Abrasive Wear Behavior of Zr-Reinforced LM13 Composites: Experimental and Machine Learning Analysis

Abrasive Wear Behavior of Zr-Reinforced LM13 Composites: Experimental and Machine Learning Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Abrasive Wear Behavior of Zr-Reinforced LM13 Composites: Experimental and Machine Learning Analysis Y P Ravitej, Madhusudhan Manjunath, Prabhakar Kuppahalli, KrantiKumar Kshaurad, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7450766/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study investigates the influence of sliding speed and zirconium (Zr) chill reinforcement on the wear behavior of LM13 alloy composites through experimental evaluation, microstructural analysis, and predictive modeling. Specimens were prepared according to ASTM standards and subjected to abrasive wear testing at varying speeds and reinforcement levels (0–12% Zr). The experimental results revealed that wear rate increased consistently with sliding speed, while the incorporation of Zr significantly reduced material loss, with 9–12% reinforcement demonstrating the highest wear resistance. Microstructural analysis using SEM confirmed the uniform dispersion of zircon particles within the LM13 matrix, enhancing hardness and load-bearing capacity, while delamination features indicated fatigue-induced wear at elevated speeds. XRD and EDS analyses further validated the crystalline structure and elemental composition, confirming the successful integration of Zr reinforcements. To complement experimental findings, machine learning techniques including Linear, Polynomial, Support Vector Regression (SVR), Decision Tree, and Random Forest models were employed to predict wear behavior. All models achieved high accuracy (R² >0.93), with Polynomial regression consistently providing the most reliable predictions, as supported by error and significance analysis. Sensitivity and feature importance studies identified Zr% as the dominant factor influencing wear resistance, while speed remained a critical contributor to wear intensification. Sliding wear behaviour Zirconium (Zr) chill reinforcement LM13 aluminium alloy composite Abrasive wear resistance Regression and machine learning models Tribological performance optimization Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 1. Introduction Wear is one of the most critical factors influencing the durability and performance of engineering materials, particularly in components subjected to sliding and rotating motion such as bearings, gears, and engine parts[ 1 ]. Excessive wear often leads to premature failure, resulting in increased maintenance costs and reduced service life. Among the various operating conditions, sliding speed plays a dominant role, as higher speeds intensify surface interactions, frictional heating, and material degradation, thereby accelerating wear[ 2 ]. To mitigate these challenges, researchers have explored the incorporation of reinforcements into base alloys[ 3 ], with zirconium (Zr) chills receiving significant attention for their ability to refine grain structure[ 4 ], enhance hardness, and improve resistance to material loss. The percentage of Zr reinforcement has been shown to directly influence wear performance[ 5 ], with higher concentrations providing substantial improvements compared to unreinforced alloys[ 6 ]. Despite these advancements, the combined effect of sliding speed and Zr reinforcement[ 7 ], [ 8 ] on wear resistance has not been thoroughly investigated[ 9 ]. Previous studies have primarily focused on either reinforcement addition or operating parameters[ 10 ], [ 11 ], [ 12 ] in isolation, leaving a gap in understanding their interaction and the trade-offs between mechanical strengthening[ 13 ] and tribological performance[ 14 ]. Furthermore, although machine learning approaches have recently been employed in materials science to model and predict complex behaviors, their application in predicting wear trends[ 1 ], [ 15 ], [ 16 ] of Zr-reinforced alloys remains relatively underexplored[ 13 ][ 17 ]. A systematic comparison of regression-based and tree-based predictive models for capturing wear characteristics under varying operating speeds and reinforcement levels is also lacking. The present study addresses these research gaps by integrating experimental analysis with computational modelling to investigate the wear behaviour of Zr-reinforced alloys under different sliding speeds. The objectives of this work are to: (i) experimentally evaluate the influence of sliding speed and varying Zr% chill reinforcement on wear performance; (ii) develop and compare multiple regression and machine learning models, including Linear, Polynomial, SVR, Decision Tree, and Random Forest, for wear prediction; (iii) perform error and significance analysis to evaluate model accuracy; (iv) carry out trend and sensitivity analysis to identify dominant factors affecting wear; and (v) establish predictive frameworks capable of optimizing reinforcement content and operating speed for enhanced tribological performance. By combining experimental validation with advanced predictive modeling, this study provides valuable insights into optimizing material composition and operating parameters, thereby contributing to the development of high-performance wear-resistant composites for demanding industrial applications. 2. Materials and methodology Specimens are prepared according to ASTM-65–81 standards. Silica sand of 200 Microns is used for the experimentation as shown in Fig. 1 , all specimens are cleaned and polished with emery paper of 1000grit size to obtain a good surface finish. A Series of Experiments are conducted by maintaining a constant 200 rpm for 30 minutes. Specimens are weighed before and after the abrasion test. Differences in weight are recorded as weight loss/wear loss. Figure 1 shows the Schematic sketch of the Sand abrasion setup, Sand abrasion set up, and Abrasive wear test specimens after the abrasive wear test respectively. 3. Results and discussions 3.1 Experimental analysis Figure 2 shows the effect of Zr% chill and speed on wear rate. Overall, the wear rate increases with the rise in speed across all compositions of Zr chill, indicating that higher speeds accelerate material wear. However, the addition of Zr chill significantly reduces the wear rate compared to 0% Zr chill. At 200 rpm, the wear rate is highest for 0% Zr and progressively decreases as the Zr% increases to 3%, 6%, 9%, and 12%. This trend remains consistent at higher speeds, where 12% Zr chill consistently exhibits the lowest wear rate. The curves also suggest that the reduction in wear rate due to Zr addition is more pronounced at higher speeds, highlighting the role of Zr chill in improving wear resistance. Thus, both speed and Zr% have a combined effect, with speed increasing wear and Zr% reducing it, demonstrating a clear trade-off between operating conditions and material reinforcement. 3.2 Microstructural analysis Figure 3 shows the scanning electron microscopy (SEM) image of the microstructural interaction between zircon particles and the LM13 alloy matrix. The zircon particles appear as distinct, relatively smooth and spherical reinforcements embedded within the matrix, while the LM13 alloy forms the surrounding continuous phase. The presence of zircon indicates effective particle incorporation during processing, contributing to grain refinement and localized strengthening. The contrast in morphology between the hard zircon reinforcements and the softer LM13 alloy highlights their role in enhancing mechanical and tribological properties. Specifically, zircon particles act as load-bearing phases that resist deformation and abrasive wear, whereas LM13 provides toughness and ductility. The image thus demonstrates the successful dispersion of zircon within LM13, which is essential for improving wear resistance and extending the material’s functional performance under sliding conditions. Figure 4 shows the SEM image highlights the delamination of the composite surface after wear, a common wear mechanism observed during sliding conditions. The layered and fragmented morphology indicates that the surface material has undergone repeated plastic deformation and shear stress, leading to the gradual peeling away of layers. Such delamination typically occurs when the subsurface experiences high cyclic stresses, causing cracks to initiate and propagate parallel to the surface until sections of the material are detached. The rough and uneven regions suggest severe material loss, which contributes to increased wear rate. This phenomenon reduces the load-bearing capacity of the composite and accelerates surface damage. The image thus provides clear evidence of fatigue-induced wear and delamination, emphasizing the importance of reinforcement and optimized operating conditions in resisting such surface failures. 