.jpg)
In mass production environments utilizing machining processes, a significant challenge arises when tools break during cutting operations. Often, the breakage goes unnoticed, leading to the continuous production of defective products due to incorrect machining.
By monitoring the condition of tools and detecting breakages in real time, it is possible to minimize the production of defective items.
MAZIN, Inc. is committed to developing AI that solves production challenges on a process-by-process basis. In machining, we focus on issues such as skill transfer and production efficiency improvement. We are working on developing algorithms for tool condition monitoring by clamping current sensors to machine tools and analyzing current values, which are correlated with cutting torque, as part of various experiments and analyses.
Machining often involves aluminum as the workpiece material. We introduce our efforts in experimenting to capture tool breakage during the machining of aluminum.
The purpose of this initiative is to determine whether tool breakage anomalies can be detected through the analysis of current waveforms obtained using current sensors during the machining of aluminum.
Experiments were conducted under the following environment and machining conditions using a new tool, a slightly damaged tool (minor breakage), and a significantly damaged tool (major breakage) to machine aluminum. The goal was to capture waveform data with a current sensor and identify differences in the current waveforms between the new tool and the tools with minor and major breakages.
.png) |
.png) |
.png) |
| New Tool | Slightly Damaged Tool (Minor Breakage) | Significantly Damaged Tool (Major Breakage) |
Upon overlaying the current waveforms obtained from machining with the new tool, the slightly damaged tool, and the significantly damaged tool, it was observed that compared to the current waveform of the new tool (represented by a blue line), the waveforms for the slightly damaged tool (orange line) and the significantly damaged tool (green line) shifted towards higher values.
Although the differences in values were small, the results suggest a high possibility of detecting anomalies using a uniquely developed algorithm. This indicates that even minor variations in the current waveforms.
.png)
In this experiment, using a tool with a diameter of φ6mm, which produces relatively low cutting torque, we observed differences in the current waveforms. It is anticipated that as the tool diameter increases, leading to higher cutting torque, the differences in current values caused by tool breakage will become more pronounced. This suggests that tool breakage during the machining of aluminum with tools larger than φ6mm in diameter could also be detectable.
Moving forward, we plan to collect data under conditions closer to those of actual machining environments to accumulate insights into the feasibility of anomaly detection.
For more detailed information about the technology or inquiries related to research and development, please contact us.