فحص بلى عدة القطع المعان بالرؤيا الحاسوبية خلال عملية الخراطة == Computer Vision Aided Inspection of Cutting Tool Wear In Turning
Author name:
علي كريم علوان العكيلي
Supervisor name:
علي عبار خليف
General topic:
Production engineering metallurgy and materials
Specific topic:
Production Engineering
Degree:
Master
University:
University of Technology
Language:
English
University location:
Baghdad
First pages:
50T166 - p.pdf
Abstract:
في الوقت الحاضر , تحظى مكائن القطع المبرمجة باھمية كبيرة في المعامل والورش التصنيعية بسبب دقتھا العالية ومرونتھا لكن لسوء الحظ لا تزال ھذه المكائن غير قادرة على التاكد من جودة السطوح المنتجة. الاتجاه نحو اتمتة عمليات القطع سير للحاجة الملحة للحصول على ج | Nowadays, computerized numerically controlled (CNC) machines have high importance in manufacturing factories and workshops due to their high accuracy and flexibility. Unfortunately they still cannot ensure the quality of machined products. The trend towards automation in machining has been driven by the need to maintain high product quality with improving production rate, these improvements can be possible by monitoring and control of machining process. In this research a vision based monitoringsystem was introduced for the on - line direct automated measurement of the cutting tool flank wear width based on a new algorithm on the basis of canny edge detection and morphological operations for the captured images.Twelve specimens were prepared using STARCHIP CNC turning with four cutting speeds (57,98,110 and 125 m/min) and three depths of cut (0.5,1 and 2 mm) and constant feed of (0.04 mm/rev). Maximum and minimum absolute error and error percentage in the maximum flank wear width was (0.058,0.002 mm) (9.7,0.54 %) respectively. The results show the effectiveness of the proposed method in the monitoring of the tool condition.Also, Vision based monitoring system was proposed for the measurement of tool nose wear which have direct interaction with the workpiece during machining. Measurement operation is done by comparing reference and worn tool images; the algorithm is composed of image preprocessing, Otsu thresholding, conforming method for the exact alignment between the two images and image subtraction is performed in order to detect the nose wear area. Morphological opening by reconstruction was used to remove isolated foreground pixels result in due to the quantization errors. The maximum and minimum error percentage in tool nose wear area (6.72,0.48 %) respectively; the experimental results show that nose wear occurs after the flank wear exceeded its maximum standard limit in roughing operations Keywords : tool wear, flank wear, nose wear image processing, automated visual inspection, tool condition monitoring.
