International Journal of Engineering Science最新文献

英文中文

Electrokinetic energy conversion efficiency in carbon nanotubes

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-23 DOI: 10.1016/j.ijengsci.2025.104263

Yongbo Liu , Yongjun Jian

The electrokinetic energy conversion (EKEC) of pressure driven flow in carbon nanotubes (CNTs) is of great interest due to its potential high conversion efficiency. The existing EKEC theories had made many simplified assumptions for this problem, such as the surface charge is fixed on the surface and can not move, the slip length is independent of pipe diameter and the surface charge density is decoupled from the solution concentration. In order to get more accurate conversion efficiency, the prior theoretical models of EKEC in CNTs are revised in this paper by focusing on the combined influence of surface charge mobility and tube diameter on slip length and conversion efficiency. In addition, the surface charge density is no longer viewed as a constant, but a function of solution concentration of the electrolyte solution inside the CNTs. Results show that considering the surface charge mobility will reduce the EKEC efficiency. However, the decrease of tube diameter could enhance the EKEC efficiency. In order to maximize the EKEC efficiency, we give the optimal values of corresponding parameters. The maximum EKEC efficiency obtained in this paper is 18.8 %, which is obtained for the pressure driven flow of a LiCl solution with a concentration of 2 mM through a CNT with a radius of 15 nm.

碳纳米管（CNTs）中的压力驱动流的电动能量转换（EKEC）因其潜在的高转换效率而备受关注。现有的 EKEC 理论对这一问题做了许多简化假设，如表面电荷固定在表面不能移动、滑移长度与管道直径无关、表面电荷密度与溶液浓度脱钩等。为了获得更精确的转换效率，本文对之前的 CNT EKEC 理论模型进行了修正，重点研究了表面电荷移动性和管道直径对滑移长度和转换效率的综合影响。此外，表面电荷密度不再被视为常数，而是 CNT 内部电解质溶液浓度的函数。结果表明，考虑表面电荷迁移率会降低 EKEC 效率。然而，减小管径可以提高 EKEC 效率。为了使 EKEC 效率最大化，我们给出了相应参数的最佳值。本文获得的最大 EKEC 效率为 18.8%，是在压力驱动下浓度为 2 mM 的氯化锂溶液流经半径为 15 nm 的 CNT 时获得的。

{"title":"Electrokinetic energy conversion efficiency in carbon nanotubes","authors":"Yongbo Liu , Yongjun Jian","doi":"10.1016/j.ijengsci.2025.104263","DOIUrl":"10.1016/j.ijengsci.2025.104263","url":null,"abstract":"<div><div>The electrokinetic energy conversion (EKEC) of pressure driven flow in carbon nanotubes (CNTs) is of great interest due to its potential high conversion efficiency. The existing EKEC theories had made many simplified assumptions for this problem, such as the surface charge is fixed on the surface and can not move, the slip length is independent of pipe diameter and the surface charge density is decoupled from the solution concentration. In order to get more accurate conversion efficiency, the prior theoretical models of EKEC in CNTs are revised in this paper by focusing on the combined influence of surface charge mobility and tube diameter on slip length and conversion efficiency. In addition, the surface charge density is no longer viewed as a constant, but a function of solution concentration of the electrolyte solution inside the CNTs. Results show that considering the surface charge mobility will reduce the EKEC efficiency. However, the decrease of tube diameter could enhance the EKEC efficiency. In order to maximize the EKEC efficiency, we give the optimal values of corresponding parameters. The maximum EKEC efficiency obtained in this paper is 18.8 %, which is obtained for the pressure driven flow of a LiCl solution with a concentration of 2 mM through a CNT with a radius of 15 nm.</div></div>","PeriodicalId":14053,"journal":{"name":"International Journal of Engineering Science","volume":"213 ","pages":"Article 104263"},"PeriodicalIF":5.7,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859748","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Utilization of the quadratic Hill yield function to model Ti-6Al-4V within multiplicative finite strain kinematics

