JPH09269215A - Superposed wave filtering method and apparatus - Google Patents

Abstract

(57) Abstract: A superposed wave filtering method and apparatus capable of extracting a reflected wave by reducing unnecessary components from a received wave and specifying the frequency of the reflected wave with high accuracy. SOLUTION: A reinforcing bar 4 is embedded in a reinforced concrete 1 so as to be parallel to an upper surface 2. The oscillation sensor 6 and the reception sensor 7 are arranged on the upper surface 2 so as to be parallel to the reinforcing bar 4 immediately above the reinforcing bar 4 along the longitudinal direction of the reinforcing bar 4. The elastic wave is oscillated from the oscillation sensor 6, and the reception sensor 7
To receive. The obtained received wave is subjected to a time series filter having a magnitude of 0 to 1 from the arrival time of the received wave to a predetermined time and a magnitude of 1 after the predetermined time, and then subjected to filtering processing. Since the components of the reflected wave indicated by the arrows 61 and 62 are extracted from the received wave by the time-series filter, and the filtering process by the frequency is applied to this component, the frequency of the reflected wave is specified with high accuracy. .

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to concrete,
The present invention relates to a superposed wave filtering method and apparatus for oscillating elastic waves in a substance such as soil or metal and filtering the received waves obtained.

[0002]

2. Description of the Related Art Conventionally, elastic waves such as ultrasonic waves have been used to measure the position of cracks in concrete, the position of underground structures, the thickness of iron, and fish hunting. For example, after the elastic wave oscillation sensor and the elastic wave reception sensor are arranged on the surface of a concrete wall or the like, the elastic wave oscillation sensor oscillates the elastic wave toward the inside of the concrete, and the elastic wave reception sensor receives the elastic wave. The reflected wave is extracted from the obtained received wave, and the fundamental frequency of this reflected wave is specified. From this fundamental frequency, the crack depth of concrete is calculated.

[0003]

However, the above-mentioned prior art has the following problems. That is, in addition to a reflected wave, a wave excited by an oscillated elastic wave (hereinafter, this wave is referred to as a forced excitation wave) propagates in the concrete. Therefore, there is a problem that the reflected wave and the forced excitation wave are superimposed to form a received wave, and the reflected wave is buried in the forced vibration wave.

FIG. 59 is a schematic diagram showing how an oscillating sensor oscillates elastic waves in concrete and a receiving sensor receives the acoustic waves. As shown in FIG. 59, a reinforced concrete 1 having an upper surface 2 and a bottom surface 8 is arranged, and a reinforcing bar 4 is embedded in the reinforced concrete 1. An oscillation sensor 6 and a reception sensor 7 are arranged on the upper surface 2 of the reinforced concrete 1.

Reinforced concrete 1 constructed in this way
When the elastic wave is oscillated from the oscillation sensor 6, the elastic wave is reflected by the reinforcing bar 4 to reach the reception sensor 7 as shown by an arrow 61 and the bottom surface 8 as shown by an arrow 62.
To reach the reception sensor 7. Further, the oscillated elastic wave forcibly excites the concrete as indicated by an arrow 63, a forced excitation wave having substantially the same frequency as this elastic wave is excited, and the forced excitation wave reaches the reception sensor 7. In order to see through an elastic body such as concrete, arrows 61,
It is necessary to measure the reflected wave indicated by 62 and specify the fundamental frequency of this reflected wave. However, the reflected wave and the forced excitation wave are superimposed to form a superimposed wave, and since this superimposed wave is measured as a received wave by the reception sensor 7, the reflected wave is buried in the forced excitation wave and vibrates in the received wave. There is a problem that it is difficult to specify the fundamental frequency of the reflected wave even when the filtering process based on the number (hereinafter referred to as the frequency filtering process) is performed. As described above, it is difficult to specify the position of the reinforcing bar from the received wave with high accuracy by simply performing the frequency filtering process on the received wave.

The present invention has been made in view of the above problems, and a reflected wave is extracted by reducing unnecessary components from a received wave and a fundamental wave frequency of the reflected wave can be specified with high accuracy. An object is to provide a filtering method and apparatus.

[0007]

According to the superposed wave filtering method of the present invention, an elastic wave oscillating sensor and an elastic wave receiving sensor are arranged in a substance through which an elastic wave propagates, and the elastic wave oscillating sensor is placed in the substance. After the elastic wave is oscillated, it is received by the elastic wave receiving sensor, and the magnitude of the obtained received wave from the arrival time of the received wave to the predetermined time is 0 to 1, and the magnitude after the predetermined time. It is characterized by applying a time-series filter with a value of 1.

The time series filter may be applied a plurality of times.

Further, it is preferable to perform filtering processing based on the frequency.

A plurality of the elastic wave oscillating sensors and the elastic wave receiving sensors are arranged, and the received waves obtained by the respective elastic wave receiving sensors are subjected to the time series filter and then added to each other. May be.

A plurality of the elastic wave oscillating sensors and the elastic wave receiving sensors are arranged, and the received waves obtained by the respective elastic wave receiving sensors are added to each other and then the time series filter is applied. Good.

