PSI - Issue 42

Kai Donnerbauer et al. / Procedia Structural Integrity 42 (2022) 738–744 Kai Donnerbauer / Structural Integrity Procedia 00 (2019) 000 – 000

740

3

Table 1. Chemical composition of AISI 347 and comparison with DIN EN 10088-1.

wt. %

C

Si

Mn

Ni

Cr

Mo

Cu

S

P

Nb

min. max.

/

/

/

9.00

17.00 19.00 17.17

/ /

/ /

/

/

/

0.08

1.00 0.41

2.00 0.71

12.00

0.015 0.003

0.045 0.029

1.00

amount 0.05

9.12

0.38

0.48

0.547

The center region of the bar material consists of a multiphase microstructure, as to be seen in Fig. 1 (a). From this unnotched fatigue testing specimens have been manufactured of which the location of the minimum cross-section is indicated as a black circle in Fig. 1 (a). Fig. 1 (b) shows a Nickel K α -EDX map (energy dispersive X-ray spectroscopy) and 1 (c) an EBSD map (electron backscatter diffraction) of the region of interest. In both mappings calculated grain boundaries in face-centered cubic (fcc) phase are represented by black lines. Brighter parts in Fig. 1 (b) have higher nickel content and appear to be interdendritic regions consisting of fcc grains. After solidification during forming processing temperature, degree of deformation or both have been too low to dissolve the dendritic solidification microstructure. This leads to a lower austenite stability inside dendrites and therefore there are already body-centered cubic (bcc) grains in the initial state, Fig. 1 (c). Besides this described microsegregation mechanism, there is most probably also macroscopic segregation present due to the microstructure gradient in solidification direction, also influencing austenite stability, Fig. 1 (a). Even heat treatments at 1120 °C, the highest temperature permissible according to DIN EN 10088-3, did not lead to a homogenization of microstructure across the bar diameter.

Fig. 1. (a) Light micrograph; (b) Nickel K α -EDX mapping, (c) EBSD phase mapping.

2.2. Experimental setups Strain-increase tests (SIT) and constant amplitude tests (CAT) were performed with servohydraulic testing system Instron 8802 with a maximum load capacity of 250 kN. Tests were carried out total strain-controlled with a strain ratio of R = -1 (fully reversed loading). To prevent a possible strain rate influence on martensitic transformation constant strain rate ε̇ = 4·10 -3 s -1 was chosen instead of a constant testing frequency. Tests were stopped at a load drop of 50 % or at 2·10 5 cycles. Experimental setups for testing in air and distilled water were designed to allow tracking of various NDT signals to evaluate their suitability for characterizing microstructural changes during fatigue loading: Testing in air, Fig. 2 (a); testing in distilled water, Fig. 2 (b). For testing in air two extensometers were used with gauge lengths of l 0 = 10 mm and l 0 = 88 mm respectively. This allows the change in temperature, the DC potential drop and the ferromagnetic phase volume to be monitored with the commercially available system Helmut Fischer Feritscope FMP30. Since tests in distilled water could only be controlled with the 88 mm extensometer, a strain correlation between both extensometers was obtained from tests in air with the same methodology as in Klein et al. (2016) and a correlated strain amplitude ε a,c was used. Furthermore, the Feritscope was substituted by a sensor, whose operating principle is based on the Hall effect. A self-developed in-situ electrochemical cell with a standard three electrode system was