Open Access

Fatigue failure under high local stresses and strains, as encountered in exceptional load cases or pressure vessel applications, remains a key design challenge. In welded joints, microstructural and geometric imperfections at the weld toes result in pre-existing short cracks and their subsequent propagation. Fatigue life assessment based on linear elastic fracture mechanics has shown high accuracy when crack growth is considered from a technical initial crack size. However, its applicability is limited when the small-scale yielding condition is violated, which is inherent to short cracks and may also occur under high loading conditions typical of the low-cycle fatigue regime. Within the German research project IBESS, a fracture mechanics-based framework for the fatigue assessment of welded joints was developed. The approach includes short crack propagation, statistical variability of weld geometry, and elastic-plastic crack growth. Welded butt joints made of high-strength structural steel and austenitic steel were tested under constant-amplitude loading, resulting in pronounced plastic strain. The results show that plasticity-related corrections significantly improve fatigue life predictions and confirm the applicability of the IBESS approach in the low-cycle fatigue regime.

High local stresses significantly reduce fatigue strength, particularly in welded joints. For the computational fatigue life assessment of such components, fracture mechanics is generally regarded as a suitable and physically well-founded approach. In practical applications, however, the use of fracture mechanics is often limited to concepts of linear elastic fracture mechanics, whose underlying assumptions are only partially fulfilled under loading conditions with pronounced plastic strain components, as typically encountered in the low-cycle fatigue regime. Against this background, a fracture-mechanics-based framework for the fatigue assessment of welded joints was developed within the German research project IBESS. The IBESS approach describes a comprehensive assessment procedure based on short crack propagation and the consideration of crack closure effects, while explicitly accounting for elastic-plastic crack growth mechanisms. Under loading conditions with high local stresses and strains, the consideration of cyclic plasticity becomes of particular importance. In this contribution, the applicability of the IBESS approach to welded butt joints made of high-strength steel under high load amplitudes is investigated. For this purpose, constant-amplitude fatigue tests were conducted, in which pronounced plastic strain components occur. Based on the experimental results, a complete fracture-mechanical assessment using the IBESS approach was performed to estimate fatigue life spanning from low to high-cycle fatigue. The results show that considering multiple crack growth and, in particular, applying plasticity-related correction leads to a significant improvement in the accuracy of fatigue life predictions under these loading conditions compared to linear elastic fracture mechanics. Thus, the fundamental applicability of the IBESS approach is confirmed also for loading conditions with pronounced plastic strain components in the low-cycle fatigue regime.

The Kitagawa-Takahashi Diagram is an important tool for describing the fatigue limit of components containing defects. The limit line in this diagram can be described using various models, but these differ significantly in the technically important region. For this reason, the KT diagram must be verified in this area through experiments. However, the stair-case method requires a large number of samples and has a long test duration. In addition, many samples with identical notches must be produced. In this work, a method is presented that allows a simpler and faster validation of the Kitagawa-Takahashi Diagram in the region of short cracks. Notches with a defined width and depth were manufactured in flat samples of a low-alloyed steel with two different heat treatments using an engraving laser. This method allows to produce very sharp notches without plastic deformation and with only a slight thermal influence on the surrounding material. The samples prepared in this way were fatigued with blockwise increasing loads until failure. Crack initiation and propagation is monitored with a direct current potential drop method. The number of cycles in each loading block is determined by the measured potential drop. When in a defined interval no potential increase and therefore no crack propagation is detected, the program switches automatically in the next loading block. For each sample, it is therefore possible to determine a stress at which a crack is arrested and a stress that causes the specimen to fail. This procedure enables a reliable and precise determination of the limit stress for the respective notch size with a low experimental effort and time consumption. The method is evaluated on a low alloyed steel in the hardened and normalized condition. In order to investigate the influence of the method used to manufacture the notch, notches of the same width and depth were introduced in the samples using both an engraving laser and Electrical Discharge Machining (EDM). The experiments showed no significant difference between the EDM and laser notches, although the laser notches were significantly sharper. The differences compared to values determined using the staircase method were slightly conservative, but still negligible. This demonstrates the suitability of the method presented here, which, due to the use of laser notches, the significantly shorter test time and the use of less specimens compared to the staircase method, results in a significant reduction in test time and thus also an enormous cost reduction.

Powder properties are considered a key factor in mechanical properties in laser additive manufacturing, although few studies have investigated the effects in laser beam directed energy deposition for metal materials (DED-LB/ M). Water atomized (WA), and gas atomized (GA) powders are frequently used but may result in different part properties due to powder properties. To examine their qualificationfor DED-LB/M, this work examines powders and mechanical properties of AISI 316 L. Also, examination techniques are compared. The results show that the powder production has no relevant influenceon porosity and Archimedian density of built parts. WA powders show good processability in the process, despite unfavorable morphology. In contrast, WA specimen reach only 10% fracture elongation in tensile testing whereas GA-based specimen achieve 30%. Tensile strength of both is above 500 MPa. The reason for the lower mechanical property values can be attributed to defects and oxides. Yet WA powders may provide a cost-effective alternative for DED-LB/M when a reduced fracture elongation is acceptable.

Additive manufacturing by DED-Arc enables the production of large and complex high-strength steel components. However, the residual stress state generated during deposition can be significantly altered when the component is separated from the substrate plate as a final manufacturing step. This study investigates the residual stress relaxation and redistribution caused by substrate detachment in DED-Arc manufactured high-strength steel hollow cuboids. The component geometry was varied in terms of height, length, and wall thickness. Longitudinal residual stresses were measured by X-ray diffraction on the side wall surfaces before and after mechanical separation from the substrate plate. In addition, 3D scanning was used to quantify the resulting component distortion. The results show that substrate detachment causes a pronounced redistribution of longitudinal residual stresses, including a reduction of tensile stresses and, in some regions, the formation of compressive residual stresses. The stress differences before and after detachment can be interpreted as a superposition of relaxed longitudinal shrinkage stresses and released bending stresses arising from inhomogeneous restraint over the build height. The sign and magnitude of the bending contribution depend strongly on the component geometry. Low-build and high-build components show opposite bending tendencies after detachment, which is attributed to the interaction between substrate restraint, component stiffness and transformation-affected upper layers. Regression analysis of the geometry variation indicates that height, length, wall thickness, and the height–length interaction significantly affect the released bending stress, while component height is the dominant factor for the normal tensile stress relaxation. The findings demonstrate that substrate detachment is a critical step for residual stress redistribution and distortion in DED-Arc manufactured high-strength steel components and that geometry tailoring is usable to influence the resulting stress state.