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-23 DOI: 10.1016/j.ijengsci.2025.104268

Jakob Platen, Johannes Storm, Michael Kaliske

In the contribution at hand, the quadratic Hill yield function is extended to a finite strain, multiplicative kinematic framework. Furthermore, it is expanded by a novel combination of linear and exponential hardening. The formulation is derived in a consistent manner. Subsequently, the capabilities to model Ti-6Al-4V in a realistic manner are demonstrated, considering the anisotropy in plastic material behavior, which is induced by additive manufacturing. Therefore, different tensile tests from the literature are simulated, and a validation is carried out. Finally, the significance of the anisotropy is demonstrated in a numerical example. The maximal force obtained in the simulation is shown to be significantly different, if the building direction is considered.

引用次数: 0

General shape transformations of thin hyperelastic shells through stress-free differential growth 通过无应力差分增长实现超弹性薄壳的一般形状变换

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-16 DOI: 10.1016/j.ijengsci.2025.104266

Zhanfeng Li , Jiong Wang

To address the needs of engineering applications, researchers often wish that the shapes of samples can be precisely controlled. The current work aims to propose a promising approach, i.e., through stress-free differential growth, to realize general shape transformations of thin hyperelastic shells. First, within the finite-strain regime, we formulate the 3D governing equations system for modeling the growth behavior of hyperelastic shells. To facilitate the derivations, it is assumed that the shell sample attains the stress-free state in its current configuration. Subsequently, through series expansions of the unknown variables, we derive the explicit analytical formulas that elucidate the intricate relationships between growth functions and the geometric quantities of general 3D target surfaces. Based on these analytical formulas, we propose a theoretical framework for controlling the shape changes of the shell sample from the reference configuration to a desired target configuration. Notably, our framework accommodates a wide array of geometric mappings, including topology transformation, conformal mapping, and isometry mapping. To promote applications of the theoretical framework, a numerical scheme is further proposed to achieve shape transformations of shell samples between complex surfaces without explicit parametric equations. Both the theoretical framework and the numerical scheme are validated through 3D finite element simulations. The results of the current work can be applied for the design of novel intelligent soft devices, which also reveal the connections between solid mechanics and differential geometry.

{"title":"General shape transformations of thin hyperelastic shells through stress-free differential growth","authors":"Zhanfeng Li , Jiong Wang","doi":"10.1016/j.ijengsci.2025.104266","DOIUrl":"10.1016/j.ijengsci.2025.104266","url":null,"abstract":"<div><div>To address the needs of engineering applications, researchers often wish that the shapes of samples can be precisely controlled. The current work aims to propose a promising approach, i.e., through stress-free differential growth, to realize general shape transformations of thin hyperelastic shells. First, within the finite-strain regime, we formulate the 3D governing equations system for modeling the growth behavior of hyperelastic shells. To facilitate the derivations, it is assumed that the shell sample attains the stress-free state in its current configuration. Subsequently, through series expansions of the unknown variables, we derive the explicit analytical formulas that elucidate the intricate relationships between growth functions and the geometric quantities of general 3D target surfaces. Based on these analytical formulas, we propose a theoretical framework for controlling the shape changes of the shell sample from the reference configuration to a desired target configuration. Notably, our framework accommodates a wide array of geometric mappings, including topology transformation, conformal mapping, and isometry mapping. To promote applications of the theoretical framework, a numerical scheme is further proposed to achieve shape transformations of shell samples between complex surfaces without explicit parametric equations. Both the theoretical framework and the numerical scheme are validated through 3D finite element simulations. The results of the current work can be applied for the design of novel intelligent soft devices, which also reveal the connections between solid mechanics and differential geometry.</div></div>","PeriodicalId":14053,"journal":{"name":"International Journal of Engineering Science","volume":"213 ","pages":"Article 104266"},"PeriodicalIF":5.7,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143835175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Design and control of multi-branch and multi-segment-based dielectric elastomer actuator and biomimetic applications

