2009 年 2009 巻 54 号 p. 49-52
Shot peening coverage is conventionally and typically measured by visual comparison with the standard image of a peened almen strip surface under 20 to 30 power magnification. There, however, are problems that indented terrain on an actual peened surface may not always give a clear enough image to compare, and that visual comparison may contain individual-oriented variability in measurement results. This article shows that probability density distributions of the peened almen strip surface profile agree with Gaussian mixture distributions, and that separation of the Gaussian mixture distributions enables coverage computation, and thereafter suggests a quantitative coverage measurement method that is independent of individual variability.