Monitoring of Freezing Water or Melting Ice in Aircraft Fuel Tanks and Fuselages by Acoustic Emission

Abstract

It is a mostly overlooked phenomenon in physical chemistry that matter produces characteristic sound emissions when passing through a phase transition. The origin of this sound is not yet fully understood, but it is certain that sudden changes in volume, cracks, friction between crystallites and other sources must be considered, but a generally accepted theory of the "sound of phase transitions" is lacking. But even if the cause is not completely clear, this phenomenon can be used to detect the presence of these substances in technical structures. This is important, for example, in the case of water and ice in aircraft fuel tanks. Water arising from contamination and condensation is frequently found in fuel tanks and regular drainage procedure are important to guarantee the safety of systems and structures. While ice dispersed in the kerosene phase can seriously hinder the combustion process, blocks of ice can cause serious crack formation as in valves and tubes. A practical problem with all drainage procedures after flights at freezing conditions is the right time to start that process. Starting too early would mean that the remaining ice cannot be removed and waiting too long is an economic problem because it increases the aircraft's downtime. Unfortunately, there is no technology yet that tells mechanics when all the water has melted. It would be beneficial for maintenance teams to find the right time to drain, and acoustic emissions from the melting ice would be the ideal tool to determine this moment. For this purpose, acoustic sensors are attached to the skin of the tank walls, and the acoustic signals from the ice are intense enough to propagate through media, the aluminum sheets and coating. In this way, the completion of the ice melt is determined by the time at which the acoustic emission stops. We present measurements on a laboratory scale, results from a realistic climate chamber on a tank model and the first results from a campaign on an operational aircraft (Airbus A330). We were able to show that relevant amounts of melted ice can be reliably determined, e.g., 30 ml of water under the kerosene phase in a replica fuel tank generate about 200,000 distinguishable signals, in our case these signals are acoustic transients. This is quite sufficient to obtain the desired information about the state of the melting ice. In addition, to address other structures, results of ice melting and water freezing processes on an aircraft fuselage model are presented.