| Authors | Chaitanya P. Kale1, P. William2 , 3 , Mohammed Almakki4, Pravin B. Khatkale1, Prashant M. Yawalkar5, Ritesh Ravindra Agwan1, Prashant V. Thokal6 |
| Affiliations |
1School of Engineering and Technology, Sanjivani University, Kopargaon, MH, India 2School of Computer Science and Technology, Karunya Institute of Technology and Sciences (Deemed University), 641114 Coimbatore, India 3Centre for Research and Consultancy, UNITAR International University, 47301 Selangor, Malaysia 4School of Engineering, Architecture and Interior Design, Amity University Dubai, P.O. Box 345019, Dubai International Academic City, United Arab Emirates 5Department of Computer Engineering, MET's Institute of Engineering, BKC, Nashik, MH, India 6Department of Electrical Engineering, Sanjivani College of Engineering, Kopargaon, MH, India |
| Е-mail | williamacads@gmail.com |
| Issue | Volume 18, Year 2026, Number 4 |
| Dates | Received 15 March 2026; revised manuscript received 15 August 2026; published online 21 August 2026 |
| Citation | Chaitanya P. Kale, P. William, et al., J. Nano- Electron. Phys. 18 No 4, 04023 (2026) |
| DOI | https://doi.org/10.21272/jnep.18(4).04023 |
| PACS Number(s) | 07.05.Tp, 81.16. – c |
| Keywords | Machine learning, MXene, Nano-material printing, Pneumatic 3D printing, Filament quality, Predictive modeling, Process optimization. |
| Annotation |
The developments of nano-material enable precise MXene fabrication, though pneumatic 3D printing still struggles with filament uniformity due to sensitivity in pressure, ink concentration, nozzle diameter and velocity. Traditional trial-and-error methods waste resources, highlighting the importance of intelligent, data-driven approaches to enhance parameters, minimize variation, and achieve consistent, high-quality MXene structures suitable for scalable applications. To address this, the research aims to develop a data-driven framework to optimize printing parameters, minimize unpredictability, and ensure uniform filament quality. MXene 3D Printing Filament Data with 5000 samples was gathered, and experimental conditions affecting filament formation, including nozzle diameter, ink concentration, temperature, printing velocity, and humidity. The dataset was standardized and cleaned to normalize feature scales and remove outliers using Min-max normalization. Dynamic Snap-Drift Cuckoo Search-driven Natural Gradient Boosting (DSDCS-NGBoost) forecasts the filament diameter using these characteristics. The DSDCS-NGBoost model outperforms all connected models, accomplishing the lowest RMSE (0.1052), MAE (0.0827), and highest R (0.9253), signifying superior predictive accuracy and reliable filament width control in MXene printing. This approach highlights the possibility of intelligent algorithms to bridge experimental precision with industrial-scale nano-manufacturing. |
|
List of References |