3.3 XRD X-ray diffraction (XRD) pattern illustrates the crystalline phases present in the sample. As shown in Fig. 5 . The sharp and well-defined peaks indicate the presence of crystalline structures, with the most intense diffraction peak appearing at approximately 2θ ≈ 45.25°, corresponding to a high-intensity reflection from the crystal planes. Such a dominant peak is often associated with the primary phase of the alloy matrix confirming its crystalline nature. The additional peaks at lower and higher angles suggest the presence of secondary phases or reinforcement particles, such as zircon, which contribute to the composite’s improved mechanical and tribological properties. The relatively high intensity of the main peak (above 4500 counts) reflects strong crystallinity, while the background scattering at lower angles indicates minor amorphous contributions or microstructural disorder. Overall, the XRD analysis confirms that the sample exhibits a well-crystallized structure with distinct phases, validating the successful incorporation of reinforcements into the alloy matrix. 3.4 Energy Dispersive X-ray Spectroscopy (EDAX) Figure 6 presents the results of an Energy Dispersive X-ray Spectroscopy (EDS) analysis conducted on the composite surface. The left image shows the electron micrograph with the selected region (Spectrum 3), while the right graph illustrates the corresponding elemental composition. The EDS spectrum reveals strong peaks for Aluminium (Al), confirming LM13 alloy as the primary matrix. Distinct peaks of Silicon (Si) and Copper (Cu) are also visible, consistent with the alloy’s composition. Additionally, the presence of Zirconium (Zr) and Oxygen (O) peaks indicates the successful incorporation of zircon particles and possible formation of zirconium oxide phases during processing. The relative intensity of the Al peak compared to other elements highlights the dominance of the base matrix, while the detection of reinforcement elements confirms their uniform distribution within the alloy. Overall, the EDS analysis validates the composite’s multi-elemental constitution, demonstrating effective reinforcement of LM13 with zircon particles that contribute to improved mechanical and tribological behaviour. Machine learning analysis 4.1 Regression analysis Figure 7 shows the relationship between wear rate and speed for different Zr% chill levels (0%, 3%, 6%, 9%, and 12%), along with regression models used to predict the trend. Across all cases, wear rate increases steadily with speed, but the extent of wear is influenced by Zr content, with higher Zr% generally lowering the wear rate. The fitted models include Linear, Polynomial, Support Vector Regression (SVR), Decision Tree, and Random Forest, with R² values indicating their accuracy. The Linear, Polynomial, and SVR models provide smooth predictions closely aligned with the experimental data, while the Decision Tree shows a stepwise pattern but achieves perfect fitting (R² = 1.000) due to overfitting to discrete points. Random Forest also performs well, balancing smoothness and accuracy. Overall, the results indicate that wear behavior follows a near-linear trend with speed, while Zr% significantly reduces wear, and advanced regression models capture the trend with high accuracy. Figure 8 compares the performance of different regression models (Linear, Polynomial, SVR, Decision Tree, and Random Forest) in predicting wear rate across varying Zr% chill levels (0%, 3%, 6%, 9%, and 12%), using R² scores as the metric. Overall, all models achieved high accuracy (R² close to 1.0), indicating strong predictive capability. Polynomial regression and Decision Tree consistently achieved perfect fitting (R² = 1.0) across most Zr% levels, highlighting their ability to capture non-linear and discrete variations effectively. Linear regression, SVR, and Random Forest also performed exceptionally well, with R² values above 0.93 in all cases, showing their robustness. Among these, the Decision Tree model stands out for consistently achieving perfect fit, though it may suffer from overfitting. The results confirm that wear rate trends are highly predictable, and the choice of model only slightly influences accuracy, with Polynomial and Decision Tree being the most precise across all Zr% conditions. 4.2 Error analysis Figure 9 presents the normalized root mean square error (NRMSE) analysis of different machine learning models (Linear, Polynomial, SVR, Decision Tree, and Random Forest) across varying Zr chill percentages (0%, 3%, 6%, 9%, and 12%). At 0% Zr, the polynomial model exhibits the lowest error, indicating a strong fit, whereas Linear and Random Forest show relatively higher errors. For 3% Zr, the errors across models are more balanced, though Polynomial and SVR maintain comparatively lower NRMSE values. At 6% Zr, all models perform almost uniformly with moderate errors, suggesting no clear dominance. For 9% Zr, Linear and Polynomial models achieve the best accuracy, while SVR and Random Forest show higher error levels. At 12% Zr, Polynomial regression again performs better than most others, though errors increase slightly compared to lower Zr levels. Overall, Polynomial regression demonstrates consistent reliability across different Zr percentages, particularly excelling at 0% and 9% Zr, while ensemble methods like Random Forest tend to have higher errors. This indicates that simpler regression-based models may be more suitable for wear rate prediction in this context compared to complex tree-based approaches. 4.3 Tukey significance plot Figure 10 shows the heatmap illustrates the Tukey HSD significance test results for comparing model errors across different machine learning approaches. It highlights the statistical significance of differences in performance between pairs of models. Linear regression shows significant differences when compared with Polynomial regression and sometimes with SVR, but it is statistically not significant against Random Forest, while comparisons with Decision Tree are categorized as misleading. Polynomial regression also shows significant differences with both Linear and Random Forest models, but no significant difference with SVR, again showing misleading results with Decision Trees. SVR generally shows no significant difference with Polynomial and Random Forest, sometimes with Linear, but misleading outcomes with Decision Tree. Random Forest, on the other hand, shows significance against Polynomial but not with Linear or SVR, while Decision Tree is consistently marked as misleading in its comparisons, suggesting its results are unreliable or inconsistent across the dataset. Overall, the heatmap indicates that Decision Tree lacks robustness in performance comparisons, while Polynomial regression and Linear regression often show significant differences, making them more distinguishable in terms of prediction error. 4.4 Trend analysis Figure 11 shows trend analysis of the relationship between wear rate and speed at different percentages of Zr chill reinforcement. Across all compositions, the wear rate increases with speed, indicating that higher rotational speeds accelerate material loss. At 0% Zr, the wear rate is the highest, demonstrating poor wear resistance. As the Zr content increases, the wear rate decreases significantly, with 12% Zr showing the lowest wear rate, implying superior resistance to wear. The polynomial fit curves closely follow the actual data points, confirming the reliability of the observed trends. Among the compositions, 9% and 12% Zr provide the most effective reduction in wear, while lower percentages such as 3% and 6% show moderate improvements. Overall, the graph highlights that increasing Zr content enhances wear resistance, and the beneficial effect becomes more pronounced at higher reinforcement levels, making 9–12% of Zr chill the most effective range for minimizing wear at varying speeds. Figure 12 illustrates the trend analysis of wear rate as a function of Zr chill percentage at different operating speeds. It is evident that for all speeds, the wear rate decreases steadily with increasing Zr content, confirming that Zr reinforcement enhances wear resistance. At 0% Zr, the wear rate is at its highest across all speeds, while at 12% Zr, the wear rate reaches its lowest, indicating the strong protective effect of Zr. Among the speeds, higher rpm values (such as 1000 rpm and 800 rpm) show consistently higher wear rates compared to lower speeds (200 rpm and 400 rpm), emphasizing that speed accelerates wear despite the presence of reinforcement. However, the reduction in wear due to Zr addition is significant at all speeds, with the polynomial fit curves aligning well with the experimental data points. 4.5 Feature importance plot Figure 13 illustrates the feature importance for wear rate prediction, highlighting the relative influence of Zr% (zirconium content) and Speed on the model’s output. Among the two parameters, Zr% plays the most dominant role, contributing nearly 70% importance, while Speed contributes around 30%. This indicates that the zirconium content in the material is the most critical factor in governing wear rate behavior, as it significantly influences the composite’s hardness, microstructure, and resistance to material loss. Figure 14 illustrates the SHAP plot of the influence of Speed on wear rate prediction, with the color gradient representing the effect of Zr%. At lower speeds (around 200–400 rpm), the SHAP values are negative, indicating a reduction in predicted wear contribution. As the speed increases beyond 600 rpm, the SHAP values gradually shift toward positive, showing that higher speeds tend to increase the wear rate contribution. Additionally, the color distribution suggests that varying Zr% interacts with speed, but the overall impact of zirconium content remains consistent across different speed ranges. 