PSI - Issue 57

J. Baumgartner et al. / Procedia Structural Integrity 57 (2024) 618–624

621

4

J. Baumgartner et al. / Structural Integrity Procedia 00 (2023) 000–000

the JWG [8]. The reduction factor for wall thickness t ref > 25 mm is evaluated by

t t attach

·

t

f ( t ) =

0 . 1

n

t ref

(1)

where t attach is the half the distance between the weld toe radii. For other joint types, the conventional equation

t

f ( t ) =

n

t ref

(2)

may be used. The thickness correction exponent n depends on the joint type and the post weld treatment. An extended table within the IIW-recommendations summarizes recommended values.

2.5. Fatigue resistant determination by testing

In case that a numerical fatigue assessment is not possible, e.g., due to a new welding process or the need of experimental proof, recommendations on the experimental procedure and the subsequent statistical evaluation are given. The complete section was revised, compared to the version from 2016. The given recommendations on the statistical evaluations are an excerpt from the B est practice guideline for the statistical analyses of fatigue results [7] developed within the WG1 of commission XIII. Whereas the main principals for the evaluation stayed same, a

more clear and comprehensive description is provided. As main improvements following items can be named:

• A check for normal distribution at start of the evaluation should be performed. If this check fails, reasons need to be identified and special care should be taken if the resulting S-N curve is used for further evaluations. • For the determination of S-N curves a full set of equations is given. As a first step, an evaluation with a free slope m should be performed and the standard deviation sdtv of m should be derived. If the evaluated slope range ( m ± stdv ) is close to a standard slope, an evaluation with a fixed fixed slope m fix can be performed. • In case two datasets should be merged and evaluated, a statistics-based procedure is proposed.

In a separate section, recommendations for the evaluation of crack propagation data is given.

2.6. Fatigue assessment

The whole chapter 4 that deals with the fatigue assessment with stress-based approaches, fracture mechanics and fatigue testing was restructured. For the stress based approaches, now, no di ff erentiation of the di ff erent approaches (nominal, structural or notch stresses) is necessary. A clear separation was introduced separating four main load cases:

1. Constant amplitude, uniaxial stress 2. Constant amplitude, multiaxial stress 3. Variable amplitude loading, uniaxial stress 4. Variable amplitude loading, multiaxial stress

New graphs have been created to guide the user in the application of the di ff erent assessment approaches, e.g., for constant amplitudes, Figure 2. One small but relevant change is the assessment of multiaxial stress states. First, the Gough-Pollard equation is recommended as standard approach for the assessment. Second, Gough-Pollard equation is extended and takes