Cutting data for automated turning tool selection in industry
This thesis is concerned with the determination of cutting parameters (cutting speed, feed rate and depth of cut) in turning operations within an industrial environment. The parameters are required for the purposes of tool selection, working with a variety of batches of different materials. Previous work of this nature, little of which has been transferred into industry, has concentrated primarily on deriving optimum cutting conditions, based on a variety of deterministic and non- deterministic approaches, with a general reliance on experimentally-derived input variables. However, this work is only suited to tool selection for a single material. Under industrial conditions tools will frequently need to be selected for more than one material, in tool/material combinations not recommended by tool makers. Consequently, the objective of the research described in this thesis was to employ existing cutting data technology and to use it as the basis for a cutting data system, suitable for multi-batch tool selection. Two companies collaborated in this work, by making available suitable personnel and the provision of shop floor facilities on their premises. The initial work concentrated on the development of an algorithmic model, based on established theory. This was then tested industrially, using the concept of shop floor approved data as a substitute for optimum cutting data. The model was found to work reasonably, but required further development to make it suitable for multi- batch tool selection. This development took three main forms: a) a reduction of input data, particularly in the number of experimentally-derived variables, b) the removal of the tool/material-specific constraints traditionally used in cutting data optimisation, c) a method of data correction based on adjustment of the mean and standard deviation of the data. Further industrial testing was carried out using the resulting system. It was demonstrated that it was possible for a relaxed system with reduced input variables and appropriate data correction to function within an industrial environment.