Silicon nitride is a ceramic material that is widely used in engineering applications due to its excellent mechanical properties. Indeed, silicon nitride is a popular material for cutting tools due to its high wear resistance, high hardness, and low coefficient of friction. It is often used in high-speed machining applications where high temperatures and pressures are generated. Moreover, silicon nitride is used in the manufacturing of bearings due to its excellent wear resistance and low coefficient of thermal expansion. It is particularly suitable for use in high-temperature and corrosive environments. Because it is biocompatible, it is commonly used in the manufacturing of biomedical implants, such as dental implants. Furthermore, silicon nitride is used in the manufacturing of electronic components, including integrated circuits and semiconductors. Silicon nitride also is widely used in the aerospace industry due to its high strength, thermal stability, and wear resistance. It is often used in the manufacturing of turbine components, such as blades and vanes, for jet engines.
One of the most important properties of silicon nitride is its fracture toughness, which is a measure of its resistance to crack propagation. Fracture toughness is defined as the ability of a material to resist the growth of a crack when a load is applied. It is a critical property in many engineering applications where the material is subjected to high stresses and may be susceptible to cracking. Silicon nitride has a high fracture toughness, which makes it an excellent material for use in applications where high strength and durability are required. The fracture toughness of silicon nitride is influenced by several factors, including its microstructure, grain size, grain-boundary characteristics, and processing conditions. In general, silicon nitride with elongated grains has higher fracture toughness than materials with fine grains. This is because elongated grains act as barriers to crack propagation. Another important factor that affects the fracture toughness of silicon nitride is the presence of residual stresses. Grain-boundary characteristics are primarily affected by the manufacturing process, and they can be significantly responsible for the fracture toughness. Therefore, it is essential to carefully control the manufacturing process to optimize the grain boundary bond strength and residual stresses in the material. Furthermore, silicon nitride’s resistance to crack propagation and failure is dependent on its fracture toughness, however this property has been challenging to analyze.
In a new study published in the peer-reviewed Journal of the American Ceramic Society, Scientists from the National Institute of Advanced Industrial Science and Technology (AIST) in Japan: Drs. Ryoichi Furushima, Yutaka Maruyama, Yuki Nakashima, Minh Chu Ngo, Tatsuki Ohji, and Manabu Fukushima used deep learning to evaluate the fracture toughness of silicon nitride which is a substance with outstanding mechanical qualities, thermal stability, and resistance
In order to create functional parts and structures, engineers need information on a material’s fracture toughness. Therefore, extensive testing is required for traditional techniques of determining fracture toughness, which is very time-consuming to provide correct conclusions due to the material’s complex microstructure. Convolutional neural networks (CNNs) are one kind of the deep learning family of learning algorithms. The goal of this research was to find a new way to employ convolutional neural networks (CNNs) to circumvent these limitations.
As a result, due to its ability to quickly and accurately analyze large data, convolutional neural networks (CNNs) have become an invaluable tool for image processing and recognition. The research team were able to develop a model that predicted silicon nitride’s fracture toughness by studying the material’s microstructural features, such as the size, shape, and distribution of the grains by using a CNN’s. By employing this novel approach, the authors were able to give precise data without the need for time-consuming and costly experimental approaches.
The impact of the AIST researchers’ study have broad ramifications for materials science and the industries that use modern ceramics. Researchers may be able to considerably quicken the development of this promising material if they use CNNs to anticipate the fracture toughness of silicon nitride. Because of this, the material may be put to greater use in a variety of fields. The approach used in this study may also be applied to the examination of other materials, opening the door to exciting new research and breakthroughs in the field of materials science.
Moreover, the use of AI and deep learning techniques such as CNNs has the potential to significantly modify the way we do research and development for novel materials. Researchers and engineers may benefit from a deeper understanding of the complex relationships between a material’s microscopic structure and its characteristics because of these state-of-the-art computational methods. This might result in the creation of new materials with specialized properties for a particular application.
In summary, authors’ work represents a significant advancement in our understanding of silicon nitride and the evaluation of its fracture toughness. In addition to streamlining and speeding up the assessment process, the novel use of CNNs to forecast fracture toughness based on microstructural traits shows potential for broader applications in materials research. These advantages may be ascribed to the usage of CNNs in the review process. In a statement to Advances in Engineering, first author Dr. Ryoichi Furushima explained “this study is a nice illustration of how advanced computational techniques may spur innovation and fill in the blanks in current issues. It offers as a great illustration of how cutting-edge computational techniques may spur innovation as we continue to push the limits of what is feasible with advanced ceramics”.
Ryoichi Furushima, Yutaka Maruyama, Yuki Nakashima, Minh Chu Ngo, Tatsuki Ohji, Manabu Fukushima, Fracture toughness evaluation of silicon nitride from microstructures via convolutional neural network , Journal of the American Ceramic Society, 2023;106:817–821.