Research On Low Velocity Impact Health Monitoring Methods For Aeronautical Materials Structures

Starting from the engineering background of urgent need of real-time and on-line monitoring impact of composites based on structural health monitoring technology in aerospace fields, This paper investigates the identification of low velocity impact load and impact energy of composite and the researches emphasize on the methods of impact localization. The main contents are as follows:Firstly, it was summarized about research background and significance of Structural Health Monitoring techniques and impact monitoring methods. Major direction of the research was proposed;Secondly, the main signal processing technique and localization algorithms of acoustic emission technology is analyzed, and the propagation characteristic of impact acoustic emission signals is studied. According to the requirements of real-time and convenience of impact location monitoring system for composite structure, the hardware and software has been designed and a PXI-bus based impact monitoring system is successfully developed. The system function has been verified and analyzed on the carbon fiber composite laminate specimen;Thirdly, approximation method of impact load history is presented, which can reduce the number of unknown parameters, thus the load identification problem is transformed to a parameter identification problem. Based on finite element method and optimization technique, load time history method of inversion is researched and the two steps are established to identify impact position by using triangulation technique and optimization technology;Fourthly, based on array signal process technique, impact localization method for composites is researched. Impact localization is divided into two parts: direction of arrival (DOA) estimation and estimation of the distances between impact source and array elements. Because impact response signals do not meet the narrow-band signal model of DOA estimation, the strategy of DOA estimation is proposed. Based on propagation characteristics of Lamb wave, the method for distance estimation of impact source is proposed. The present methodology is verified on several composite specimens;Fifthly, based on independent component analysis (ICA) technology, it is researched that the localization of two sources that simultaneously impact on structure. Two neural networks based on information maximization (informax) principle are analyzed. The feed-back network is used to separate the convolved mixtures of two impact sources, and it iterative algorithm is deduced. The experiment on aluminum beam shows that multiple impact sources locating can be solved by ICA;Sixthly, based on principal analysis impact damage and energy recognition of composite is researched. Because of great difference of frequency spectrum of response signals induced by different impact energy, spectrums of sensors response signals are used as the major feature for identifying impact damage and impact energy. Based on principal analysis, a new method is established to detect occurrence of impact damage, and the method of identifying the value of impact energy is proposed based on principal analysis and BP neural network.Finally, summary of the whole works is drawn. The problems met in the research and the continuing directions are presented.