No abstract

No abstract

A B S T R A C T The surface damage evolution under stress is often analysed by images of long-distance microscopes. Usually hundreds of images are obtained during the fatigue process. To analyse this huge number of images automatically, a new image tool is presented. This new image tool is included in free statistic software so that a statistical analysis of the damage evolution is easily possible. In particular several specific damage parameters can be calculated during the fatigue process. Some of these specific damage parameters are compared statistically here with simple damage parameters using images of two specimens under different stress levels at different time points of the fatigue process. It is shown that the specific damage parameters discriminate between the two different damage evolutions in an earlier stage than the simple parameters. They are also less influenced by different brightness and scales of the images and show other desirable properties of a damage parameter.CumL = cumulated length of detected cracks (μm) FGV.Auto = fraction of grey values between 0 and an automatic threshold (%) FGV.140 = fraction of grey values between 0 and 140 (%) MaxL = maximum length of detected cracks (μm) MeanL = mean length of detected cracks (μm) MGV = mean grey value NoC = number of detected cracks R = stress ratio S a = stress amplitude (MPa) σ max = maximum stress (MPa)

No abstract

No abstract