Tool wear condition monitoring is an important component of mechanical processing automation, and accurately identifying the wear status of tools can improve processing quality and production efficiency. This paper studied a new deep learning model, to identify the wear status of tools. The force signal was transformed into a two-dimensional image using continuous wavelet transform (CWT), short-time Fourier transform (STFT), and Gramian angular summation field (GASF) methods. The generated images were then fed into the proposed convolutional neural network (CNN) model for further analysis. The calculation results show that the accuracy of tool wear state recognition proposed in this paper was above 90%, which was higher than the accuracy of AlexNet, ResNet, and other models. The accuracy of the images generated using the CWT method and identified with the CNN model was the highest, which is attributed to the fact that the CWT method can extract local features of an image and is less affected by noise. Comparing the precision and recall values of the model, it was verified that the image obtained by the CWT method had the highest accuracy in identifying tool wear state. These results demonstrate the potential advantages of using a force signal transformed into a two-dimensional image for tool wear state recognition and of applying CNN models in this area. They also indicate the wide application prospects of this method in industrial production.

The initial reaction mechanism of energetic materials under impact loading and the role of crystal properties in impact initiation and sensitivity are still unclear. In this paper, we report reactive molecular dynamics simulations of shock initiation of 1,3,5-trinitroperhydro-1,3,5-triazine (RDX) crystals containing a cube void. Shock-induced void collapse, hot spots formation and growth, as well as spalling are revealed to be dependent on the shock velocity. The void collapse times are 1.5 and 0.7 ps, for the shock velocity of 2 and 4 km·s^–1, respectively. Results indicate that the initial hot spot formation consists of two steps: one is the temperature rise caused by local plastic deformation and the other is the temperature increase resulting from the collision of upstream and downstream particles during the void collapse. Whether hot spots will continue to grow or quench depends on sensitive balance between energy release caused by local physical and chemical reactions and various heat dissipation mechanisms. In our simulations, hot spot would grow for U_p = 4 km·s^–1; hot spot is weak to some extent for U_p = 2 km·s^–1. The tensile wave reflected by the shock wave after reaching the free surface causes the spalling, which depends on the initial shock velocity. Typical spalling occurs for the shock velocity 2 km·s^–1, while the tensile wave induces the microsplit region in RDX crystals in the case of U_p = 4 km·s^–1. Chemical reactions are studied for Rankine–Hugoniot shock pressures P_s = 14.4, 57.8 GPa. For the weak shock, there is almost no decomposition reaction of the RDX molecules near the spalling region. On the contrary, there are large number of small molecule products, such as H₂O, CO₂, NO₂, and so forth, around the microsplit regions for the strong shock. The ruptures of N–NO₂ bond are the main initial reaction mechanisms for the shocked RDX crystal and are not affected by shock strength, while the microsplit slows down the decomposition rate of RDX. The work in this paper can shed light on a thorough understanding of thermal ignition, hot spot growth, and other physical and chemical phenomena of energetic materials containing voids under impact loading.