PSI - Issue 14

Filin V.Yu. et al. / Procedia Structural Integrity 14 (2019) 758–773 Author name / Structural Integrity Procedia 00 (2018) 000–000

770

13

3.6 Validity of test results in respect of crack shape.

Base metal specimens normally show crack tunneling (Fig. 8,b). On contrary, specimens welded with a symmetrical groove tend to no crack growth at the mid-thickness (close to weld root) as there are compressive residual welding stress (RWS) there, Fig. 8,a. Some hints are applicable like chevron notch (for base metal), reverse side compression before cracking (for straight notch), cracking at a higher minimum to maximum cyclic load ratio, and the last but not the least – local side compression that is a standard procedure for welded specimens but helpful even for base metal, Fig. 8,c, it allows to get much straighter crack front.

a c Figure 8 — Invalid crack shape in a welded specimen ( a ), base metal specimen ( b ), and valid crack shape after local side compression ( c ). 4. Ductile to brittle transition. Critical temperatures found by different methods Low-alloyed steels are still the most widely used material for large-scale welded structures, they have a body centered cubic crystalline structure (BCC) known to have a ductile to brittle transition at lowering temperatures. In engineering practice a term “crack arrest temperature” (CAT) exists that may be related to the minimum design temperature for a certain steel of a particular thickness, in addition to the fracture toughness requirement for welded joints where a crack usually arises. Crack arrest by base metal occurs as an energy required for propagation of the central cleavage tunneling part of the crack front arises with the width increase of side areas (shear lips) deformed in the conditions of plane stress. This effect may be directly observed in a full-size test (e.g. ESSO test as per IACS W31) but it is extremely expensive, needs high-capacity equipment and large specimens. So a lot of indirect test methods have been developed but the most related criteria are still qualitative. The present study uses fracture mechanics to substantiate correlations of test outcomes with CAT. In Russian shipbuilding and offshore practice two test methods are used: NDT (nil-ductility temperature) and TKB (critical temperature with big specimens). Both tests are performed at three-point bend but NDT uses a free falling weight while TKB specimens are tested at a slow displacement rate. NDT is generally performed as per ASTM E208 developed by Pellini in 1950s. A non-machined side of a specimen with a brittle deposit is bent to nearly the yield strength. If a crack starting from brittle deposit extends to a specimen side, the specimen is “broken”. NDT is the maximum temperature at which specimens break. TKB is estimated as a temperature corresponding to 70% shear portion in fracture of full-thick notched bars. Generally, it is a fracture appearance test performed at different temperatures. Now both test procedures are modified to fit modern low-alloyed steels. The studies of Ilyin et al. (2018) are based on LEFM approach as it deals with cleavage behaviour of material and low test temperatures. A cleavage crack arrest condition is b

Ia I K K  .

(34)