Pub Date : 2026-01-13DOI: 10.1016/j.matlet.2026.140109
Neelima Singh , Mohit Agarwal , Deepa Negi
The current study explores the use of numerical simulation to investigate the Pb-free perovskite absorber layer (PAL) RbGeBr3 based PSC using SCAPS 1D. The impact of various HTLs and ETLs is analyzed, and the optimized structure is SrCu2O2/RbGeBr3/ZnOS/FTO enhances the PCE to 23.90%. Further, the effect of the defect density and the thickness of the RbGeBr3 is analyzed which enhances PCE to 26.02%. These notable findings demonstrate RbGeBr3 as a substantial potential for developing sustainable, eco-friendly, and highly efficient PSC technology.
{"title":"Numerical simulation of lead-free RbGeBr3 based perovskite solar cell","authors":"Neelima Singh , Mohit Agarwal , Deepa Negi","doi":"10.1016/j.matlet.2026.140109","DOIUrl":"10.1016/j.matlet.2026.140109","url":null,"abstract":"<div><div>The current study explores the use of numerical simulation to investigate the Pb-free perovskite absorber layer (PAL) RbGeBr<sub>3</sub> based PSC using SCAPS 1D. The impact of various HTLs and ETLs is analyzed, and the optimized structure is SrCu<sub>2</sub>O<sub>2</sub>/RbGeBr<sub>3</sub>/ZnOS/FTO enhances the PCE to 23.90%. Further, the effect of the defect density and the thickness of the RbGeBr<sub>3</sub> is analyzed which enhances PCE to 26.02%. These notable findings demonstrate RbGeBr<sub>3</sub> as a substantial potential for developing sustainable, eco-friendly, and highly efficient PSC technology.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"408 ","pages":"Article 140109"},"PeriodicalIF":2.7,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145976141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140097
Qingsheng Chen , Jinhui Li , Gaoliang Tao , Zhilin Xiong , Sanjay Nimbalkar
Improving the strength and durability of cemented soil foundations in coastal regions is a significant concern that must be tackled in marine engineering projects. This study utilized various amounts of Nano-SiO2 (NS) and Carbon Fiber (CF) to collaboratively enhance the strength and stability of cemented soil subjected to seawater corrosion. Following a curing period and immersion in seawater for 60 days, the unconfined compressive strength (qu) and the strength growth rate (Re) of the samples were assessed. Analyses using scanning electron microscopy (SEM) and X-ray diffraction (XRD) demonstrated that the interactions between NS, CF, and cement resulted in a compact microstructure capable of resisting the penetration of corrosive ions. In seawater conditions, the mixture containing 5% NS and 2% CF showed the best performance.
{"title":"Durability and microstructural characteristics of Nano-SiO₂–Carbon Fiber composites under seawater environment","authors":"Qingsheng Chen , Jinhui Li , Gaoliang Tao , Zhilin Xiong , Sanjay Nimbalkar","doi":"10.1016/j.matlet.2026.140097","DOIUrl":"10.1016/j.matlet.2026.140097","url":null,"abstract":"<div><div>Improving the strength and durability of cemented soil foundations in coastal regions is a significant concern that must be tackled in marine engineering projects. This study utilized various amounts of Nano-SiO<sub>2</sub> (NS) and Carbon Fiber (CF) to collaboratively enhance the strength and stability of cemented soil subjected to seawater corrosion. Following a curing period and immersion in seawater for 60 days, the unconfined compressive strength (q<sub>u</sub>) and the strength growth rate (R<sub>e</sub>) of the samples were assessed. Analyses using scanning electron microscopy (SEM) and X-ray diffraction (XRD) demonstrated that the interactions between NS, CF, and cement resulted in a compact microstructure capable of resisting the penetration of corrosive ions. In seawater conditions, the mixture containing 5% NS and 2% CF showed the best performance.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140097"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973784","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140096
Jialong Li , Zhichao Ma , Kaizhou Han , Xiaoxu Liu , Bin Sheng , Beile Cao , Kuizhe Feng , Congcong Zhu , Jiaxin Kou , Jie Kang
All-organic P(VDF-HFP)-based composites were fabricated by simply incorporating PCVE. Compared with the pristine P(VDF-HFP), the P/C composites present an high maximum discharge energy density (Ue) of 18.6 J cm−3 at 700 kV/mm (the electric field approaching their breakdown strength (Eb), Eb = 715.6 kV/mm), attributable to concomitant gains in Eb and optimized polarization behavior. Hybrid DFT calculations reveal electrostatic interactions between the P(VDF-HFP) and PCVE chains the intermolecular spacing and compresses the relaxation space of the composites. This kind of electrostatic interactions between polymer chains spontaneously reduced the residual polarization and increased the potential barriers of carrier transporting, thereby reducing energy loss. Therefore, the P/C composites display supressed residual polarization and high breakdown strength, finally excellent energy storage performances. This letter provides a simple yet effective strategy of preparing high energy storage density polymer-based dielectric via molecular-level electrostatic interactions, and the generality of this strategy is further supported in other system.
