This study constructs four functionalized graphene/nitrile-butadiene rubber (FGNS/NBR) molecular dynamics (MD) models under the COMPASS force field to evaluate mechanical properties and abrasion rates. For each composite, 50 independent shear simulations (200 samples) were run; averages of force-field energy components over the post–steady-state segment built a bonded/non-bonded energy-fingerprint library. Feature screening relied on the consensus across multiple importance measures, yielding 17 descriptors. A Random Forest regressor (RFR) was trained, and mechanistic interpretation combined SHapley Additive explanations with mutual information. The MD analyses indicate that functionalization generally enhances mechanical properties and improves wear resistance. Subsequently, the RFR explanatory model further revealed three principal determinants of wear performance—Fe layer van der Waals energy, total bond length after shearing, and the torsion–stretch energy of FGNS—and indicated that the interplay between chain extension and interfacial adsorption constitutes the dominant anti-wear pathway. Overall, the study introduces an interpretable comparative method focused on relative trends across functionalizations, offering a useful analytic perspective for small-difference systems and for considering functional group and grafting choices.
HolmbergKKivikytö-ReponenPHärkisaariP, et al.Global energy consumption due to friction and wear in the mining industry. Tribol Int2017; 115: 116–139.
2.
FriedrichKSchlarbAK. Tribology of polymeric nanocomposites: friction and wear of bulk materials and coatings. 2nd ed.Elsevier (Tribology and Interface Engineering Series), 2013, vol 55.
3.
ZhaiWBaiLZhouR, et al.Recent progress on wear-resistant materials: designs, properties, and applications. Adv Sci2021; 8(11): e2003739.
4.
SongSZhuZDuS, et al.Research on polymer wear under water conditions: a review. Lubricants2024; 12(9): 312.
5.
MyshkinNKPetrokovetsMIKovalevAV. Tribology of polymers: adhesion, friction, wear, and mass-transfer. Tribol Int2005; 38(11–12): 910–921.
6.
GeorgakilasVOtyepkaMBourlinosAB, et al.Functionalization of graphene: covalent and non-covalent approaches, derivatives and applications. Chem Rev2012; 112(11): 6156–6214.
7.
GovindarajPSokolovaASalimN, et al.Distribution states of graphene in polymer nanocomposites: a review. Compos B Eng2021; 226: 109353.
8.
QamarSRamzanNAleemW, et al.Graphene dispersion, functionalization techniques and applications: a review. Synth Met2024; 307: 117697.
JenjobRSeidiFCrespyD. Recent advances in polymerizations in dispersed media. Adv Colloid Interface Sci2018; 260: 24–31.
11.
WangCLanYYuW, et al.Preparation of amino-functionalized graphene oxide/polyimide composite films with improved mechanical, thermal and hydrophobic properties. Appl Surf Sci2016; 362: 11–19.
12.
RafieeRRabczukTPouraziziR, et al.Challenges of the modeling methods for investigating the interaction between the CNT and the surrounding polymer. Adv Mater Sci Eng2013; 2013: 1–10.
13.
RafieeRMahdaviM. Characterizing nanotube–polymer interaction using molecular dynamics simulation. Comput Mater Sci2016; 112: 356–363.
14.
ShenXZhengQKimJ-K. Rational design of two-dimensional nanofillers for polymer nanocomposites toward multifunctional applications. Prog Mater Sci2021; 115: 100708.
15.
AllenMPTildesleyDJ. Computer simulation of liquids. 2nd ed.Oxford University Press, 2017.
16.
LiYWangQWangS. A review on enhancement of mechanical and tribological properties of polymer composites reinforced by carbon nanotubes and graphene sheet: molecular dynamics simulations. Compos B Eng2019; 160: 348–361.
17.
HeJTangHWangC. Advances of molecular dynamics simulation in tribochemistry and lubrication investigations: a review. J Ind Eng Chem2023; 126: 1–19.
18.
YinTWangGGuoZ, et al.Molecular dynamics simulation on polymer tribology: a review. Lubricants2024; 12(6): 205.
19.
SunH. COMPASS: an ab initio force-field optimized for condensed-phase applications—Overview with details on alkane and benzene compounds. J Phys Chem B1998; 102(38): 7338–7364.
20.
YangZGuoZYuanC, et al.Tribological behaviors of composites reinforced by different functionalized carbon nanotube using molecular dynamic simulation. Wear2021; 476: 203669.
21.
LvCXueQXiaD, et al.Effect of chemisorption structure on the interfacial bonding characteristics of graphene–polymer composites. Appl Surf Sci2012; 258(6): 2077–2082.
22.
CuiJZhaoJWangS, et al.A comparative study on enhancement of mechanical and tribological properties of nitrile rubber composites reinforced by different functionalized graphene sheets: molecular dynamics simulations. Polym Compos2021; 42(1): 205–219.
23.
LiangCShuaiCWangX. Molecular dynamics simulation and experimental study of friction and wear characteristics of carbon nanotube-reinforced nitrile butadiene rubber. Lubricants2024; 12(7): 261.
24.
DongZTangHYangB, et al.Molecular dynamics simulation and experimental study of the mechanical and tribological properties of GNS-COOH/PEEK/PTFE composites. Polymers2024; 16(18): 2572.
25.
TalapatraADattaD. Experimental and molecular dynamics simulation based investigation to understand tribological performance of graphene reinforced thermoplastic polyurethane (Gr/TPU) nanocomposites. Tribol Int2024; 196: 109703.
