a.Lab of Precision Manufacturing, Institute of Intelligent Machines, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Changzhou 213164, China
b.School of Machinery and Automation, Wuhan University of Science and Technology, Wuhan 430081, China
Milling force is key to the understanding of cutting mechanism and the control of machining process. Traditional milling force models have limited prediction accuracy due to their simplified conditions and incomplete knowledge contained for model construction. On the other hand, due to the lack of guidance from physics, the data-driven models lack interpretability, making them challenging to generalize to practical applications. To meet these difficulties, a deep network model guided by milling dynamics is proposed in this study to predict the instantaneous milling force and spindle vibration under varying cutting conditions. The model uses a milling dynamics model to generate data sets to pre-train the deep network and then integrates the experimental data for fine-tuning to improve the model’s generalization and accuracy. Additionally, the vibration equation is incorporated into the loss function as the physical constraint, enhancing the model’s interpretability. A milling experiment is conducted to validate the effectiveness of the proposed model, and the results indicate that the physics incorporated could improve the network learning capability and interpretability. The predicted results are in good agreement with the measured values, with an average error as low as 2.6705%. The prediction accuracy is increased by 24.4367% compared to the pure data-driven model.
HajduD, AstarloaA, KovacsI, DombovariZ. The curved uncut chip thickness model: a general geometric model for mechanistic cutting force predictions. Int J Mach Tools Manuf2023;188:104019. . 10.1016/j.ijmachtools.2023.104019
[2]
LiZM, SongQH, JinPJ, LiuZQ, WangB, MaHF. Chatter suppression techniques in milling processes: a state of the art review. Chin J Aeronaut2024;37 (7):1‒23. . 10.1016/j.cja.2023.10.001
[3]
BiróI, SzalayT, GeierN. Effect of cutting parameters on section borders of the empirical specific cutting force model for cutting with micro-sized uncut chip thickness. Procedia CIRP2018;77:279‒82. . 10.1016/j.procir.2018.09.015
[4]
MagliaroJ, SheryP, SmithR, AltenhofW, BeehE, StraßburgerP, et al. A semi-empirical analytical model to predict the axial cutting force of AZ31B magnesium extrusions. Thin-Walled Struct2018;131:360‒72. . 10.1016/j.tws.2018.06.032
[5]
MaoJ, ZhaoM, WeiXC, ZhangLQ. A material constitutive model-based prediction method for flank milling force considering the deformation of workpiece. J Manuf Processes2022;84:403‒13. . 10.1016/j.jmapro.2022.09.053
[6]
ZhouT, CuiH, WangY, YangWJ, HeL. Multi-physics analytical modeling of the primary shear zone and milling force prediction. J Mater Process Technol2023;316:117949. . 10.1016/j.jmatprotec.2023.117949
[7]
LiKX, ZhuKP, MeiT. A generic instantaneous undeformed chip thickness model for the cutting force modeling in micromilling. Int J Mach Tools Manuf2016;105:23‒31. . 10.1016/j.ijmachtools.2016.03.002
[8]
AslanD, AltintasY. Prediction of cutting forces in five-axis milling using feed drive current measurements. IEEE/ASME Trans Mechatron2018;23(2):833‒44. . 10.1109/tmech.2018.2804859
[9]
SchmuckerB, WangCP, ZaehMF, ErkorkmazK. Wide-bandwidth cutting force monitoring via motor current and accelerometer signals. CIRP Ann2023;72(1):309‒12. . 10.1016/j.cirp.2023.04.012
[10]
HouMX, CaoHR, LiQ, ShiJH. Dynamic force identification in milling based on IRLS using acceleration signals. Measurement2023;217:113096. . 10.1016/j.measurement.2023.113096
[11]
PostelM, AslanD, WegenerK, AltintasY. Monitoring of vibrations and cutting forces with spindle mounted vibration sensors. CIRP Ann2019;68(1):413‒6. . 10.1016/j.cirp.2019.03.019
[12]
MostaghimiH, ParkCI, KangG, ParkSS, LeeDY. Reconstruction of cutting forces through fusion of accelerometer and spindle current signals. J Manuf Processes 2021;68(Pt A):990‒1003.
