Tool deflection detection and error analysis with image processing method in machining

In CNC machines, which have an important place in the machining process, proactive approaches that provide the opportunity to intervene during the operation come to the fore in order to ensure the quality of the parts produced. Along with this process, the development of imaging and computer technology provides the opportunity to take control of the products during the manufacturing phase and to produce them correctly at once. However, there are some factors that result from variable machining parameters, tool wear and material internal structure and cause surface and form errors during machining. It is seen that these factors, which cause time and material loss in the manufacturing sector, where mass production gains more importance due to increasing consumption, are big problems that need to be solved before the product leaves the counter. These factors, which adversely affect the machining process, increase the forces that the cutting tools are exposed to in general, disrupt the standard working conditions and reduce their cutting capabilities. In this context, one of the disruptive factors frequently seen in machining is tool deflection, which reduces the ability to remove metal and occurs as a result of forces opposite to the machining direction of the tool. Tool deflection, which causes surface, form and geometric errors in machining, is a factor that must be detected during manufacturing and eliminated by changing the machining parameters. In the study, tool deflection measurement with the image processing method has been proposed as an alternative to the currently used tool deflection measurement methods, with its simultaneous, fast and low cost features. Within the scope of the thesis, a method has been developed that instantly detects the tool deflection seen in CNC milling and hole drilling processes, which is a common application in machining methods, by image processing method. In the developed method, the cutting tool was recorded with an industrial camera during the milling and drilling operations. Force measurements were made simultaneously. By using the image processing method, the amount of bending was calculated as a result of the opposite forces that the tool is exposed to. The obtained data were compared with the finite element analysis (FEA) results modeled by force measurement. The model was applied to the femoral component of the knee prosthesis used in the medical field, and the results were interpreted. The thesis work was supported by the BAP unit of İnönü University. This study is a guide for researchers who want to obtain numerical data from dynamic images in Mechanical Engineering applications.