Research On The Wall Thickness By Simulation Prediction And Experimental Investigation During The Power Spinning Of Cone Parts

Metal spinning as an important method of continuous partial plastic forming,because of its great deformation conditions, good performance, high precision andmaterial utilization, is widely used in the manufacture of civilian, defense and aerospaceproducts. Cone as a typical shape of shaped pieces, its forming rules and deformationmechanisms can be directly applied to relevant studies of other shaped pieces. Based onelastic-plastic explicit ABAQUS/Explicit finite element platform, this article analyzesstress and strain distribution during cone parts power spinning, and especiallyresearches the distribution characteristics of wall thickness and the impacts on the wallthickness distribution of a variety of process parameters. The main research contentsand conclusions as following:Firstly, based on the analysis of each factor, a3D finite element model, which isnot only in line with actual production conditionsbut also takesthe computationalefficiency and accuracy into account, is built. The build model is calculated on theABAQUS/Explicit finite element platform. Using solid section and the correspondingelements, the distribution law and mechanism of stress and strain areanalyzed during theforming process. The results showed that: during the forming process, the maximumvalue of stress always lies in the contact area of the blank and rollers, and at the end ofthe forming process,the stress of die fillet also reaches maximum value;The distributionof strain is always in a loop.At the beginning,the maxima occurs at the contact area ofblank and rollers, as the spinning process on, the maximum value is graduallytransferred to the formed area.At the end of forming process,the maxima loop of strainbreaks into two sub-extremum loops, one near the small end of the die, another near therollers.Secondly, under the shell section properties and the corresponding elements, thedistribution of wall thickness was investigated through two aspects of the thicknesscontours and finite element nodes during forming process. The results show that: byanalyzing the thickness contours, it is found that the distribution of thickness is a loop.Through introducing the mean absolute relative error(AARE),which can characterize theuniformityof wall thickness, we find that the wall thickness distributes uniformly in thecircumferential direction, however extremely unequally along the mold generatrixdirection, and in a distribution situation that the wall thickness distributes highlyin middle of die, lowly at the two ends; In the scale of finite element nodes,the thicknessof each node was calculated respectively in circumferential direction and axial directionusing statistical methods. The results show that the uniformity is very small incircumferential direction but extremely large in axial direction. For above conclusion,the uniformity of wall thickness in axial direction should be made a further study. Then,for this phenomenon, the wall thickness accuracy and uniformity in axial direction werestudied by a series of orthogonal simulation experiments.A set of optimal combinationof process parameters level was found through range analysis.Based on this group ofprocess parameters, the results of orthogonal experiments were optimized usingresponse surface methodology and validated by the finite element simulation.Simulationresults indicate that the optimal combination of process parameters is feasible.Finally, based on the results optimized through range analysis and response surfacemethodology, the corresponding spinning simulations and tests were taken, and the testsresults were compared to the optimal simulation results under the same die.Because ofthe limit of test conditions, part of the process parameters were not set as the optimizedresults, but permissible conditions werecompletely carried out as the optimizedresults.Tests were completed at different deviation rate, and the thickness of obtainedspinning coneswas measuredthrough three-dimensional scan tests. The simulationverification shows that the results optimized through orthogonal experiments andresponse surface methodology are reliable;The results of spinning tests show that thewall thickness distribution of cone power spinning obtained from simulations is wellagree with actual production, and because of the large springback value of cone powerspinning, the larger negative presupposed deviation rate should be taken when in theactual production. In conclusion, the overall research ideas of this paper are feasible andcan be applied to the relevant research of other shaped pieces.