Symbolic regression accelerates the discovery of quantitative relationships in rubber material aging

Abstract

Polymer materials, especially rubber, play an indispensable role in modern life and manufacturing. However, their aging and deterioration pose serious challenges to their stability and service life. Unexpected aging can lead to the deterioration of the physical and chemical properties of materials, thereby triggering a series of safety hazards and environmental pollution issues. Exploring the correspondence between the microscopic characteristics and macroscopic properties of materials during the aging process helps researchers deeply understand and control the aging process of materials. Symbolic regression (SR) algorithm, as a machine learning method with strong interpretability, plays an important role in exploring the quantitative relationship of data in scientific fields. This method has a strong potential for discovering the intrinsic quantitative relationships within the experimental data of material aging. In this study, we propose a comprehensive evaluation framework for SR, aiming to identify SR algorithms that are truly suitable for aging experimental data. Furthermore, by integrating characterization data of aging experiments, we conduct further validation and knowledge discovery with the selected method. The results obtained from our experimental data demonstrate a strong consistency with those of the proposed evaluation framework. Notably, this research methodology exhibits extensibility and can serve as a guiding light for the discovery of knowledge and the elucidation of mechanisms within other realms of polymer materials and diverse material systems.

Keywords

Symbolic regression algorithm / microscopic and macroscopic properties / rubber / materials aging / knowledge discovery

References

[1] Hao J,Wang J.An XPS investigation of thermal degradation and charring of cross-linked polyisoprene and polychloroprene.Polym Degrad Stab2001;71:305-15

[2] Wang J,Zhu J.An XPS study of the thermal degradation and flame retardant mechanism of polystyrene-clay nanocomposites.Polym Degrad Stab2002;77:249-52

[3] Moo-Tun NM,Uribe-Calderon JA.Thermo-oxidative aging of low density polyethylene blown films in presence of cellulose nanocrystals and a pro-oxidant additive.Polym Bull2018;75:3149-69

[4] Cai G,Jiang D.Degradation of fluorinated polyurethane coating under UVA and salt spray. Part II: Molecular structures and depth profile.Prog Org Coat2018;124:25-32

[5] Goliszek M,Sevastyanova O,Chabros A.Investigation of accelerated aging of lignin-containing polymer materials.Int J Biol Macromol2019;123:910-22

[6] Zuo P,Shirinbayan M,Bakir F.Multiscale physicochemical characterization of a short glass fiber–reinforced polyphenylene sulfide composite under aging and its thermo-oxidative mechanism.Polym Adv Technol2019;30:584-97

[7] Tripathy SP,Dwivedi KK,Ghosh S.Degradation in polytetrafluoro ethylene by 62 mev protons.Radiat Eff Defects Solids2002;157:303-10

[8] Khanna ND,Bhalla TC.Effect of biodegradation on thermal and crystalline behavior of polypropylene–gelatin based copolymers.J Appl Polym Sci2010;118:1476-88

[9] Nouh SA,El Hussieny HM.Modification induced by alpha particle irradiationin Makrofol polycarbonate.J Appl Polym Sci2008;109:3447-51

[10] Lei Z,Mattson CN.Aerosol acidity sensing via polymer degradation.Anal Chem2020;92:6502-11

[11] Gu X,Nguyen T,Yebassa D.Characterization of polyester degradation using tapping mode atomic force microscopy: exposure to alkaline solution at room temperature.Polym Degrad Stab2001;74:139-49

[12] Park B.Effects of acid hydrolysis on microstructure of cured urea-formaldehyde resins using atomic force microscopy.J Appl Polym Sci2011;122:3255-62

[13] Mouaci S,Saidi-Amroun N.Oxidative degradation and morphological properties of gamma-irradiated isotactic polypropylene films.Micro Nano Lett2017;12:478-81

[14] Moudoud M,Lamrous O,Touam T.Physical ageing of insulating polystyrene from dielectric properties measurements and structural analysis.Mater Res Express2019;6:095324

[15] Saba N,Alothman OY.A review on dynamic mechanical properties of natural fibre reinforced polymer composites.Constr Build Mater2016;106:149-59

[16] Hu X,Liu X,Huang Y.Temperature and frequency dependent rheological behaviour of carbon black filled natural rubber.Plast Rubber Compos2013;42:416-20

[17] Al-itry R,Maazouz A.Improvement of thermal stability, rheological and mechanical properties of PLA, PBAT and their blends by reactive extrusion with functionalized epoxy.Polym Degrad Stab2012;97:1898-914

[18] Shimamura H.Mechanical properties degradation of polyimide films irradiated by atomic oxygen.Polym Degrad Stab2009;94:1389-96

