Comparative assessment of supervised machine learning algorithms for predicting geometric characteristics of laser cladded inconel 718

Aug 4, 2026

Upload by Oliver

Laser cladding is a surface modification and coating process where laser–powder–substrate interactions create a complex mapping from process parameters to clad layer quality. This study uses fast-evolving supervised machine learning models to learn that mapping for Inconel 718 cladding on an A286 substrate with a full factorial design covering 64 experimental groups. Variance analysis and contour/surface plots assess how laser power, powder feed rate, and scanning speed affect width, height, and dilution rate. Model performance is compared using index of merit based on MSE, MAE, and R².

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages19

Reading Time48 min

AccessFree to download

Abstract

Introduction

Laser cladding and process-parameter effects

Complexity of mapping between parameters and clad quality

Materials Research Express PAPER • OPEN ACCESS Comparative assessment of supervised machine learning algorithms for predicting geometric characteristics of laser cladded inconel 718 To cite this article: Hao Yang et al 2024 Mater. Res. Express 11 046516 View the article online for updates and enhancements. You may also like -Out-of-equilibrium gene expression fluctuations in the presence of extrinsic noise Marta Biondo, Abhyudai Singh, Michele Caselle et al. -Microstructure and Mechanical Properties of Cermet Composite Coating on TC4 Surface Lijuan Zheng, Yinkai Xie, Kuo Zhang et al. -Study on microstructure and mechanical properties of Ni60 + WC/Ni35/AISI1040 functional surface gradient structure of remanufacturing chute plate for the mining scraper by a low cost high power CO2 laser cladding technique J Luo, J J Gao, S W Gou et al. This content was downloaded from IP address [155.185.75.56](155.185.75.56) on 12/11/2024 at 12:24 Mater. Res. Express11(2024)046516 [https:](https://doi.org/10.1088/2053-1591/ad4006)[//](https://doi.org/10.1088/2053-1591/ad4006)[doi.org](https://doi.org/10.1088/2053-1591/ad4006)[/](https://doi.org/10.1088/2053-1591/ad4006)[10.1088](https://doi.org/10.1088/2053-1591/ad4006)[/](https://doi.org/10.1088/2053-1591/ad4006)[2053-1591](https://doi.org/10.1088/2053-1591/ad4006)[/](https://doi.org/10.1088/2053-1591/ad4006)[ad4006](https://doi.org/10.1088/2053-1591/ad4006) PAPER Comparative assessment of supervised machine learning algorithms OPENACCESS for predicting geometric characteristics of laser cladded inconel718 RECEIVED 24November2023 Hao Yang1, HeranGeng1, Marco Alfano2,3 and Junfeng Yuan 1,2 R8EVAISpDil2024 1 School ofMechatronic Engineering, China University ofMining and Technology, Xuzhou221116, People’s Republic ofChina 2 Department of Mechanical and Mechatronics Engineering, University of Waterloo, 200 University Avenue West, Waterloo, ON N2L ACCEPTED FOR PUBLICATION 3G1, Canada 17April2024 3 Dipartimento di Scienze e Metodi dell’Ingegneria, Università di Modena e Reggio Emilia, Via Amendola 2, Padiglione Morselli, 42122 PUBLISHED Reggio Emilia, Italy 29April2024 [E-mail:yuanjf@cumt.edu.cn andyuanjfacademia@outlook.com](E-mail:yuanjf@cumt.edu.cn andyuanjfacademia@outlook.com) Original content from this Keywords: laser cladding, IN718, machine learning, performance prediction, processing variables work maybe used under the terms ofthe Creative CommonsAttribution4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Abstract Laser cladding, an innovative surface modiﬁcation and coating preparation process, has emerged asa research hotspot in material surface modiﬁcation and green remanufacturing domains. In the laser cladding process, the interaction between laser light, powder particles, and the substrate results ina complicated mapping connection between process parameters and clad layer quality. This work aims to shed light on this mapping using fast evolving machine learning algorithms. A full factorial experimental design was employed tocladInconel718powder on anA286substrate comprising64 groups. Analysis ofvariance, contour plots, and surface plots were used to explore the effects oflaser power, powder feeding rate, and scanning speed on the width, height, and dilution rate ofthe cladding. The performance ofthe predictive models was evaluated using the indexofmerit (IM), which includes mean square error (MSE), mean absolute error (MAE), and coefﬁcient ofdetermination (R2) . By comparing the performance ofthe models, it was found that the Extra Trees, Random forest regression, Decision tree regression, and XGBoost algorithms exhibited the highest predictive accuracy. Speciﬁcally, the Extra Trees algorithm outperformed other machine learning models in predicting the cladding width, while theRFR algorithm excelled in predicting the associate

FAQs

What problem does the study address in laser cladding?

It addresses the complex relationship between laser cladding process parameters and the geometric quality of the resulting clad layer.

How were the experimental conditions organized?

A full factorial experimental design was used, forming 64 experimental groups for cladding Inconel 718 powder on an A286 substrate.

Which machine learning algorithms showed the best predictive accuracy?

Extra Trees, Random Forest Regression, Decision Tree Regression, and XGBoost achieved the highest predictive accuracy, with Extra Trees best for width, RFR best for height, and DTR best for dilution rate.

Aug 4, 2026

Upload by Oliver