Research On Straightening Process Of Thick-walled Pipe Fittings Based On Part Characteristics Identification

Thick-walled fittings are mainly tube parts with the ratio of outer diameter and wall thickness less than 20.These fittings are susceptible to various deformations due to stresses during daily transportation,processing and use.Therefore,in order to repair the bending deformation,it is necessary to straighten the fittings.Compared with the roll straightening method,the pressure straightening method has the advantages of controlled accuracy,easy operation and a wide range of straightening applications.The purpose of this paper is to identify the material parameters of thick-walled fittings using finite element analysis and neural network for thick-walled fittings to reduce the influence of material parameters on the subsequent straightening process.The parameters affecting the final straightening results,including material parameters(modulus of elasticity,yield limit),geometric parameters(wall thickness,internal diameter)and process parameters(downward pressure,pivot point distance),are also discussed and the influence law is analyzed.Finite element analysis and multi-objective optimization are used to find the optimal size of straightening process parameters(amount of straightening press,).The foremost research task of this paper is divided into the next parts.1.A brief overview of the development status of the straightening process and equipment is given.The basic principle of straightening of pipe fittings is explained on the basis of the three-point back bending principle of electrodynamics.2、Based on the elastic-plastic theory,studied the relationship between stress and deformation in the process of straightening,got the mathematical relationship between the load and deflection.The deflection model of load in the course of straightening is derive.3、Based on the neural network,a recognition model of material parameters with thick-walled pipe fittings as the research object is established.The phenomenon of fluctuation of material parameters of specific parts is studied with the modulus of elasticity and yield limit as the final identification objects.4、The parameters affecting the final straightening results in the process of pressure straightening are analyzed by finite element simulation study,and the influence law of the parameters is analyzed.5.The multi-objective optimization model of the straightening effect,radial-based neural network agent model and multi-island genetic optimization algorithm(MIGA)were established to optimize the parameters of the straightening process with the design variables of the following pressure volume and pivot point distance,and the equivalent residual stress and residual deformation after straightening as the optimization design objectives,and the optimization results were verified by experiments.The results show that the optimized straightening process parameters can significantly improve the quality of straightening.A load deflection model and finite element analysis model for thick-walled fittings and a material parameter identification model for fittings were established by combining the analysis of elastic-plasticity theory.The straightening effect is optimized with multiple objectives and the results are verified.This further improves the straightening efficiency and achieves the purpose of improving the straightening quality.