JPH0473746B2 - - Google Patents

Description

Detailed Description of the Invention (Industrial Application Field) The present invention is an ultrasonic automatic flaw detector that can automatically detect internal defects in porcelain products, cast products, etc. where forest echoes occur in the ultrasonic flaw detection output waveform. It is about the method.

(Conventional technology) Porcelain products and cast products have coarse crystal structures and high porosity, so they have poor ultrasound penetration. It is normal to cover up the disadvantages of However, when the flaw detection frequency is increased, forest-like echoes as shown in Figure 1 occur due to the influence of reflected waves from many minute interfaces such as grain boundaries inside the product, indicating the presence of internal defects in the product. In this case, as shown in FIGS. 2 and 3, the defective waveform 2 will be included in the normal waveform 1 constituting the forest echo. Therefore, in order to perform ultrasonic flaw detection on porcelain products such as solid insulators, an inspector visually observes the ultrasonic flaw detection output waveform on a CRT and discerns whether defective waveform 2 is included in normal waveform 1. (for example, "Insulator" published by the Institute of Electrical Engineers of Japan, June 15, 1983, P.384-385), but since the shape of the forest echo is not constant depending on where the probe is applied, it is difficult to perform accurate inspection. There was a problem that required me to rely on my many years of experience.

Therefore, as shown in Figure 4, slice level 3 is set to a level higher than the maximum value of normal waveform 1 included in the forest echo, and anything exceeding this level is automatically determined to have a defect. This method also uses defective waveform 2, which has a lower level than the maximum value of normal waveform 1, as shown in Figure 3.
There was a drawback that correct judgment could not be made when .

In addition, an attempt was made to extract the defective waveform 2 by directly comparing the forest echo consisting only of the normal waveform 1 shown in Fig. 1 and the forest echo including the defective waveform 2 shown in Figs. 2 and 3. However, by slightly changing the position of the probe, the waveform of the forest echo consisting only of normal waveform 1 may change as shown in Figure 5, for example, and the shape of normal waveform 1 may change. Since these are not constant, it has been impossible to detect defective waveform 2 by directly comparing them.

(Problems to be solved by the invention) The present invention solves these conventional problems,
This method was developed with the aim of creating an automatic ultrasonic flaw detection method that can automatically and reliably detect defective waveforms that are lower than their maximum level even if they are included in forest echoes.

(Means for Solving the Problems) The present invention divides the time between the transmission pulse of the original waveform of ultrasonic flaw detection and the bottom reflection pulse into n To sections, and divides the ultrasonic flaw detection output waveform into each To section. The waveform is converted by inputting it to a peak hold circuit that holds the peak value of each To interval, and then inputting this converted waveform to a sample hold circuit that samples the value of the final part of each To interval to obtain a sample waveform. The present invention is characterized in that the presence or absence of a defect is detected by determining the difference between the sample waveform obtained from the ultrasonic flaw detection output waveform that does not contain any defects.

The invention will now be described in detail along with the circuitry used to implement it. In the circuit of Fig. 11,
The left half of the dotted line is the peak hold circuit, and the right half is the sample hold circuit. The peak hold circuit consists of an operational amplifier (OP1), an operational amplifier (OP2), a diode (D), a capacitor (C1), a transistor (TR1), and (TR2), and the ultrasonic flaw detection output waveform is connected to the terminal of the operational amplifier (OP1). ( _ei )
is input, and a clock pulse (Cp ₁ ) as shown in FIG. 12 is input to the base of the transistor (TR1). This clock pulse (Cp ₁ ) is emitted for a very short time only at the beginning of the To interval, which is the time between the ultrasonic flaw detection transmission pulse and the bottom reflection pulse, and is divided into n equal parts. It also has the role of opening the gate of the transistor (TR2) and grounding the voltage of the capacitor (C1). When the above-mentioned forest echo is input to such a peak hold circuit as the ultrasonic flaw detection output waveform (e _i ), as the waveform rises, the same waveform is output to the output terminal of the operational amplifier (OP1). ,
While the waveform rises, the voltage of the capacitor (C1) increases similarly, but the ultrasonic flaw detection output waveform (e _i )
After reaching the maximum value, the voltage of the capacitor (C1) is maintained as it is, and if a higher level waveform is input, it will be maintained at its maximum level. The voltage of this capacitor (C1) becomes the output of the operational amplifier (OP2) as it is, so until the aforementioned clock pulse (Cp ₁ ) is input, that is, until the end of each To interval, the output terminal of the operational amplifier (OP2) is To
The peak value input during the interval will be output as the peak hold conversion output (e _i ). Therefore, when the forest echo consisting of only the normal waveform shown in Figure 1 is input as the ultrasonic flaw detection output waveform (e _i ) to this peak hold circuit, the peak hold conversion output (e _p ) will be as shown in Figure 6. Moreover, if the forest echo consisting only of the normal waveform shown in FIG. 5 is input, the peak hold conversion output (e _p ) as shown in FIG. 7 is obtained. Furthermore, when the forest echo including the defective waveform 2 shown in FIG. 3 is input, the 8th
As shown in the figure, an output with a high peak value in the second To section is obtained.

