Automatic leveling procedure by use of the spring method in measurement of three-dimensional surface roughness

Leveling of specimen surfaces is very important in measurement of surface roughness. If the surface is not leveled, the measured roughness has large distortion and less vertical measurement range. It is convenient to utilize some automatic leveling procedures instead of manual leveling which needs longer adjustment time. In automatic leveling, a new algorithm is proposed, which is named the spring method superior to the least square method. The spring method has an advantage that a part of tentative data points is used to calculate the surface inclination, so the obtained results are less influenced by local pits for example. As examples, the spring method was applied to actual engineered surfaces, which were milled, shot-peened, and ground surfaces, and also an artificial ditched surface. The results went well for the calculation of the surface inclinations and consequently the specimen surfaces were leveled with less distortion and large vertical measurement range can be achieved. It is also found the least square method is a special case of the spring method with using all sampling data points. That means the spring method is a comprehensive procedure including the least square method. This must become a very strong and robust method in automatic leveling algorithm.