PSI - Issue 57

12

Kashif Kamran Toor et al. / Procedia Structural Integrity 57 (2024) 772–784 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

783

• Both shell models, A and B showed comparable results to the corresponding solid weld models. The shell model with weld toe location excluding the plate thickness, Model A, is a conservative approach. The shell model B where the weld toe is modelled at the correct location showed also conservative results compared to the solid weld model representing the actual weld geometry including the weld at both the edge and the sides of the plate. The hot spot S-N curve was used in these analyses. • The effective notch stress method is typically employed where the hot spot stress method is not able to provide reliable results from simplified models. This method predicted a fatigue life of 173 years for Model A and 2506 years for model B. The effective notch stress model also validates that the oblique shell formulation can be applied for the analysis of such connections.

6. Conclusion Based on the performed work the following conclusions are made: •

Fatigue damage at the weld toe of a cope hole has been calculated using the hot spot stress methodology and is found to be highly sensitive to including the weld geometry and stiffness. Different modelling approaches can be applied to evaluate the effect of including weld on hot spot stress calculation. The reliability of modelling approaches is increased by increasing the model accuracy. The FE model accuracy can also be increased by adapting the model to the as-built documentation and it is highly recommended to model the correct weld toe position in the local geometry in the connections. • The plated joints are conventionally modelled using the mid surface shell element type approach where plate intersections represent welded joints without the weld profile itself. This approach is widely accepted and also normally accepted by certifying authorities like DNV and is recommended by IIW. However, it is also said that this approach should be applied carefully for the hot spots where the stress gradient at the weld toe is significant. In this case the read-out distance is at 4, 8 and 12 mm from the weld toe and quadratic extrapolation is applied to calculate the structural hot spot stress. This type of hot spot is referred to as type ‘b’ in DNV and IIW recommendations. The mid surface shell type modelling approach for the type ‘b’ hot spot is found to provide very conservative hot spot stress when the weld is not included in the model. It is observed that the analysis accuracy can be significantly improved by including the weld profile in the FE model • The cope-hole welded details can be included in the model using oblique shell elements. This modelling technique has been verified by using a detailed solid modelling technique. The oblique shell elements can be used for accurate stiffness representation without reducing the computational efficiency. FE model reliability can be increased by a detailed FE model based on a solid model including the weld profile and the notch at weld toe. The sub-modelling technique is found to provide an efficient method to model the weld geometry based on the effective notch stress method. • The methodology is not entirely novel, however, the study highlights the significance of selecting the right approach for assessing fatigue damage at weld toes, emphasizing the sensitivity of hot spot stress to weld geometry and stiffness. The research purposes the use of oblique shell elements for accurate modelling of weld stiffness and encourages the use of sub-modelling techniques for enhanced reliability.

References

DNV-RP-C203, 2016. “ Fatigue Design of offshore structures ”.