4.6 Sensitivity analysis The sensitivity plots shown in Fig. 15 illustrate the effect of Speed and Zr% reinforcement on the predicted wear rate. The left plot shows that as Speed increases from 200 rpm to 1000 rpm, the predicted wear rate rises almost linearly, indicating a strong positive correlation between speed and material wear. Higher speeds intensify surface interactions and frictional heating, thereby accelerating wear. In contrast, the right plot shows that increasing Zr% reinforcement has the opposite effect: the predicted wear rate decreases significantly, especially beyond 6–8% reinforcement, where a sharp reduction in wear is observed. This demonstrates that Zr% enhances the material’s resistance by improving hardness and structural stability, thus minimizing material loss. Together, these results highlight that speed promotes wear, while zirconium reinforcement mitigates it, underscoring the importance of optimizing both parameters for improved tribological performance. 4.7 Predictive analysis Figure 16 shows the 3D surface plot illustrates the predicted wear rate as a function of cutting speed (rpm) and Zr% chill addition, modeled using a polynomial regression (degree = 2). The surface indicates a gradual increase in wear rate with rising cutting speed, suggesting that higher rpm intensifies material wear. The effect of Zr% chill is comparatively moderate, with wear rates showing slight variation across the 0–12% range, indicating that zirconium addition plays a secondary role compared to speed. The plotted data points align closely with the polynomial-fitted surface, confirming a good model fit and predictive accuracy. Overall, the analysis highlights that while Zr% chill influences wear marginally, cutting speed is the dominant factor affecting wear rate, with higher speeds consistently leading to increased wear. The residual plot for the polynomial regression model (degree = 2) shows the difference between the predicted and actual wear rate values as shown in Fig. 17 . The residuals are scattered randomly around the red dashed line (zero error line), without displaying any systematic pattern or trend. This indicates that the polynomial regression model has captured the underlying relationship effectively and that the errors are mostly random in nature. The residuals are small in magnitude (mostly within ± 0.0008), confirming good predictive accuracy. Overall, the residual plot suggests that the polynomial model provides a reasonably good fit for predicting wear rate. 4. Conclusion The present study investigated the combined effect of sliding speed and zirconium (Zr) chill reinforcement on the wear behavior of LM13-based composites through experimental testing, microstructural characterization, and predictive modeling. The results clearly demonstrated that wear rate increases with sliding speed, highlighting the detrimental influence of higher rpm on surface integrity. However, the incorporation of Zr chills significantly improved wear resistance, with 9–12% reinforcement showing the most effective reduction in material loss. SEM analysis confirmed the uniform distribution of zircon particles within the LM13 matrix, which acted as hard load-bearing phases, while delamination features revealed the fatigue-induced wear mechanisms at higher speeds. XRD and EDS analyses further validated the presence of zircon phases and confirmed their role in strengthening the composite microstructure. Regression and machine learning models, including Linear, Polynomial, SVR, Decision Tree, and Random Forest, achieved high predictive accuracy, with Polynomial regression consistently providing reliable fits and error analysis confirming its robustness. Sensitivity and feature importance studies established that Zr% was the dominant factor governing wear resistance, while speed remained a critical contributor to wear intensification. Overall, this integrated approach demonstrated that optimizing zirconium reinforcement can effectively counteract the adverse effects of high sliding speeds, thereby extending the service life of the material. The findings contribute valuable insights for the design of advanced wear-resistant composites and establish predictive frameworks that can guide industrial applications under varying operating conditions. Declarations Conflict of Interest The authors declare that there is no conflict of interest regarding the publication of this paper. Funding statement: There are no funds received the research on this article. Author Contribution Author Contribution StatementRavitej Y P: Conceptualization, Methodology, Experimental Investigation, Data Curation, Writing – Original Draft, Corresponding Author.Madhusudhan Manjunath: Supervision, Guidance on Methodology, Validation, Critical Review of Manuscript.Prabhakar Kuppahalli: Experimental Design Support, Data Analysis, Resources.KrantiKumar Kshaurad: Machine Learning Modeling, Data Interpretation, Visualization.Bindhushree B S: Microstructural Characterization (SEM, XRD, EDS), Data Curation.Balachandra Halemani: Review and Editing, Technical Input on Wear Analysis.Rajeev Kumar Gupta: Writing – Review & Editing, Validation of Computational Results.Manish Kumar Mishra: Statistical and Error Analysis, Support in Predictive Modeling.Jayatirtha Patil: Validation, Results Interpretation, Support in Trend Analysis.Ramakumar BVN: Contribution to Materials Selection, Tribological Testing.Nithyananda B: Resources, Supervision, Project Administration, Final Review.All authors have read and approved the final version of the manuscript and agree to be accountable for all aspects of the work. Acknowledgement Its a unique article containing experimental, Microstructural and machine learning approach to the abrasive wear problem Data Availability Statement The data supporting the findings of this study are available from the corresponding author, Ravitej Y P, upon reasonable request. Graphs, raw measurements, and model code used for regression analysis can be provided to interested researchers for further exploration or validation. References Deepak V, Abhilash O, Ravitej YP, Veerachari, Abhinandan L (2021) Design and development of progressive tool for mold tag, AIP Conf. Proc. , vol. 2316, no. February. 10.1063/5.0038385 Prakash RY, Venkata B (2025) Experimental, FEA, and machine learning studies on wear behavior of LM13 aluminum hybrid composites reinforced with zircon and graphite. 10.1177/13506501251367549 32 0 2 4, pp. 0–116 Ravitej YP, Keshavamurthy R, Adarsha H, Kumar VV, Bavan DS, Asdaque PMGB (2025) Effect of Zircon on Aluminum/Graphite Alloy Hybrid Composite and Wear Characterization with Load: Experimental and ANN Test. J Bio- Tribo-Corrosion 11(1):1–13. 10.1007/s40735-025-00944-7 Ravitej YP, Chandra BT, Adarsha H, Chandrashekar A, R PC (2024) Analysis of Wear Parameters of Chill Casted LM13 / Zircon / Carbon Hybrid Composites Using Experimental and Statistical Approach. 77–88. 10.52209/2706-977X Sandeep D et al (2021) February,., CFD simulation of transonic turbulent flow past NACA 0012 aerofoil, AIP Conf. Proc. , vol. 2316, no. 10.1063/5.0036890 Ravitej YP, Mohan CB, Anantha Prasad MG (2021) Corrosion analysis of LM13/ZrSiO4 /C Hybrid Metal Matrix Composites, J. Corros. Sci. Eng. , vol. 23, no. November Prem Chand R et al (2021) Characterization of banana and e glass fiber reinforced hybrid epoxy composites, Mater. Today Proc. , vol. 46, no. xxxx, pp. 9119–9125. 10.1016/j.matpr.2021.05.402 Naveen Kumar S et al (2021) Fabrication and characterization of hardness and microstructure of large sized Al2014-SiC composite, Mater. Today Proc. , vol. 46, no. xxxx, pp. 9102–9106. 10.1016/j.matpr.2021.05.398 Chand RP, Ravitej YP, Kumar N, Veeracahri, Gowda NA (2057) Characterization of banana and bagasse fiber reinforced hybrid epoxy composites, AIP Conf. Proc. , vol. no. January, 2019. 10.1063/1.5085624 Prem Chand R, Halemani BS, Chandrasekhar KM, Ravitej YP, Hemanth Raju T, Udayshankar S (2021) Investigation and analysis for mechanical properties of banana and e glass fiber reinforced hybrid epoxy composites, Mater. Today Proc. , vol. 47, no. xxxx, pp. 2509–2515. 10.1016/j.matpr.2021.05.044 Ravitej YP, Swaroop V, Ramesh S, Adarsha H, Veerachari, and, Nischith (2018) Finite element analysis of mild steel - Rubber sandwich composite material. IOP Conf Ser Mater Sci Eng 376(1). 10.1088/1757-899X/376/1/012040 Prakash RY, Boganarasimhaiah MC, Gangadharaiah APM (2021) Fabrication, micro structural and mechanical characterization of Zircon Particles (ZrSiO4) reinforced Aluminum alloy (MMCs). Indian J Eng Mater Sci 28(6):549–553. 10.56042/ijems.v28i6.40173 Maity D, Siddheshwar PG, Saha S (1998) Adv Fluid Mech Turbomach. 10.1007/978-3-642-72157-1 Ravitej YP, Mohan CB, Ananthaprasad MG (2022) Dry Sliding Friction and Wear Behavior of LM13/Zircon/Carbon (HMMC’s): An Experimental, Statistical and Artificial Neural Network Approach. Tribol Ind 44(3):374–393. 