{"title":"Electrostatic interaction induced high energy storage performance of all-organic PVDF-based composites","authors":"Jialong Li , Zhichao Ma , Kaizhou Han , Xiaoxu Liu , Bin Sheng , Beile Cao , Kuizhe Feng , Congcong Zhu , Jiaxin Kou , Jie Kang","doi":"10.1016/j.matlet.2026.140096","DOIUrl":"10.1016/j.matlet.2026.140096","url":null,"abstract":"<div><div>All-organic P(VDF-HFP)-based composites were fabricated by simply incorporating PCVE. Compared with the pristine P(VDF-HFP), the P/C composites present an high maximum discharge energy density (<em>U</em><sub><em>e</em></sub>) of 18.6 J cm<sup>−3</sup> at 700 kV/mm (the electric field approaching their breakdown strength (<em>E</em><sub><em>b</em></sub>), <em>E</em><sub><em>b</em></sub> = 715.6 kV/mm), attributable to concomitant gains in <em>E</em><sub><em>b</em></sub> and optimized polarization behavior. Hybrid DFT calculations reveal electrostatic interactions between the P(VDF-HFP) and PCVE chains the intermolecular spacing and compresses the relaxation space of the composites. This kind of electrostatic interactions between polymer chains spontaneously reduced the residual polarization and increased the potential barriers of carrier transporting, thereby reducing energy loss. Therefore, the P/C composites display supressed residual polarization and high breakdown strength, finally excellent energy storage performances. This letter provides a simple yet effective strategy of preparing high energy storage density polymer-based dielectric via molecular-level electrostatic interactions, and the generality of this strategy is further supported in other system.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140096"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973793","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140104
Guihao Sun, Chuanhai Li, Yao Li, Zihao Zhao, Jun Liang, Yuxiu Zhang
The creep behaviors of Al-Cu-Mg alloys after T4 and T6 treatments were investigated. The results showed that the precipitates introduced during T6 treatment easily coarsened and dissolved, which loosened the continuous strengthening effect. In contrast, the gradual generation of precipitates during creep in the T4 condition could compensate for the softening effect caused by precipitate coarsening, thereby providing a continuous strengthening effect. Thus, the T4 sample exhibited greater creep resistance than the T6 sample.
{"title":"Dependence of creep resistance on the precipitation time node of Al-Cu-Mg alloys","authors":"Guihao Sun, Chuanhai Li, Yao Li, Zihao Zhao, Jun Liang, Yuxiu Zhang","doi":"10.1016/j.matlet.2026.140104","DOIUrl":"10.1016/j.matlet.2026.140104","url":null,"abstract":"<div><div>The creep behaviors of Al-Cu-Mg alloys after T4 and T6 treatments were investigated. The results showed that the precipitates introduced during T6 treatment easily coarsened and dissolved, which loosened the continuous strengthening effect. In contrast, the gradual generation of precipitates during creep in the T4 condition could compensate for the softening effect caused by precipitate coarsening, thereby providing a continuous strengthening effect. Thus, the T4 sample exhibited greater creep resistance than the T6 sample.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140104"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140107
Bingyan Ou , Daibo Zhu , Zhirui Liu , Weihao zhang , Canwei Zhang , Tan Han , Yang liu
This study investigates the effect of solution temperatures on the aged-state mechanical properties of 0.03 mm-diameter C17200 copper beryllium micro-wires. At 680 °C, fine grains, high-density grain boundaries, and residual undissolved γ phase yield higher initial tensile strength but restrict subsequent age-hardening. In contrast, 840 °C solution treatment forms a supersaturated solid solution (lower as-solution-treated strength); after aging, high-density nanoscale γ’ precipitates are uniformly distributed in the matrix, enabling significant precipitation strengthening and tensile strength approaching 1200 MPa. These findings show that for microscopic-scale CuBe micro-wires, the contribution of full precipitation strengthening to overall mechanical improvement exceeds that of grain boundary strengthening.
