The response of a mechanical system depends on “effort” excitations (forces and torques) applied to the system.
Many applications exist in which process performance is specified in terms of forces and torques. Examples include machine-tool operations, such as grinding, cutting, forging, extrusion, and rolling; manipulator tasks, such as parts handling, assembly, engraving, and robotic fine manipulation; and actuation tasks, such as locomotion.
The forces and torques present in a dynamic system are generally functions of time. Performance monitoring and evaluation, failure detection and diagnosis, testing, and control of mechatronic systems can depend heavily on accurate measurement of associated forces and torques. One example in which force (and torque) sensing can be very useful is a drilling robot.
The drill bit is held at the end effector by the gripper of the robot and the work piece is rigidly fixed to a support structure by clamps.
Although a displacement sensor (such as a potentiometer or a differential transformer) can be used to measure drill motion in the axial direction, this alone does not determine the drill performance. Depending on the material properties of the work piece (e.g., hardness) and the nature of the drill bit (e.g., degree of wear), a small misalignment or slight deviation in feed (axial movement) or speed (rotational speed of the drill) can create large normal (axial)
and lateral forces and resistance torques.
This can create problems such as excessive vibrations, uneven drilling, excessive tool wear, and poor product quality and eventually may lead to a major mechanical failure.
By sensing the axial force or motor torque, for example, and using the information to adjust process variables (speed, feed rate, etc.), or even to provide warning signals and eventually stop the process, can significantly improve the system performance.
Another example in which force sensing is useful is in nonlinear feedback control (or, feedback linearization technique or FLT) of mechanical systems such as robotic manipulators. Since both force and torque are effort variables, the term force may be used to represent both these variables.
This generalization is adopted here except when discrimination might be necessary.
Force Causality Issues
One important application of force (and torque) sensing is in the area of control. Since forces are variables in a mechanical system, their measurement can lead to effective control.
There are applications in which force control is invaluable. This is particularly evident in situations where a small error in motion can lead to the generation of large forces, which is the case, for example, in parts assembly operations. In assembly, a slight misalignment (or position error) can cause jamming and generation of damaging forces. As another example, consider-high precision machining of a hard work piece.
A slight error in motion could generate large cutting forces, which might lead to unacceptable product quality or even to rapid degradation of the machine tool. In such situations, measurement and control of forces seem an effective way to improve the system performance.
First, we shall address the force control problem from a generalized and unified point of view. The concepts introduced here will be illustrated further by examples.