The optimization of bellows convolutions in bellows pump for better stress distribution

Abstract

Bellows pump has been widely used for chemical transportation with excellent elasticity and seal ability of the core component bellows. Nevertheless, for repeated alternating compression and expansion, bellows convolutions, that is, the maximum stress points, are at risk of fatigue rupture, which leaves a safety hazard for practical applications. Though a special bellows shape with non-constant wall thickness was proposed as an effective method to strengthen the convolutions, it also affects other characteristics of bellows, such as stiffness, single-stroke displacement, and size, which are equally important for the bellows pump. This article designs a fine-tuned convolution structure on the premise of avoiding its effect on volume properties of bellows. Limiting the value of variables by the stability of bellows stiffness, the structure parameters are optimized with the combination of the finite element method (FEM) and whale optimization algorithm. To ensure the accuracy of the FEM, the stress–strain curve of bellows material and the bellows axial stiffness is measured. Finally, digital image correlation is employed to obtain the strain at convolutions before and after the optimization, and the significant reduction of the strain verifies the effectiveness of the optimization. We believe that this method is valuable for optimizing stress distribution and guiding the design of longer-life pump bellows.