| Summary: | In refrigerated spaces, the inside air is cooled by a heat sink either by forced ornatural convection. The first situation is usually found on refrigerated storeswhile the second one is more frequent on small apparatus like domestichousehold refrigerators or on isothermal boxes to transport medical products,like vaccines or medicines. On the mentioned refrigerated spaces it is notfrequent to monitored the inside air temperatures. So, the knowledge of the airtemperature field inside them is almost unknown and often it can be found largeair temperature gradients inside, which can put in risk the stored products.Usually there is only one thermostat located on some appropriate place whosebulb senses the temperature around it, assuming that the remainder inside airhas the same temperature. As will be seen later on, in refrigerated spaces thereis a wide air temperature field that must be known in order to better locate theperishable or other products inside, regarding their specific storagetemperature. In order to accomplish this desiderate it was used on this work acommercial household refrigerator that was monitored with thermocouples onseveral points. The measured temperatures were then compared with the onesobtained from two different simulation tools, the Fluent and the other one basedon an Artificial Neural Network (ANN) with supervised learning performed usinga Genetic Algorithm (GA) supported by an elitist strategy. It was possible toconclude that, at least in this case, the last one presents a lower absolute error- 0.8K - when compared with the first one - 1K -and also that the airtemperature fields inside are more consistent with the reality.
|
|---|