PSI - Issue 28

S. Henschel et al. / Procedia Structural Integrity 28 (2020) 1369–1377 S. Henschel et al. / Procedia Structural Integrity 00 (2020) 000–000

1371

3

1700

80

60

1600

40

casting

1500

20

Temperature / °C

Activitity O / ppm

1400

0

0

10

20

Time / min

Fig. 1. Temperature and oxygen activity as a function of time.

Fig. 2. Drawing of the CTS specimen.

casting. Details can be found in Wetzig et al. (2020). The melt eventually solidified in a Al 2 O 3 / mullite crucible. The total amount of steel was approximately 100 kg. The solidified steel had a cylindrical shape (radius R = 155 mm, height H = 140 mm). In order to close shrinkage porosity, which is typically found, the cylindrical casting was hot-isostatically pressed. Compact tension shear (CTS) specimen were machined according to the drawing in Figure 2. The specimens were then heat treated. This treatment consisted of austenitizing (840 °C, 20 min, vacuum), quenching in a stream of He (equal to quenching in oil), and tempering (450 °C, 1 h, N 2 ). The relatively low tempering temperature was chosen to promote brittle fracture in the fracture toughness tests. The specimen thickness was set to be 8 mm. The other dimensions of the specimen suggested by Richard (1985) were scaled in order to fit to the loading capacity of the applied testing machine ( F max = 100 kN) by considering the expected maximum fracture toughness of the material. Subsequently, the diameter of the six loading pins was in creased. Furthermore, the loading device had to fit to the loading clevis typically used for compact tension specimens. The loading device for the CTS specimens and the loading pins, see Figure 3, were made of high-strength maraging steel (1.2709). With this material, it was estimated that the pins withstand a maximum force of 95 kN. At a loading angle α = 60 ◦ , this corresponds to a force of 85 kN to be applied by the test machine. Each of the upper and lower parts of the loading device consisted of two parts. Hence, the e ff ort of milling was reduced while keeping the central alignment of the specimen. Furthermore, with this approach it is possible to test specimens of di ff erent thickness without introducing a bending moment at the six loading pins. The specimens were tested at loading angles of 0 to 45 ◦ . The stress intensity factors K I and K II were determined by applying a finite element analysis (Abaqus 6.14). The mesh, the loading and the boundary conditions are given in Figure 4. Eight-node quadrilateral elements (plane strain) were applied. Around the crack tip, special crack tip elements with collapsed nodes and middle nodes at quarter-point positions to account for the stress singularity were used. From this analysis, the equations for the stress intensity factors were revealed, see Equations (1) and (2). The holes in the specimen were larger in diameter and may form a di ff erent stress field in combination with the notches at the back face. Hence, the e ff ect of the notch length was analyzed with the finite element model, see Figure 5. It was observed that only for a / W 0 . 5 K I is increased with increased notch length. Hence, lower critical forces at fracture are expected.

F √ π a cos α WB (1 − a / W ) F √ π a sin α WB (1 − a / W )

11 . 17 − 44 . 01 A − 1 . 014 + 20 . 56 A − 64 . 83 A 2 0 . 031 − 2 . 741 A 2 . 185 + 0 . 399 A − 23 . 88 A 2

K I =

(1)

K II =

(2)

with A = a / ( W − a ) and 0 . 5 ≤ a / W ≤ 0 . 7