Abstract: Large number of testing results of the welded joints treated by ultrasonic peening are analyzed and summarized. Difference in fatigue design between as welded joints and joints treated by ultrasonic peening are discussed. The conclusions are: ① Material strength has influence, to a certain extent, on the treatment effect on fatigue performance of welded joints treated by ultrasonic peening. As the material strength increases, the treatment effect will be better. Consequently, it is safe to use the fatigue design S—N curves gained from the fatigue testing results of the low strength steel. ② The S—N curves’ slope m of welded joints treated by ultrasonic peening are much bigger than 3.0 (the sampling range is 6~23). It is not appropriate to regulate forcedly that the statistics should be processed accord-ing to m=3.0, while 10 is recommended. ③ After the treat-ment of ultrasonic peening, the fatigue strength has something to do with the mean stress. Fatigue stress range that joints can support will decrease as the stress ratio R increases. Hence, it is suggested that the maximum stress should be used in the fa-tigue design of the welded joints treated by ultrasonic peening. ④ When the fatigue design of the welded joints treated by ultrasonic peening is processed using the maximum stress, the relationship between the mean stress of the applied load and the fatigue stress FAT of the welded joints treated by ultrasonic peening is: , –1≤R≤0.5.

Key words: Fatigue design, Material strength, Mean stress, Ultrasonic peening, Welded joint