[13]
SalehiM, AlbertelliP, GolettiM, RipamontiF, TomasiniG, MonnoM. Indirect model based estimation of cutting force and tool tip vibrational behavior in milling machines by sensor fusion. Procedia CIRP2015;33:239‒44. . 10.1016/j.procir.2015.06.043
AttiaH, SadekA, AltintasA, MatsubaraA, UmbrelloD, WegenerK, et al. Physics based models for characterization of machining performance—a critical review. CIRP J Manuf Sci Technol2024;51:161‒89. . 10.1016/j.cirpj.2024.04.008
[17]
ZhouJ, ZhouYH, WangBC, ZangJY. Human-cyber-physical systems (HCPSs) in the context of new-generation intelligent. Engineering2019;5(4):624‒36. . 10.1016/j.eng.2019.07.015
[18]
PengDF, LiHK, DaiYB, WangZD, OuJY. Prediction of milling force based on spindle current signal by neural networks. Measurement2022;205:112153. . 10.1016/j.measurement.2022.112153
[19]
TaoY, LiZW, HuP, ChenFW, JuBF, ChenYL. High-accurate cutting forces estimation by machine learning with voice coil motor-driven fast tool servo for micro/nano cutting. Precis Eng2023;79:291‒9. . 10.1016/j.precisioneng.2022.11.014
[20]
LiBC, LiuTK, LiaoJ, FengCH, YaoLC, ZhangJY. Non-invasive milling force monitoring through spindle vibration with LSTM and DNN in CNC machine tools. Measurement2023;210:112554. . 10.1016/j.measurement.2023.112554
[21]
RabieeAH, TahmasbiV, QasemiM. Experimental evaluation, modeling and sensitivity analysis of temperature and cutting force in bone micro-milling using support vector regression and EFAST methods. Eng Appl Artif Intell2023;120:105874. . 10.1016/j.engappai.2023.105874
[22]
DarshanS, DesaiKA, BhattacharyyaA. Neural network-based model parameter estimation for end milling of carbon fiber reinforced polymer (CFRP) composites. Procedia CIRP2023;117:365‒70. . 10.1016/j.procir.2023.03.062
[23]
ZhangJJ, GaoRX. Deep learning-driven data curation and model interpretation for smart manufacturing. Chin J Mech Eng2021;34(1):71. . 10.1186/s10033-021-00587-y
[24]
Navarro-DeviaJH, ChenY, DaoDV, LiHZ. Chatter detection in milling processes—a review on signal processing and condition classification. Int J Adv Manuf Technol 2023;125(9‒10):3943‒80. . 10.1007/s00170-023-10969-2
[25]
ZhongRY, XuX, KlotzE, NewmanST. Intelligent manufacturing in the context of industry 4.0: a review. Engineering2017;3(5):616‒30. . 10.1016/j.eng.2017.05.015
XuK, LiYG, ZhangJC, ChenGX. ForceNet: an offline cutting force prediction model based on neuro-physical learning approach. J Manuf Syst2021;61:1‒15. . 10.1016/j.jmsy.2021.08.001
[28]
AgarwalA, DesaiKA. Amalgamation of physics-based cutting force model and machine learning approach for end milling operation. Procedia CIRP2020;93:1405‒10. . 10.1016/j.procir.2020.04.102
[29]
XieJJ, HuPC, ChenJH, HanWS, WangRH. Deep learning-based instantaneous cutting force modeling of three-axis CNC milling. Int J Mech Sci2023;246:108153. . 10.1016/j.ijmecsci.2023.108153
[30]
ZhuY, ZiY, LiJ, XuJ. PhysiCausalNet: a causal-and physics-driven domain generalization network for cross-machine fault diagnosis of unseen domain. IEEE Trans Industr Inform2024;20(6):8488‒98. . 10.1109/tii.2024.3369240
[31]
LiX, ZhangW. Physics-informed deep learning model in wind turbine response prediction. Renew Energy2022;185:932‒44. . 10.1016/j.renene.2021.12.058
[32]
HanachiH, YuWN, KimYI, LiuJ, MechefskeCK. Hybrid data-driven physics-based model fusion framework for tool wear prediction. Int J Adv Manuf Technol 2019;101(9‒12):2861‒72.
