Deep learning–enhanced prediction of microstructure and porosity evolution in additive-manufactured membrane coatings for harsh environments

This study investigates the capability of additive manufacturing (AM) to produce thick coatings functioning as multifunctional membranes with enhanced barrier, transport, and mechanical properties for harsh operating environments. The primary objective was to evaluate how deposition technique and microstructural optimisation influence porosity, diffusion resistance, corrosion protection, and thermal stability. A combined methodology was implemented, integrating experimental testing of laser cladding, thermal spraying, and direct energy deposition (DED) with mathematical models for permeability, diffusion, and thermal conductivity. Laser cladding demonstrated the densest structures, achieving porosity levels below 2% and reducing gas permeability to 1.2 × 10⁻¹⁵ m², nearly an order of magnitude lower than thermal spraying (1.1 × 10⁻¹⁴ m²). Corrosion testing showed nickel-based cladded coatings reached rates as low as 0.0025 mm/year, representing a 90% reduction compared to uncoated substrates (0.026 mm/year). Thermal barrier evaluation of YSZ coatings indicated a conductivity of 0.95 W/m·K at 1200 °C, corresponding to a 38% reduction in heat flux across 1.2 mm-thick layers. Ultrasonic spray post-treatment reduced surface roughness by up to 55% and biofilm accumulation by nearly half. Error analysis confirmed deviations within ± 6%. These results confirm that AM thick coatings function as functional membranes, offering selective transport regulation, structural durability, and sustainability across the aerospace, energy, and marine sectors.

Citation

Fajingbesi, A. A., Malachi, I. O., Omigbodun, F. T., Oluwabiyi, E. O., Oni, A. A., & Adeyekun, F. P. (2026). Deep learning–enhanced prediction of microstructure and porosity evolution in additive-manufactured membrane coatings for harsh environments. International Journal of Advanced Manufacturing Technology, 142(3-4), 1971-1982. https://doi.org/10.1007/s00170-025-17226-8

Publisher

Springer

Journal

International Journal of Advanced Manufacturing Technology

URI

http://hdl.handle.net/10034/629862

DOI

10.1007/s00170-025-17226-8

Type

Article

Description

© The Author(s) 2025.
The version of record of this article, first published in [International Journal of Advanced Manufacturing Technology], is available online at Publisher’s website: http://dx.doi.org/10.1007/s00170-025-17226-8

ISSN

0268-3768

EISSN

1433-3015

Additional Links

https://link.springer.com/article/10.1007/s00170-025-17226-8