This last measurement was used with feedback to control the strain rate in the specimen center.
The specimen thickness should be 1.2 mm in order to maximize the strain range at the specimen center.
For this purpose different circular notch sizes (2, 4.5, 7, 9 mm) were drilled at the specimen center.
The authors have previously optimized the cruciform geometry for biaxial fatigue initiation, using a revolved spline to reduce the specimen center thickness.
At the beginning, the load–strain diagrams at the specimen center showed the domination of bending strains, tension on one surface and compression on the other.
Whereas the GTN model confirms damage nucleation in the specimen center of the uniaxial tensile bars, as observed in experiments, the progressive damage model predicts damage starts from the specimen surface where the plastic strain reaches its maximum.
The design, control, and operation of the loading frame are discussed in detail, and experimental data is provided to validate the effectiveness of the control system with respect to specimen center shifting.
It was concluded that the disc bending fatigue test can derive the fatigue life under an equi-biaxial loading condition, for which strain range is measured at the specimen center.
From an experimental point of view, we found that a video-extensometer based on marker tracking cannot efficiently measure the strain rate in the specimen center at large strain levels when the markers are greatly very deformed.
Specifically, crack wake roughness will lead to impeded transport in the crack growth direction, this leads to the development of an irregular crack front due to environment enhanced crack extension occurring on the crack flanks where the transport distance is low and slow crack progression at the specimen center.