[33]
RahimiMH, HuynhHN, AltintasY. On-line chatter detection in milling with hybrid machine learning and physics-based model. CIRO J Manuf Sci Technol2021;35:25‒40. . 10.1016/j.cirpj.2021.05.006
[34]
CorsonG, KarandikarJ, SchmitzT. Physics-informed Bayesian machine learning case study: integral blade rotors. J Manuf Processes2023;85:503‒14. . 10.1016/j.jmapro.2022.12.004
[35]
XiongJW, FinkO, ZhouJ, MaYZ. Controlled physics-informed data generation for deep learning-based remaining useful life prediction under unseen operation conditions. Mech Syst Signal Process2023;197:110359. . 10.1016/j.ymssp.2023.110359
[36]
LuWJ, WangY, ZhangMQ, GuJW. Physics guided neural network: remaining useful life prediction of rolling bearings using long short-term memory network through dynamic weighting of degradation process. Eng Appl Artif Intell2024;127(Pt B):107350. . 10.1016/j.engappai.2023.107350
[37]
VaishnavS, AgarwalA, DesaiKA. Machine learning-based instantaneous cutting force model for end milling operation. J Intell Manuf2020;31(6):1353‒66. . 10.1007/s10845-019-01514-8
[38]
FinkeldeyF, SaadallahA, WiederkehrP, MorikK. Real-time prediction of process forces in milling operations using synchronized data fusion of simulation and sensor data. Eng Appl Artif Intell2020;94:103753. . 10.1016/j.engappai.2020.103753
[39]
WangJC, ZouB, LiuMF, LiYS, DingHJ, XueK. Milling force prediction model based on transfer learning and neural network. J Intell Manuf2021;32(4):947‒56. 85. . 10.1007/s10845-020-01595-w
[40]
LiuYP, AltintasY. Predicting the position-dependent dynamics of machine tools using progressive network. Precis Eng2022;73:409‒22. . 10.1016/j.precisioneng.2021.10.010
[41]
HaJM, FinkO. Domain knowledge-informed synthetic fault sample generation with health data map for cross-domain planetary gearbox fault diagnosis. Mech Syst Signal Process2023;202:110680. . 10.1016/j.ymssp.2023.110680
[42]
TaiCY, AltintasY. A physics-based model-data-driven method for spindle health diagnosis-part III: model training and fault detection. J Manuf Sci Eng2024;146(8):081006. . 10.1115/1.4065227
[43]
AlbertsM, St. JohnS, JaredB, KarandikarJ, KhojandiA, SchmitzT, et al. Chatter detection in simulated machining data: a simple refined approach to vibration data. Int J Adv Manuf Technol2024;132:4541‒57. . 10.1007/s00170-024-13590-z
ChenKD, ZhangH, van de WouwN, DetournayE. An alternative approach to model the dynamics of a milling tool. J Sound Vibrat2024;569:117940. . 10.1016/j.jsv.2023.117940
[48]
ZhaoMZ, WeiHL, MaoYM, ZhangCD, LiuTT, LiaoWH. Predictions of additive manufacturing process parameters and molten pool dimensions with a physics-informed deep learning model. Engineering2023;23:181‒95. . 10.1016/j.eng.2022.09.015
[49]
ZaghbaniI, SongmeneV. Estimation of machine-tool dynamic parameters during machining operation through operational modal analysis. Int J Mach Tools Manuf 2009;49(12‒13):947‒57.
[50]
WangYS, FangHZ, JinJY, MaGJ, HeX, DaiX, et al. Data-driven discovery of stochastic differential equations. Engineering2022;17:244‒52. . 10.1016/j.eng.2022.02.007
[51]
ChakrabortyS, TomsettR, RaghavendraR, HarborneD, AlzantotM, CeruttiF, et al. Interpretability of deep learning models: a survey of results. In: Proceedings of the 2017 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computed, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/ CBDCom/IOP/SCI); 2017 Aug 4‒8; San Francisco, CA, USA. Piscataway: IEEE; 2018. p. 1‒6. . 10.1109/uic-atc.2017.8397411