Machining vibrations - How can they be identified and addressed

In 1907, in his seminal work on the art of metal cutting, Frederick Winslow Taylor said: "Chatter is the most obscure and delicate of all problems facing the machinist". In 2018, studies on the subject reported that vibration remains a major issue for machining, and that many manufacturers still identify it as one of the most limiting factors in the process.

The costs associated with this problem are rarely quantified, and the machinist naturally anticipates them, especially when machining particularly flexible parts or tools.

The car manufacturer Renault was able to put a figure of precisely three million per year on this cost for cylinder blocks machined in series. In this case, the extra cost was linked to premature tool wear, and in 2002, for example, represented exactly €0.35 per part, or €120,000 per year.

Estimates show that the majority of extra costs are linked to lost productivity and time lost for set-ups or rework, followed by tool and machine wear, and finally scrap parts.

The solutions found by machinists are often the result of trial and error and the fruit of experience: modify the speed, change the tool, increase the number of passes, clamp the part differently, fit rubber elements, etc. The result is a significant reduction in productivity.

The theory of "stability lobes", which appeared in the 1950s, seemed to provide a global solution, but it's clear that it's not so easy to apply and doesn't solve the majority of problems.

So, as Taylor already said, there is still a crying lack of robust predictive methods and a lack of overall logic to tackle the problem in concrete terms.

The first objective of this article is to present the many facets of the problem, which is most often characterized by characteristic vibration noises and degraded surface finish, but also by tool breakage or premature spindle wear. It is important to be able to pinpoint these events and their links with vibratory phenomena, in order to get to the root of the problem. It also shows how these vibrations can be measured in practice, and how they can provide early warning indicators to anticipate problems.

The second objective is to present models of the phenomenon of chatter, which is a particular category of machining vibration (the one most often involved in the problems mentioned), in order to highlight the main influencing parameters. Machining is an operation that involves a large number of parameters, but it is possible to group them together, prioritize them and identify the parameters to be systematically monitored.

The third aim of this article is to show that there are a number of tried-and-tested...