Integração de visão computacional e fotogrametria para medição automática de anéis de pistão

Abstract

The piston ring is a critical component in internal combustion engines whichs plays a vital role by sealing the gap between pistons and cylinder walls. This sealing is indispensable to prevent gas leakage from the combustion chamber and to adequately control lubricant distribution. With the evolution of engines, operating with increasingly compression ratios and lower lubricant viscosities to maximize performance and effiency, precision in piston ring manufacturing has become imperative. Thus, the manufacturing of these elements entails designing rings that meet the complexities of the combustion system, manufactured with rigorous dimensional accuracy. In this manufacturing process, ensuring that the rings conform to the project’s error margins is essential to avoid rejections during quality analysis and subsequent rework. This necessitates, in conjunction with process qualification, a reliable measurement system to optimize production and guarantee highquality products. In this context, this study examines the presently employed manual measurement method and proposes a distinct automated approach aimed at enhancing measurement consistency and reducing waste. The central proposition is the development of an alternative measurement method based on image processing. This approach offers several advantages, including optimized automation, process agility, and the absence of direct contact with the measured components. The research explores the design of a machine vision system that measures piston rings through images, ensuring the required precision. The study integrates established computer vision techniques with novel contributions, including modifications in subpixel edge detection, estimation of contour angles using zero phase filters, and adjustments in the system’s execution flow. Additionally, the method employs photogrammetry techniques to map image points to the real world, correcting image distortions, inclinations, and relative distances between measurement points and the camera, as well as accounting for part height, throught the implementation of a proposed height compensation method. The developed prototype is validated through tracking tests using standard blocks and comparisons with conventional equipment. The results attest to the prototype’s ability to maintain tracking across various size, height, position, and inclination variations. Remarkably, the proposed automated method stands out compared to manual measurements, particularly in assessing ring opening. Ultimately, the implementation of this innovative method has the potential to substantially optimize the piston ring manufacturing process, enhancing consistency and agility to meet the essential quality requirements for these vital components in internal combustion engines.