7th International Conference on Advanced Computational Engineering and Experimenting - ACE-X 2013, Madrid (Spain). 01-04 July 2013
Summary:
Nowadays, the industry is increasingly replacing traditional materials with composite materials. This is due to the excellent mechanical properties that can be achieved and the weight reduction they can provide. The use of composite materials has extended significantly in several industrial sectors, like aerospace engineering, automotive, renewable energies (windturbine blades), etc. However, defects may result from imperfect manufacturing process, e.g. a poor bonding between the fibres and matrices, embedded foreign objects, excessive porosity, etc. Flaws may also occur from in-service use such as fatigue, impacts, environmental exposure, etc. Several Non-Destructive Testing (NDT) techniques have been developed to achieve an adequate inspection of composite parts. These NDT techniques are applied in production as well as during maintenance. The most applied NDT method is Ultrasonic testing, US (pulse echo, transmission and phased array). Other methods are shearography, thermography and optical fiber sensing. In this work, Frequency Response Functions (FRFs) and Sound Pressure Level (SPL) were used to detect structural faults in GFRP. Damage in structures causes small changes in the frequency resonances. US technics require the evaluation of the object in numerous small sections. In Global Faults Non-destructive testing (FRF and SPL) the fault detection procedure requires only a global measurement in the structural component in operational conditions, which decreases the cost considerably since do not require very large number of measurements. In this effort a neural network as a global fault diagnosis detector for structural mechanical components will be applied. The research has been applied in GFRP beams with cuts of different deepness in order to simulate possible faults. Acoustic signals are used as neuronal network input. In some experiments carried out previously, vibration signals were obtained for a steel beam of rectangular section where faults were simulated by saw cuts of different depths [1,2]. A Levenberg-Marquardt backpropagation algorithm is used for training a supervised fully connected feedforward neural network [3,4]. This paper focuses on the ability of an acoustic non-destructive method to be able to detect defects in composite materials.
Publication date: 2013-07-01.
Citation:
Y. Ballesteros, J. Rodríguez, L. Molisani, N. Ponso, J.C. del Real-Romero, Application of acoustic NDT methods to detect damage on composite GFRP structures, 7th International Conference on Advanced Computational Engineering and Experimenting - ACE-X 2013, Madrid (Spain). 01-04 July 2013.