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-12 DOI: 10.1016/j.ijengsci.2025.104271

Haihao Ci, Zhan-Sheng Guo

Dielectric elastomer actuators (DEAs) based on the dielectric elastomer minimum energy structure (DEMES) exhibit excellent dynamic shape deformation and fast response characteristics, making them widely applicable in flexible actuators, smart grippers, and biomimetic devices. While existing studies have explored the fundamental mechanisms of DEMES, most focus on simple structural types or single application scenarios. However, a systematic approach for designing, modeling, and computing versatile, complex DEAs remains underexplored. In this work, the nonlinear motion equations for an electroelastic cantilever plate are derived based on Hamilton's principle to capture the nonlinear behavior of DEAs. Multi-structured single-segment DEAs are designed to investigate the influence of the driven shape and geometric configuration of the frame on actuator performance. Additionally, complex deformations and grasping strategies for four-branch single-segment and six-segment DEAs are analyzed in detail, considering the actuation of individual branches and parts of each segment. An octopus-inspired variable-stiffness DEA is designed to grasp polygonal objects of varying sizes and shapes using diverse voltage control strategies. Different hand gestures are mimicked by independently controlling each segment and finger of a human hand-based DEA. The accuracy of the three-segment and six-segment DEMES is validated by comparison with rotary joint experiments and a microsatellite gripper, respectively. Additionally, a duty cycle voltage strategy that enables the variation of the flapping angle of a rotary joint to closely match real flight conditions is proposed. The results confirm the independence and controllability of multi-branch and multi-segment DEAs in complex tasks. This research offers new insights and methodologies to advance the field of flexible actuators, highlighting the vast application potential of DEAs in multi-task and multi-objective operations.

{"title":"Design and control of multi-branch and multi-segment-based dielectric elastomer actuator and biomimetic applications","authors":"Haihao Ci, Zhan-Sheng Guo","doi":"10.1016/j.ijengsci.2025.104271","DOIUrl":"10.1016/j.ijengsci.2025.104271","url":null,"abstract":"<div><div>Dielectric elastomer actuators (DEAs) based on the dielectric elastomer minimum energy structure (DEMES) exhibit excellent dynamic shape deformation and fast response characteristics, making them widely applicable in flexible actuators, smart grippers, and biomimetic devices. While existing studies have explored the fundamental mechanisms of DEMES, most focus on simple structural types or single application scenarios. However, a systematic approach for designing, modeling, and computing versatile, complex DEAs remains underexplored. In this work, the nonlinear motion equations for an electroelastic cantilever plate are derived based on Hamilton's principle to capture the nonlinear behavior of DEAs. Multi-structured single-segment DEAs are designed to investigate the influence of the driven shape and geometric configuration of the frame on actuator performance. Additionally, complex deformations and grasping strategies for four-branch single-segment and six-segment DEAs are analyzed in detail, considering the actuation of individual branches and parts of each segment. An octopus-inspired variable-stiffness DEA is designed to grasp polygonal objects of varying sizes and shapes using diverse voltage control strategies. Different hand gestures are mimicked by independently controlling each segment and finger of a human hand-based DEA. The accuracy of the three-segment and six-segment DEMES is validated by comparison with rotary joint experiments and a microsatellite gripper, respectively. Additionally, a duty cycle voltage strategy that enables the variation of the flapping angle of a rotary joint to closely match real flight conditions is proposed. The results confirm the independence and controllability of multi-branch and multi-segment DEAs in complex tasks. This research offers new insights and methodologies to advance the field of flexible actuators, highlighting the vast application potential of DEAs in multi-task and multi-objective operations.</div></div>","PeriodicalId":14053,"journal":{"name":"International Journal of Engineering Science","volume":"212 ","pages":"Article 104271"},"PeriodicalIF":5.7,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143820575","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A large-deformation investigation into the electromechanically coupled sensing performances of flexible nanoparticle-reinforced composite stretch sensors

