PSI - Issue 39

Hannes Panwitt et al. / Procedia Structural Integrity 39 (2022) 20–33 Author name / Structural Integrity Procedia 00 (2019) 000–000

22

3

M t,max maximum torsional moment M t,ol

torsional moment for single overload

number of cycles

N

number of cycles at the last DIC picture exponent of the threshold function

N end

n th

stress ratio

R

stress ratio of axial load

R axial

tensile strength

R m

yield strength for 0.2% plastic strain

R p0.2

R torsion stress ratio of torsional load th cw

threshold value for the crack width method threshold value for the maximum principal strain method

th ε 1

th ε 1,end strain threshold value at the last DIC picture th ε 1,start strain threshold value at the first DIC picture ε 1 first principal strain ϕ phase shift angle ϕ 0 kinking angle ν Poisson’s ratio σ N,max maximum nominal axial stress τ N,max maximum nominal torsional stress DIC digital image correlation HSA high strain area SEN single-edge notched SIF stress intensity factor

2. Automated crack detection and measuring by Gehri et al. The approach of Gehri et al. (2020) utilizes the principal strain distribution on the surface of a specimen, which is obtained with digital image correlation. Firstly, a constant strain threshold value th ε 1 is defined. Secondly, for each load step all areas, where the first principal strain exceeds this threshold value, are determined (Fig. 1). These areas are referred to as high strain areas (HSA) in the following. By combining the HSA of all load steps a single HSA is obtained. This final combination is essentially a binary image, which contains the information, where the principal strain threshold value is exceeded at any time during the test and which can be processed by morphological operations. By morphological thinning, the HSA is skeletonized to a line with a width of a single pixel. This line is the detected final crack pattern, where intersections between branches are labelled as nodes. As this method uses the first principal strain, instead of e.g., the strain in tensile direction, it is independent on the crack orientation and therefore also suitable for mixed mode applications. After the crack path has been determined as a whole, the displacement field obtained by the DIC is used to measure crack kinematics, such as crack width and crack slip. In essence, for the crack width measurement virtual extensometers are placed perpendicular to each point of the crack by means of reference points. The distance between these reference points should be chosen as small as possible, but is limited by the subset size. From the displacement field the displacements of the virtual extensometers are calculated for each picture and load step, respectively. As the extensometers are placed as close as possible to the crack flanks, the displacement of the extensometers can be treated as the crack width. Furthermore, the orientation of the branches is calculated from the coordinates of multiple points within a moving observation window. A detailed description for all calculations as well as an accuracy analysis can be found in Gehri et al. (2020). In addition, Gehri et al. have made their program code freely available as an open-source MATLAB tool.