Session: Student Poster Competition

Submission Number: 185605

Reproducibility and Human Error in Hardness Measurements: A Comparative Study of Independent User Data 

This study examines the influence of human error on hardness measurements of (Laser Impact Welded) LIW samples by evaluating the reproducibility and consistency of results obtained by two independent users. The work addresses a broader need within materials characterization to understand operator-dependent variability, particularly in experimental techniques that involve manual procedures and are sensitive to local microstructural heterogeneity. Because hardness testing is often used as a proxy for mechanical performance and microstructural evolution, ensuring measurement reliability across users is critical for both research and industrial applications.

Hardness testing was performed independently by both users following the same testing protocol, and the resulting datasets were statistically analyzed using paired two-sample t-tests at a 95% confidence level. This approach allowed for direct comparison of user measurements while accounting for paired experimental conditions. Of the six datasets evaluated, only one dataset—corresponding to the Titanium to Mg-6Al interface—exhibited a statistically significant difference between users (p = 0.0374). The remaining five datasets demonstrated strong agreement, with overlapping uncertainties and no statistically meaningful differences, indicating a high degree of reproducibility.

Measurement uncertainty was quantified by comparing the differences between mean hardness values obtained by each user. The largest discrepancy was observed in the Ti to Mg-9Al dataset, while the smallest difference occurred in the Al to Mg-6Al dataset. Potential contributors to variability include differences in manual indentation technique, equipment setup and alignment, surface preparation quality, material defects, calibration conditions, elastic recovery effects, and indentation spacing relative to microstructural features.

Overall, the results suggest that the hardness testing method is largely robust and reproducible across operators, while also highlighting specific material interfaces where user-dependent variability may arise. These findings support more confident interpretation of hardness data when assessing microstructural changes in LIW systems.

Presenting Author: Kai Hubbard North Carolina A & T State University

Presenting Author Biography: Kai Hubbard is a graduate student in the Department of Mechanical Engineering at North Carolina Agricultural and Technical State University, where he also earned his Bachelor of Science degree in Mechanical Engineering during the Summer of 2023. His current research focuses on maximizing efficiency and accuracy in material evaluation by using the Vickers Hardness (VH) test—the simplest, most cost-effective, and fastest method available. This approach enables the precise measurement of mechanical behavior, structural changes, and data-driven prediction modeling, ensuring reliable insights for optimizing material performance. This work aims to develop 3D crystal plasticity models that accurately depict the deformation of metals and subsequently optimize their microstructure. This research integrates the Vickers Hardness Test, Excel Statistical analysis, and Electron Backscatter Diffraction (EBSD).
Kai’s interest in engineering started with Hot Wheels toy cars, next Remote Controlled (R/C) Cars, and then full-sized automobiles. He is fascinated by how machines work and how efficient they have become. As a member of the HAMMER ERC and a researcher in Dr. Sankar’s lab, he has contributed to multiple projects, including Laser Impact Welding, the process of rolling, polishing, and studying the microstructure of lightweight metals. In the future, he wants to work for Toyota and then plans to open his car shop to service his community.

Authors:

Kai Hubbard North Carolina A & T State University
Dr. Svitlana Fialkova North Carolina A & T State University