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-12 DOI: 10.1016/j.ijengsci.2025.104265

Xiaodong Xia , Yu Su , Chuang Feng , George J. Weng

In contrast to the conventional strain sensors under the small-deformation condition, the large-deformation analysis on the flexible nanocomposite-reinforced stretch sensors remains to be investigated. In this research, an extended multi-field coupled homogenization model has been developed to illustrate the nonlinear stretch sensing capacities of flexible nanoparticle-reinforced composite sensors. In this analysis, the stretch-dependent pseudo-moduli and conductivity are chosen as the dual homogenization parameters of current stretch sensing analysis. The predicted resistance change ratio and stretch sensitivity factor are consistent with the experimental data of silver nanoparticle/PDMS nanocomposite sensors over a broad range of stretch loading. The nonlinear stretch sensing performance is attributed to the shape deformation of nanoparticles and significant variation of tunneling distance. The uncovered stretch sensing capacities can provide the directions to optimize flexible nanoparticle-reinforced composite sensors in the area of electronic skin.

{"title":"A large-deformation investigation into the electromechanically coupled sensing performances of flexible nanoparticle-reinforced composite stretch sensors","authors":"Xiaodong Xia , Yu Su , Chuang Feng , George J. Weng","doi":"10.1016/j.ijengsci.2025.104265","DOIUrl":"10.1016/j.ijengsci.2025.104265","url":null,"abstract":"<div><div>In contrast to the conventional strain sensors under the small-deformation condition, the large-deformation analysis on the flexible nanocomposite-reinforced stretch sensors remains to be investigated. In this research, an extended multi-field coupled homogenization model has been developed to illustrate the nonlinear stretch sensing capacities of flexible nanoparticle-reinforced composite sensors. In this analysis, the stretch-dependent pseudo-moduli and conductivity are chosen as the dual homogenization parameters of current stretch sensing analysis. The predicted resistance change ratio and stretch sensitivity factor are consistent with the experimental data of silver nanoparticle/PDMS nanocomposite sensors over a broad range of stretch loading. The nonlinear stretch sensing performance is attributed to the shape deformation of nanoparticles and significant variation of tunneling distance. The uncovered stretch sensing capacities can provide the directions to optimize flexible nanoparticle-reinforced composite sensors in the area of electronic skin.</div></div>","PeriodicalId":14053,"journal":{"name":"International Journal of Engineering Science","volume":"212 ","pages":"Article 104265"},"PeriodicalIF":5.7,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143820576","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

On the extension of the concept of rheological connections to a finite deformation framework using multiple natural configurations

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-12 DOI: 10.1016/j.ijengsci.2025.104267

Tarun Singh, Sandipan Paul

The constitutive behaviors of materials are often modeled using a network of different rheological elements. These rheological models are mostly developed within a one-dimensional small strain framework. One of the key impediments of extending these models to a three-dimensional finite deformation setting is to determine how the different types of connections, i.e., a series and a parallel connection, are incorporated into the material models. The primary objective of this article is to develop an appropriate strategy to address this issue. We show that both the series and the parallel connection between two rheological elements can be modeled within a multiple natural configurations framework without changing or introducing new configurations. The difference in a series and a parallel connection is manifested in the ratio of the stress powers expended during the deformations of the associated rheological elements. Finite deformation version of some well-known rheological models have been used to demonstrate the utility of the proposed theory.

