JPS5822701B2 - Accorstates query discrimination device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an acoustic emission (hereinafter abbreviated as AE) discriminator, and particularly to a discriminator that can effectively remove noise having frequency characteristics different from signals.

The AE measurement method is a method of receiving elastic waves generated by materials or structures themselves to evaluate the strength of the test objects, inspect for defects, etc.

In this measurement, a conventional problem has been how to distinguish noise generated during measurement from signals.

As one measure to eliminate this noise, a method using a frequency filter has conventionally been used, but it is not very effective.

The conventional method will be explained with reference to FIG.

The figure shows an example of monitoring crack growth by AE measurement in a tensile fatigue test, in which test specimen 1 is
It is assumed that the jigs 4 and 5 are installed through the jigs 4 and 5.

The jig 4 is fixed, and a tensile load is repeatedly applied to the jig 5.

A sensor 6 for detecting the AE multiplied signal is installed on the test piece 1, and its output voltage is amplified and impedance-converted by a preamplifier 7, passed through a filter 8, and further amplified by a main amplifier 9, and subjected to signal processing. circuit 1
The filter 8 that processes the signal at 0 and outputs it to the recorder 11 has a so-called frequency filtering function for noise removal.

Further, the signal processing circuit has functions such as an AE effective value conversion circuit, an AE event generating circuit, and an AEIJ down counting circuit.

In the case of such a tensile fatigue test, the notch part 1 of the test piece 1
As the crack propagates, a sudden type AE double signal 8 is generated from c, as shown in FIG. 2a.

In addition, at that time, the pins 2 and 3 should be installed using parts or jigs 4 and 5.
Due to the installation of the motor, mechanical noise en often occurs at a timing different from that of the AE multiplied signal s as shown in FIG. 2a.

The amplitude and frequency characteristics of these AE multiplied signal noises vary depending on factors such as the shape, material, and size of the test piece and testing machine, but the characteristics of the signal and noise often differ.
Noise discrimination is performed using this.

Figure 3 shows a frequency analysis of so-called AE multiplied signal noise.

An example of the results will be shown, and when the frequency range of the noise en is lower than that of the AE multiplied signal 8 (based on the frequency fc), the filter 8 has the characteristics (H
, P, F) by using bypass filters.
Attenuates noise to some extent.

I can do it.

However, in actual cases, as shown in FIG. 3, the frequency regions of the AE multiplied signal 8 and the noise en are not completely separated, but rather usually overlap with each other.

Therefore, when noise with large amplitude comes in, the second
figure.

As shown in b, the output from draining the sap does not reach the level where the AE multiplied signal can be identified.

Furthermore, as long as the same frequency component of the AE multiplied signal is mixed in the noise, that frequency component will remain no matter how many stages of filtering are applied.

SUMMARY OF THE INVENTION An object of the present invention is to eliminate the drawbacks of the prior art as described above and to provide an acoustic emission discriminator that provides effective AE discrimination.

That is, in order to increase the effect of removing interference noise in AE measurement, the present invention provides a frequency filter that passes the AE multiplied signal frequency/waveband, and calculates the ratio of its output signal to the output before filtering. It has an arithmetic function that multiplies the proportional coefficient by the f liquid drain signal.

FIG. 4 shows an embodiment of the present invention, in which the output of the preamplifier I is divided into two groups, a numerator system and a denominator system, and the numerator system is a signal filter 8a1 main molecule system whose passband is the AE multiplied signal main frequency band. The amplifier 9a1 consists of a molecular detection circuit 12a that performs rectification and smoothing, and its output voltage Ea is determined by a multiplication/division circuit 130.
Applied to input terminal a.

On the other hand, the denominator system consists of a denominator main amplifier 9b and a denominator detection circuit 12b that performs rectification and smoothing, and its output voltage Eb
is applied to the input terminal of the multiplication/division circuit 130.

Further, the output voltage ea of the molecular main amplifier 9a is also applied to the input terminal C of the multiplication/division circuit 130.

The multiplication/division circuit 13 is an analog calculation circuit that can perform calculations as shown in the following equation.

Next, we will discuss the noise removal effect of this method.

