APPLICATION OF SEMICONDUCTOR STRAIN GAUGES IN DEVICES FOR MEASURING CUTTING FORCES

M. F. POLETIKA, V. A. KRASILNIKOV

In studying the processes of metal cutting, various types of devices are used to measure the components of the cutting force; a number of specific requirements are imposed on such devices. One of these is the possibility, during measurement, of reproducing the conditions of the actual production process. To satisfy this requirement, the rigidity of the tool mounting and fastening in the dynamometer (and of the dynamometer on the machine tool) must be approximately the same as under production conditions—that is, when the tool is clamped in an ordinary holder. This requirement is most fully met by elastic-electric devices (dynamometers), in which various types of electrical sensors are used as the transmitting element. The most widely used have been inductive, wire, and foil sensors, which convert deformation of the elastic element into an electrical signal whose magnitude depends on the sensitivity of these sensors. To obtain an electrical signal sufficient for recording with sensors of this type, it is necessary either to increase the magnitude of the deformation or to use various kinds of amplifiers.

In order to eliminate these shortcomings, it is necessary to use more sensitive strain sensors. Such sensors include semiconductor strain gauges, already mastered by domestic industry, whose strain-sensitivity coefficient is 50–60 times higher than that of wire and foil sensors. The properties and methods of working with these strain gauges are described in the literature [1, 2, 3].

In the cutting laboratory of Tomsk Polytechnic Institute, various types of dynamometers equipped with inductive, wire, foil, and semiconductor strain sensors were investigated. As a result of this work, a two-component dynamometric table using semiconductor strain gauges and having increased rigidity was designed, manufactured, and thoroughly tested (Fig. 1).

The table consists of an elastic element 1, fastened to a base 2. The elastic element is the same as in the known [4] elastic dynamometers; however, its rigidity has been increased by a factor of 2–3, and the wire strain gauges have been replaced by semiconductor ones. The greatest displacements (deformations) of the table in the direction of the cutting-force components $P_z$ and $P_y$ do not exceed, respectively, 30 and 5 µm.

The load in the direction of the force $P_y$ is perceived by the strain gauges $(R_1; R_2; R_3; R_4; R'_1; R'_2; R'_3; R'_4)$, with the outer ones being stretched and the inner—

…stresses are compressed. The horizontal component \(P_z\) is recorded by strain gauges glued to the side faces of the elastic element \((R_5; R_6; R_7; R_8; (R'_5; R'_6; R'_7; R'_8)\).

Fig. 1

The sensors are connected in a bridge circuit (Fig. 2) with the following calculation: on the one hand, to increase the sensitivity of the bridge circuit as much as possible and to provide temperature compensation, and on the other—to ensure automatic elimination of the influence of the components \(P_z\) and \(P_y\) on each other and to obtain independence of the readings from the point of application of the forces on the working surface of the table.

The magnitude of the current flowing through the semiconductor strain gauge is limited by the self-heating limit; when it is exceeded, the strain-sensitivity coefficient decreases and the operation of the circuit becomes unstable. The value of this limit can be raised by improving heat removal or by switching to pulsed power. In our design, the massive elastic element serves as such a peculiar radiator. This made it possible, with a resistance of the silicon semiconductor strain gauge of 100 ohms, to pass through it a current of 0.04 A. The circuit is powered by direct current from storage batteries; this made it possible to get rid…

from the complex stabilizing apparatus required when supplying rectified current from the mains. An oscilloscope vibrator and a microammeter are connected into the measuring diagonal of the bridges. The bridge is balanced by two fixed resistances \(R\) and one variable resistance \(R_b\).

Fig. 2

The calibration graph of the dynamometric table is shown in Fig. 3. The graph is rectilinear and passes through the origin. Multiple repeated calibrations over the course of half a year showed that the maximum discrepancy of the calibration straight lines does not exceed 3% and does not de-

Fig. 3

depends on the point of application of the force on the table surface. The maximum load for the components of the cutting force is 500 kg.

Despite all the advantages of semiconductor strain resistors over other strain gauges, they have a number of serious shortcomings. These are, above all, their still high cost. The increased fragility of the gauges requires very delicate and careful handling during gluing and installation. Semiconductor gauges currently produced have a large spread in resistance (up to 1.5%), which makes balancing bridge circuits difficult. In addition, the strain-sensitivity coefficient depends on the properties of the crystal; therefore, an identical strain-sensitivity coefficient is found in a batch of strain resistors made from a single wafer. The difference in the ohmic resistance of the soldering points of the lead wires and of the crystal itself makes it difficult to use alternating current to power the bridges. The imbalance that arises in this case is practically impossible to eliminate.

But even with these shortcomings, direct-current supply, careful selection of semiconductor strain resistors by resistance and strain sensitivity, and accuracy in gluing made it possible to obtain a stable, reliable, and easy-to-operate electrical circuit.

Experience in working with semiconductor strain resistors has shown that they are promising for equipping instruments used in the study of metal-cutting processes.

REFERENCES

1. Semiconductor strain gauges. Ed. by M. Din, Energiya Publishing House, Moscow—Leningrad, 1965.

2. L. S. Il’inskaya, A. N. Podmarkov. Semiconductor strain gauges. Energiya Publishing House, Moscow—Leningrad, 1966.

3. B. S. Trukhachev, N. P. Udalov. Semiconductor strain transducers. Energiya Publishing House, Moscow—Leningrad, 1968.

4. M. F. Poletika. Instruments for measuring cutting forces and torques. Mashgiz, Moscow—Leningrad, 1963.

5. G. I. Smagin, K. A. Nassonov, P. N. Obukhov, G. P. Kersha, G. N. Guk. A dynamometer using semiconductor strain gauges for measuring small torques. Machine Tools and Instrument, No. 4, 1967.