引用次数: 0

Nonlinear theory and finite element analysis of cylindrical deformation of hyperelastic thick rods

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-10 DOI: 10.1016/j.ijengsci.2025.104269

J. Chróścielewski, A. Sabik, W. Witkowski

This paper presents a theory and finite element formulation for the plane strain problem of hyperelastic bodies undergoing cylindrical bending using one-dimensional C⁰ elements called ROD. The paper addresses several aspects characteristic of thick shells experiencing finite elastic deformations, which are not typically encountered in the analysis of classical thin shells. These include significant deformations through the shell's thickness in highly nonlinear range, the numerical implementation of various nonlinear material behaviors, the impact of the chosen reference surface (load) location on the solutions, and the treatment of related boundary conditions. The obtained results are compared with those obtained by commercial code finite element analysis Abaqus system to show the effectiveness of the formulation.

引用次数: 0

Dynamic behaviour of state-based peridynamic media through analysis of potential fields

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-09 DOI: 10.1016/j.ijengsci.2025.104261

Subrata Mondal, Anasuyakumari Maram, Sudarshan Dhua

Unlike classical theories, which rely on local interactions and differential equations, peridynamic theory employs integro-differential equations to describe the mechanics of materials and structures. This distinctive approach allows peridynamics to naturally incorporate long-range forces and discontinuities, such as cracks, which are challenging to handle using classical partial differential equations. In this study, a novel approach for the implementation of nonlocal Helmholtz-Hodge decomposition is used to decompose the displacement field into components that are divergence-free and curl-free. State-based peridynamics, which was introduced to overcome the limitations of bond-based peridynamics, has been considered in this work. As a consequence, two governing equations involving integrals are derived, which are associated with potential fields. An analysis of the propagation of waves has been performed to determine the dispersion relation for both longitudinal and transverse waves. The general solutions for initial-value problems are derived, and the closed-form expression for Green’s function in terms of potential fields is obtained. This study considers Gaussian, exponential, and constant functions as nonlocality functions. Graphs are demonstrated to illustrate the variations in frequency, phase velocity with changes in the size of the horizon (which represents nonlocal length) and nonlocality functions. The validation is achieved both analytically and numerically by ensuring classical correspondence in the limit of the nonlocality parameters approach zero. This work implements nonlocal Helmholtz decomposition, a robust framework that simplifies the study and provides comprehensive insight while analysing vector fields.

{"title":"Dynamic behaviour of state-based peridynamic media through analysis of potential fields","authors":"Subrata Mondal, Anasuyakumari Maram, Sudarshan Dhua","doi":"10.1016/j.ijengsci.2025.104261","DOIUrl":"10.1016/j.ijengsci.2025.104261","url":null,"abstract":"<div><div>Unlike classical theories, which rely on local interactions and differential equations, peridynamic theory employs integro-differential equations to describe the mechanics of materials and structures. This distinctive approach allows peridynamics to naturally incorporate long-range forces and discontinuities, such as cracks, which are challenging to handle using classical partial differential equations. In this study, a novel approach for the implementation of nonlocal Helmholtz-Hodge decomposition is used to decompose the displacement field into components that are divergence-free and curl-free. State-based peridynamics, which was introduced to overcome the limitations of bond-based peridynamics, has been considered in this work. As a consequence, two governing equations involving integrals are derived, which are associated with potential fields. An analysis of the propagation of waves has been performed to determine the dispersion relation for both longitudinal and transverse waves. The general solutions for initial-value problems are derived, and the closed-form expression for Green’s function in terms of potential fields is obtained. This study considers Gaussian, exponential, and constant functions as nonlocality functions. Graphs are demonstrated to illustrate the variations in frequency, phase velocity with changes in the size of the horizon (which represents nonlocal length) and nonlocality functions. The validation is achieved both analytically and numerically by ensuring classical correspondence in the limit of the nonlocality parameters approach zero. This work implements nonlocal Helmholtz decomposition, a robust framework that simplifies the study and provides comprehensive insight while analysing vector fields.</div></div>","PeriodicalId":14053,"journal":{"name":"International Journal of Engineering Science","volume":"212 ","pages":"Article 104261"},"PeriodicalIF":5.7,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143799827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Similarity solutions and wave interactions in a rarefied polyatomic gas

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-03 DOI: 10.1016/j.ijengsci.2025.104262