{"title":"Effect of solution treatment temperature on the mechanical properties of age-hardened C17200 CuBe micro-wires","authors":"Bingyan Ou , Daibo Zhu , Zhirui Liu , Weihao zhang , Canwei Zhang , Tan Han , Yang liu","doi":"10.1016/j.matlet.2026.140107","DOIUrl":"10.1016/j.matlet.2026.140107","url":null,"abstract":"<div><div>This study investigates the effect of solution temperatures on the aged-state mechanical properties of 0.03 mm-diameter C17200 copper beryllium micro-wires. At 680 °C, fine grains, high-density grain boundaries, and residual undissolved γ phase yield higher initial tensile strength but restrict subsequent age-hardening. In contrast, 840 °C solution treatment forms a supersaturated solid solution (lower as-solution-treated strength); after aging, high-density nanoscale γ’ precipitates are uniformly distributed in the matrix, enabling significant precipitation strengthening and tensile strength approaching 1200 MPa. These findings show that for microscopic-scale Cu<img>Be micro-wires, the contribution of full precipitation strengthening to overall mechanical improvement exceeds that of grain boundary strengthening.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"408 ","pages":"Article 140107"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140103
S.J. Caraguay , T.S. Pereira , M. Pereira , F.A. Xavier
Specifications for coating metallic components typically require surface preparation to achieve a defined surface roughness. However, the relationship between surface roughness characteristics and corrosion resistance remains poorly understood, as most specifications rely on a single roughness parameter, typically Rz. Nanosecond laser surface texturing has emerged as an effective alternative to conventional pretreatment methods, producing well-defined microstructures that can improve interfacial stability under corrosive conditions. In this study, laser-textured steel surfaces were characterized using several 3D roughness parameters (Sa, S10z, Ssk, Sku, Sdr, and Vvc) to establish correlations between surface morphology and corrosion propagation resistance of epoxy-coated samples. Results demonstrate that parameters such as Sa, Sdr, and Vvc correlate inversely with coating delamination. Surfaces with higher Ssk and platykurtic (Sku < 3) profiles exhibited reduced corrosion propagation. These findings demonstrate that selected 3D roughness parameters can complement conventional specifications and support optimization of laser surface texturing for corrosion protection.
{"title":"Correlation between 3D roughness parameters and coating stability on laser-textured surfaces","authors":"S.J. Caraguay , T.S. Pereira , M. Pereira , F.A. Xavier","doi":"10.1016/j.matlet.2026.140103","DOIUrl":"10.1016/j.matlet.2026.140103","url":null,"abstract":"<div><div>Specifications for coating metallic components typically require surface preparation to achieve a defined surface roughness. However, the relationship between surface roughness characteristics and corrosion resistance remains poorly understood, as most specifications rely on a single roughness parameter, typically Rz. Nanosecond laser surface texturing has emerged as an effective alternative to conventional pretreatment methods, producing well-defined microstructures that can improve interfacial stability under corrosive conditions. In this study, laser-textured steel surfaces were characterized using several 3D roughness parameters (S<sub>a</sub>, S<sub>10z</sub>, S<sub>sk</sub>, S<sub>ku</sub>, S<sub>dr</sub>, and V<sub>vc</sub>) to establish correlations between surface morphology and corrosion propagation resistance of epoxy-coated samples. Results demonstrate that parameters such as S<sub>a</sub>, S<sub>dr</sub>, and V<sub>vc</sub> correlate inversely with coating delamination. Surfaces with higher S<sub>sk</sub> and platykurtic (S<sub>ku</sub> < 3) profiles exhibited reduced corrosion propagation. These findings demonstrate that selected 3D roughness parameters can complement conventional specifications and support optimization of laser surface texturing for corrosion protection.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140103"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140105
Zicheng Wu , Yi Zhou , Ninian Sing Kok Ho , Tong Liu , John Hock Lye Pang
The spatial placement of parts on the baseplate during laser powder bed fusion (LPBF) can lead to inconsistent fatigue performance in 316L stainless steel. This study establishes a quantitative link between part placement, defect distribution, and fatigue life within a single LPBF build by combining X-ray computed tomography and stress-controlled fatigue testing. The specimens positioned downstream of the shielding gas flow exhibited higher porosity (0.009–0.016%) than those located upstream (0.001–0.005%), resulting in shorter fatigue life cycles due to crack-defect coalescence, as validated by fractographic analysis. In contrast, upstream specimens showed longer life and failed primarily from geometric stress concentration features. These results show that the placement of parts on the baseplate influences crack propagation and fatigue failure mechanisms in LPBF 316L, highlighting the importance of optimal part placement and shielding gas flow control in applications where fatigue performance is crucial.