Now, if the attenuation rate of the noise voltage en by the signal filter 8a is Kn, the noise voltage enf of draining is as follows: enf=Kn@en ・・・・・・・・・・・・・・・
・(2), where the detection outputs of en and enf are respectively En and Enf, and using equation (1), the output voltage en
o becomes.

Also, En(t)σe n(t), Enf(t)ff
Since e and f(t), equation (3) can be expressed as the following equation.

On the other hand, since eno in the conventional method is eno=enf-Kn6en° song (5),
Comparing equations (4) and (5), it can be seen that the noise removal effect of the present invention is Kn times better.

Comparing this using practical values, we find that the effective value is e821v, en-2V, Kn = 0.3, and the attenuation rate of the signal filter 8a for the AE multiplied signal is Ks.
Assuming that K s = 0.9, in the case of the conventional method, and assuming that the signal filter 8 has the same characteristics as the signal filter 8a, e ==KsXe =0.9V −
(6) SOS e = K n x e = 0.6 V
-=・(7)nOn 5, e8o is the AE multiplied signal stream liquid extraction output voltage, eno is the noise stream liquid extraction output voltage, and e8o and en.

are almost at the same level and cannot be distinguished.

Next, in the case of the present invention, the S/N of the conventional method is 1.5, whereas that of the present invention is 4.5, and discrimination is sufficiently possible and the noise removal effect is large.

In the case of this method, since the condition Eb)Ea)0 is satisfied, if the multiplication/division circuit 13 has a fast response characteristic, ea can maintain an almost faithful waveform.

FIG. 5 shows another embodiment of the invention.

The present invention is characterized in that a noise filter 8b for passing the noise frequency band is provided before the denominator main amplifier, and by doing so, the noise removal effect can be further enhanced.

Furthermore, as another embodiment, the noise removal effect can be enhanced by increasing the power of the denominator of the multiplication/division circuit 13.

That is, the following equation can be used instead of equation (1).

Here, n=1, 2, 3-..., m=1, 2
, 3...

In each embodiment, the multiplication/division circuit does not have the role of signal discrimination, but the multiplication/division circuit may be provided with discrimination comparison means.

This can also be achieved by providing a discrimination comparison means separately from this multiplication/division circuit and inputting the multiplication/division output to this means.

In any case, the discrimination and comparison means can be achieved by comparing successively input signals with a reference level (or by comparing the magnitudes of successively input signals with each other).

As described above, according to the present invention, the reliability of the data is significantly higher than that of the conventional method, and therefore, when the AE measurement technique is used for product inspection or abnormality diagnosis of equipment, the effect is significant.

[Brief explanation of the drawing]

Figure 1 is a diagram of a conventional example of AE measurement, Figure 2 is a diagram showing an example of the generation of the AE multiple signal noise waveform before and after the tear fluid of the frequency filter, and Figure 3 is an example of the frequency analysis results of the AE multiple signal and noise. FIG. 4 shows an embodiment of the present invention, and FIG. 5 shows another embodiment. 6...Sensor, 8a...Signal filter, 13...
・Multiplication/division circuit.

Claims (1)

[Claims] 1. A sensor for detecting an acoustic emission signal, a filter means for amplifying and filtering the output of the sensor, a first means for amplifying and detecting a signal obtained from the means, and the above-mentioned sensor. A second means for amplifying and detecting the output, outputs a and b obtained from the first and second means, and output C obtained from the filter means are input,
An acoustic emission discrimination device comprising a calculation means for calculating a-c/b, and a means for discriminating between normal acoustic emission and noise based on the output of the calculation means. 2. Claim 1, wherein a noise filter means for passing a noise frequency band is provided between the second means and the sensor, and the output of the filter means is input to the second means. Acoustic emission discriminator as described in Section 2. 3. The acoustic emission discriminator according to claim 1, wherein the first and second means are set such that their outputs are powers of their respective inputs. 4. Claim 1, in which normal acoustic emissions and noise are discriminated from the outputs of the arithmetic means, by comparing the magnitudes of the outputs successively output from the arithmetic means. Acoustic emission discriminator as described in Section 2. 5. A patent claim in which the output from the calculation means is compared with a predetermined reference level to discriminate between normal acoustic emissions and noise from the output of the calculation means, and the discrimination is performed based on the magnitude relationship. The acoustic emission discriminator according to item 1.