The Gurson model [8] is a widely known micro-mechanical model for ductile fracture. With modifications by many authors, however, the Gurson model can only simulate void nucleation and growth but not predict ductile fracture. In the modified Gurson model proposed by Z. Zhang [1], ductile fracture is linked to one single void nucleation parameter. Once this void nucleation parameter has been determined, the remaining characteristic length parameter, which describes the inclusion spacing, can be fitted from fracture mechanics tests. The void nucleation parameter and the resulting length parameter are the transferable parameters for ductile fracture. THE MODIFIED GURSON MODEL The modified Gurson combines the Gurson-Tvergaard [2] model and the void coalescence criterion by Thomason [3] . Two simple void nucleation models representing two extreme situations were used. The cluster nucleation model assumes that all the initial voids nucleate suddenly when the plastic strain level, pε , has reached a certain critical value, p cε . This condition can be written as: p p c p o nucleation d f df ε ε ε δ ) ( − = (1) of is the initial void volume fraction that has to be fitted. ) ( p c p ε ε δ − is the unitary impulse function (Kronecker function) and the position of the impulse are determined by the critical value p cε . In this work the commonly used critical value 0 = p cε , is adopted. The second nucleation model proposes that voids nucleate continuously during plastic loading. The nucleation rate is constant. Such simple continuous nucleation model can be written as: df Ad nucleation p = ε (2) The constant A is the damage parameter to be fitted. Equations 1 and 2 have greatly simplified the nucleation modelling and reduced the number of the unknowns of the nucleation process into one. TENSILE TESTING The material investigated was a welded joint in a 70 mm TMCP steel plate, welded by double-sided SAW. Yield strength level was 500 MPa and the main alloying elements were C (0.07%), Mn (1.5%), Ni (0.43%), Al (0.035%) and Nb (0.020%). Microstructure of the steel was polygonal ferrite and bainite. Areas subjected to investigation were the base metal (BM) the heat-affected zone (HAZ) and the weld metal (WM). Ductile crack initiation behavior was determined by using the multi specimen approach, including both smooth and notched round bar tensile specimens. HAZ and WM were tested in longitudinal and transversal direction, base material in longitudinal direction only. Both smooth and notched specimens had a cross sectional diameter of 6.0 mm. Four different notch geometries with notch radiuses of 3.0 mm, 2.0 mm, 1.0 mm and 0.4 mm respectively, were prepared to represent different levels of stress triaxiality. The tensile specimens were extracted from four different locations with respect to the plate surface as shown in Figure 1. Notch bottom of the HAZ transversal specimens was located 1.0 mm outside the horizontal fusion line. Tensile testing was conducted in a 250 kN INSTRON 1126 testing machine. The crosshead speed during the tests was 0.01 mm/s for the smooth specimens, and 0.005 mm/s for the notched specimens. Accurate measurement of diameter reduction during the tensile test is essential for the establishment of the Bridgman corrected true stress-strain curve.
RkJQdWJsaXNoZXIy MjM0NDE=