{"title":"Influence of part placement on defect distribution and fatigue failure mechanism in laser powder bed fused 316L stainless steel","authors":"Zicheng Wu , Yi Zhou , Ninian Sing Kok Ho , Tong Liu , John Hock Lye Pang","doi":"10.1016/j.matlet.2026.140105","DOIUrl":"10.1016/j.matlet.2026.140105","url":null,"abstract":"<div><div>The spatial placement of parts on the baseplate during laser powder bed fusion (LPBF) can lead to inconsistent fatigue performance in 316L stainless steel. This study establishes a quantitative link between part placement, defect distribution, and fatigue life within a single LPBF build by combining X-ray computed tomography and stress-controlled fatigue testing. The specimens positioned downstream of the shielding gas flow exhibited higher porosity (0.009–0.016%) than those located upstream (0.001–0.005%), resulting in shorter fatigue life cycles due to crack-defect coalescence, as validated by fractographic analysis. In contrast, upstream specimens showed longer life and failed primarily from geometric stress concentration features. These results show that the placement of parts on the baseplate influences crack propagation and fatigue failure mechanisms in LPBF 316L, highlighting the importance of optimal part placement and shielding gas flow control in applications where fatigue performance is crucial.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140105"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973790","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-12DOI: 10.1016/j.matlet.2026.140106
Yanzhao Yang , Lei Chen , Chunbo Jin , Hongfei Cai , Haifeng Zhai , Sheng Xu , Risu Na , Fu Chen , Shiyu Liang , Xian Dong
This study addresses the pronounced fatigue-life scatter and unresolved micro-mechanisms of additively manufactured (AM) 316L steel under complex loading by comparing AM and conventionally manufactured (CM) counterparts subjected to periodic overloads with different intervals. EBSD-based metrics, Kernel average misorientation (KAM) evolution, residual stress (RS) profiles, and crack-path observations reveal that localized softening and strain concentration drive early crack initiation in AM, while CM benefits from RS shielding and delayed cracking. The results clarify why AM steels show reduced life with increasing overload severity and offer practical insight for fatigue design and life prediction of AM components.
{"title":"Fatigue cyclic behavior and microstructural evolution of additively and conventionally manufactured 316L stainless steel under periodic overloads","authors":"Yanzhao Yang , Lei Chen , Chunbo Jin , Hongfei Cai , Haifeng Zhai , Sheng Xu , Risu Na , Fu Chen , Shiyu Liang , Xian Dong","doi":"10.1016/j.matlet.2026.140106","DOIUrl":"10.1016/j.matlet.2026.140106","url":null,"abstract":"<div><div>This study addresses the pronounced fatigue-life scatter and unresolved micro-mechanisms of additively manufactured (AM) 316L steel under complex loading by comparing AM and conventionally manufactured (CM) counterparts subjected to periodic overloads with different intervals. EBSD-based metrics, Kernel average misorientation (KAM) evolution, residual stress (RS) profiles, and crack-path observations reveal that localized softening and strain concentration drive early crack initiation in AM, while CM benefits from RS shielding and delayed cracking. The results clarify why AM steels show reduced life with increasing overload severity and offer practical insight for fatigue design and life prediction of AM components.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140106"},"PeriodicalIF":2.7,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973771","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-11DOI: 10.1016/j.matlet.2026.140102
Runze Yang , Jinyue Zhang
Continuous carbon fiber reinforced polymers (C-CFRPs) fabricated by 3D printing hold great potential for lightweight structural applications, yet their internal defect characteristics remain insufficiently understood. In this study, X-ray computed tomography (X-ray CT) combined with a U-Net deep learning segmentation approach was employed to automatically detect and quantitatively analyze pores in 3D-printed C-CFRPs. The results show that defects are periodically distributed across layers, dominated by ellipsoidal pores and net-shaped voids. The long axes of ellipsoidal pores align with the fiber direction, and pores located in the upper layers exhibit greater elongation. Compared to vacuum-assisted resin infusion molding (VARIM) composites, the porosity of printed Specimen was nearly 30 times higher, highlighting the limitations of layer bonding and densification in additive manufacturing. This work demonstrates the efficiency of deep learning-based defect characterization and provides new insights for process optimization and quality assessment of 3D-printed structures.
