on the general assumption that heterogeneous nucleation in boiling systems is exclusively seeded by vapor trapping cavities. Due to flooding, few sites are active on the brass surface and at most two are active on the stainless steel surface during gas nucleation experiments. However, nucleation sites readily form in large concentration on both the brass and stainless steel surfaces during pool boiling. The pool boiling nucleation site densities for ethanol on rough and mirror polished brass surfaces are also compared. It shows that there is not a significant difference between the measured nucleation site densities on the smooth and rough surfaces. These results suggest that, in addition to vapor trapping cavities, another mechanism must exist to seed vapor bubble growth in boiling systems.

Bubble incipience in artificial cavities manufactured from silicon has been studied using gas nucleation and pool boiling. Moderately wetting water and highly wetting ethanol have both been used as the bulk fluid with cylindrical cavities, as well as those with a triangle, square, and rectangle shape cross section. Nominal cavity sizes range from 8 to 60 μm. The incipience conditions observed for water using both gas nucleation and pool boiling suggest that bubble initiation originates from a concave meniscus. Cornwell’s contact angle hysteresis theory for vapor trapping cavities is used to explain the gas nucleation results. The pool boiling results are more difficult to explain. Using ethanol, cavities appeared to be completely flooded and were not activated using either gas nucleation or pool boiling. Using water and gas nucleation cavities were almost always activated, provided the incipience criterion was satisfied; in contrast cavities in pool boiling with water activated with different superheats during different experiments. The difference in incipience behavior between gas nucleation and pool boiling with water is explained based on vapor trapping and thermal suppression considerations. Based on limited experimental results, it appears that the backpressure does not influence gas bubble incipience, provided the pressure difference is the same. The experimental results presented affirm the theory of heterogeneous nucleation from vapor trapping cavities provided contact angle hysteresis and vapor trapping are fully accounted for. However, the results also suggest that the theoretical considerations required for a deterministic model for incipience from vapor trapping cavities during boiling is more complex than previously hypothesized.

Bubble incipience in artificial cavities manufactured from silicon has been studied using gas nucleation and pool boiling. Moderately wetting water and highly wetting ethanol have both been used as the bulk fluid with cylindrical cavities, as well as those with a triangle, square, and rectangle shape cross section. Nominal cavity sizes range from 8to60μm. The incipience conditions observed for water using both gas nucleation and pool boiling suggest that bubble initiation originates from a concave meniscus. Cornwell’s contact angle hysteresis theory for vapor-trapping cavities is used to explain the gas nucleation results. The pool boiling results are more difficult to explain. Using ethanol, cavities appeared to be completely flooded and were not activated using either gas nucleation or pool boiling. Using water and gas nucleation, cavities were almost always activated, provided the incipience criterion was satisfied; in contrast cavities in pool boiling with water activated with different superheats during different experiments. The difference in incipience behavior between gas nucleation and pool boiling with water is explained based on vapor-trapping and thermal suppression considerations. Based on limited experimental results, it appears that the backpressure does not influence gas bubble incipience, provided the pressure difference is the same. The experimental results presented affirm the theory of heterogeneous nucleation from vapor-trapping cavities provided contact angle hysteresis and vapor trapping are fully accounted for. However, the results also suggest that the theoretical considerations required for a deterministic model for incipience from vapor-trapping cavities during boiling is more complex than previously hypothesized.

It has been well established that the rate of heat transfer associated with boiling systems is strongly dependent on the nucleation site density. Over many years attempts have been made to predict nucleation site density in boiling systems using a variety of techniques. With the exception of specially prepared surfaces, these attempts have met with little success. This paper presents an experimental investigation of nucleation site density measured on roughly polished brass and stainless steel surfaces for gas nucleation and pool boiling over a large parameter space. The fluids used for this study, distilled water and ethanol, are moderately wetting and highly wetting, respectively. Using distilled water it has been observed that the trends of nucleation site density versus the inverse of the critical radius are similar for pool boiling and gas nucleation. The nucleation site density is higher for gas nucleation than for pool boiling. An unexpected result has been observed with ethanol as the heat transfer fluid, which casts doubt on the general validity of heterogeneous nucleation theory. Due to flooding, few sites are active on the brass surface and at most two are active on the stainless steel surface during gas nucleation experiments. However, nucleation sites readily form in large concentration on both the brass and stainless steel surfaces during pool boiling. The nucleation site densities for the rough and mirror polished brass surfaces are also compared. It shows that there is no large difference for the measured nucleation site density.