Influence of material configurations in the design of filament-wound composite pressure vessels with combined load investigated using deep learning evolutionary algorithms
Restricted accessResearch articleFirst published online 2026
Influence of material configurations in the design of filament-wound composite pressure vessels with combined load investigated using deep learning evolutionary algorithms
This study focuses on the analysis and optimization of filament-wound composite pressure vessels loaded with a combination of mechanical and thermal loading, which occurs during the curing process and subsequent cooling to the ambient temperature. The state of stress in the pressure vessel is determined using the analytical description based on the classical lamination theory supplemented by netting theory. The analytical description is used for generating robust input data for the data-driven deep learning evolutionary algorithm EvoDN2, to carry out machine learning followed by many-objective optimization, to compute the optimal tradeoff between the nominal pressure and weight of the pressure vessel, maintaining a high level of safety. The influence of material configurations on the design of pressure vessels is also studied, where four material configurations of E-glass/DA 4518U, T300/N5208, AS4/3501-6, and Kevlar 49/CYCOM 919 are considered. The results indicate that the usage of aramid fiber in the design has its limits, whereas the materials based on carbon-epoxy show an impressive performance and great strength-to-weight ratio.
AzeemMYaHHAlamMA, et al.Application of filament winding technology in composite pressure vessels and challenges: a review. J Energy Storage2022; 49: 103468.
2.
KhanSKumarA. Failure analysis in advance cylindrical composite pressure vessel under pressure & temperature for hydrogen storage: a comprehensive review. Polym Compos2025; 46: 2933–2973.
3.
PetrieEM. Handbook of adhesives and sealants. 2nd ed. New York: McGraw-Hill, 2007.
4.
ReifsniderKL. Fatigue of composite materials. Blacksburg: Elsevier, 2012.
KangYParkGKwakH, et al.Optimal design of curing process for manufacturing a high pressure hydrogen vessel (type 4). J Mech Sci Technol2023; 37: 3495–3505.
7.
AlamSYandekGRLeeRC, et al.Design and development of a filament wound composite overwrapped pressure vessel. Compos Part C2020; 2: 100045.
8.
RafieeRShahzadiRSperespH. Structural optimization of filament wound composite pipes. Front Struct Civ Eng2022; 16: 1056–1069.
9.
Van BavelBShishkinaOVandepitteD, et al.Reliability-based composite pressure vessel design optimization with cure-induced stresses and spatial material variability. Comput Methods Appl Mech Eng2024; 432: 117463.
10.
WangQQinHJiaL, et al.Failure prediction and optimization for composite pressure vessel combining FEM simulation and machine learning approach. Compos Struct2024; 337: 118099.
11.
VondráčekDPadovecZMarešT, et al.Optimization of dome shape for filament wound pressure vessels using data-driven evolutionary algorithms. Mater Manuf Process2023; 38: 1899–1910.
12.
VondráčekDPadovecZMarešT, et al.Analysis and optimization of junction between cylindrical part and end dome of filament wound pressure vessels using data driven evolutionary algorithms. Proc Inst Mech Eng Part C J Mech Eng Sci2024; 238: 2348–2361.
13.
VondráčekDPadovecZMarešT, et al.Complex design and analysis of filament wound composite pressure vessels using data driven evolutionary algorithms. Proc Inst Mech Eng Part E J Process Mech Eng2024: 09544089241238410.
14.
VondráčekDPadovecZMarešT, et al.Analysis of thermoelastic stresses in filament wound composite pressure vessels using evolutionary deep learning and many-objective optimisation. Philos Mag Lett2025; 105: 2476499.
15.
VondráčekDPadovecZMarešT, et al.Application of data-driven optimized evolutionary learning in the design of filamnet-wound conical pressure vessels. SAMPE J2026; 62: 12–18.
16.
PadovecZ. Analytical and numerical analysis of filament-wound toroidal pressure vessel loaded by internal pressure and temperature field. Acta Polytech2025; 65: 442–453.
17.
KeršysAKeršienėN. Žiliukas A. Experimental research of the impact response of E-glass/epoxy and carbon/epoxy composite systems. Mater Sci2010; 16: 324–329.
18.
IdrisiAHMouradA-HIAbdel-MagidBM, et al.Investigation on the durability of E-glass/epoxy composite exposed to seawater at elevated temperature. Polymers (Basel)2021; 13: 2182.
19.
Saif Aldeen GhafelHNamerNSAliAJ. Mechanical properties of multi-layer woven E-glass/epoxy in variable fiber-mat directions. J Tech2023; 5: 52–60.
20.
UnderwoodJChaitRSmithC, et al.Fracture mechanics: seventeenth volume. West Conshohocken: ASTM International, 1986.
21.
HerakovichCTGregoryMABeuth JrJ. Crack growth direction in unidirectional off-axis graphite epoxy. 1984.
22.
HerakovichCTAboudiJLeeSW, et al.Damage in composite laminates: effects of transverse cracks. Mech Mater1988; 7: 91–107.
AdamsDFAdamsLG. Tensile impact tests of AS4/3501-6 and S2/3501-6 unidirectional composites and the 3501-6 epoxy matrix. J Compos Mater1990; 24: 256–268.
DavidPMarešTChakrabortiN. Evolutionary multi-objective optimization of truss topology for additively manufactured components. Mater Manuf Process2023; 38: 1922–1931.
34.
MaláAPadovecZMarešT, et al.A method for designing filament-wound composite frame structures using a data-driven evolutionary optimisation algorithm EvoDN2. Philos Mag Lett2023; 103: 2272975.
35.
MaláAPadovecZMarešT, et al.The stacking sequence optimisation of a filament wound composite bicycle frame using the data-driven evolutionary algorithm EvoDN2. Philos Mag Lett2024; 104: 2347899.
36.
KropíkBMarešTMaláA, et al.Multi-objective optimization of composite kayak paddle shaft through evolutionary deep learning. Mater Manuf Process2025; 40: 893–901.
37.
MahantaBKKumarSPathakSK, et al.Machine learning-based prediction of high-entropy alloys for hydrogen storage with optimized thermodynamic and kinetic parameters. J Energy Storage2025; 139: 118865.
38.
EbrahimibaghaDArroyo ArmidaSDattaS, et al.Machine learning based models to investigate the thermoelectric performance of carbon nanotube-polyaniline nanocomposites. Comput Mater Sci2024; 232: 112601.
39.
ChakrabortiN. Data-driven evolutionary modeling in materials technology. 1st ed. Boca Raton, London, New York: CRC Press, Taylor & Francis Group, 2023.
40.
ChakrabortiN. The advent of the deep learning evolutionary algorithm EvoDN2 and its recent applications. Comput Methods Mater Sci2025; 25: 41–55.
41.
MothilalMKumarA. Predictive modeling of ultimate tensile strength in dissimilar friction stir welded aluminum alloys via machine learning approach. Philos Mag Lett2025; 105: 2472669.
