Preform Design And Finite Element Analysis Of Turbine Blade Hot Forging Process

Francis turbine blade is one of the key components of hydraulic generator, and it has complicated configurations and varied thickness. It is a difficult and time-consuming task to manufacture such products,there are also lots of technical bottlenecks in the whole manufacturing process. Traditional design, so called trial-and-error method, inherits disadvantages such as immature process plan, high investment risk and repeated trials. In the dissertation, the research was carried out about the blade hot forging process based on finite element analysis, and new ideas such as hybrid method of blade preform design were brought forward. Process design and multiple-objective optimization were also studied, the results by theory analysis, optimization calculating and trial manufacture were compared and verified. The research achievements are listed as follows:Using the CAD model of X shape Francis turbine blade, a new horizontal positioning method was put forward under the common coordinate of the upper die closure. The new original point was the intersection point between the inlet edge and the lower ring, and the X anchor point as reference positioning point was the intersection point between the outlet edge and the upper ring. In order to ensure effective cooling effects, a mathematical model of heat transfer from die to water was set up. Calculation results show that the flow velocity should be larger than 970 mm/s so as to keep the die temperature around 300℃during forming process.A hybrid method for turbine blade perform design integrating One-step FEA and Equi-potential Lines Method was proposed, while One-step FEA was used to predict the profile of the unfold billet and Equi-potential Lines Method provided the thickness distribution of featured sections. The hybrid membrane/shell element was used by integrating CST3 and DKT6 elements to consider bending effect. Configuration calculations of the unfold blade billet were improved with a higher design efficiency.To keep the variation of forming load center within predefined tolerance scope during the whole forging process, a new algorithm to dynamically calculate the forming load center for the whole process of non-uniform thickness metal thermal forming was put forward based on rigid plastic FEA simulation, The formulations were developed and coded into a user routine on software DEFORM 3D V 6.1. The history of forming load center could be calculated, which is helpful for the unbalanced load control and optimization.Based on thermal compression test, a new two-segment function was put forward to characterize the flow stress behavior of 0Cr13Ni5Mo during ascend and steady stage. A sub-function utilized the influence coefficient method and the other sub-function adopted the Arrhenius equation, statistic analysis proves the reliability of the new model clearly.The modified Archard model was introduced to evaluate the effects of process parameters on die wear, the effect of temperature to the parameters in Archard model was considered. It concluded that the most severe die wear is located at the corresponding area where the outlet and lower ring intersect, and the forming temperature has most sensitive effect to the die wear.To minimize forming load and energy consumption, multi-objective optimization method was proposed, and new sub-functions were established with constraint conditions. By linear weighing-sum method, the multi-objective optimization was converted into a linear programming problem with the evaluation equation. Process parameters were optimized with SQP algorithm. The optimized load was 6.21 MN, much lower than the limit value of the press (10 MN); the optimized energy consumption was 18.63 MJ, reduced by 19% compared with the original one (23 MJ); and the optimized product efficiency was 39.67 min per piece, increased by 17%. The optimization design was evaluated by comparing the thickness distribution error between the optimized result and original design with that between plant try-out result and original design, and the relative thickness error was reduced down to 15% from 25%, which demonstrated the effective optimization advantage.