10.24874/ti.1223.11.21.03 Ravitej YP, Mohan CB, Ananthaprasad MG (2022) Evaluation of copper chill on tribological behaviour of LM13/ZrSiO 4 /C hybrid metal matrix composites. J Mines Met Fuels 69(12):142. 10.18311/jmmf/2021/30143 Ravitej YP, Mohan CB, Ananthaprasad MG (2021) Effect of Reinforcement and Copper Chill on LM13/ZrSiO4/C Hybrid Metal Matrix Composites (HMMCS)-An Experimental and Statistical Analysis. China’s Refract 30(4):12–18. 10.19691/j.cnki.1004-4493.2021.04.003 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7450766","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":510695685,"identity":"610d2ae2-459a-45ea-8c8c-3db8527c8655","order_by":0,"name":"Y P Ravitej","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA70lEQVRIiWNgGAWjYBAC+wYGhgMMDGwMDOxAFoOBBWEtBkD1Bw6AtPAcAHEliNMCsoaBQSIBTBKh5fgZw8MfKvgSt0s+v7rhR4EEA397dwJeLfY9OQYHDpxhS9w5O6fsZg/QYRJnzm7Aq8WOIS3hwME2tsQNt3PSbvAAtRhI5OLXYsz/DKjlH1DLzTNpN/8Qo8VwRvKBAwcbgFpusB+7TZQtBjceA71yjM14Z08O220ZAwkegn4xOJ/Y/KGi5pjsdvbjz26++WMjx9/ei18LFBxz3MDAYwBi8RCjHARq7A0Y2B8Qq3oUjIJRMApGGAAAsM9UHZ+i3V0AAAAASUVORK5CYII=","orcid":"","institution":"Dayananda Sagar University","correspondingAuthor":true,"prefix":"","firstName":"Y","middleName":"P","lastName":"Ravitej","suffix":""},{"id":510695687,"identity":"df2b9aa9-67b4-4fd4-b3ab-1c5587876bbf","order_by":1,"name":"Madhusudhan Manjunath","email":"","orcid":"","institution":"Dayananda Sagar College of Engineering","correspondingAuthor":false,"prefix":"","firstName":"Madhusudhan","middleName":"","lastName":"Manjunath","suffix":""},{"id":510695689,"identity":"4ca32a6a-d179-41b8-a7a6-abc470cabcf3","order_by":2,"name":"Prabhakar Kuppahalli","email":"","orcid":"","institution":"Dayananda Sagar College of Engineering","correspondingAuthor":false,"prefix":"","firstName":"Prabhakar","middleName":"","lastName":"Kuppahalli","suffix":""},{"id":510695690,"identity":"d636a02e-f995-4d7b-be02-78fbdfe1f312","order_by":3,"name":"KrantiKumar Kshaurad","email":"","orcid":"","institution":"Dayananda Sagar College of Engineering","correspondingAuthor":false,"prefix":"","firstName":"KrantiKumar","middleName":"","lastName":"Kshaurad","suffix":""},{"id":510695691,"identity":"65fc22c8-a864-45a3-911b-f0900bb6385c","order_by":4,"name":"B S Bindhushree","email":"","orcid":"","institution":"Dayananda Sagar College of Engineering","correspondingAuthor":false,"prefix":"","firstName":"B","middleName":"S","lastName":"Bindhushree","suffix":""},{"id":510695692,"identity":"fe4f3d3b-a6ad-47a8-b1da-4a45f94fa670","order_by":5,"name":"Balachandra Halemani","email":"","orcid":"","institution":"KLE technological university","correspondingAuthor":false,"prefix":"","firstName":"Balachandra","middleName":"","lastName":"Halemani","suffix":""},{"id":510695693,"identity":"80694b51-7cd3-43ae-bef1-3c7f833895fb","order_by":6,"name":"Rajeev Kumar Gupta","email":"","orcid":"","institution":"Dayananda Sagar University","correspondingAuthor":false,"prefix":"","firstName":"Rajeev","middleName":"Kumar","lastName":"Gupta","suffix":""},{"id":510695694,"identity":"d922a067-081f-40b9-92fd-e8283f62637c","order_by":7,"name":"Manish Kumar Mishra","email":"","orcid":"","institution":"Dayananda Sagar University","correspondingAuthor":false,"prefix":"","firstName":"Manish","middleName":"Kumar","lastName":"Mishra","suffix":""},{"id":510695695,"identity":"9d165b5a-8e7b-4881-8c24-6d675572d679","order_by":8,"name":"Jayatirtha Patil","email":"","orcid":"","institution":"Jyothy institute of technology","correspondingAuthor":false,"prefix":"","firstName":"Jayatirtha","middleName":"","lastName":"Patil","suffix":""},{"id":510695696,"identity":"cbca7759-63cd-4400-abd8-0b53d534cb4f","order_by":9,"name":"Ramakumar BVN","email":"","orcid":"","institution":"Dayananda Sagar University","correspondingAuthor":false,"prefix":"","firstName":"Ramakumar","middleName":"","lastName":"BVN","suffix":""},{"id":510695697,"identity":"59c2cedc-c01b-453e-a24c-91565ae1ac89","order_by":10,"name":"Nithyananda B","email":"","orcid":"","institution":"Shri Dharmasthala Manjunatheshwara Institute of Technology","correspondingAuthor":false,"prefix":"","firstName":"Nithyananda","middleName":"","lastName":"B","suffix":""}],"badges":[],"createdAt":"2025-08-25 07:23:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7450766/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7450766/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":90942200,"identity":"d1ce8447-b1ec-4a76-be97-b5d8a53f05c4","added_by":"auto","created_at":"2025-09-09 18:41:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":77405,"visible":true,"origin":"","legend":"\u003cp\u003eAbrasion wear test rig\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/c1f39e2ec828e00f58246460.png"},{"id":90942203,"identity":"df94a459-69c0-4b55-b658-b8321d698ad8","added_by":"auto","created_at":"2025-09-09 18:41:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":386249,"visible":true,"origin":"","legend":"\u003cp\u003eZr% chill and speed on wear rate\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/b884540ed19b455a7335034b.png"},{"id":90942399,"identity":"a4b1dec6-a2f3-4c1a-826a-e961e0ad17c5","added_by":"auto","created_at":"2025-09-09 18:49:27","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1424862,"visible":true,"origin":"","legend":"\u003cp\u003eMicrostructural interaction between zircon particles and the LM13 alloy matrix\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/59cd7b43442997aca946d885.png"},{"id":90942207,"identity":"bf0a5a42-1d43-47d5-a851-a158cfb9b359","added_by":"auto","created_at":"2025-09-09 18:41:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2134681,"visible":true,"origin":"","legend":"\u003cp\u003eSEM image highlights the delamination of the composite surface after wear\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/a84978ab16de0d0cf202ac62.png"},{"id":90942201,"identity":"a4d0af99-b467-4cc6-942d-c4f90363c0c5","added_by":"auto","created_at":"2025-09-09 18:41:27","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":6767,"visible":true,"origin":"","legend":"\u003cp\u003eX-ray diffraction (XRD) pattern for 9wt.% of Zircon\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/3b7ac07708d16b256192bc06.png"},{"id":90942400,"identity":"001b01dd-461a-42b4-b5af-7606da5f1178","added_by":"auto","created_at":"2025-09-09 18:49:27","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":388087,"visible":true,"origin":"","legend":"\u003cp\u003eEnergy Dispersive X-ray Spectroscopy (EDS) analysis on 9wt.% of Zircon composite specimen\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/736ca97d43d74febddaebc7e.png"},{"id":90942402,"identity":"b96be858-ee91-4981-9ad5-5674ef85a1c1","added_by":"auto","created_at":"2025-09-09 18:49:27","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":623279,"visible":true,"origin":"","legend":"\u003cp\u003eDifferent regression models of Zircon contents\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/9627eb4bd05fd57b74df90cd.png"},{"id":90943085,"identity":"34fe6d15-b088-4325-8ab9-aa145dc895da","added_by":"auto","created_at":"2025-09-09 19:05:27","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":225620,"visible":true,"origin":"","legend":"\u003cp\u003ePerformance of different regression models\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/182aa1edeac33e4cbd044ac2.png"},{"id":90942409,"identity":"8929ac74-6419-4365-8e50-207fab00e215","added_by":"auto","created_at":"2025-09-09 18:49:28","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":373283,"visible":true,"origin":"","legend":"\u003cp\u003eNormalized root means square error (NRMSE) analysis of different machine learning models\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/c11c557a32f7829c760b0a8c.png"},{"id":90942217,"identity":"309df5f0-9eb4-4d4d-923c-04397089c97b","added_by":"auto","created_at":"2025-09-09 18:41:27","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":366692,"visible":true,"origin":"","legend":"\u003cp\u003eHeatmap illustrates the Tukey HSD significance\u003c/p\u003e","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/faaffde8d892a6ffc1545e6b.png"},{"id":90942222,"identity":"3679cff4-1347-4c5c-9e75-80da1bd4ca19","added_by":"auto","created_at":"2025-09-09 18:41:28","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":474888,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between wear rate and speed at different percentages of Zr chill reinforcement\u003c/p\u003e","description":"","filename":"floatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/9ca295bc25ecaeaf9592d677.png"},{"id":90942224,"identity":"25b29952-7d77-4867-ad20-07dbed863115","added_by":"auto","created_at":"2025-09-09 18:41:28","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":480663,"visible":true,"origin":"","legend":"\u003cp\u003eTrend analysis of wear rate as a function of Zr chill percentage\u003c/p\u003e","description":"","filename":"floatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/dbd2c36a1e5e6859e30bd471.png"},{"id":90942976,"identity":"8689e410-a1d6-4107-8d67-b62f437d714b","added_by":"auto","created_at":"2025-09-09 18:57:28","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":179341,"visible":true,"origin":"","legend":"\u003cp\u003eFeature importance for wear rate prediction\u003c/p\u003e","description":"","filename":"floatimage13.