Page 1089 - The Mechatronics Handbook

With respect to tool fracture, the system should be able to detect both small fractures, “chipping” phe-
nomena, and catastrophic failure of a tool. Although prediction of such failures would be desirable, it is
problematic whether this is a practical goal, at least in the near future. The number of variables, which
determine the actual occurrence of tool fracture together with their complex interactions, and in many
instances their underlying stochastic nature, make reliable prediction capabilities, at best, a long-term
prospect in tool monitoring systems.
Cutting Tool Failure Monitoring Techniques
Tool condition monitoring systems are based upon either direct or indirect methods of quantifying the
magnitude of tool failure.
The direct methods are those that utilize effects caused directly by tool failure. The direct methods,
usually performed by means of optical, radiometric, pneumatic, or contact sensors can be effectively
applied to the offline measurement of tool wear or breakage. However, such direct means of measuring
tool failure have generally been found to be difficult to apply in practical shop floor applications. This
is particularly true in those situations requiring online (real-time) monitoring capability.
Indirect methods of sensing tool failure depend upon the measurement of parameters, which are
indirectly related to the condition of the cutting edge. For example, the cutting forces generated during
a machining operation are dependent upon the condition of the tool’s cutting edge. Generally, as the tool
edge wears the generated cutting forces increase. Thus, measurement of the cutting forces present during
a machining operation provides an indication of the tool condition, i.e., increasing cutting forces indicates
increasing wear. In reality the relationship can be very complex. Other parameters that have been studied
to determine their suitability as indicators of cutting tool failure include spindle motor current, acoustic
emissions, cutting tool temperature, and noise and vibration signals. It is also possible to measure cutting
forces directly and then relate these values to the condition of the cutting tool. In fact, this is one of the
more common indirect tool wear monitoring methods. It has been reported that cutting force signals are
more sensitive to tool wear than vibration or power measurements. The general reliability of force mea-
surements is another reason for their popularity in tool condition monitoring applications. To use cutting
force measurements for practical tool monitoring systems, there is a need to relate these forces to the
state of tool condition online. However, since the measured cutting forces are affected by both cutting
edge condition and changes in cutting conditions (feed rate, cutting speed, and depth of cut), the detection
of tool failure using measurements of these forces becomes quite challenging in practice.

System Characteristics
Whether a tool condition monitoring system employs direct or indirect measures of tool failure (auto-
mated, computer-based system or not), it must include a number of common features if it is to be truly
practical. Figure 39.4 shows the block diagram of a generalized tool condition monitoring system (Braun,
1986).
In the measurement section, the physical parameter (or possibly, parameters) of interest is converted
to a form that is appropriate for further manipulation by the system (generally a digitized representation
of an electric analog signal).
Within the processing section, various techniques are implemented in order to suppress noise, compress
information, and emphasize important features of the acquired signal. Typical methods include analog

FIGURE 39.4 Block diagram of a generalized tool condition monitoring system.

©2002 CRC Press LLC