R. Roberti

The Influence of Notch Root Radius and Austenitizing Temperature on Fracture Appearance of As-Quenched Charpy-V Type AISI 4340 Steel Specimens

Charpy-V type samples either step-quenched from 1200 °C or directly quenched from the usual 870 °C temperature, fractured by a slow bend test procedure, have been fractographically examined. Their notch root radius,ρ, ranged from almost zero (fatigue precrack) up to 2.0 mm. The fracture initiation process at the notch differs according to root radius and heat treatment. Conventionally austenitized samples withρ values larger than 0.07 mm approximately (ρeff) always display a continuous shear lip formation along the notch surface, whereas specimens with smaller notches do not exhibit a similar feature. Moreover, shear lip width in specimens withρ >ρeff is linearly related to the applied J-integral at fracture. In high temperature austenitized samples similar shear lips are almost nonexistent. The above findings, as well as overall fractographic features, are combined to explain why blunt notch AISI 4340 steel specimens display a better fracture resistance if they are conventionally heat treated, whereas fatigue precracked samples show a superior fracture toughness when they are step-quenched from 1200 °C. Variations of fracture morphologies with the notch root radius and heat treating procedures are associated with a shift toward higher Charpy transition temperatures under the combined influence of decreasing root radii and coarsening of the prior austenitic grain size at high austenitizing temperatures.

Influence of Notch Root Radius on Ductile Rupture Fracture Toughness Evaluation with Charpy-V Type Specimens

Abstract The blunt notch fracture toughness of four types of carbon-manganese steel (ASTM A516 grade 70) has been determined by J -integral tests on Charpy-V type samples with different values of notch root radius, ρ. J-ρ plots, determined using specimens with a notch depth to width ratio, a/w, equal to 0.5, have shown the existence of a limiting ρ value ( ρ e ff ) below which applied J -intergral values at fracture initiation are constant. These ρ eff values have been seen to depend only on second-phase particle distribution and not on their volume fraction or on the steel ferritic grain size. The procedure for deriving J -integral values at the onset of stable crack growth from J resistance curves in the case of notches has also been discussed. Experiments with Charpy specimens with a/w = 0.2 do not allow the derivation of meaningful J - ρ plots. In all cases, a ductile fracture criterion based on the constancy of the notch tip strain at rupture initiation has been proved when ρ > ρ eff .

Interrelation among microstructure, crack-tip blunting, and ductile fracture toughness in mild steels

ABSTRACT After a review of various types of models proposed up to now to illustrate the nucleation of ductile fractures from sharp cracks, the mechanical and microstructural material parameters controlling the phenomenon are discussed at the light of the mechanisms of crack-tip blunting and void nucleation at second-phase particles that precede real crack advance. A link is established between sharp and blunt notch specimen behavior. Finally, a new mathematical model incorporating previously identified critical parameters is derived and verified by comparing calculated and experimental values of fracture toughness of a number of mild steels.

WIDTH INFLUENCE ON R-CURVES OBTAINED WITH THIN SHEET TEST-PIECES BEARING BLUNT CENTRAL NOTCHES

Crack growth resistance curves obtained from 2 series of central notch AISI 4340 steel samples, quenched and tempered at 250°C, have been compared. A series was characterized by a 76.2 mm width and a notch length/width ratio of 0.3, the other by a 25.4 mm width and a notch length/width ratio of 0.4. Notch root radius (ρ) was varied between 0 (fatigue pre-crack) and 0.8 mm. Specimen thickness was 1.6 mm for both series. Keff-Δaeff curves show that the behavior of the two types of samples is similar, provided that the net section stress is somewhat smaller than the yield stress. Also, the Ki-/p and Jc- ρ series of data are clustered around the same interpolating lines as far as σn < hσσyσ h has been seen to vary between 0.93 and 0.99, according to the specimen width, and can be conservatively set at 0.9 for standardization purposes.