Fatigue failure and structural integrity analysis of a 3-inch butterfly valve shaft using finite element modelling and microstructural evaluation

A 3-inch butterfly valve shaft failed at a water treatment plant after 50 cycles, despite having a 50000 cycles design life.To identify failure processes, a comprehensive methodology was used, which included visual inspection, microstructural analysis, and finite element analysis (FEA) to assess failure mechanisms under cyclic loading.The shaft, made of SS 316L steel with a tensile yield strength of 205 MPa and ultimate tensile strength of 515 MPa, exhibited stress concentrations in transition zones, with maximum von Mises stress reaching 506.15 MPa and total deformation of 2.6 mm.Fatigue life analysis identified regions experiencing as few as 50 cycles before failure, emphasizing the need for geometric optimization.Microstructural analysis using scanning electron microscopy (SEM) confirmed fatigue crack initiation at machining marks and propagation along grain boundaries, aligning with FEA predictions.The analysis incorporated boundary conditions, including fixed supports at flange ends, an applied moment of 27572 N.mm and an operating pressure of 1.76 Mpa.The Goodman mean stress theory was employed for fatigue life prediction, revealing von Mises stress peaks of 427.78 MPa, equivalent alternating stress of 676.84 MPa under fully reversed loading, and critical deformation of 0.081 mm.Damage analysis pinpointed failure-prone regions with a maximum damage index exceeding 1e5, limiting the design life to approximately 1.76 million cycles in critical areas.To enhance fatigue resistance, future work proposes replacing SS 316L with XM19 steel or increasing the shaft diameter and validating its performance under simulated operational conditions and cyclic testing contains continually opening and closing the valve under test pressure to assess the leakage and operating torque as per API 609 standard.These insights are vital for improving butterfly valve durability and ensuring long-term operational reliability in industrial applications.