26.
YangBLiYWangS, et al.Aminosilane modified graphene oxide for reinforcing nitrile butadiene rubber: experiments and molecular dynamic simulations. Compos Sci Technol2023; 235: 109956.
27.
DongZLiuYTangH, et al.A study on the mechanism of enhanced friction behavior in functionalized CNT-reinforced PEEK/PTFE composites. Polym Eng Sci2025; 65(8): 4186–4197.
28.
YuanQLiYWangS, et al.A molecular dynamics simulation study on enhancement of mechanical and tribological properties of nitrile-butadiene rubber with varied contents of acrylonitrile. Polymers2023; 15(18): 3799.
29.
ChoderaJD. A simple method for automated equilibration detection in molecular simulations. J Chem Theory Comput2016; 12(4): 1799–1805.
30.
FuruyaTKogaT. Molecular simulation of structure formation and rheological properties of mixtures of telechelic and monofunctional associating polymer. J Polym Sci B Polym Phys2018; 56(18): 1251–1264.
31.
ZuoZLiangLBaoQ, et al.Molecular dynamics calculation on the adhesive interaction between the polytetrafluoroethylene transfer film and iron surface. Front Chem2021; 9: 740447.
32.
ZhangSLuoJLiuY. Tribo-informatics approaches in tribology research: a review. Wear2023; 518–519: 204723.
33.
WuXWangZYeZ, et al.AI for tribology: present and future. Friction2024; 12: 1345–1376.
34.
YaoSHeWChenY, et al.Molecular dynamics simulations and microscopic analysis of the damping performance of hindered phenol AO-60/nitrile-butadiene rubber composites. RSC Adv2014; 4(105): 61062–61072.
35.
AltmannAToloşiLSanderO, et al.Permutation importance: a corrected feature importance measure. Bioinformatics2010; 26(10): 1340–1347.
36.
SweattAJYuJBiY, et al.What makes a good prediction? Feature importance and beginning to open the black box of machine learning in genetics. Annu Rev Genet2022; 56: 133–164.
37.
HastieTTibshiraniRFriedmanJ. The elements of statistical learning: data mining, inference, and prediction. 2nd ed.Springer, 2009.
38.
BreimanL. Random forests. Mach Learn2001; 45(1): 5–32.
39.
RaschkaS. Model evaluation, model selection, and algorithm selection in machine learning. arXiv preprint arXiv:1811.12808, 2018.
40.
HutterFHoosHHLeyton-BrownK. An efficient approach for assessing hyperparameter importance. Proc Int Conf Mach Learn2014; 32(1): 754–762.
41.
HuangBFFBoutrosPC. The parameter sensitivity of random forests. BMC Bioinf2016; 17: 331.
42.
ShapleyLS. A value for n-person games. In: KuhnHWTuckerAW (eds). Contributions to the Theory of Games II. Princeton University Press, 1953, pp. 307–317.
43.
LundbergSMLeeS-I. A unified approach to interpreting model predictions. In: Advances in Neural Information Processing Systems 30 (NeurIPS 2017), Long Beach, CA, USA, 4–9 December 2017, 2017.
44.
ShannonCE. A mathematical theory of communication. Bell Syst Tech J1948; 27(3): 379–423.
45.
WangLYeYLuX, et al.Prussian blue nanocubes on nitrobenzene-functionalized reduced graphene oxide and its application for H2O2 biosensing. Electrochim Acta2013; 114: 223–232.
46.
AbrahamsonJTSongCHuJH, et al.Synthesis and energy release of nitrobenzene-functionalized single-walled carbon nanotubes. Chem Mater2011; 23(20): 4557–4562.
47.
PengCZengFYuanB, et al.An MD simulation study to the indentation size effect of polystyrene and polyethylene with various indenter shapes and loading rates. Appl Surf Sci2019; 492: 579–590.
48.
MyshkinNKKovalevAV. Adhesion and friction of polymers. In: SinhaSKBriscoeBJ (eds). Polymer Tribology. Imperial College Press, 2009, pp. 3–37.
49.
AcarÇGŽundaA. Tribological aspects of graphene and its derivatives. Lubricants2025; 13(6): 232.
VitralE. Stretch formulations and the poynting effect in nonlinear elasticity. Int J Non Lin Mech2023; 148: 104293.
52.
HorganCOMurphyJG. Poynting effects in soft elastic materials: a review of recent results. J Elast2025; 157: 35.
53.
ZhaoXKongCYangY, et al.Reversed-torsion-type crush energy absorption structure and its inexpensive partial-heating torsion manufacturing method based on origami engineering. J Manuf Sci Eng2022; 144(6): 061001.
54.
MortazaviBSilaniMPodryabinkinEV, et al.First-principles multiscale modeling of mechanical properties in graphene/borophene heterostructures empowered by machine-learning interatomic potentials. Adv Mater2021; 33(35): 2102807.
55.
SamaniegoEAnitescuCGoswamiS, et al.An energy approach to the solution of partial differential equations in computational mechanics via machine learning: concepts, implementation and applications. Comput Methods Appl Mech Eng2020; 362: 112790.
56.
EshaghiMSAnitescuCThombreM, et al.Variational physics-informed neural operator (VINO) for solving partial differential equations. Comput Methods Appl Mech Eng2025; 437: 117785.
