No abstract

An experimental and numerical research program is executed to investigate the fatigue lifetime increase due to grinding-type post-weld treatment method. Plate edge gusset, longitudinal attachment and T-type joint specimens are investigated using fatigue test results, numerical models and mathematical statistical tools. The favourable impact of grinding on fatigue lifetime of the analysed joints is commonly known for researchers. However, considering the lifetime increasing effect is mainly based on experimental results. In the current research program, numerical models for as-welded and post-weld-treated specimens are developed and the lifetime increasing effect of weld treatment is investigated using effective notch stress (ENS) method. Fatigue tests are executed on as-welded and weld-treated specimens, and their results are comparatively evaluated based on numerical calculations and measured lifetimes. The results prove that the fatigue lifetime of weld treated (ground) details can be predicted with high accuracy by applying modified weld geometry in the numerical model and ENS method. Based on the current investigation, it is proved that the advantage of weld geometry treatment method comes mainly from the modification of the weld geometry at the fatigue critical point. Results also show that the same S-N curve can be applied for both as-welded and ground specimens, if the modified weld geometry is implemented in the numerical model.

The focus of the current paper is the investigation of different local stress calculation methods for fatigue design. To evaluate the accuracy and applicability of the hot-spot and effective notch stress (ENS) fatigue assessment methods an experimental and numerical research program is carried out at the BME Department of Structural Engineering, in Hungary. Two types of structural details are tested using 32 test specimens, a cruciform and a gusset plate type joint. The test results and the accompanying numerical calculation results are statistically evaluated and compared. Analysing the results of the fatigue tests and numerical calculations, it is concluded, that both the hot-spot and the ENS method could be used for fatigue design of the cruciform joints according to the recommendation of the IIW (Hobbacher in Recommendations for fatigue design, Springer, Paris, 2004). However, the rules of the ENS method needs improvement, if used for the design of gusset plate type joint. According to the current results, the ENS method can be adjusted to make applicable for the gusset plate joints as well, recommendation for that is presented in the current paper.

There are contradictional information in the previous studies regarding the relationship between the fatigue strength and the yield strength in case of steel structures. Several previous test result show that the fatigue strength of as-welded specimens do not depend on the yield strength of the base material. However, several studies show that the fatigue detail class of HSS structure can be higher than for structures made form NSS material for the same structural detail. Collecting the previous fatigue test results of cruciform and flange gusset structural details the effect of the yield strength on the fatigue behaviour is studied. Additional laboratory test program with 32 test specimens is performed to investigate the fatigue behaviour of structural details typically used in bridges produced from HSS and NSS material. For cruciform welded joints the test results show clear correlation between the yield strength and the fatigue life time. For flange gusset joints this relationship cannot be observed. In addition, post weld treatment of the flange gusset plate joint is studied. Using the results of the fatigue tests the fatigue lifetime increase applying the grinding type post-weld treatment is evaluated.

No abstract