PSI - Issue 45

Teresa Magoga et al. / Procedia Structural Integrity 45 (2023) 28–35 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

31

4

production than can be expected in a set of fatigue specimens. This is relevant because the welding procedure in a shipyard is different from that in a controlled laboratory (Kecsmar and Shenoi 2004). Most of the S-N curves comprise three slopes: m 1 is the inverse slope of the S- N curve for N ≤ 5x10 6 cycles, m 2 is inverse slope for 5x10 6 < N ≤ 10 8 cycles, and any stress cycles below the cut-off limit (10 8 cycles) are assumed to be non-damaging. The detail categories are expressed in terms of the reference value of fatigue strength at 2x10 6 cycles (  c ) and m 1 .

Table 1: Description, photograph, schematic, reference strain gauge, load direction, FL m normalised by FL m at ID-1, and possible analogous Eurocode 9 details for each welded detail of interest

Eurocode 9 details (  c -m 1 ) 7.4.3 ( 32-3.4 ) 7.5 ( 18-3.4 ) 9.3 ( 12-3.4 ) 7.2.1 ( 50-4.3 ) 7.2.2 (4 0-4.3 ) 11.3 ( 36-3.4 ) 5.5 ( 40-4.3 ) 7.6 ( 36-3.4 ) 9.1 ( 28-3.4 ) 9.2 ( 25-3.4 ) 11.3 ( 36-3.4 ) 7.2.1 ( 50-4.3 ) 7.2.2 ( 40-4.3 )

Ref. strain gauge; direction

FL m / FL m at ID 1

ID Description

Photograph

Schematic

CROSS-SECTION

Pillar (hollow tube) joined to an end plate by a bevel-butt circumferential weld

1.0



A; vertical

1

25%

45 

Butt weld between a relatively thick insert and thinner deck plating

B; longitudinal

0.72  31%

2

Built-up beam comprised of web plating of different thickness

N/A – no defect reports

C; longitudinal

3

t 1

t 2

Longitudinal plating joined to transverse plating via a double fillet weld

B; longitudinal

1.2  9.3%

4

Butt weld in girder flange, between different thicknesses

D; longitudinal

1.7  10.3%

5

t 1

t 2

The fatigue relationship for endurance between 10 5 and 5x10 6 cycles, where N

i is the number of cycles to failure

for the ith stress range  i , is defined by Equation 1: =2×10 6 ( ∆ ∆ ) 1

(1)