Research on Predicting Model of Surface Roughness for Ground Engineering Ceramics

Grinding is the mostly leading machining technology for engineering ceramics. The quality of ground surface can be evaluated by various roughness parameters. And a textural analysis method based on Gray Level Co-occurrence Matrix was employed in researching the roughness of ground ceramics. The relationship between texture features and roughness was investigated through a series of surface images of engineering ceramics collected by a digital microscope. The sampling offset and total gray levels of surface images were determined firstly. Then 4 GLCMs were built up to calculate the average of texture features. And 6 parameters were fixed as main texture parameters. Furthermore, the multiple nonlinear regression theory was used to assess the relation between the texture features and roughness Ra. By statistic test and comparison, the deviation of calculated Ra and actual Ra is less than 0.25. It is shown that this relation is much satisfactory and the method may be suitable for quickly measuring the roughness of ground ceramics.