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/2b1365273aa45b1eafe3953c.png"},{"id":90942239,"identity":"d72a541c-f5a4-41d2-8614-f4b800b12aea","added_by":"auto","created_at":"2025-09-09 18:41:28","extension":"png","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":196353,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP plot of Speed on wear rate prediction\u003c/p\u003e","description":"","filename":"floatimage14.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/556e22426e197654fc05b62c.png"},{"id":90942972,"identity":"4a422737-158e-4a8a-a832-3a9227b1700f","added_by":"auto","created_at":"2025-09-09 18:57:28","extension":"png","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":325083,"visible":true,"origin":"","legend":"\u003cp\u003eSensitivity plots for effect of Speed and Zr% reinforcement\u003c/p\u003e","description":"","filename":"floatimage15.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/13fa42db331f65f01810deab.png"},{"id":90942414,"identity":"8181db0b-5505-4bef-b067-6307494090e1","added_by":"auto","created_at":"2025-09-09 18:49:28","extension":"png","order_by":16,"title":"Figure 16","display":"","copyAsset":false,"role":"figure","size":445944,"visible":true,"origin":"","legend":"\u003cp\u003e3D surface plot for the predicted wear rate\u003c/p\u003e","description":"","filename":"floatimage16.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/047bfe417ce23e0bd4f73e7b.png"},{"id":90942232,"identity":"b725bb83-6a54-4620-bbdb-f8bfa2e4b101","added_by":"auto","created_at":"2025-09-09 18:41:28","extension":"png","order_by":17,"title":"Figure 17","display":"","copyAsset":false,"role":"figure","size":318515,"visible":true,"origin":"","legend":"\u003cp\u003eResidual plot for the polynomial regression model\u003c/p\u003e","description":"","filename":"floatimage17.png","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/2fe4e19dfbe74df0de778ce1.png"},{"id":91934618,"identity":"80a9f927-81b5-4055-9fcc-29716550b654","added_by":"auto","created_at":"2025-09-23 02:40:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9054678,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7450766/v1/8aa9c241-ee7f-4187-890a-ebdcf8735bbc.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Abrasive Wear Behavior of Zr-Reinforced LM13 Composites: Experimental and Machine Learning Analysis","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eWear is one of the most critical factors influencing the durability and performance of engineering materials, particularly in components subjected to sliding and rotating motion such as bearings, gears, and engine parts[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Excessive wear often leads to premature failure, resulting in increased maintenance costs and reduced service life. Among the various operating conditions, sliding speed plays a dominant role, as higher speeds intensify surface interactions, frictional heating, and material degradation, thereby accelerating wear[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. To mitigate these challenges, researchers have explored the incorporation of reinforcements into base alloys[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], with zirconium (Zr) chills receiving significant attention for their ability to refine grain structure[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], enhance hardness, and improve resistance to material loss. The percentage of Zr reinforcement has been shown to directly influence wear performance[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e], with higher concentrations providing substantial improvements compared to unreinforced alloys[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eDespite these advancements, the combined effect of sliding speed and Zr reinforcement[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] on wear resistance has not been thoroughly investigated[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Previous studies have primarily focused on either reinforcement addition or operating parameters[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] in isolation, leaving a gap in understanding their interaction and the trade-offs between mechanical strengthening[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] and tribological performance[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Furthermore, although machine learning approaches have recently been employed in materials science to model and predict complex behaviors, their application in predicting wear trends[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] of Zr-reinforced alloys remains relatively underexplored[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e][\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. A systematic comparison of regression-based and tree-based predictive models for capturing wear characteristics under varying operating speeds and reinforcement levels is also lacking. The present study addresses these research gaps by integrating experimental analysis with computational modelling to investigate the wear behaviour of Zr-reinforced alloys under different sliding speeds. The objectives of this work are to: (i) experimentally evaluate the influence of sliding speed and varying Zr% chill reinforcement on wear performance; (ii) develop and compare multiple regression and machine learning models, including Linear, Polynomial, SVR, Decision Tree, and Random Forest, for wear prediction; (iii) perform error and significance analysis to evaluate model accuracy; (iv) carry out trend and sensitivity analysis to identify dominant factors affecting wear; and (v) establish predictive frameworks capable of optimizing reinforcement content and operating speed for enhanced tribological performance. By combining experimental validation with advanced predictive modeling, this study provides valuable insights into optimizing material composition and operating parameters, thereby contributing to the development of high-performance wear-resistant composites for demanding industrial applications.\u003c/p\u003e"},{"header":"2. Materials and methodology","content":"\u003cp\u003eSpecimens are prepared according to ASTM-65\u0026ndash;81 standards. Silica sand of 200 Microns is used for the experimentation as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, all specimens are cleaned and polished with emery paper of 1000grit size to obtain a good surface finish. A Series of Experiments are conducted by maintaining a constant 200 rpm for 30 minutes. Specimens are weighed before and after the abrasion test. Differences in weight are recorded as weight loss/wear loss. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the Schematic sketch of the Sand abrasion setup, Sand abrasion set up, and Abrasive wear test specimens after the abrasive wear test respectively.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"3. Results and discussions","content":"\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Experimental analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the effect of Zr% chill and speed on wear rate. Overall, the wear rate increases with the rise in speed across all compositions of Zr chill, indicating that higher speeds accelerate material wear. However, the addition of Zr chill significantly reduces the wear rate compared to 0% Zr chill. At 200 rpm, the wear rate is highest for 0% Zr and progressively decreases as the Zr% increases to 3%, 6%, 9%, and 12%. This trend remains consistent at higher speeds, where 12% Zr chill consistently exhibits the lowest wear rate. The curves also suggest that the reduction in wear rate due to Zr addition is more pronounced at higher speeds, highlighting the role of Zr chill in improving wear resistance. Thus, both speed and Zr% have a combined effect, with speed increasing wear and Zr% reducing it, demonstrating a clear trade-off between operating conditions and material reinforcement.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Microstructural analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows the scanning electron microscopy (SEM) image of the microstructural interaction between zircon particles and the LM13 alloy matrix. The zircon particles appear as distinct, relatively smooth and spherical reinforcements embedded within the matrix, while the LM13 alloy forms the surrounding continuous phase. The presence of zircon indicates effective particle incorporation during processing, contributing to grain refinement and localized strengthening. The contrast in morphology between the hard zircon reinforcements and the softer LM13 alloy highlights their role in enhancing mechanical and tribological properties. Specifically, zircon particles act as load-bearing phases that resist deformation and abrasive wear, whereas LM13 provides toughness and ductility. The image thus demonstrates the successful dispersion of zircon within LM13, which is essential for improving wear resistance and extending the material\u0026rsquo;s functional performance under sliding conditions.