Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Doctor of Philosophy

Program

Chemical and Biochemical Engineering

Supervisor

Zhu Jesse

2nd Supervisor

Zhang Hui

Co-Supervisor

Abstract

Powder coatings have emerged as an eco-friendly and efficient alternative to traditional liquid coatings, owing to their solvent-free composition and superior performance in impact resistance and corrosion resistance. However, the particle size distribution (PSD) of powder coatings remains a critical challenge, as wide PSDs adversely affect powder flowability, deposition efficiency, and the appearance of the coating films. This dissertation addresses the wide PSDS of powder coatings by introducing innovative modifications to air classifying mill (ACM) and applying advanced analytical methods to explore the impacts of PSD on powder coatings.

The research focuses on narrowing PSDs through two primary interventions. Firstly, self- designed ACM classifiers with serrated edges were developed to improve classification precision, achieving a significant reduction in span values (< 0.2) and enhancing both powder flowability and film quality. Secondly, a novel secondary air guider (SAG) modification for ACM cyclones was introduced, effectively reducing ultra-fine and ultra-large particle fractions, as confirmed by experimental tests and computational fluid dynamics (CFD) simulations.

Beyond system optimization, this study introduces two novel PSD parameters, E5 and E80, to capture the distinct impacts of ultra-fine and ultra-large particles. Comprehensive experiments revealed that E5 strongly correlates with powder flowability and deposition efficiency, while E80 mainly affects thickness uniformity and film performance.

Finally, this research pioneers the application of machine learning in the powder coating domain. Artificial neutral network (ANN) was utilized to systematically quantify the intricate relationships between PSD and the gloss and surface roughness (Ra) of coating films. The analysis revealed that ultra-large particles negatively correlate with gloss and positively correlate with Ra, indicating that ultra-large particles significantly cause gloss loss and increase surface roughness. On the contrary, higher fractions of ultra-fine particles contributed to smoother surface and higher gloss values. These findings aligned with experimental results, highlighting that minimizing ultra-large particles (smaller E80) is essential for enhancing surface aesthetics and achieving smoother coatings.

This dissertation contributes to the powder coating research field by advancing PSD control methodologies, introducing innovative classification techniques, and pioneering the use of machine learning for analysis. These findings offer practical pathways for improving the quality of powder coatings, with implications for broader industrial adoption.

Summary for Lay Audience

Powder coatings are widely used in products such as car parts, household appliances, and construction materials due to their environmentally friendly nature, durability, and absence of harmful solvents. However, achieving high-quality powder coatings involves significant challenges. The particle size and particle size distribution (PSD) — the uniformity of particle sizes — are critical factors affecting application, adhesion, and appearance. A broad PSD, with excessive ultra-large or ultra-fine particles, can lead to poor flowability, inferior surface defects and deteriorated film performances.

This research focuses on addressing these challenges by enhancing the production process of powder coatings. The key components of air classifying mills (ACMs), commonly used in the industry for grinding and classifying particles, were redesigned to improve their performance. A novel design of classifiers and the introduction of secondary air guides significantly improved the uniformity of particle sizes. These advancements led to coatings with narrower particle size distributions, improved flowability, and enhanced surface quality.

Additionally, the effects of particle size distribution on powder coating appearance were systematically analyzed through experiments and machine learning methods. The findings revealed that ultra-large particles (> 80 µm) tend to cause surface roughness and loss of gloss, while ultra-fine particles contributed to smoother surfaces and higher gloss. By identifying these relationships, strategies were developed to optimize particle size distributions for specific applications, improving coating efficiency and quality.

This work not only improves the performance and sustainability of powder coatings but also provides valuable insights into optimizing production methods. The advancements have the potential to reduce waste, improve production efficiency, and support the broader application of powder coatings across industries such as automotive, construction, and consumer goods. These innovations represent a step forward in creating high-performance, environmentally sustainable coating solutions.

Download

Available for download on Sunday, December 20, 2026

Share

COinS