Effect of Undercut Depth on Fatigue Strength of Welds

Abstract

Occurrence of undercutting in welded joint is inevitable in some welding position such as horizontal, verticul or overhead. Undercutting, must affect the fatigue strength of welded joint as it constitutes a notch. From this standpoint, the relation between the undercut depth and the fatigue strength was experimentally studied on the test pieces, which were taken from butt welded joints and fillet welded plates of mild steel.
Test pieces were prepared according to Fig. 2 in the shape as shown by Fig. 1 ; types and undercut depths of test pieces are listed in Table 2. The undercut depth was measured using the instrument shown in Fig. 3. Fatigue tests were conducted under pulsating tension. The testing machine was a 20t capacity hydraulic type. The tests were carried out up to 5 × 10⁶ cycles. Understandably, the failure originated at the root of undercut in all test pieces.
The results of fatigue tests are shown in Figs. 5 and 6, and the 2 × 10⁶ fatigue strength value (range of stress), obtained through interpolation and extrapolation, were plotted on a probability paper (Figs. 7 and 8). Then the mean value and the values of 5% probability and 95% probability of failure were plotted against the undercut depth in Figs. 9 and 10. Considerable variances are observed in the S-N curves of Figs. 5 and 6. These variances are represented by the fattened portions of vertical lines in Figs. 9 and 10; they are particularly wide in the case of the butt welded joint with a 0 mm undercut depth, and in that of -the fillet welded plate with a 0.5 mm undercut depth. Perhaps it so seems on account of the number of test pieces being limited.
As seen from Fig. 9, the fatigue strength decreases in butt welded joints with an increase in the undercut depth ; namely, it is seriously influenced by the undercut depth. By contrast, it is some what less influenced in the case of fillet welded plate, but ceases to decline beyond 0.3 mm. Perhaps there exists already a stress concentration at the toe of fillet and the notch effect in this area must be mitigated. It is due to an imperfect machining that no improvement on fatigue strength was observed in the machined test piece with a 0.46 mm undercut.