Título

EFFECTS OF CRACK RELATIVE DEPTH ON THE EXPERIMENTAL EVALUATION OF FRACTURE TOUGHNESS OF ASTM A516 GR 70 STEEL ON THE DUCTILE-TO-BRITTLE TRANSITION TEMPERATURE USING SE(B) SPECIMENS

EFFECTS OF CRACK RELATIVE DEPTH ON THE EXPERIMENTAL EVALUATION OF FRACTURE TOUGHNESS OF ASTM A516 GR 70 STEEL ON THE DUCTILE-TO-BRITTLE TRANSITION TEMPERATURE USING SE(B) SPECIMENS

Autoria

Donato, Gustavo Henrique Bolognesi; Dantas, Sullivan de Souza

DOI

10.5151/2594-5327-23612

Downloads

Baixar Artigo 140 Downloads

Resumo

Fracture testing of structural steels is usually based on standardized specimens (e.g.: C(T), SE(B)). Current standards usually recommend deep cracks (a/W ≥ 0.45) to guarantee high levels of stress triaxiality and therefore critical J and CTOD values in the case of Elastic Plastic Fracture Mechanics (EPFM). This approach relies upon the notion that similitude concept is valid since plasticity is limited and SSY conditions are respected. However, pressure vessels and pipelines present membrane stresses combined to shallow cracks and develop low stress triaxiality favoring plasticity. In these cases toughness data from deep cracked specimens can potentially underestimate the load-carrying capacity of real structures (being conservative). As an alternative, shallow cracked specimens can reproduce low triaxiality structures and in some cases more accurately predict failure. The problem is that these geometries develop significant plasticity ahead of the crack tip and fracture toughness results can become geometry-dependent, demanding biparametric methodologies (e.g.: J-Q, J-T). Results from the literature regarding fracture toughness of ferritic structural steels on the ductile-to-brittle transition temperature as a function of crack depth are relatively scarce and when available present very large scatter. To better quantify toughness of a ASTM A516 Gr70 steel under different triaxiality levels, this work investigates the effects of shallow (a/W ≈ 0.20) and deep cracks (a/W ≈ 0.50) on fracture toughness using SE(B) specimens. Approximately 20 specimens (18 mm thick) were tested for each condition at ~ -75ºC and results were evaluated using J-Q theory and Weibull statistics. FE models provided Q values and therefore triaxiality levels. All fracture results could be very well described using two-parameters Weibull distributions. Shallow cracks slightly overestimated toughness (~ 8%), presented larger scatter and provided more noncritical data. In addition, ~ 90% of the fracture toughness data obtained from shallowcracked samples violated the deformation limits of EPFM (validity of J integral as a single parameter to describe crack-tip stress fields), demanding bi-parametric approaches.

 

Fracture testing of structural steels is usually based on standardized specimens (e.g.: C(T), SE(B)). Current standards usually recommend deep cracks (a/W ≥ 0.45) to guarantee high levels of stress triaxiality and therefore critical J and CTOD values in the case of Elastic Plastic Fracture Mechanics (EPFM). This approach relies upon the notion that similitude concept is valid since plasticity is limited and SSY conditions are respected. However, pressure vessels and pipelines present membrane stresses combined to shallow cracks and develop low stress triaxiality favoring plasticity. In these cases toughness data from deep cracked specimens can potentially underestimate the load-carrying capacity of real structures (being conservative). As an alternative, shallow cracked specimens can reproduce low triaxiality structures and in some cases more accurately predict failure. The problem is that these geometries develop significant plasticity ahead of the crack tip and fracture toughness results can become geometry-dependent, demanding biparametric methodologies (e.g.: J-Q, J-T). Results from the literature regarding fracture toughness of ferritic structural steels on the ductile-to-brittle transition temperature as a function of crack depth are relatively scarce and when available present very large scatter. To better quantify toughness of a ASTM A516 Gr70 steel under different triaxiality levels, this work investigates the effects of shallow (a/W ≈ 0.20) and deep cracks (a/W ≈ 0.50) on fracture toughness using SE(B) specimens. Approximately 20 specimens (18 mm thick) were tested for each condition at ~ -75ºC and results were evaluated using J-Q theory and Weibull statistics. FE models provided Q values and therefore triaxiality levels. All fracture results could be very well described using two-parameters Weibull distributions. Shallow cracks slightly overestimated toughness (~ 8%), presented larger scatter and provided more noncritical data. In addition, ~ 90% of the fracture toughness data obtained from shallowcracked samples violated the deformation limits of EPFM (validity of J integral as a single parameter to describe crack-tip stress fields), demanding bi-parametric approaches.

Palavras-chave

Crack depth; Shallow cracks; SE(B) specimens; Fracture toughness; ASTM A516 Gr70 steel.

Crack depth; Shallow cracks; SE(B) specimens; Fracture toughness; ASTM A516 Gr70 steel.