In this way, the waveform of the peak-hold conversion output (e _p ), which is the output of the peak-hold circuit, is the sixth
Since there are differences as shown in FIG. 7 and FIG. 7, automatic detection of defective waveform 2 is impossible at this stage. Therefore, this peak-hold conversion output (e _p ) is further input to a sample-and-hold circuit. As shown in Figure 11, the sample and hold circuit consists of a field effect transistor (FET), a capacitor (C2), and an operational amplifier (OP3). (Cp ₂ ) is input. This clock pulse (Cp ₂ ) is generated only during the last _TSA of each To interval, and the other To- _TSA times are taken as the sample and hold time T _HO . The sample and hold circuit uses a field effect transistor (FET) only when this clock pulse (Cp ₂ ) is input.
conducts _and captures the peak hold conversion output ( _e It has the function of outputting as _(SH ).
And the peak hold conversion output (e _p ) like this
is captured at the end of each To interval, so whether the waveform of the peak-hold conversion output (e _p ) is shown in Figure 6 or Figure 7, the sample Both hold conversion outputs (e _SH ) have the same waveform as shown in Figure 10, and if the waveform of the peak hold conversion output (e _p ) is as shown in Figure 8, the sample hold conversion output The waveform of (e _SH ) is as shown in FIG. Therefore, it is possible to find the difference between the waveform in FIG. 10, which shows a normal case, and the waveform in FIG. 9, which includes an abnormality, and automatically determine that a defect exists when the difference exceeds a tolerance value. It becomes possible. In order to perform such a comparison, for example, the probes are sequentially brought into contact with the end faces of a solid insulator, etc. at a pitch of about 2 to 3 mm, and two adjacent
All you have to do is find the difference in the ultrasonic flaw detection output waveforms at the points. As mentioned above, if the position of the probe is slightly changed, the waveform of the forest echo consisting of only the normal waveform 1 changes as shown in Figs. 1 and 5, but in the present invention, these changes are By passing it through the peak hold circuit and the sample hold circuit, it can be converted to the same waveform shown in Figure 10, so the difference in the output of the sample hold circuit will be zero in the normal part, and the difference in the output of the sample hold circuit will be zero in the part containing the defect. This means that the difference will show a large value.

(Effects of the Invention) As is clear from the above description, the present invention allows the ultrasonic flaw detection output waveform to pass through a peak hold circuit and a sample hold circuit, thereby detecting defects at a level lower than the maximum level during forest echoes. Even if a waveform is included, it is automatically extracted to ensure reliable detection. Therefore, the present invention greatly contributes to the development of industry as an ultrasonic automatic flaw detection method particularly suitable for inspecting internal defects in articles that produce forest echoes, such as porcelain products and cast products.

[Brief explanation of drawings]

Figure 1 is a waveform diagram showing a forest echo consisting of only normal waveforms, Figure 2 is a waveform diagram showing a forest echo including a defective waveform, and Figure 3 is a waveform diagram showing a forest echo including a defective waveform at a level lower than the maximum level. FIG. 4 is a waveform diagram showing the slice level in the conventional method. FIG. 5 is a waveform diagram showing other shapes of forest echoes consisting only of normal waveforms. FIGS. 6 to 8 are respectively Waveform diagrams of peak-hold conversion output corresponding to Figures 1, 5, and 3, Figure 9 is the waveform diagram of the peak-hold conversion output corresponding to Figure 8, and Figure 10 is the sample-hold conversion output corresponding to Figures 6 and 7. 11 is a circuit diagram of a peak hold circuit and a sample hold circuit used in the present invention, and FIGS. 12 and 13 are waveform diagrams of a clock pulse (Cp ₁ ) and a clock pulse (Cp ₂ ). e _i ...Ultrasonic flaw detection output waveform, e _p ...Peak hold conversion output, e _SH ...Sample hold conversion output.

Claims (1)

[Claims] 1 Divide the time between the transmitted pulse of the original waveform of ultrasonic flaw detection and the bottom reflected pulse into n To intervals,
The ultrasonic flaw detection output waveform is input into a peak hold circuit that holds its peak value for each To interval and converted, and this converted waveform is then input into a sample hold circuit that samples the final value of each To interval. An automatic ultrasonic flaw detection method characterized in that the presence or absence of a defect is detected by obtaining a sample waveform using a sample waveform obtained from an ultrasonic flaw detection output waveform that does not include a defect.