Optimizing Solid Microneedle Design: A Comprehensive ML-Augmented DOE Approach.

Area of Science:

• Biomaterials Engineering
• Mechanical Engineering
• Computational Modeling

Background:

• Microneedles (MNs) are micrometer-scale needles used in drug delivery, skincare, and health monitoring.
• MNs provide a minimally invasive, painless, and cost-effective alternative to hypodermic needles with reduced tissue damage.

Purpose of the Study:

• To explore and optimize four distinct MN designs: cone, tapered cone, square-based pyramid, and triangular-based pyramid.
• To enhance MN durability for various physiological conditions through design optimization.

Main Methods:

• Investigated four MN shapes under compressive, buckling, and bending load conditions.
• Utilized a Design of Experiments approach within ANSYS Workbench to analyze geometric parameters (base/tip dimensions, height, taper angle).
• Developed mathematical and response surface models, integrating machine learning (ML) for multiobjective optimization.

Main Results:

• Established a comprehensive framework for evaluating MN performance based on deformation, buckling load, factor of safety (FOS), and bending stress.
• Generated regression models, sensitivity charts, and response curves to guide optimization.
• Identified optimized geometrical designs for the four MN shapes.

Conclusions:

• The study presents a novel ML-augmented optimization framework for microneedle design.
• The findings offer valuable insights for creating durable MNs suitable for diverse physiological applications.
• Optimized MN designs ensure structural integrity under various loading scenarios.