Advanced structural design of engineering components utilizing an artificial neural network and GNDO algorithm DOI
Ali Rıza Yıldız, Betül Sultan Yıldız

Materials Testing, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 26, 2024

Abstract In today’s competitive environment, the lightweighting of vehicle components is under intense study. While some these studies focus on material modification, a very important part focuses same material. The most widely used techniques in light-weight are topology, topography, size, shape optimization, and metaheuristic algorithms. This work introduces novel hybrid generalized normal distribution optimization (GNDO) simulated annealing algorithm (GNDO-SA) adapted to optimize component made aluminum which aims minimize weight while ensuring that stress constraints met. A combination latin hypercube sampling (LHS) artificial neural network generate mathematical equations governing for objective/constraint optimization. These findings highlight effectiveness superiority GNDO-SA method problems.

Язык: Английский

Reconstruction of residual stresses in additively manufactured Inconel 718 bridge structures using contour method DOI Creative Commons
Fatih Uzun, Alexander M. Korsunsky

The International Journal of Advanced Manufacturing Technology, Год журнала: 2025, Номер unknown

Опубликована: Март 25, 2025

Язык: Английский

Процитировано

0

Improving the strength properties of PLA acetabular liners by optimizing FDM 3D printing: Taguchi approach and finite element analysis validation DOI Creative Commons
Nejmeddine Layeb, Najoua Barhoumi, István Oldal

и другие.

The International Journal of Advanced Manufacturing Technology, Год журнала: 2025, Номер unknown

Опубликована: Март 7, 2025

Язык: Английский

Процитировано

0

Dimensional and geometric deviation modelling for polycarbonate parts fabricated by fused filament fabrication-a machine learning approach DOI
Faheem Faroze, Faheem Faroze, Ajay Batish

и другие.

International Journal on Interactive Design and Manufacturing (IJIDeM), Год журнала: 2025, Номер unknown

Опубликована: Апрель 5, 2025

Язык: Английский

Процитировано

0

Mechanical behavior of composite pipe structures under compressive force and its prediction using different machine learning algorithms DOI
İlyas Bozkurt

Materials Testing, Год журнала: 2024, Номер unknown

Опубликована: Дек. 3, 2024

Abstract Thanks to machine learning algorithms, the performance of composites with high energy absorption capacity can be predicted accuracy rates a small number data. The aim this study is experimentally and numerically determine crushing performances glass/epoxy composite pipe structures under compressive force predict their compression behavior help different algorithms. In study, pipes (peak (PF), peak displacement (PFD), mean (MCF), specific (SEA), total inner (TIE)) were determined for specimen thicknesses, lengths, mesh sizes, numbers integration points, diameters ( D ), directions (axial radial). Additionally, maximum strength values estimated Linear Regression (LR), K-Nearest Neighbors (KNN), Artificial Neural Networks (ANN) data taken from ANN algorithm found more reliable in estimating PF TIE values, an rate 92 %. When determining MCF value, it was that obtained LR than other 80

Язык: Английский

Процитировано

3

Design optimization for inner core of crash box for vehicle based on NPR/PU structure DOI

Guan Zhou,

Jinyu Ren,

Yingxin Hu

и другие.

Mechanics of Advanced Materials and Structures, Год журнала: 2024, Номер unknown, С. 1 - 14

Опубликована: Ноя. 14, 2024

Vehicle crashworthiness is a critical aspect of the passive safety domain in passenger cars, and crash boxes play significant role vehicle collisions. Currently, predominantly utilized vehicles are primarily simple thin-walled structures, which exhibit average energy-absorbing capabilities. To enhance collision safety, this article proposes an inner core filled with negative Poisson's ratio (NPR) structure polyurethane (PU) material to design box. Initially, double-arrow type NPR selected as framework, serving filling material. This combination forms A analysis conducted on three types boxes, examining differences their performance indicators detail demonstrate superiority proposed design. Subsequently, variables that significantly influence evaluation metrics were identified through extreme value difference analysis, these designated parameters for subsequent optimization process. Finally, Neighborhood Cultivation Genetic Algorithm (NCGA) Non-dominated Sorting Algorithm-II (NSGA-II) employed algorithms optimal design, results two determined separately using Normal Boundary Intersection (NBI) method, then compared determine overall solution. The simulation indicate NSGA-II optimized NPR/PU box provides substantial advantages performance. After optimization, exhibits reduced collapse displacement maximum peak force other along enhanced specific energy absorption capacity. These findings designed improves vehicle's event collision. offers valuable theoretical insights support development exploration automotive boxes.

Язык: Английский

Процитировано

1

Enhanced crashworthiness performance of auxetic structures using artificial neural network and geyser inspired algorithm DOI
Betül Sultan Yıldız, Ali Rıza Yıldız,

Cihan Yakupoğlu

и другие.

Materials Testing, Год журнала: 2024, Номер 67(2), С. 353 - 360

Опубликована: Дек. 18, 2024

Abstract This study focuses on the optimum design of an auxetic energy absorber intended for automobile applications. The material chosen this is SCGA27D galvanized steel. research proposes utilization artificial neural network-assisted metaheuristic optimizing structural components. geyser inspired algorithm (GEA), ship rescue algorithm, and mountain gazelle are employed to optimize absorber. objective problem obtain optimal geometry while simultaneously reducing mass meeting absorption constraints. findings demonstrate that both GEA steel exhibit exceptional capabilities in designing vehicle structures.

Язык: Английский

Процитировано

1

Advanced structural design of engineering components utilizing an artificial neural network and GNDO algorithm DOI
Ali Rıza Yıldız, Betül Sultan Yıldız

Materials Testing, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 26, 2024

Abstract In today’s competitive environment, the lightweighting of vehicle components is under intense study. While some these studies focus on material modification, a very important part focuses same material. The most widely used techniques in light-weight are topology, topography, size, shape optimization, and metaheuristic algorithms. This work introduces novel hybrid generalized normal distribution optimization (GNDO) simulated annealing algorithm (GNDO-SA) adapted to optimize component made aluminum which aims minimize weight while ensuring that stress constraints met. A combination latin hypercube sampling (LHS) artificial neural network generate mathematical equations governing for objective/constraint optimization. These findings highlight effectiveness superiority GNDO-SA method problems.

Язык: Английский

Процитировано

0