Structure Damage Detection Based On Wavelet Analysis And High-pass Filtering Approach

The small structural damage is not easy to find, especially when damage liesinside the structure and is not visible. If the unqualified components used in industrialand engineering, it is very easy to create security problems. On the other hand, if thequalified components are also easy to produce fatigue crack and other damage in thelong-term work and/or harsh working conditions. In industrial and engineeringapplications, the accidents caused by structural damage often occur, such as aircraftand fighter crash, long-term service bridge fracture and so on. These accidents causedsignificant damage for people’s lives and property. So damage detection technology,especially the nondestructive testing technique based on vibration has been studiedintensively.With an example of a simple beam structure, a cracked beam model isestablished. The natural frequency and mode shape of the cracked beam can beobtained under different boundary conditions. Then wavelet based damage detectionmethod was introduced in detail. Furthermore, based on modern digital signalprocessing technology, a relatively new damage detection method based on high-passfiltering is proposed. First, the mode shape of the cracked beam is applied for damagedetection by high-pass filtering. This article also presents the operating deflectionshape of structure in forced vibration for structural damage detection. The high-passfiltering method and the wavelet method are applied to operating deflection shape todetect damage qualitatively and quantitatively. Numerical calculation results showthat operating deflection shape is suitable for the damage detection Finally theexperiments of crack beam and the damage wing of an aircraft were carried out toverify the proposed high-pass filtering based damage detection method.The main content of this thesis is as follows:1. Based on recursive method, a model for Euler-Bernouli beams with arbitrarycracks is established. The dynamic system parameters such as natural frequencies andmode shapes of the cracked beam under different boundary conditions can beobtained.2. The damage detection principle of wavelet analysis is presented. Damagedetection of beams with single and two cracks are studied based on one dimensionalcontinuous wavelet transform and one dimensional discrete wavelet decomposition. 3. In this thesis, the idea based on high-pass filtering in damage detection isproposed, and the feasibility and accuracy of this method is verified by numericalcalculation.4. The forced vibration problem for the cracked beam is inverstigated. Then theoperating deflection shapes are extracted from the the frequency response functions.Furthermore, wavelet analysis and high-pass filtering are applied in operatingdeflection shapes to detect the locations and depths of the crackes.5. Experimental modal analysis was carried out for clamped-free andcalmped-clamped beams with one and two cracks, respectively. The experimentalmode shapes and operating deflection shapes are used to detect the locations anddepths of the caracks based on high-pass filtering method, and the realistic feasibilityof the damage detection method based on high-pass filter was validated.6. The experimental damage detection for a cracked wing of the aircraft wascarried out by using operating deflection shapes and high-pass filtering method.The numerical and experimental results show that the mode shapes and operatingdeflection shape can be applied for structural damage detection. The main advantagesof the proposed method is without the need of the information of the completestructure numerical calculation (such as central difference method) for mode shape.The high-pass filtering based damage detection method can detect the structuraldamage, with accurate, convenient, and efficient characteristics. it is of greatsignificance for the online damage detection in further.