Modeling And Control Of Springback In Heat-Assisted Bending Of High-Strength Titanium Tube

To meet the ever-growing demands on high working pressure,lightweight and long service life of tube systems in high-end equipment such as aircrafts and engines,it urgently needs to achieve precision manufacture of high-strength titanium tube(HSTT)components with small-radius curvature structures.Towards such kind of hard-to-deform material and difficult-to-fabricate structure,heat-assisted bending(HAB)can excavate the deformation potential and improve the forming limits,becoming a promising manufacturing technology for such kind of high-performance tubular components.However,the higher strength-to-modulus ratio leads to an extremely significant springback upon bending even under the heat-assisted forming conditions,which seriously reduces the dimension accuracy of products and further affects assembly quality and service safety.The thermal-mechanical coupling loading,under the joint action of locally thermal fields and multi-tool constraints,makes the unloading behavior in the HSTT heat-assisted bending process more complicated.Therefore,accurate prediction and control of unloading springback have grown to be a challenging problem in precision manufacture of high-performance titanium tubular components.To this end,taking the HSTT heat-assisted bending as the typical case in this thesis,a systematic and in-depth investigation on the fundamentally scientific problems,with focus on springback rules,mechanisms,modeling and control,is conducted by using theoretical,finite element(FE)and experimental approaches.The main contents and results are summarized as follows:Based on the tension/compression tests at various temperatures,the evolving tension-compression(T-C)asymmetric work hardening effect of HSTT with deformation temperature and plastic strain is experimentally clarified: the hardening effect of tension-compression stress states trends from “T﹥C” to “C﹥T” with strain accumulation,and the T-C strength difference weakens with the increase of temperature.Based on the tension loading-unloading-reloading tests at various temperatures,the nonlinear unloading behavior and its temperature-dependent evolution are revealed: both unloading stress-strain relations and strain-related modulus reduction effect are nonlinear,and the nonlinearities are weakened at the higher temperatures.A continuum dislocation dynamics-based model for nonlinear unloading is constructed,and the underlying mechanisms of temperature-dependent nonlinear unloading is explained from the points of dynamic evolutions of reversible mobile dislocation density and dislocation mean free path within the physical modeling framework.A constitutive modeling framework for coupling the temperature-dependent asymmetric plasticity and nonlinear unloading elasticity is established.The plasticity model is constructed based on an asymmetric/anisotropic yield criterion associated with the temperature/strain-related interpolation approach,thus achieving the description of temperature-dependent asymmetry/anisotropy evolution of yielding and work hardening in thermal-mechanical loading.For the unloading process,a temperature/strain-related unloading chord modulus evolution model is firstly derived,and then the multi-surface theory for defining the different elastic fields during unloading process is employed to model the nonlinear unloading behavior at different temperatures.Based on the implicit Euler algorithm and the judging algorithm of unloading elastic fields,the coupled elasticity-plasticity model is numerically implemented.From analyzing the characteristics of the through-process of heat-assisted bending of high-strength titanium tube,the key FE modeling issues for heating/heat-transferring,heat-assisted bending and thermal-mechanical coupling unloading are identified,and a combined Implicit-Explicit-Implicit thermal-mechanical coupling FE model system for heating-bending-unloading is established.An experimental platform including multi-tool system and heating-control system is developed,and thus the manufacture of the high-strength titanium bent tubular parts with R=1.5D small radius is realized.Combing the experiments and FE simulation,the prediction capabilities of various combinations of plasticity and unloading elasticity constitutive models are evaluated,indicating that the coupling model for asymmetric plasticity and nonlinear elasticity can significantly improve the whole-process prediction accuracy with the wall-thinning error ≤ 2.15%,cross-section flattening error ≤ 0.65%,and springback error ≤ 0.20°.The influences of bending angle,bending radius and the non-uniform local thermal fields on the springback as well as residual stress distribution of unloaded bent tube in HSTT heat-assisted bending are clarified based on FE simulation and experiments.An analytical model for neutral layer shifting(NLS)in tube bending considering the T-C asymmetry,geometry parameters and multi-tool-induced constraints is constructed based on the axial force equilibrium.On this basis,an analytical springback modeling framework with coupling the nonlinear unloading and residual stress after unloading is established.Using the analytical models in combination with FE simulation and typical experiments,the underlying mechanisms of NLS and springback for tube bending are extensively explored from analyzing the coupled influence of material properties,geometry parameters,tool/process parameters and residual stress distribution on the force/moment equilibrium upon bending and unloading.The contributions of springback minimization strategy and compensation strategy to the global control of springback,and their influences on forming quality in HSTT heat-assisted bending are analyzed and evaluated,indicating that the springback minimization strategy can hardly contribute to the effective control of springback.Based on this,two selectable compensation methods are developed for different manufacturing needs,including: a “One Trail + FEM(OT-FEM)” based inverse compensation method is proposed to meet the large batch manufacture of bent tubular parts with simple structures,in which the compensation error under typical bending condition “R=2D,θ=135°” is less than 0.15°;and a machine learning based flexible compensation method is proposed to meet the small batch manufacture of bent tubular parts with diverse angular/radius structure features,in which the maximum error is less than 0.50° and average error is less than 0.20° under a wide bending range“R=1.5～3D,θ=5～180°”.Toward the needs of Industry 4.0 manufacturing,a concept strategy of online close-loop-control for springback in tube bending is proposed based on smart tooling design and big-data-driven modeling.