Research On Multiple Signal Classification Based Structural Health Monitoring For Composite Structures

Composite structures have been increasingly used in various aircraft structures due to their superior stiffness and weight characteristics. However, the composite structures are susceptible to low velocity impacts which can cause internal damage including matrix cracking, delamination and fiber breakage. Such barely visible damage caused can lead to significant reduction of local structural strength. Therefore, the growing applications of composite materials have attracted much interest in developing structural health monitoring(SHM) system to ensure the safety of these applications.This research presents a near-field structural health monitoring method based on the 2-D multiple signal classification(MUSIC) method for the composite structure. The main works and innovations are as follows.(1) Aim at the near-field source localization problem of MUSIC algorithm in structural health monitoring application, a near-field 2D-MUSIC processing algorithm is presented for impact monitoring on composite structure.To verify the proposed method, two experiments are performed on composite structures. The localization results are in good agreement with the actual impact occurring positions in near-field area.(2) To deal with the anisotropy performance of the complex composite, the 2-D near-field MUSIC processing algorithm is also improved by taking into account the existing anisotropy performance. These validation results show that the algorithm which considers the anisotropy performance of the structure has a better localization result.(3) Since that the performance of the MUSIC algorithm largely depends on the successful determination of the number of impacts, eigenvalue threshold and Gerschgorin disk theorem based methods are applied for estimating the number of impacts firstly. And then multiple impacts localization method based on the near-field 2-D MUSIC method is present for the composite structure.(4) Though ordinary MUSIC algorithms can be used to distinguish multiple sources, they request the signals from these sources are independent with each other. By using the active damage monitoring method, since all damage scattered signals are the damage reflection signals of the same active excitation one, they are all correlated. To distinguish multiple damages in the structure, a correlated near-field 2D-MUSIC method based on spatial smoothing algorithm is also proposed to solve the correlation problem of multiple damage detection.(5) Aim at the blind area of impact localization using uniform linear array, omni-directional impact localization method using plum blossom array is presented for composite structures. The plum blossom array based impact localization method has certain essential advantages: first, it can cover the whole azimuth range of monitored structure. Second, the velocity of Lamb wave propagating in the structure can be estimated by impact signal itself.