This paper describes an experimental investigation of machine tool chatter, in which the machine tool structure was replaced by a model, two-degrees-of-freedom structure with adjustable and consistent vibration characteristics. Primarily non-regenerative chatter, and secondarily regenerative chatter, were investigated for orthogonal cutting of an aluminium alloy with both conventional and restricted-contact cutting tools. The results are presented in the form of stability charts; these show the limiting widths of cut which can be machined without chatter, for given sets of machining and structural conditions. For non-regenerative chatter, it was found that the limiting width of cut: increases with a decrease in the structure's cross-receptance between the directions normal and tangential to the cut surface; increases with a decrease in cutting speed, but in a manner depending on the structural characteristics; is substantially independent of the mean undeformed chip thickness; increases by at least 25 per cent if contact is restricted to a length approximately equal to the undeformed chip thickness. For regenerative chatter it was found that the limiting width of cut: was approximately one half of the limiting width for non-regenerative chatter, for the otherwise similar machining and structural conditions investigated; increases with a decrease in cutting speed; increases by at least 25 per cent if contact is restricted to a length approximately equal to the undeformed chip thickness. Theoretical predictions of non-regenerative chatter with a conventional tool, based on independent measurements of machining force oscillations during tool vibration, agree well with experimental results. For regenerative chatter with a conventional tool, the theory was based on the superposition of machining force oscillations arising from tool vibration and from removing a wavy surface. The predictions were in error at low cutting speeds, indicating that the force oscillations are not superposable at this condition.
