Springback Prediction of TA18 Tube Bending Based on Kriging Model and Active Learning

XIAO Huimin; DING Qingguo; PAN Lin; ZHANG Weisheng

doi:10.7643/issn.1672-9242.2026.04.009

PDF(7476 KB)

Equipment Environmental Engineering ›› 2026, Vol. 23 ›› Issue (4) : 90-103. DOI: 10.7643/issn.1672-9242.2026.04.009

Aviation and Aerospace Equipment

Springback Prediction of TA18 Tube Bending Based on Kriging Model and Active Learning

XIAO Huimin¹, DING Qingguo^2,*, PAN Lin², ZHANG Weisheng¹

Author information +

History +

Abstract

To address the challenges of insufficient springback prediction accuracy and the reliance on trial-and-error for process optimization caused by strong nonlinearity and multi-parameter coupling during the numerical control (NC) bending of aero-engine TA18 titanium alloy thin-walled tubes, the work aims to investigate methods for predicting bending springback mechanical behavior and compensating for deformation. An efficient springback prediction and compensation method was established based on orthogonal experiment data-driven approaches combined with parametric finite element (FE) simulation by incorporating an improved Kriging surrogate model and an active learning algorithm. Specifically, the contribution of each process parameter to springback was firstly analyzed via orthogonal experiments to construct a range feature weighting matrix. This matrix was then embedded into the Kriging surrogate model, effectively integrating physical prior knowledge into the modeling process. To reduce prediction variance at the boundaries of the parameter space, a parametric FE simulation model was established as the data foundation. Furthermore, an active learning framework based on a variance-maximization sampling strategy was designed to adaptively supplement samples in regions of high uncertainty. Finally, a corresponding springback compensation algorithm was developed and experimentally validated. Compared to the linear regression model based on orthogonal data, the method proposed in this work reduced the root mean square error (RMSE) of springback prediction for TA18 tube material from 0.99° to 0.50°, and the average relative error decreased from 2.40% to 0.83%. Additionally, the computational efficiency was improved by three orders of magnitude compared to pure FE simulation. This work establishes an efficient prediction and compensation framework for the bending springback of TA18 titanium alloy tubes by integrating physical prior knowledge with data-driven active learning. This method not only resolves the issue of model overfitting under small sample conditions but also optimizes computational resource allocation through variance-maximization sampling, providing a technical pathway for the precision manufacturing of aero-engine ducts that balances accuracy and efficiency.

Key words

TA18 titanium alloy / Kriging model / active learning / springback prediction / springback compensation / CNC bending

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

XIAO Huimin, DING Qingguo, PAN Lin, ZHANG Weisheng. Springback Prediction of TA18 Tube Bending Based on Kriging Model and Active Learning[J]. Equipment Environmental Engineering. 2026, 23(4): 90-103 https://doi.org/10.7643/issn.1672-9242.2026.04.009

