68 - Structural Health Monitoring of composite aerospace structures with Acoustic Emission

摘要

While composite materials have many advantages due to their lightweight potential impact events orForeign Object Damage (FOD) is critical for composite and many other lightweight and thin aerospacestructures. FOD can lead to unscheduled maintenance in the cases of hail damage, tool drop, ramprash or even loss in the cases of tire debris (Concorde Accident July 25th, 2000) or insulation debris(Columbia Space Shuttle Feb. 1st, 2003). Impact damage reduces the static load capability and thefatigue life of a structure. The cost of an aircraft on ground is about 40 k$ per hour. Thus a decision onunscheduled inspection and repair or a return to flight must be taken fast and responsibly. Nondestructiveinspection combined with numerical analysis is the state of the art. A detailed numericalanalysis of predamaged parts may take several days or weeks. Novel fast Structural Health Monitoring(SHM) and predictive maintenance tools can support the necessary decision making process.Impact damage generates characteristic acoustic signals that can be detected and analyzed byacoustic emission systems during the event. This was investigated in the Clean Sky Program in theGreen Regioanl (GRA) platform* by Fraunhofer LBF (Laboratory for “Betriebsfestigkeit” – structuraldurability). A fast but simple analytical model was developed that can analyze certain extractedacoustic features. This model was trained with 50 composite plates clamped to simulate a stringer bayeach. The specimen were subjected to different impact energies and locations and correspondingAcoustic Emission (AE) features, damage sizes as well as the compression load after impact werederived from these tests. After this training the system could analyze impact events in near-real-timeand present estimations on impact energy levels, location, damage size, mechanical properties,delamination growth as well as the remaining fatigue life under a given load level.The project closed the loop from data acquisition with a commercial AE system, via the assessment ofthe structural properties based on sensor records to the prognosis of this structural health and itspresentation in near real-time after impact events. Unfortunately the high scatter of results regardingimpact testing as well as of the extracted acoustic signals affects the reliability of the system so far butwith approaches from big data methods AE may become an interesting sensor type for predictivemaintenance of sporadic failure types.