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows the SEM image highlights the delamination of the composite surface after wear, a common wear mechanism observed during sliding conditions. The layered and fragmented morphology indicates that the surface material has undergone repeated plastic deformation and shear stress, leading to the gradual peeling away of layers. Such delamination typically occurs when the subsurface experiences high cyclic stresses, causing cracks to initiate and propagate parallel to the surface until sections of the material are detached. The rough and uneven regions suggest severe material loss, which contributes to increased wear rate. This phenomenon reduces the load-bearing capacity of the composite and accelerates surface damage. The image thus provides clear evidence of fatigue-induced wear and delamination, emphasizing the importance of reinforcement and optimized operating conditions in resisting such surface failures.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e3.3 XRD\u003c/h2\u003e\u003cp\u003eX-ray diffraction (XRD) pattern illustrates the crystalline phases present in the sample. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. The sharp and well-defined peaks indicate the presence of crystalline structures, with the most intense diffraction peak appearing at approximately 2θ\u0026thinsp;\u0026asymp;\u0026thinsp;45.25\u0026deg;, corresponding to a high-intensity reflection from the crystal planes. Such a dominant peak is often associated with the primary phase of the alloy matrix confirming its crystalline nature. The additional peaks at lower and higher angles suggest the presence of secondary phases or reinforcement particles, such as zircon, which contribute to the composite\u0026rsquo;s improved mechanical and tribological properties. The relatively high intensity of the main peak (above 4500 counts) reflects strong crystallinity, while the background scattering at lower angles indicates minor amorphous contributions or microstructural disorder. Overall, the XRD analysis confirms that the sample exhibits a well-crystallized structure with distinct phases, validating the successful incorporation of reinforcements into the alloy matrix.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e\u003cb\u003e3.4 Energy Dispersive X-ray Spectroscopy (EDAX)\u003c/b\u003e\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents the results of an Energy Dispersive X-ray Spectroscopy (EDS) analysis conducted on the composite surface. The left image shows the electron micrograph with the selected region (Spectrum 3), while the right graph illustrates the corresponding elemental composition. The EDS spectrum reveals strong peaks for Aluminium (Al), confirming LM13 alloy as the primary matrix. Distinct peaks of Silicon (Si) and Copper (Cu) are also visible, consistent with the alloy\u0026rsquo;s composition. Additionally, the presence of Zirconium (Zr) and Oxygen (O) peaks indicates the successful incorporation of zircon particles and possible formation of zirconium oxide phases during processing. The relative intensity of the Al peak compared to other elements highlights the dominance of the base matrix, while the detection of reinforcement elements confirms their uniform distribution within the alloy. Overall, the EDS analysis validates the composite\u0026rsquo;s multi-elemental constitution, demonstrating effective reinforcement of LM13 with zircon particles that contribute to improved mechanical and tribological behaviour.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eMachine learning analysis\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Regression analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e shows the relationship between wear rate and speed for different Zr% chill levels (0%, 3%, 6%, 9%, and 12%), along with regression models used to predict the trend. Across all cases, wear rate increases steadily with speed, but the extent of wear is influenced by Zr content, with higher Zr% generally lowering the wear rate. The fitted models include Linear, Polynomial, Support Vector Regression (SVR), Decision Tree, and Random Forest, with R\u0026sup2; values indicating their accuracy. The Linear, Polynomial, and SVR models provide smooth predictions closely aligned with the experimental data, while the Decision Tree shows a stepwise pattern but achieves perfect fitting (R\u0026sup2; = 1.000) due to overfitting to discrete points. Random Forest also performs well, balancing smoothness and accuracy. Overall, the results indicate that wear behavior follows a near-linear trend with speed, while Zr% significantly reduces wear, and advanced regression models capture the trend with high accuracy.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e compares the performance of different regression models (Linear, Polynomial, SVR, Decision Tree, and Random Forest) in predicting wear rate across varying Zr% chill levels (0%, 3%, 6%, 9%, and 12%), using R\u0026sup2; scores as the metric. Overall, all models achieved high accuracy (R\u0026sup2; close to 1.0), indicating strong predictive capability. Polynomial regression and Decision Tree consistently achieved perfect fitting (R\u0026sup2; = 1.0) across most Zr% levels, highlighting their ability to capture non-linear and discrete variations effectively. Linear regression, SVR, and Random Forest also performed exceptionally well, with R\u0026sup2; values above 0.93 in all cases, showing their robustness. Among these, the Decision Tree model stands out for consistently achieving perfect fit, though it may suffer from overfitting. The results confirm that wear rate trends are highly predictable, and the choice of model only slightly influences accuracy, with Polynomial and Decision Tree being the most precise across all Zr% conditions.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Error analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e presents the normalized root mean square error (NRMSE) analysis of different machine learning models (Linear, Polynomial, SVR, Decision Tree, and Random Forest) across varying Zr chill percentages (0%, 3%, 6%, 9%, and 12%). At 0% Zr, the polynomial model exhibits the lowest error, indicating a strong fit, whereas Linear and Random Forest show relatively higher errors. For 3% Zr, the errors across models are more balanced, though Polynomial and SVR maintain comparatively lower NRMSE values. At 6% Zr, all models perform almost uniformly with moderate errors, suggesting no clear dominance. For 9% Zr, Linear and Polynomial models achieve the best accuracy, while SVR and Random Forest show higher error levels. At 12% Zr, Polynomial regression again performs better than most others, though errors increase slightly compared to lower Zr levels. Overall, Polynomial regression demonstrates consistent reliability across different Zr percentages, particularly excelling at 0% and 9% Zr, while ensemble methods like Random Forest tend to have higher errors. This indicates that simpler regression-based models may be more suitable for wear rate prediction in this context compared to complex tree-based approaches.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Tukey significance plot\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e shows the heatmap illustrates the Tukey HSD significance test results for comparing model errors across different machine learning approaches. It highlights the statistical significance of differences in performance between pairs of models. Linear regression shows significant differences when compared with Polynomial regression and sometimes with SVR, but it is statistically not significant against Random Forest, while comparisons with Decision Tree are categorized as misleading. Polynomial regression also shows significant differences with both Linear and Random Forest models, but no significant difference with SVR, again showing misleading results with Decision Trees. SVR generally shows no significant difference with Polynomial and Random Forest, sometimes with Linear, but misleading outcomes with Decision Tree. Random Forest, on the other hand, shows significance against Polynomial but not with Linear or SVR, while Decision Tree is consistently marked as misleading in its comparisons, suggesting its results are unreliable or inconsistent across the dataset. Overall, the heatmap indicates that Decision Tree lacks robustness in performance comparisons, while Polynomial regression and Linear regression often show significant differences, making them more distinguishable in terms of prediction error.