Pub Date : 2026-01-08DOI: 10.1186/s40712-025-00392-8
Alexander Wolff, Lukas Mueller, Steffen Klingel, Marco Rahm
This work presents a transfer learning approach for the forward design of composite unit cells of metasurfaces and frequency-selective surfaces (FSS). As a specific application, we target reconfigurable intelligent surfaces (RIS) for wireless communication. We demonstrate that forward models for dual-dipole unit cells can be trained using significantly fewer data by reusing pre-trained models of simpler, single-dipole structures. Our architecture reduces the required training data by up to a factor of 25 while maintaining a mean squared error (MSE) on the order of (10^{-2}). After establishing the forward model for the dual-dipole RIS, we use it in an inverse design framework to synthesize a composite 2-bit RIS unit cell operating at 26.5 GHz. The resulting RIS provides phase modulation in a 270(^circ) range by tuning the reverse bias voltage of integrated varactor diodes. Numerical simulations confirm the validity of the proposed approach, establishing transfer learning as a data-efficient and practical method for the design of composite unit cell structures.
{"title":"Transfer learning for data-efficient modeling of composite metasurface and FSS unit cells","authors":"Alexander Wolff, Lukas Mueller, Steffen Klingel, Marco Rahm","doi":"10.1186/s40712-025-00392-8","DOIUrl":"10.1186/s40712-025-00392-8","url":null,"abstract":"<div><p>This work presents a transfer learning approach for the forward design of composite unit cells of metasurfaces and frequency-selective surfaces (FSS). As a specific application, we target reconfigurable intelligent surfaces (RIS) for wireless communication. We demonstrate that forward models for dual-dipole unit cells can be trained using significantly fewer data by reusing pre-trained models of simpler, single-dipole structures. Our architecture reduces the required training data by up to a factor of 25 while maintaining a mean squared error (MSE) on the order of <span>(10^{-2})</span>. After establishing the forward model for the dual-dipole RIS, we use it in an inverse design framework to synthesize a composite 2-bit RIS unit cell operating at 26.5 GHz. The resulting RIS provides phase modulation in a 270<span>(^circ)</span> range by tuning the reverse bias voltage of integrated varactor diodes. Numerical simulations confirm the validity of the proposed approach, establishing transfer learning as a data-efficient and practical method for the design of composite unit cell structures.</p></div>","PeriodicalId":592,"journal":{"name":"International Journal of Mechanical and Materials Engineering","volume":"21 1","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s40712-025-00392-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146082615","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The accumulation of space charges in high-voltage cable insulation layers significantly compromises electrical performance and accelerates material degradation. This study proposed a novel strategy to inhibit charge injection by incorporating highly conductive annealed MXene nanosheets into the semiconductive shielding material. Results demonstrate the introduction of MXene remarkably reduce positive temperature coefficient (PTC) of composites. With MXene content of 0.5 wt% (carbon black (CB) of 31.5 wt%), the PTC of composite decreases to 3.6 from 5.74 (pure CB sample), indicating enhanced thermal stability. Furthermore, MXene modified materials effectively suppress space charges accumulation, reducing the charge density in the insulation layer to 13.38 C/m3 (MXene content of 0.5 wt%) compared to 20.44 C/m3 for unmodified sample. These enhancements are attributed to MXene’s dual role in forming efficient conductive networks and introducing deep charge traps. This study sheds new light on the development of advanced semiconductive shielding materials for high-voltage cable applications.
{"title":"Impact of MXene on the positive temperature coefficient (PTC) properties of semiconductive composites and space charge accumulation in insulation","authors":"Jiana Hu, Jian Qiao, Yun Chen, Bidong Tu, Ruyun Yang, Jiayao Wang, Shengyi Li, Yaodong Zhang, Peng Xiao, Liheng Yang, Dabing Chen, Wei Yang","doi":"10.1186/s40712-025-00391-9","DOIUrl":"10.1186/s40712-025-00391-9","url":null,"abstract":"<div><p>The accumulation of space charges in high-voltage cable insulation layers significantly compromises electrical performance and accelerates material degradation. This study proposed a novel strategy to inhibit charge injection by incorporating highly conductive annealed MXene nanosheets into the semiconductive shielding material. Results demonstrate the introduction of MXene remarkably reduce positive temperature coefficient (PTC) of composites. With MXene content of 0.5 wt% (carbon black (CB) of 31.5 wt%), the PTC of composite decreases to 3.6 from 5.74 (pure CB sample), indicating enhanced thermal stability. Furthermore, MXene modified materials effectively suppress space charges accumulation, reducing the charge density in the insulation layer to 13.38 C/m<sup>3</sup> (MXene content of 0.5 wt%) compared to 20.44 C/m<sup>3</sup> for unmodified sample. These enhancements are attributed to MXene’s dual role in forming efficient conductive networks and introducing deep charge traps. This study sheds new light on the development of advanced semiconductive shielding materials for high-voltage cable applications.</p></div>","PeriodicalId":592,"journal":{"name":"International Journal of Mechanical and Materials Engineering","volume":"21 1","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s40712-025-00391-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146082614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}