Contributions to the study of pool boiling and spray evaporation on horizontal tubes

Refrigeration cannot stay aside from the environmental and energy challenges that humanity is about to face in the coming years. Montreal’s Protocol and later revisions marked the beginning of usage restrictions of CFCs and HCFCs, due to environmental issues such as ozone layer depletion and global warming. R134a and other HFCs will be soon phased out due to their global warming potential (GWP). Natural refrigerants such as CO2, ammonia or hydrocarbons appear as interesting alternatives from environmental and performance points of view. The high pressures of CO2 systems, the toxicity of ammonia or the flammability of hydrocarbons are important disadvantages of these fluids. Natural refrigerants combined with more efficient systems should be the investigation line followed in the future.\n\t\t\t\t Falling film evaporators, also known as spray evaporators, have been widely employed in petrochemical industry, desalination processes and OTEC (Ocean Thermal Energy Conversion) systems. The experience in other fields such as heat pumps and refrigeration is limited, but falling film evaporators appear as an interesting alternative to flooded evaporators due to potential benefits in terms of refrigerant charge reduction and heat transfer improvement. In addition, the boiling temperature increase caused by hydrostatic head in flooded evaporators is avoided, the temperature approach between refrigerant and cooled fluid decreases and the efficiency of the cycle improves and the evaporators can be of smaller size. The main drawback of falling film evaporators is that the design of the distribution system is critical and, if incorrect, may cause heat transfer deterioration due to dryout of the falling film. \n\t\t\t\t Falling film evaporators in refrigeration systems are heat exchangers with a shell-and-tube structure. Spray nozzles or other spreading devices distribute liquid refrigerant over the first rows of tubes of a tube bundle. Part of the refrigerant boils on the top row, cooling the fluid flowing inside the tubes, and the rest forms a film that flows to the following row. This boiling and flowing process occurs from one row to the next one. The exceeding refrigerant is collected at the bottom of the evaporator and recirculated to the distribution unit (with intermediate conditioning steps if needed).\n\t\t\t\t A large number of parameters affect the performance of falling film evaporators, but authors disagree about the effect of each of them. Those with a higher influence are heat flux, film flow rate, geometry of the tube, refrigerant properties and distribution system. The use of enhanced tubes enhances heat transfer if compared to plain tubes. In addition, most enhanced tubes delay film breakdown, maintaining the surface wet. Only those geometries that limit liquid axial movement, such as low-finned tubes with a high concentration of fins, should be avoided in these systems. \n\t\t\t\t An experimental setup available in the laboratory was redesigned in order to allow spray evaporation tests. These tests consist in distributing the liquid refrigerant, in conditions very close to saturation, on the tested tube/s, simulating the situation that occurs in spray evaporators. \n\t\t\t\t The main modification developed was to include a liquid distribution system to distribute the liquid refrigerant. The system was designed to allow testing different types of spreading devices. Among the different possibilities, full cone nozzles have been selected. The equipment permits different spacings between the nozzles, as well as choosing between two positions for the tube that receives the refrigerant from the nozzles. In addition, it allows comparing the heat transfer coefficients obtained when the tested tube receives the liquid refrigerant directly from the nozzles and when the liquid refrigerant falls by gravity from a conditioning tube above the tested one.\n\t\t\t\t The experimental setup does not follow a typical vapour compression cycle. Instead, pool boiling/spray evaporation and condensation occur at the same pressure and refrigerant flows from one shell-and-tube heat exchanger to the other due to the differences of density between liquid and vapour refrigerant. This configuration allows, on the one hand, testing very different conditions and refrigerants and, on the other hand, discarding the effect of lubricants on the heat transfer coefficients determined.