Dia Zeidan , Pabitra Kumar Pradhan , Manoj Pandey

The present work describes the development of symmetry analysis for a rarefied polyatomic gas with a focus on the Extended Thermodynamics six independent fields (ET6) model. Rarefied polyatomic gas phenomena in Extended Thermodynamics derive mechanisms for several engineering fields, such as aerospace and mechanical engineering. In this work, symmetry analysis is developed by utilizing the one-dimensional optimal system of the corresponding ET6 model to establish the inequivalence class of group invariant solutions. The model deploys the mass density, the velocity, the absolute temperature, and the dynamics pressure to explicitly derive the transport equations and investigate the evolution of characteristic shock and discontinuity waves. A thorough investigation of the interaction between shock and weak discontinuity waves is presented. It enables complete and specific expressions for the amplitudes of transmitted and reflected waves.

引用次数: 0

Comparison of semi-empirical models, symbolic regression, and machine learning approaches for prediction of tensile strength in steels

IF 5.7 1区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

International Journal of Engineering Science

Pub Date : 2025-04-01 DOI: 10.1016/j.ijengsci.2025.104247

Gerfried Millner , Gabriel Kronberger , Manfred Mücke , Lorenz Romaner , Daniel Scheiber

We employ data-driven models to predict the tensile strength of steel coils using information on their chemical composition and process parameters. The dataset contains extensive chemical analyses, diverse process parameters, and the characterized tensile strength as target property. We compare prediction quality and traceability of the predictions of pure machine learning models and physics-informed models. To introduce physical knowledge, we combine models from literature knowledge with symbolic regression and compare the physics-inspired models to machine learning models. In contrast to classic black-box models, symbolic regression provides mathematical equations for the estimation of the target value, facilitating straightforward interpretation. To analyze the predictions from classic black-box machine learning models, we use feature importance analysis with SHAP and contrast the obtained feature impacts with physics-based model parameters. We find that for the present use case, Artificial Neural Networks perform best, while the physics-infused models from symbolic regression allow for better interpretability.

{"title":"Comparison of semi-empirical models, symbolic regression, and machine learning approaches for prediction of tensile strength in steels","authors":"Gerfried Millner , Gabriel Kronberger , Manfred Mücke , Lorenz Romaner , Daniel Scheiber","doi":"10.1016/j.ijengsci.2025.104247","DOIUrl":"10.1016/j.ijengsci.2025.104247","url":null,"abstract":"<div><div>We employ data-driven models to predict the tensile strength of steel coils using information on their chemical composition and process parameters. The dataset contains extensive chemical analyses, diverse process parameters, and the characterized tensile strength as target property. We compare prediction quality and traceability of the predictions of pure machine learning models and physics-informed models. To introduce physical knowledge, we combine models from literature knowledge with symbolic regression and compare the physics-inspired models to machine learning models. In contrast to classic black-box models, symbolic regression provides mathematical equations for the estimation of the target value, facilitating straightforward interpretation. To analyze the predictions from classic black-box machine learning models, we use feature importance analysis with SHAP and contrast the obtained feature impacts with physics-based model parameters. We find that for the present use case, Artificial Neural Networks perform best, while the physics-infused models from symbolic regression allow for better interpretability.</div></div>","PeriodicalId":14053,"journal":{"name":"International Journal of Engineering Science","volume":"212 ","pages":"Article 104247"},"PeriodicalIF":5.7,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143737960","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

下一页尾页

类型

全部化学•材料生命科学医学物理工程技术环境•农林材料科学地球科学法学管理学化学环境科学与生态学计算机科学教育学经济学农林科学人文科学生物学数学物理与天体物理心理学综合性期刊其他工业工程理学历史学农学文学信息工程

数据库

全部 ACS Publications Elsevier ieeexplore Springer The Royal Society of Chemistry Wiley

期刊

International Journal of Engineering Science

全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.

﹀