[19] Starkova O,Karl CW,Burlakovs J.Modelling of environmental ageing of polymers and polymer composites-durability prediction methods.Polymers2022;14:907 PMCID:PMC8912441

[20] Plota A.Lifetime prediction methods for degradable polymeric materials - a short review.Materials2020;13:4507 PMCID:PMC7599543

[21] Zhang X,Chen X,Xiao S.Theoretical study on decomposition mechanism of insulating epoxy resin cured by anhydride.Polymers2017;9:341 PMCID:PMC6418942

[22] Ismail AE,Grest GS.Diffusion of small penetrant molecules in polybutadienes.Mol Phys2011;109:2025-33

[23] Vuković F.Moisture ingress at the molecular scale in hygrothermal aging of fiber-epoxy interfaces.ACS Appl Mater Interfaces2020;12:55278-89

[24] Meng X,Yang R.Computational and experimental study on the mechanism of CO₂ production during photo-oxidative degradation of poly(butylene adipate-co-terephthalate): differences between PBA and PBT segments.Macromolecules2023;56:7749-62

[25] Doblies A,Fiedler B.Prediction of thermal exposure and mechanical behavior of epoxy resin using artificial neural networks and fourier transform infrared spectroscopy.Polymers2019;11:363 PMCID:PMC6419193

[26] Yuan W,Tamura R.Revealing factors influencing polymer degradation with rank-based machine learning.Patterns2023;4:100846 PMCID:PMC10724228

[27] Larché J,Thérias S.Photooxidation of polymers: relating material properties to chemical changes.Polym Degrad Stab2012;97:25-34

[28] Li X,Ye L.Stress photo-oxidative aging behaviour of polyamide 6.Polym Int2012;61:118-23

[29] Pourmand P,Furó I.Deterioration of highly filled EPDM rubber by thermal ageing in air: kinetics and non-destructive monitoring.Polym Test2017;64:267-76

[30] Neffe AT,Alteheld A.Controlled change of mechanical properties during hydrolytic degradation of polyester urethane networks.Macromol Chem Phys2010;211:182-94

[31] Kaiser E,Brunton SL.Sparse identification of nonlinear dynamics for model predictive control in the low-data limit.Proc Math Phys Eng Sci2018;474:20180335 PMCID:PMC6283900

[32] Schmidt M.Distilling free-form natural laws from experimental data.Science2009;324:81-5

[33] Kamienny PA,Lample G. End-to-end symbolic regression with transformers. arXiv 2022, arXiv:2204.10532. Available online: https://doi.org/10.48550/arXiv.2204.10532. (accessed on 21 Mar 2025)

[34] Valipour M,Panju M. SymbolicGPT: a generative transformer model for symbolic regression. arXiv 2021, arXiv:2106.14131. Available online: https://doi.org/10.48550/arXiv.2106.14131. (accessed on 21 Mar 2025)

[35] Udrescu SM.AI Feynman: a physics-inspired method for symbolic regression.Sci Adv2020;6:eaay2631 PMCID:PMC7159912

[36] Petersen BK,Mundhenk TN,Kim SK. Deep symbolic regression: recovering mathematical expressions from data via risk-seeking policy gradients. arXiv 2019, arXiv:1912.04871. Available online: https://doi.org/10.48550/arXiv.1912.04871. (accessed on 21 Mar 2025)

[37] Landajuela M,Yang J. A unified framework for deep symbolic regression. In: Proceedings of the 36nd International Conference on Neural Information Processing Systems, New Orleans, United States of America. 2022. Available from: https://proceedings.neurips.cc/paper_files/paper/2022/file/dbca58f35bddc6e4003b2dd80e42f838-Paper-Conference.pdf. (accessed on 2025-03-21)

[38] La Cava W,Burlacu B. Contemporary symbolic regression methods and their relative performance. arXiv 2021, arXiv:2107.14351. Available online: https://doi.org/10.48550/arXiv.2107.14351. (accessed on 21 Mar 2025)

[39] Wang Y,Rondinelli JM.Symbolic regression in materials science.MRS Commun2019;9:793-805

[40] He M.Machine learning and symbolic regression investigation on stability of MXene materials.Comput Mater Sci2021;196:110578

[41] Abdusalamov R,Itskov M.Automatic generation of interpretable hyperelastic material models by symbolic regression.Numer Methods Eng2023;124:2093-104

[42] Wang S.Interpretable catalysis models using machine learning with spectroscopic descriptors.ACS Catal2023;13:7428-36

[43] Matsubara Y,Igarashi R. Rethinking symbolic regression datasets and benchmarks for scientific discovery. arXiv 2022, arXiv:2206.10540. Available online: https://doi.org/10.48550/arXiv.2206.10540. (accessed on 21 Mar 2025)