arXiv subjects

Hannah Zhang

Publications and source records attributed to Hannah Zhang.

4 recordsLinked to original sources

arXiv2026

Towards standardisation of average grain size measurement of additively manufactured microstructures using EBSD

Additively manufactured (AM) alloys have heterogeneous microstructures with broad grain size distributions and highly anisotropic and/or non-convex grain shapes. AM components can have complex geometries and porosity which may affect the local microstructure. Currently there is no electron backscatter diffraction (EBSD)-based grain size measurement standard suitable for typical AM materials. An interlaboratory comparison study was conducted to find out what grain size metrics and summary statistics are currently used to describe average grain size. Participants were asked to measure and report the average grain size from the same EBSD map dataset. Detailed reports have been published in Reference [1]. Based on these results, we have tested and propose recommendations for a new standard for measuring average grain size in AM materials. The present work demonstrates the suitability and limitations of the proposal across several different Ni and Al AM components.

physics.app-ph↗

arXiv2025

Nemotron-Flash: Towards Latency-Optimal Hybrid Small Language Models

Efficient deployment of small language models (SLMs) is essential for numerous real-world applications with stringent latency constraints. While previous work on SLM design has primarily focused on reducing the number of parameters to achieve parameter-optimal SLMs, parameter efficiency does not necessarily translate into proportional real-device speed-ups. This work aims to identify the key determinants of SLMs' real-device latency and offer generalizable principles and methodologies for SLM design and training when real-device latency is the primary consideration. Specifically, we identify two central architectural factors: depth-width ratios and operator choices. The former is crucial for small-batch-size latency, while the latter affects both latency and large-batch-size throughput. In light of this, we first study latency-optimal depth-width ratios, with the key finding that although deep-thin models generally achieve better accuracy under the same parameter budget, they may not lie on the accuracy-latency trade-off frontier. Next, we explore emerging efficient attention alternatives to evaluate their potential as candidate building operators. Using the identified promising operators, we construct an evolutionary search framework to automatically discover latency-optimal combinations of these operators within hybrid SLMs, thereby advancing the accuracy-latency frontier. In addition to architectural improvements, we further enhance SLM training using a weight normalization technique that enables more effective weight updates and improves final convergence. Combining these methods, we introduce a new family of hybrid SLMs, called Nemotron-Flash, which significantly advances the accuracy-efficiency frontier of state-of-the-art SLMs, e.g., achieving over +5.5% average accuracy, 1.3x/1.9x lower latency, and 18.7x/45.6x higher throughput compared to Qwen3-1.7B/0.6B, respectively.

cs.LG↗

arXiv2023

Estimation of fatigue life of TiN coatings using cyclic micro-impact testing

This paper studies the behaviour of a thin titanium nitride (TiN) coating (1.5 um thick) on a tool steel substrate material under dynamic and cyclic impacts through an approach combining experimental testing and computational modelling. Dynamic impact testing was used to investigate the load-dependent dynamic hardness and assess the energy-dissipation capabilities of the coating system. In cyclic impact tests, the materials experienced permanent plastic deformation in each cycle, ultimately leading to coating failure. Chemical analysis identified an interlayer between the coating and the substrate, while cross-sectional analysis revealed the extent of coating damage due to cycling and impact load. A three-dimensional map was constructed, connecting the acceleration load, sensed depth, and cycles to the coating failure, and an empirical equation used to characterize the relationship between the depth and cycles to failure. The computational model examined the traction component distribution during loading and unloading, with a focus on normal and shear tractions. These findings suggested the potential significance of normal traction in the interfacial fatigue failure due to impact.

physics.app-ph↗

arXiv2022

Deformation mechanism of WC single crystals under nanoindentation: Effects of surface defects and orientation on pop-in and hysteresis

Nanoindentation was carried out on pure tungsten carbide (WC) on the basal (0001) and prismatic (1010) planes, using Berkovich and spherical indenters, in both single load and multi-load testing. The work focuses on correlating the load-displacement curves, including elastic to plastic deformation, size effect and hysteresis with the deformation behaviour of WC. With different specimen preparation processes, the elastic to plastic deformation started at different threshold loads: This observation was found to be due to the variation in surface dislocation density. Staircase deformation was observed thought to be caused by dislocation motion and the formation of slip bands; sudden displacement discontinuities in the load-displacement response - associated with dislocation loop nucleation - occurred at, or near the theoretical shear strength. Furthermore, discontinuities in load-displacement curves were also used to confirm that hysteresis loops were a result of plastic deformation, as they when the loading was purely elastic.

cond-mat.mtrl-sci↗