Page 70 - IJAMD-2-3

International Journal of AI for
Materials and Design

ORIGINAL RESEARCH ARTICLE
Data-driven optimization of biaxial shrinkage

and stability in electrospun membranes via
machine learning and Monte Carlo simulation

Shiyu He 1,2 , Chentong Gao 2,3 , Runzhi Lu 2,4 ,Fei Xiao * , Li Cong Huang 6 ,
1,5
and Wei Min Huang *
2
1 State Key Laboratory of Metal Matrix Composites, School of Materials Science and Engineering,
Shanghai Jiao Tong University, Shanghai, China
2 School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore
3 College of Aerospace Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing,
Jiangsu, China
4 School of Civil Engineering, Southeast University, Nanjing, Jiangsu, China
5 Department of Computer Science, Institute of Medical Robotics, Shanghai Jiao Tong University,
Shanghai, China
6 School of Computing, National University of Singapore, Singapore

Abstract
*Corresponding authors:
Fei Xiao Controlling shrinkage behavior in electrospun membranes is critical for applications
(xfei@sjtu.edu.cn) that require precise dimensional or mechanical performance. However, experimental
Wei Min Huang variability and limited datasets often hinder the development of robust process
(mwmhuang@ntu.edu.sg)
models. This study introduces a data-driven framework that combines machine
Citation: He S, Gao C, Lu R, learning with Monte Carlo simulation to enable both accurate and stable shrinkage
Xiao F, Huang LC, Huang WM.
Data-driven optimization of control in electrospinning using a small experimental dataset. Multiple regression
biaxial shrinkage and stability models were trained to predict biaxial shrinkage ratios and their variability, with
in electrospun membranes via support vector regression and extreme gradient boosting showing the best
machine learning and Monte Carlo
simulation. Int J AI Mater Design. performance for accuracy and stability prediction, respectively. Feature importance
2025;2(3):64-77. analysis revealed applied voltage and thermoplastic polyurethane concentration
doi: 10.36922/IJAMD025260022 as the dominant parameters. A Monte Carlo-based optimization strategy was
Received: June 26, 2025 employed to identify process parameter sets that achieve target shrinkage ratios
while minimizing output variability. The proposed approach enables multi-objective
Revised: August 15, 2025
optimization in low-data, high-variability manufacturing environments, offering
Accepted: August 21, 2025 practical insights into precision fabrication of stimulus-responsive membranes.
Published online: September 9,
2025
Keywords: Electrospinning; Shrinkage stability; Machine learning; Monte Carlo
Copyright: © 2025 Author(s). simulation; Process parameter optimization
This is an Open-Access article
distributed under the terms of the
Creative Commons Attribution
License, permitting distribution,
and reproduction in any medium, 1. Introduction
provided the original work is
properly cited. Electrospinning has become a key nanofabrication technique in biomedical and
Publisher’s Note: AccScience engineering applications due to its simplicity, material adaptability, and ability to produce
Publishing remains neutral with continuous fibers with diameters ranging from nanometers to micrometers. Electrospun
1-3
regard to jurisdictional claims in
published maps and institutional membranes, such as polyvinyl alcohol, poly(lactic acid), poly(lactide-co-glycolide), are
affiliations. widely employed in tissue engineering, drug delivery, smart materials, flexible electronics,

Volume 2 Issue 3 (2025) 64 doi: 10.36922/IJAMD025260022