PSI - Issue 17

Ashu Garg et al. / Procedia Structural Integrity 17 (2019) 456–463 Ashu Garg et al. / Structural Integrity Procedia 00 (2019) 000 – 000

458

3

fabricated from H13 die steel with two different pin profiles namely cylindrical (CYL) and threaded with three intermittent flat faces (TIF). The concave shoulder (3º taper) of diameter 20 mm and pin length of 5.9 mm were same for both tool pins. For TIF tool pin first full threaded pin profile of pitch 1 mm was prepared and then three flat faces were cut at 120° apart. For tensile testing, samples were cut longitudinally through stir zone (parallel to welding direction) and transversely (perpendicular to welding direction) from two sets of welded plates (Fig. 1(a)).

(a)

(b)

AA7075-T651 (RS)

100

R 6

30

2.5 60°

6

25

100

Welding direction

AA6061-T6 (AS)

10

10

(d)

(c)

Fixed

Cross-section of weld prepared with CYL tool pin and modeled region for FE analysis

Displacement

Fixed

Cross-section of weld prepared with TIF tool pin and modeled region for FE analysis

Displacement

Fig. 1. (a) Schematic illustration of tensile, notch tensile specimen cut from welded plate; (b) testing in UTM machine with high speed camera arrangement; (c) weld cross-section modeled for FE analysis; (d) applied boundary conditions for FE analysis.

For notch tensile test, specimen was cut transversely to the welding direction and creating a notch at the weld center (in the stir zone). The dimensions of tensile and notch tensile specimens are shown in Fig. 1(a). The uniaxial tensile testing of samples (at room temperature) were conducted on universal testing machines (Z050, ZwickRoell) at 0.5 mm/min cross-head speed to measure ultimate tensile strength (UTS) and to measure the yield strength (YS) (at 0.2% offset strain) contact type extensometer of 25 mm gauge length was used. In-situ imaging for observing crack initiation and propagation was captured using high speed camera (V7.3, Phantom) at 1000 fps as shown in Fig. 1(b). Failure of FSW joints under tensile loading were studied using finite element (FE) analysis software ABAQUS. For a realistic FE model both geometrically and mechanically, different regions such as base metal (BM) (both AA6061-T6 and AA7075-T651), stir zone (SZ), thermo-mechanically affected zone (TMAZ) as well as heat affected zone (HAZ) on both advancing side (AA6061) and retreating side (AA7075) of joint was considered. In addition, tunnel/void region was also incorporated in FE model. The dimensions of SZ, TMAZ, HAZ and tunnel regions of the actual sample were measured using optical microscope (Axio Imager.M2m, Zeiss). The microscopic views of the joint cross-section obtained are shown in Fig. 1(c) and corresponding FE model is also shown in Fig. 1(c). To assign the material properties, tensile testing of both base metals (AA6061-T6 and AA7075-T651) as well as the longitudinal tensile specimen cut from the weld SZ was conducted. Later, strain hardening law ( = ) was established from the true stress-strain curve of base metals and longitudinal specimen to calculate the value of K (hardening constant) and n (strain hardening exponent). To model the failure of the tensile and notch specimens in FE analysis, the ductile damage model was used and the values of fracture strain, stress triaxiality and strain rate were assigned to the FE model. The fracture strain (equation 1) was calculated measuring fracture sample dimensions in necking region and calculating the strains in thickness ( ) and width ( ) direction. = √ 2 3 × { 2 + 2 + [−( + )] 2 } (1)