Abstract
Bolted flanges are widely used to connect pipelines in many industries. To assure sealability of a flange, generation of proper preloads in the bolts during the assembly process is critical. However, in the existing standard practice, identical torques are typically applied to all bolts to assemble the flange. Due to elastic interactions between the bolts, tightening one bolt can alter the tensile loads in the other bolts. Hence, the resultant preloads can vary significantly. Even with an improved makeup sequence, the variation in the bolt preloads can be still substantial, as high as 60%. This could pose a risk of leakage. When the bolted flange works under non-benign conditions, such as vibration, pressure and temperature variation, the risk could become even higher.
This paper introduces a new methodology to greatly enhance the preload assurance in bolted flanges with an optimized assembly procedure, which is enabled with advanced numerical modeling. A significantly improved uniform distribution of bolt preloads is achieved by optimizing the makeup torques, which is implemented by using physical test data as input and uniformly distributed preloads as the target function. The complexity of the elastic interactions between the flange, the sealing gasket, and the bolts presents uncertainties for the numerical model for quantitative prediction of the torque distribution that is required to yield uniform resultant bolt preloads. This paper resolves this modeling limitation through iterations between modeling and testing. These iterations calibrate and finally validate the model to generate the optimized makeup torque distribution which then leads to improved bolt preload uniformity. Based on the tests conducted on two different sizes of API flanges, 3-API-15K and 5-API-10K, the final preload distribution variation has been reduced to around 30% by utilizing the optimized makeup torque distributions.
Research on Design and Magnet Assembly Process of Multivariate and Multi-Roll Permanent Magnetic Separator