References

[1] LI H, YANG H, YAN J, et al.Numerical Study on Deformation Behaviors of Thin-Walled Tube NC Bending with Large Diameter and Small Bending Radius[J]. Computational Materials Science, 2009, 45(4): 921-934.
[2] YANG H, LI H, ZHANG Z Y, et al.Advances and Trends on Tube Bending Forming Technologies[J]. Chinese Journal of Aeronautics, 2012, 25(1): 1-12.
[3] LI H, YANG H, SONG F F, et al.Springback Characterization and Behaviors of High-Strength Ti-3Al-2.5V Tube in Cold Rotary Draw Bending[J]. Journal of Materials Processing Technology, 2012, 212(9): 1973-1987.
[4] LI H, YANG H, SONG F F, et al.Springback Nonlinearity of High-Strength Titanium Alloy Tube Upon Mandrel Bending[J]. International Journal of Precision Engineering and Manufacturing, 2013, 14(3): 429-438.
[5] MA J, LI H, HE Z R, et al.Complex Unloading Behavior of Titanium Alloy in Cold and Thermal-Mechanical Working[J]. International Journal of Mechanical Sciences, 2022, 233: 107672.
[6] JIANG Z Q, YANG H, ZHAN M, et al.Coupling Effects of Material Properties and the Bending Angle on the Springback Angle of a Titanium Alloy Tube during Numerically Controlled Bending[J]. Materials & Design, 2010, 31(4): 2001-2010.
[7] 鄂大辛, 王险峰. 管材无芯弯曲中回弹规律的研究[J]. 锻压技术, 2006, 31(5): 68-71.
E D X, WANG X F. Study on the Back-Bounding Law in the Process of Coreless Tube Bending[J]. Forging & Stamping Technology, 2006, 31(5): 68-71.
[8] AL-QURESHI H A, RUSSO A. Spring-back and Residual Stresses in Bending of Thin-Walled Aluminium Tubes[J]. Materials & Design, 2002, 23(2): 217-222.
[9] AL-QURESHI H A. Elastic-Plastic Analysis of Tube Bending[J]. International Journal of Machine Tools and Manufacture, 1999, 39(1): 87-104.
[10] MA J, LI H, FU M W.Modelling of Springback in Tube Bending: A Generalized Analytical Approach[J]. International Journal of Mechanical Sciences, 2021, 204: 106516.
[11] ZHAN M, WANG Y, YANG H, et al.An Analytic Model for Tube Bending Springback Considering Different Parameter Variations of Ti-Alloy Tubes[J]. Journal of Materials Processing Technology, 2016, 236: 123-137.
[12] LIANG C, LUO Y D, LIANG J C, et al.Analytical Springback Assessment and Compensation in 3D Multi-Point Flexible Stretch Bending Forming[J]. The International Journal of Advanced Manufacturing Technology, 2023, 129(1): 197-206.
[13] SAFDARIAN R.Investigation of Tube Fracture in the Rotary Draw Bending Process Using Experimental and Numerical Methods[J]. International Journal of Material Forming, 2020, 13(4): 493-516.
[14] WANG Z L, LIN Y C, QIU L M, et al.Spatial Variable Curvature Metallic Tube Bending Springback Numerical Approximation Prediction and Compensation Method Considering Cross-Section Distortion Defect[J]. The International Journal of Advanced Manufacturing Technology, 2022, 118(5): 1811-1827.
[15] XIAO J, XU X F, ZUO D W, et al.Influences on the Extended Length and Performance of Push Bending for the GH4169 Superalloy Tube with Small Bending Radius of 1D[J]. The International Journal of Advanced Manufacturing Technology, 2021, 112(1): 107-119.
[16] PENG M L, XU X F, FAN Y B, et al.Process Parameters Optimization of Differential Heating Push-Bending for Small Bending Radius Tube Based on Orthogonal Experiments[J]. The International Journal of Advanced Manufacturing Technology, 2024, 133(7): 3399-3414.
[17] WELO T, MA J, BLINDHEIM J, et al.Flexible 3D Stretch Bending of Aluminium Alloy Profiles: An Experimental and Numerical Study[J]. Procedia Manufacturing, 2020, 50: 37-44.