{"title":"X-ray CT analysis of internal defects in 3D-printed continuous carbon fiber composites using deep learning","authors":"Runze Yang , Jinyue Zhang","doi":"10.1016/j.matlet.2026.140102","DOIUrl":"10.1016/j.matlet.2026.140102","url":null,"abstract":"<div><div>Continuous carbon fiber reinforced polymers (C-CFRPs) fabricated by 3D printing hold great potential for lightweight structural applications, yet their internal defect characteristics remain insufficiently understood. In this study, X-ray computed tomography (X-ray CT) combined with a U-Net deep learning segmentation approach was employed to automatically detect and quantitatively analyze pores in 3D-printed C-CFRPs. The results show that defects are periodically distributed across layers, dominated by ellipsoidal pores and net-shaped voids. The long axes of ellipsoidal pores align with the fiber direction, and pores located in the upper layers exhibit greater elongation. Compared to vacuum-assisted resin infusion molding (VARIM) composites, the porosity of printed Specimen was nearly 30 times higher, highlighting the limitations of layer bonding and densification in additive manufacturing. This work demonstrates the efficiency of deep learning-based defect characterization and provides new insights for process optimization and quality assessment of 3D-printed structures.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140102"},"PeriodicalIF":2.7,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-11DOI: 10.1016/j.matlet.2026.140100
Min Xue , Shaoxuan Wang , Lijuan Kong
Existing machine-learning-based studies on compressive strength prediction of alkali-activated materials (AAM) mainly focus on model accuracy. Most studies are limited to a single curing age or single-factor analysis. As a result, the interpretable quantitative characterization of multi-parameter coupling effects and their age-dependent evolution remains insufficient. To address these limitations, a database containing 529 AAM mix designs was established in this study. A genetic-algorithm-optimized neural network was developed for multi-age compressive strength prediction and interpretable quantitative analysis. Six key variables—CaO, SiO2, Al2O3, water-to-binder ratio, activator modulus, and alkali equivalent—were used as model inputs. Three models were trained for 3, 7and 28 days, achieving R2 values above 0.88. Shapley Additive Explanation (SHAP) analysis revealed that SiO2 predominantly affects early-age strength, Al2O3 governs later-age strength, and CaO contributes across all curing stages. Early strength is maximized when SiO2 is below 35%, while the optimal Al2O3 content for 28 days strength is approximately 15%. The optimal synergistic ratio expands to SiO2 content is 25% ∼ 35%, Al2O3 content is 15% ∼ 20%, and CaO content is 15% ∼ 30%. These findings provide insights into precursor design and activator selection for high-performance AAM.
{"title":"Effect of precursor chemical composition on the compressive strength of alkali-activated materials based on interpretable artificial neural networks","authors":"Min Xue , Shaoxuan Wang , Lijuan Kong","doi":"10.1016/j.matlet.2026.140100","DOIUrl":"10.1016/j.matlet.2026.140100","url":null,"abstract":"<div><div>Existing machine-learning-based studies on compressive strength prediction of alkali-activated materials (AAM) mainly focus on model accuracy. Most studies are limited to a single curing age or single-factor analysis. As a result, the interpretable quantitative characterization of multi-parameter coupling effects and their age-dependent evolution remains insufficient. To address these limitations, a database containing 529 AAM mix designs was established in this study. A genetic-algorithm-optimized neural network was developed for multi-age compressive strength prediction and interpretable quantitative analysis. Six key variables—CaO, SiO<sub>2</sub>, Al<sub>2</sub>O<sub>3</sub>, water-to-binder ratio, activator modulus, and alkali equivalent—were used as model inputs. Three models were trained for 3, 7and 28 days, achieving R<sup>2</sup> values above 0.88. Shapley Additive Explanation (SHAP) analysis revealed that SiO<sub>2</sub> predominantly affects early-age strength, Al<sub>2</sub>O<sub>3</sub> governs later-age strength, and CaO contributes across all curing stages. Early strength is maximized when SiO<sub>2</sub> is below 35%, while the optimal Al<sub>2</sub>O<sub>3</sub> content for 28 days strength is approximately 15%. The optimal synergistic ratio expands to SiO<sub>2</sub> content is 25% ∼ 35%, Al<sub>2</sub>O<sub>3</sub> content is 15% ∼ 20%, and CaO content is 15% ∼ 30%. These findings provide insights into precursor design and activator selection for high-performance AAM.</div></div>","PeriodicalId":384,"journal":{"name":"Materials Letters","volume":"407 ","pages":"Article 140100"},"PeriodicalIF":2.7,"publicationDate":"2026-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145973768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号