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e4.4 Trend analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e shows trend analysis of the relationship between wear rate and speed at different percentages of Zr chill reinforcement. Across all compositions, the wear rate increases with speed, indicating that higher rotational speeds accelerate material loss. At 0% Zr, the wear rate is the highest, demonstrating poor wear resistance. As the Zr content increases, the wear rate decreases significantly, with 12% Zr showing the lowest wear rate, implying superior resistance to wear. The polynomial fit curves closely follow the actual data points, confirming the reliability of the observed trends. Among the compositions, 9% and 12% Zr provide the most effective reduction in wear, while lower percentages such as 3% and 6% show moderate improvements. Overall, the graph highlights that increasing Zr content enhances wear resistance, and the beneficial effect becomes more pronounced at higher reinforcement levels, making 9\u0026ndash;12% of Zr chill the most effective range for minimizing wear at varying speeds.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e illustrates the trend analysis of wear rate as a function of Zr chill percentage at different operating speeds. It is evident that for all speeds, the wear rate decreases steadily with increasing Zr content, confirming that Zr reinforcement enhances wear resistance. At 0% Zr, the wear rate is at its highest across all speeds, while at 12% Zr, the wear rate reaches its lowest, indicating the strong protective effect of Zr. Among the speeds, higher rpm values (such as 1000 rpm and 800 rpm) show consistently higher wear rates compared to lower speeds (200 rpm and 400 rpm), emphasizing that speed accelerates wear despite the presence of reinforcement. However, the reduction in wear due to Zr addition is significant at all speeds, with the polynomial fit curves aligning well with the experimental data points.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e4.5 Feature importance plot\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003e illustrates the feature importance for wear rate prediction, highlighting the relative influence of Zr% (zirconium content) and Speed on the model\u0026rsquo;s output. Among the two parameters, Zr% plays the most dominant role, contributing nearly 70% importance, while Speed contributes around 30%. This indicates that the zirconium content in the material is the most critical factor in governing wear rate behavior, as it significantly influences the composite\u0026rsquo;s hardness, microstructure, and resistance to material loss.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003e illustrates the SHAP plot of the influence of Speed on wear rate prediction, with the color gradient representing the effect of Zr%. At lower speeds (around 200\u0026ndash;400 rpm), the SHAP values are negative, indicating a reduction in predicted wear contribution. As the speed increases beyond 600 rpm, the SHAP values gradually shift toward positive, showing that higher speeds tend to increase the wear rate contribution. Additionally, the color distribution suggests that varying Zr% interacts with speed, but the overall impact of zirconium content remains consistent across different speed ranges.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e4.6 Sensitivity analysis\u003c/h2\u003e\u003cp\u003eThe sensitivity plots shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003e illustrate the effect of Speed and Zr% reinforcement on the predicted wear rate. The left plot shows that as Speed increases from 200 rpm to 1000 rpm, the predicted wear rate rises almost linearly, indicating a strong positive correlation between speed and material wear. Higher speeds intensify surface interactions and frictional heating, thereby accelerating wear. In contrast, the right plot shows that increasing Zr% reinforcement has the opposite effect: the predicted wear rate decreases significantly, especially beyond 6\u0026ndash;8% reinforcement, where a sharp reduction in wear is observed. This demonstrates that Zr% enhances the material\u0026rsquo;s resistance by improving hardness and structural stability, thus minimizing material loss. Together, these results highlight that speed promotes wear, while zirconium reinforcement mitigates it, underscoring the importance of optimizing both parameters for improved tribological performance.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e4.7 Predictive analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003e shows the 3D surface plot illustrates the predicted wear rate as a function of cutting speed (rpm) and Zr% chill addition, modeled using a polynomial regression (degree\u0026thinsp;=\u0026thinsp;2). The surface indicates a gradual increase in wear rate with rising cutting speed, suggesting that higher rpm intensifies material wear. The effect of Zr% chill is comparatively moderate, with wear rates showing slight variation across the 0\u0026ndash;12% range, indicating that zirconium addition plays a secondary role compared to speed. The plotted data points align closely with the polynomial-fitted surface, confirming a good model fit and predictive accuracy. Overall, the analysis highlights that while Zr% chill influences wear marginally, cutting speed is the dominant factor affecting wear rate, with higher speeds consistently leading to increased wear.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe residual plot for the polynomial regression model (degree\u0026thinsp;=\u0026thinsp;2) shows the difference between the predicted and actual wear rate values as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e17\u003c/span\u003e. The residuals are scattered randomly around the red dashed line (zero error line), without displaying any systematic pattern or trend. This indicates that the polynomial regression model has captured the underlying relationship effectively and that the errors are mostly random in nature. The residuals are small in magnitude (mostly within \u0026plusmn;\u0026thinsp;0.0008), confirming good predictive accuracy. Overall, the residual plot suggests that the polynomial model provides a reasonably good fit for predicting wear rate.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Conclusion","content":"\u003cp\u003eThe present study investigated the combined effect of sliding speed and zirconium (Zr) chill reinforcement on the wear behavior of LM13-based composites through experimental testing, microstructural characterization, and predictive modeling. The results clearly demonstrated that wear rate increases with sliding speed, highlighting the detrimental influence of higher rpm on surface integrity. However, the incorporation of Zr chills significantly improved wear resistance, with 9\u0026ndash;12% reinforcement showing the most effective reduction in material loss. SEM analysis confirmed the uniform distribution of zircon particles within the LM13 matrix, which acted as hard load-bearing phases, while delamination features revealed the fatigue-induced wear mechanisms at higher speeds. XRD and EDS analyses further validated the presence of zircon phases and confirmed their role in strengthening the composite microstructure. Regression and machine learning models, including Linear, Polynomial, SVR, Decision Tree, and Random Forest, achieved high predictive accuracy, with Polynomial regression consistently providing reliable fits and error analysis confirming its robustness. Sensitivity and feature importance studies established that Zr% was the dominant factor governing wear resistance, while speed remained a critical contributor to wear intensification. Overall, this integrated approach demonstrated that optimizing zirconium reinforcement can effectively counteract the adverse effects of high sliding speeds, thereby extending the service life of the material. The findings contribute valuable insights for the design of advanced wear-resistant composites and establish predictive frameworks that can guide industrial applications under varying operating conditions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eConflict of Interest\u003c/h2\u003e\u003cp\u003eThe authors declare that there is no conflict of interest regarding the publication of this paper.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding statement:\u003c/h2\u003e\u003cp\u003eThere are no funds received the research on this article.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eAuthor Contribution StatementRavitej Y P: Conceptualization, Methodology, Experimental Investigation, Data Curation, Writing \u0026ndash; Original Draft, Corresponding Author.Madhusudhan Manjunath: Supervision, Guidance on Methodology, Validation, Critical Review of Manuscript.Prabhakar Kuppahalli: Experimental Design Support, Data Analysis, Resources.KrantiKumar Kshaurad: Machine Learning Modeling, Data Interpretation, Visualization.Bindhushree B S: Microstructural Characterization (SEM, XRD, EDS), Data Curation.Balachandra Halemani: Review and Editing, Technical Input on Wear Analysis.Rajeev Kumar Gupta: Writing \u0026ndash; Review \u0026amp; Editing, Validation of Computational Results.Manish Kumar Mishra: Statistical and Error Analysis, Support in Predictive Modeling.Jayatirtha Patil: Validation, Results Interpretation, Support in Trend Analysis.Ramakumar BVN: Contribution to Materials Selection, Tribological Testing.Nithyananda B: Resources, Supervision, Project Administration, Final Review.All authors have read and approved the final version of the manuscript and agree to be accountable for all aspects of the work.