[18] GUO X Z, CHENG X, XU Y, et al.Finite Element Modelling and Experimental Investigation of the Impact of Filling Different Materials in Copper Tubes during 3D Free Bending Process[J]. Chinese Journal of Aeronautics, 2020, 33(2): 721-729.
[19] LI H, MA J, LIU B Y, et al.An Insight into Neutral Layer Shifting in Tube Bending[J]. International Journal of Machine Tools and Manufacture, 2018, 126: 51-70.
[20] 王鹏鹏, 程子詹, 凌强, 等. 基于深度学习神经网络的铝合金型材回弹预测[J]. 锻压技术, 2024, 49(7): 105-111.
WANG P P, CHENG Z Z, LING Q, et al.Springback Prediction of Aluminum Alloy Profile Based on Deep Learning Neural Network[J]. Forging & Stamping Technology, 2024, 49(7): 105-111.
[21] 占少伟, 龚俊杰, 韦源源, 等. 基于DPSO-BP神经网络的V形自由折弯成形角度和回弹预测[J]. 锻压技术, 2023, 48(8): 151-157.
ZHAN S W, GONG J J, WEI Y Y, et al.Prediction on V-Shaped Free Bending Angle and Springback Based on DPSO-BP Neural Network[J]. Forging & Stamping Technology, 2023, 48(8): 151-157.
[22] ZHANG P F, FANG Z L, LI L Y, et al.A GAN-SVR Prediction Method of the Metal Tube-Bending Rebound with Small Samples[J]. Journal of Sensors, 2023, 2023(1): 6616607.
[23] 孙振彪. 小直径铝合金管机器人弯曲成形起皱缺陷调控研究[D]. 南京: 南京航空航天大学, 2023.
SUN Z B.Research on Wrinkle of Small Diameter Aluminum Alloy by the Robot-Based Tube Bending Process[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2023.
[24] 郭小农, 李根, 曾强, 等. 基于响应面法的多拱度铝合金节点板冲压回弹预测与补偿[J]. 湖南大学学报(自然科学版), 2025, 52(5): 28-40.
GUO X N, LI G, ZENG Q, et al.Springback Prediction and Compensation of Multi-Camber Aluminum Alloy Gusset Joint Plates Using Response Surface Method[J]. Journal of Hunan University (Natural Sciences), 2025, 52(5): 28-40.
[25] GAO Y, LI Q, SHI M X, et al.Springback Analysis of Free Bending Based on Uniform Design[C]// 2010 International Conference on Mechanic Automation and Control Engineering. Wuhan. IEEE, 2010: 3811-3814.
[26] LI Z Y, WANG Z L, ZHANG S Y, et al.Springback Active Prediction-Compensation Framework: Difficult-to-Manufacturing Metal Tubes Intelligent Bending Based on Alert Collaborative Sand Cat Swarm Algorithm[J]. The International Journal of Advanced Manufacturing Technology, 2025, 137(3): 1683-1704.
[27] KLEIJNEN J P C. Regression and Kriging Metamodels with Their Experimental Designs in Simulation: A Review[J]. European Journal of Operational Research, 2017, 256(1): 1-16.
[28] LOU H Z, STELSON K A.Three-Dimensional Tube Geometry Control for Rotary Draw Tube Bending: Part 1—Bend Angle and Overall Tube Geometry Control[C]//Manufacturing Engineering. Orlando, Florida, USA. American Society of Mechanical Engineers, 2000: 613-620.
[29] ZHANG S Y, FU M Y, WANG Z L, et al.Springback Prediction Model and Its Compensation Method for the Variable Curvature Metal Tube Bending Forming[J]. The International Journal of Advanced Manufacturing Technology, 2021, 112(11): 3151-3165.
[30] SIMONETTO E, GHIOTTI A, BRUSCHI S.In-Process Measurement of Springback in Tube Rotary Draw Bending[J]. The International Journal of Advanced Manufacturing Technology, 2021, 112(9): 2485-2496.
[31] CHEN P, LI J S, LI C C, et al.Research on the Hot-Temperature Rheological Behavior and Microstructural Evolution of TA18 Titanium Alloy[J]. Materials Today Communications, 2025, 43: 111810.
[32] 鄂大辛, 周大军. 金属管材弯曲理论及成形缺陷分析[M]. 北京: 北京理工大学出版社, 2016: 367-426.
E D X, ZHOU D J. Metal Tube Bending: Theory and Forming Defects Analysis[M]. Beijing: Beijing Institute of Technology Press, 2016: 367-426.
[33] LOOKMAN T, BALACHANDRAN P V, XUE D Z, et al.Active Learning in Materials Science with Emphasis on Adaptive Sampling Using Uncertainties for Targeted Design[J]. npj Computational Materials, 2019, 5: 21.