Optimization of fatigue life of the seismic vibrator baseplate considering the coupling effect of welding residual stress

The seismic vibrator baseplate (SVB) is a welded structure. Ignoring welding residual stress (WRS) and only considering cyclic stress can decrease the accuracy of fatigue life assessments. This study aims to investigate the fatigue life prediction and optimization of SVB under the influence of WRS. A welding model of SVB and an SVB-ground excitation model were developed, and the stress distribution of SVB was determined using the finite element method. A fatigue life prediction study that incorporates WRS, based on a modified S-N curve, was conducted. A multi-objective fatigue life optimization analysis was performed using genetic algorithms. The results shows that the backing weld (BW) of the SVB is the most susceptible area to fatigue failure. Under the combined effects of WRS and working stress, the fatigue life of the SVB is 8.14 years. When considering WRS, the prediction results align well with field data, showing an error margin of less than 5 %. WRS is identified as a critical factor influencing the fatigue life of SVB, and optimizing welding parameters can effectively reduce WRS, thus extending the service life. The optimal welding parameters were found to be welding interpass temperature of 105 °C and welding speed of 10.23 mm/s. After optimization, the fatigue life of the SVB increased to 10.23 years, a 25.68 % improvement over the pre-optimization condition. Previous studies did not consider the influence of WRS, whereas this work integrates it into the fatigue analysis, producing a modified S-N curve that enhances the accuracy of life prediction. Furthermore, multi-objective optimization algorithms and fusion decision-making methods were applied to the fatigue life optimization of the SVB, offering novel research approaches and methodologies. This study provides theoretical guidance for the structural design and optimization of the SVB, aligning with practical engineering needs and showing substantial application value.