\n\t\t\t\t The test rig is prepared for tubes of nominal external diameters up to 20 mm, but we tested tubes of nominal external diameter 3/4” (19.05 mm), which are widely extended in shell-and-tube heat exchangers. We chose tubes with plain and enhanced external surfaces, and the material depends on the refrigerant used (compatibility refrigerant – material). \n\t\t\t\t The compatibility with ammonia was the main consideration during the selection of materials for building the experimental setup. Thus, we used stainless steel (AISI-316L) in almost every component. \n\t\t\t\t As previously mentioned, due to the working principle of the experimental facility, a wide range of condensation and evaporation conditions can be tested. We chose liquid temperatures between 0 and 10 ºC for our pool boiling and spray evaporation tests, common temperatures in water chillers. This range of temperatures allowed using water as secondary fluid both for condensing the vapour refrigerant at the condenser tubes and for vaporising the liquid at the evaporator tubes. The use of water is very convenient since its properties are very accurately determined using the temperatures measured. \n\t\t\t\t The experimental facility stabilises the refrigerant pool temperature or the liquid distribution temperature (boiling saturation pressure), for pool boiling and spray evaporation tests, respectively. The distributed liquid refrigerant flow rate is also controlled accurately, maintaining the rest of the conditions constant. The temperatures and flow rates of the secondary fluids can also be regulated and stabilised at the values needed for each test. \n\t\t\t\t The experimental setup allowed measuring the conditions (temperature, pressure and flow rate) of the refrigerant and secondary fluids. Several sensors were also included to determine the conditions of the distributed refrigerant. \n\t\t\t\t We designed an experimental methodology to obtain pool boiling and spray evaporation heat transfer coefficients. The methodology is based on determining the different thermal resistances of the overall heat transfer process that occurs at the tubes. \n\t\t\t\t The design of experiments has focused on studying the boiling heat transfer coefficients under temperature conditions close to those found at water chillers, and with the largest range of heat fluxes possible. We have conceived a specific experimental methodology to analyse the influence of the impingement effect on the heat transfer coefficients with distribution of the liquid refrigerant. \n\t\t\t\t Pool boiling tests consisted in registering the values measured by the different sensors of the test rig, keeping constant the mean heating water temperature and flow rate through the tube and the refrigerant pool temperature. Except for special sets of experiments, a group of tests (constant pool temperature) started with the maximum mean heating water temperature possible. When stationary conditions were achieved, data were registered for a minimum of 15 minutes. After that, the mean heating water temperature was lowered and fixed at the next testing condition. When finished a group of tests, a new was began at another refrigerant pool temperature, repeating the procedure explained in the previous lines. \n\t\t\t\t In spray evaporation experiments there are two new parameters to be considered: the relative position between the tested tube and the distribution tube and the liquid refrigerant distributed mass flow rate. The relative position of the tested tube and the distribution tube was introduced as a new experimental variable because other authors observed that the liquid droplet impingement effect could enhance heat transfer. Thus, we should expect differences in the heat transfer performance between those tubes that receive refrigerant directly from the distribution devices and those that are wetted by the excess liquid from the row of tubes placed above them. Therefore, we have designed two different spray evaporation tests to illustrate these two possibilities. The first tests, named ST tests, consist in distributing the refrigerant directly from the nozzles to the tested tube, being the distance from the tip of the nozzle to the tube axis 59 mm. In the second tests, named SB tests, the refrigerant is distributed on the same tube, which works as a conditioning tube, forming a film that falls to the tube placed underneath (distance of 45 mm between tube axes). No heating water circulates through the conditioning tube to prevent liquid refrigerant from vaporising before falling to the tested tube.\n\t\t\t\t We started with ST tests. The liquid refrigerant distribution temperature was fixed and, for each group of tests, the mean heating water temperature started at the maximum value achievable by the experimental test rig. Then, the liquid refrigerant distribution flow rate was set at the maximum of the experimental set points considered, which was different for each of the refrigerants tested. When stationary conditions were achieved, data were registered for a minimum of 15 minutes. After that, the liquid refrigerant distribution flow rate was lowered and the process repeated. When all the flow rates were tested, the mean heating water temperature was lowered and fixed at the next testing condition and the sequence of tests was repeated. Once finished ST tests, we repeated the whole process with SB tests.\n\t\t\t\t Independently of the kind of tests, the heating water flow rate was kept as high as possible to minimise the inner thermal re

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.