Özet:
Polymer composites fail through complex damage mechanisms. It is not easy to determine stress levels for onset of various damage mechanisms with a single uniaxial tension test, since their stress-strain responses do not provide a clear yield point or stiffness degradation during loading. Acoustic Emission (AE) is an important technique used to detect damage in composite materials. An AE signal is an ultrasonic wave resulting from the sudden release of the strain energy when damage initiates and contains information about the damage mode. General conclusions in literature for the correlation of damage modes with corresponding AE characteristics are relied on interpretations rather than direct observation of damage modes. In this thesis, damage progression in Carbon Fibre Reinforced Plastic composites are investigated using AE technique. Optical instruments are used to obtain reliable correlations with damage modes and the AE events. First unidirectional laminates are tested. Artificial defects in the form of slits are incorporated at certain plies during manufacturing to stimulate damage in desired sequence. Tension tests are stopped at certain stress levels before the ultimate strength and specimen edges are investigated with optical microscope to identify damage modes and correlate with AE characteristics. Then results are compared with predictions of a progressive damage model implemented using Finite Element Micromechanical Model and a very good consistency is achieved. In the second part, Digital Image Correlation (DIC) and in-situ edge observation are applied simultaneously during the tension tests of different quasi-isotropic laminates. They provide robust evidences for damage mode correlations. The k-means++ clustering algorithm is used to group similar AE events. It is seen that damage progression and their AE characteristics change with lay-up sequence. The results obtained in this thesis put the reliability of AE based damage mode classifications, widely adopted in literature, in question and a new classification scheme is proposed.
