As part of this thesis an optical method for crack growth detection of plastic specimens under simultaneous mechanical strain (variable load) and media exposure (for example, liquid hydrocarbons) was developed. To implement this approach, both a customized software and a suitable image processing system (including hardware) was developed. The realized system allows automatic data acquisition (load cycle, crack length, crack opening, etc.) in real time. Sinusoidal force signal was applied to the specimen under test and the pictures were taken at maximum displacement in a user-defined time interval.

The software-based technical analysis (measurement including distortion correction) was carried out with the calibration of the recorded (crack growth) images using a reference image (calibration grid) with a known geometry. A graphical representation based on the detected crack growth rate was utilized (da/dN) for a defined media temperature environment on fracture mechanics parameters (stress intensity factor K). This representation helps to estimate kinetic crack propagation values, which are used for lifetime calculation of components under similar stresses (strain or ambient medium).