Generation and Assessment of Random Surface Texture over a Wide Area

Abstract

Product surface textures are designed to improve their aesthetic, tactile, and mechanical quality. Surface texture manufactured with microrough patterns over a wide area differs from common geometric product machining. We have proposed generating a surface texture with regular patterns by milling. Here we propose generating a random-pattern surface texture using image processing. Digital surface-texture data consists of “real” three-Dimensional (3D) machining information. Wide-area digital surface-texture data such as scattered point data, Initial Graphics Exchange Specifications (IGES), and Standard Triangulated Language (STL) require humongous memory. The complexity and area of surface texture processed to generate tool paths is limited by computational considerations and generating the tool path for a widearea surface texture is time-consuming, so we propose generating random wide-area-pattern surface texture without the need for wide-area digital texture data. Instead, this uses only wide-area image data and narrowarea digital data. A wide-area tool path is generated by image quilting, which creates a patchwork in which patches represent both image and digital data for narrow-area surface texture, reducing surface distortion for patch boundaries. This paper introduces the generation of random pattern texture and machined samples assessing patch-boundary distortion.