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eIts a unique article containing experimental, Microstructural and machine learning approach to the abrasive wear problem\u003c/p\u003e\u003ch2\u003eData Availability Statement\u003c/h2\u003e\u003cp\u003eThe data supporting the findings of this study are available from the corresponding author, Ravitej Y P, upon reasonable request. Graphs, raw measurements, and model code used for regression analysis can be provided to interested researchers for further exploration or validation.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDeepak V, Abhilash O, Ravitej YP, Veerachari, Abhinandan L (2021) Design and development of progressive tool for mold tag, \u003cem\u003eAIP Conf. Proc.\u003c/em\u003e, vol. 2316, no. February. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1063/5.0038385\u003c/span\u003e\u003cspan address=\"10.1063/5.0038385\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrakash RY, Venkata B (2025) Experimental, FEA, and machine learning studies on wear behavior of LM13 aluminum hybrid composites reinforced with zircon and graphite. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/13506501251367549\u003c/span\u003e\u003cspan address=\"10.1177/13506501251367549\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e32 0 2 4, pp. 0\u0026ndash;116\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Keshavamurthy R, Adarsha H, Kumar VV, Bavan DS, Asdaque PMGB (2025) Effect of Zircon on Aluminum/Graphite Alloy Hybrid Composite and Wear Characterization with Load: Experimental and ANN Test. J Bio- Tribo-Corrosion 11(1):1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s40735-025-00944-7\u003c/span\u003e\u003cspan address=\"10.1007/s40735-025-00944-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Chandra BT, Adarsha H, Chandrashekar A, R PC (2024) Analysis of Wear Parameters of Chill Casted LM13 / Zircon / Carbon Hybrid Composites Using Experimental and Statistical Approach. 77\u0026ndash;88. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.52209/2706-977X\u003c/span\u003e\u003cspan address=\"10.52209/2706-977X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSandeep D et al (2021) February,., CFD simulation of transonic turbulent flow past NACA 0012 aerofoil, \u003cem\u003eAIP Conf. Proc.\u003c/em\u003e, vol. 2316, no. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1063/5.0036890\u003c/span\u003e\u003cspan address=\"10.1063/5.0036890\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Mohan CB, Anantha Prasad MG (2021) Corrosion analysis of LM13/ZrSiO4 /C Hybrid Metal Matrix Composites, \u003cem\u003eJ. Corros. Sci. Eng.\u003c/em\u003e, vol. 23, no. November\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrem Chand R et al (2021) Characterization of banana and e glass fiber reinforced hybrid epoxy composites, \u003cem\u003eMater. Today Proc.\u003c/em\u003e, vol. 46, no. xxxx, pp. 9119\u0026ndash;9125. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.matpr.2021.05.402\u003c/span\u003e\u003cspan address=\"10.1016/j.matpr.2021.05.402\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNaveen Kumar S et al (2021) Fabrication and characterization of hardness and microstructure of large sized Al2014-SiC composite, \u003cem\u003eMater. Today Proc.\u003c/em\u003e, vol. 46, no. xxxx, pp. 9102\u0026ndash;9106. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.matpr.2021.05.398\u003c/span\u003e\u003cspan address=\"10.1016/j.matpr.2021.05.398\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChand RP, Ravitej YP, Kumar N, Veeracahri, Gowda NA (2057) Characterization of banana and bagasse fiber reinforced hybrid epoxy composites, \u003cem\u003eAIP Conf. Proc.\u003c/em\u003e, vol. no. January, 2019. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1063/1.5085624\u003c/span\u003e\u003cspan address=\"10.1063/1.5085624\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrem Chand R, Halemani BS, Chandrasekhar KM, Ravitej YP, Hemanth Raju T, Udayshankar S (2021) Investigation and analysis for mechanical properties of banana and e glass fiber reinforced hybrid epoxy composites, \u003cem\u003eMater. Today Proc.\u003c/em\u003e, vol. 47, no. xxxx, pp. 2509\u0026ndash;2515. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.matpr.2021.05.044\u003c/span\u003e\u003cspan address=\"10.1016/j.matpr.2021.05.044\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Swaroop V, Ramesh S, Adarsha H, Veerachari, and, Nischith (2018) Finite element analysis of mild steel - Rubber sandwich composite material. IOP Conf Ser Mater Sci Eng 376(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1757-899X/376/1/012040\u003c/span\u003e\u003cspan address=\"10.1088/1757-899X/376/1/012040\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrakash RY, Boganarasimhaiah MC, Gangadharaiah APM (2021) Fabrication, micro structural and mechanical characterization of Zircon Particles (ZrSiO4) reinforced Aluminum alloy (MMCs). Indian J Eng Mater Sci 28(6):549\u0026ndash;553. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.56042/ijems.v28i6.40173\u003c/span\u003e\u003cspan address=\"10.56042/ijems.v28i6.40173\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMaity D, Siddheshwar PG, Saha S (1998) Adv Fluid Mech Turbomach. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-642-72157-1\u003c/span\u003e\u003cspan address=\"10.1007/978-3-642-72157-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Mohan CB, Ananthaprasad MG (2022) Dry Sliding Friction and Wear Behavior of LM13/Zircon/Carbon (HMMC\u0026rsquo;s): An Experimental, Statistical and Artificial Neural Network Approach. Tribol Ind 44(3):374\u0026ndash;393. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.24874/ti.1223.11.21.03\u003c/span\u003e\u003cspan address=\"10.24874/ti.1223.11.21.03\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Mohan CB, Ananthaprasad MG (2022) Evaluation of copper chill on tribological behaviour of LM13/ZrSiO\u003csub\u003e4\u003c/sub\u003e/C hybrid metal matrix composites. J Mines Met Fuels 69(12):142. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18311/jmmf/2021/30143\u003c/span\u003e\u003cspan address=\"10.18311/jmmf/2021/30143\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRavitej YP, Mohan CB, Ananthaprasad MG (2021) Effect of Reinforcement and Copper Chill on LM13/ZrSiO4/C Hybrid Metal Matrix Composites (HMMCS)-An Experimental and Statistical Analysis. China\u0026rsquo;s Refract 30(4):12\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.19691/j.cnki.1004-4493.2021.04.003\u003c/span\u003e\u003cspan address=\"10.19691/j.cnki.1004-4493.2021.04.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Sliding wear behaviour, Zirconium (Zr) chill reinforcement, LM13 aluminium alloy composite, Abrasive wear resistance, Regression and machine learning models, Tribological performance optimization","lastPublishedDoi":"10.21203/rs.3.rs-7450766/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7450766/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study investigates the influence of sliding speed and zirconium (Zr) chill reinforcement on the wear behavior of LM13 alloy composites through experimental evaluation, microstructural analysis, and predictive modeling. Specimens were prepared according to ASTM standards and subjected to abrasive wear testing at varying speeds and reinforcement levels (0\u0026ndash;12% Zr). The experimental results revealed that wear rate increased consistently with sliding speed, while the incorporation of Zr significantly reduced material loss, with 9\u0026ndash;12% reinforcement demonstrating the highest wear resistance. Microstructural analysis using SEM confirmed the uniform dispersion of zircon particles within the LM13 matrix, enhancing hardness and load-bearing capacity, while delamination features indicated fatigue-induced wear at elevated speeds. XRD and EDS analyses further validated the crystalline structure and elemental composition, confirming the successful integration of Zr reinforcements. To complement experimental findings, machine learning techniques including Linear, Polynomial, Support Vector Regression (SVR), Decision Tree, and Random Forest models were employed to predict wear behavior. All models achieved high accuracy (R\u0026sup2; \u0026gt;0.93), with Polynomial regression consistently providing the most reliable predictions, as supported by error and significance analysis. Sensitivity and feature importance studies identified Zr% as the dominant factor influencing wear resistance, while speed remained a critical contributor to wear intensification.\u003c/p\u003e","manuscriptTitle":"Abrasive Wear Behavior of Zr-Reinforced LM13 Composites: Experimental and Machine Learning Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-09 18:41:23","doi":"10.21203/rs.3.rs-7450766/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2f69f164-f172-4c99-92b8-c7164c5b5e7d","owner":[],"postedDate":"September 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-09-23T02:40:14+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-09 18:41:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7450766","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7450766","identity":"rs-7450766","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}