Experimental Studies of the Phase-Generator Circuit of a Flaw Detector

V. V. Volodarskii, V. K. Zhukov

(Presented by the joint scientific seminar of the sector of electromagnetic methods of inspection and the Department of Information and Measuring Technology)

The most important task facing researchers working in the field of flaw detection is to increase the resolving power of flaw detectors. To this end, the designs of primary eddy-current transducers and the circuits for processing the information obtained from them are being improved.

One of the circuits for processing information from an eddy-current transducer is the phase-generator circuit (PGC) investigated in this work. At present, the phase-generator circuit (PGC) is beginning to be widely used [1—4] owing to its high noise immunity and high sensitivity.

According to [5], the potential noise immunity in the case of phase modulation, which is what occurs in a PGC, is slightly lower than in the case of frequency modulation and is an order of magnitude or more higher than for amplitude modulation.

To illustrate the high sensitivity of a PGC, the following example [2] may be cited. The use of a PGC for sorting capacitors made it possible to sort capacitors differing from one another in capacitance by four-millionths of a percent.

The use of a PGC for flaw-detection purposes makes it possible to implement the principle of self-comparison automatically, which is difficult or practically impossible for many methods of inspection by the eddy-current method. In addition, the flaw-detector circuit is considerably simplified.

Let us consider what a PGC used for flaw-detection purposes represents.

The phase-generator circuit (Fig. 1) consists of two generators: G1 and G2 with circuits L1, C1 and L2, C2, coupled through mutual inductance M. The generators operate in the frequency-locking mode.

The essence of the frequency-locking phenomenon consists in the mutual synchronization of two mutually coupled generators operating at very low—

Fig. 1. Phase-generator circuit

… frequencies, that is, owing to the presence of coupling through the mutual inductance \(M\), and when the generator frequencies are close, one generator (for example, G1) forces the other (G2) to generate at the frequency of the first generator.

When the natural frequencies of the generator circuits are equal, i.e., when the condition \(L_1C_1 = L_2C_2\) is satisfied, the phase shift between the generator voltages is zero.

With inductive detuning of one of the circuits by an amount \(\Delta L\), the generators continue to operate at one frequency, but a phase shift \(\varphi\) arises between their voltages.

Synchronization is disrupted at \(\varphi = \pm 90^\circ\).

The sensitivity of the PGC varies significantly depending on the degree of coupling of the generators. Moreover, the sensitivity will be the higher, the smaller the coupling coefficient of the generator circuits,

\(k = \dfrac{M}{\sqrt{L_1L_2}}\).

The graph of the dependence of the magnitude of the phase shift for various coupling coefficients under asymmetric inductive detuning \(\Delta L\) is shown in Fig. 2, a.

Labels in Fig. 2: vertical axis \(\varphi\), degrees; in (a) horizontal axis \(\Delta L/L, \%\), curves \(k = 0.001\), \(k = 0.0025\), \(k = 0.005\); in (b) horizontal axis \(\Delta r/r, \%\).

Fig. 2. Dependence of the phase shift on the change in inductance (a) and active resistance (b) of the sensor

In addition to the sensitivity of the PGC to a change in the inductance of the circuit, a dependence is observed of the magnitude of the phase shift on the change in the active resistance of the sensor.

The plot of the dependence $\varphi=f\left(\dfrac{\Delta r}{r}\right)\big|_{k=\mathrm{const}}$ for asymmetric tuning of the generators is given in Fig. 2, b.

Studies were also carried out with the aim of determining the possibility of practical application of the PGC for flaw detection of elongated products.

The purpose of these experimental studies was:

1) to determine the dependence of the sensitivity of the PGC on the distance between the transducers and on the degree of their shielding;

2) to determine the optimal distance between the transducers;

3) to study the detectability of surface defects in turbine-blade blanks using a phase-generator flaw detector (PGD) with through-type transducers.

The first question, which is of great practical importance, is the question of the influence of the distance between the transducers on the sensitivity of the PGD.

As the distance between the transducers increases, the sensitivity of the flaw detector increases, since the coupling coefficient decreases. However, operation with large distances between the transducers is undesirable because this increases the influence of interfering factors, such as changes in the electromagnetic properties of the specimens (electrical conductivity $\gamma$ and magnetic permeability $\mu$), as well as changes in the geometrical dimensions along the axis of the specimen. When the transducers are located close together, the influence of changes in $\gamma$ and $\mu$ on the phase change will be considerably smaller, since their gradient is small. However, at a small distance between the transducers, the coupling between the generators becomes strong and, consequently, the sensitivity is low.

To reduce the coupling and increase the sensitivity, shields were used. Shields of various forms were investigated: a short-circuited wire ring, copper plates, and closed-type cylindrical shields.

As the studies showed, the best results are obtained when closed-type cylindrical shields are used. The use of these shields made it possible to increase significantly the sensitivity of the PGD at distances between the transducers not exceeding its radius. The optimal distance between the transducers, which makes it possible to obtain high sensitivity and to detune from nonuniformity of the electromagnetic properties and changes in the geometrical dimensions along the axis of the specimen, is approximately equal to the radius of the transducer.

As noted earlier, the main purpose of the experimental studies is to investigate the detectability of surface defects with the aid of the PGD.

To resolve the question of defect detectability, artificial defects of the same type, in the form of holes of different depth and diameter, were made in the specimen. This method of evaluating electromagnetic flaw-detection equipment was first proposed by Dr. Förster [6] and is now becoming increasingly widespread. The distance between defects was chosen so as to exclude the influence of neighboring defects on the test results. The different defect sizes made it possible to reveal the pattern of change in the output signals as a function of defect geometry.

The block diagram of the PGD for inspecting surface defects in turbine-blade blanks is shown in Fig. 3.

The following designations of elements are used in the block diagram:

D1, D2 — parametric through-type transducers connected into the generator circuits;

G1, G2 — generators;

Fig. 3. Block diagram of a phase-generator flaw detector

FD — phase detector;
F — filter;
CE — control element for automatic tuning of the generators;
SU — selecting device.

From the output of circuit 1, pulse signals were taken, indicating the presence of defects in the specimens.

A KSP-4 chart recorder was connected to the output of circuit 2, by means of which the signals from defects were recorded on chart paper.

From examination of the diagrams obtained it is evident that the smallest defect in the tested specimens, with a depth of 0.5 mm and a diameter of 1 mm, is detected at a signal-to-noise ratio of no worse than five, which is quite satisfactory.

The experimental studies carried out showed that the PGC has high sensitivity to changes in the reactive and active resistances of the transducer and good noise immunity, as a result of which it can be recommended for use in electromagnetic flaw detectors.

REFERENCES

1. K. S. Polyulyakh et al. Automation of production processes in mechanical engineering and instrument making. Issue 4, L’vov, Publishing House of L’vov University, 1966.

2. Klutke F. “Arch. für technisches Messen,” Januar, 1961, No. 300.

3. B. Z. Mikhlin. High-frequency capacitive and inductive transducers. Moscow, Gosenergoizdat, 1960.

4. K. S. Polyulyakh. Galvanometric amplifier with a generator phase converter. Izmeritel’naya tekhnika, 1966, No. 9.

5. V. A. Kotel’nikov. Theory of potential noise immunity. Moscow, Gosenergoizdat, 1956.

6. Förster F. Die magnetische und elektromagnetische Sortentrennung von Stahlhalbzeug und Massenteilen, “Arch. Eisenhüttenw,” 25, 1954, 383.