A superposed wave filtering processing apparatus according to the present invention includes an elastic wave oscillating sensor, an elastic wave receiving sensor, and a time series of a received wave received by the elastic wave receiving sensor which is time-series filtered to form a modified received wave. The time-series filter has a size from 0 to 1 from the arrival time of the received wave to a predetermined time, and a size after the predetermined time is 1.

It is preferable to have a filtering processing unit for filtering the modified received wave by frequency.

In the superposed wave filtering method according to the present invention, elastic waves are oscillated in the substance and the received waves received are time-series filtered. If there is a substance that reflects elastic waves in the substance (hereinafter, this substance is called a reflector), the input elastic wave is reflected by the reflector to form a reflected wave, and this reflected wave propagates and is received. Reach the sensor.
Since the reflected wave is an elastic wave generated by resonance, its vibration continues for a long time. Further, the input elastic wave causes the wave to be forcibly excited in the substance, and the forced excitation wave propagates in the substance and reaches the reception sensor. Unlike the reflected wave, the forced excitation wave decays and disappears in a short time. A reception wave is formed by superposing the forced excitation wave and the reflected wave. The time-series filter has a magnitude of 0 to 1 from the arrival time of the received wave to a predetermined time and a magnitude of 1 after the predetermined time, and the time-series filter is applied to this received wave to The waveform of the wave (received wave) is modified (hereinafter, the received wave subjected to the time series filter is referred to as a modified received wave). Then, the portion of the received wave from when the received wave reaches the receiving sensor to when a predetermined time elapses is reduced. This portion contains many components of the forced excitation wave, and when the modified reception wave is obtained by reducing this portion, the modified reception wave has the components of the forced excitation wave reduced. For this reason,
Since the modified received wave is composed of substantially reflected wave components, the frequency of the reflected wave can be specified. As described above, in the present invention, the received wave is subjected to the time-series filter and the forced excitation wave in the received wave is removed, so that the frequency of the reflected wave in the received wave can be specified with high accuracy.

The above-mentioned predetermined time may be appropriately determined in consideration of the time required for the forced excitation wave to decay.

In the superposed wave filtering method of the present invention, the received wave may be repeatedly subjected to a time series filter. Thereby, the component of the forced excitation wave in the received wave can be further reduced.

When the above-mentioned modified received wave is subjected to the frequency filtering process, unnecessary components are deleted from the modified received wave and necessary reflected wave components are amplified. Thereby, the frequency of the reflected wave in the received wave can be specified with higher accuracy.

Further, when the received wave is repeatedly subjected to the time series filter and the obtained modified received wave is subjected to the frequency filtering processing, the frequency of the reflected wave can be specified with higher accuracy.

In the superposed wave filtering method of the present invention, the elastic wave oscillation sensor and the elastic wave reception sensor are
A plurality of each may be arranged. In this case, the sensors are arranged so that the shortest distance between the elastic wave oscillation sensor and the elastic wave reception sensor and the intended reflector is constant.
For example, if a long bar is placed in this concrete so that it is parallel to the top surface of the concrete,
An elastic wave oscillation sensor and an elastic wave reception sensor are paired, and these are arranged on the upper surface of the pair of concrete along the longitudinal direction of the reinforcing bar. With this arrangement, the shortest distance between each elastic wave oscillation sensor and elastic wave reception sensor and the reinforcing bar is
It will be constant. After arranging each elastic wave oscillating sensor and elastic wave receiving sensor in this way, each receiving sensor receives an elastic wave, and a time-series filter is applied to each of the obtained received waves to obtain a modified received wave for each received wave. obtain. The obtained modified received wave is mainly formed of reflected waves, but since various reflected waves are superposed, it is relatively difficult to specify the reflected wave of the target reflecting object. However, since the distance between the target reflector and each received wave sensor is constant, the frequency of the reflected wave by this reflector is constant, while the distance between other reflected objects and the received wave sensor is constant. Since the distance of is different for each receiving sensor, the frequency of the reflected wave due to other reflectors is different for each receiving sensor. Therefore, in order to utilize this property, the modified reception waves are added to each other (hereinafter, the waves thus obtained are referred to as synthetic reception waves).

Then, since the frequency of the reflected wave of the target reflecting object is invariable, the intensity increases substantially in proportion to the added number of modified received waves. On the other hand, the intensity of the reflected waves of the other reflectors remains substantially constant because the frequency changes. Therefore, when the combined received wave is Fourier transformed, in the obtained Fourier spectrum, the spectrum value of the reflected wave of the target reflector becomes large. Thereby, the frequency of this reflected wave can be specified. In this way, when a plurality of elastic wave oscillation sensors and elastic wave reception sensors are arranged to obtain a combined received wave, the frequency of the reflected wave of the target reflector is easily specified,
It is not necessary to perform frequency filtering processing on this combined received wave. However, it goes without saying that if the combined received wave is subjected to frequency filtering, it becomes easier to specify the frequency of the reflected wave of the target reflector. In the case described above, a time-series filter is applied to each received wave, a modified received wave is obtained for each received wave, and the modified received waves are added together,
Although the synthetic received wave is obtained, the synthetic wave may be obtained by adding the received waves to each other and then subjecting the obtained sum wave to a time series filter. The former combined received wave and the latter combined received wave are the same.

A superposed wave filtering processing apparatus according to the present invention is for carrying out the above-mentioned superposed wave filtering processing method. After the elastic wave oscillation sensor inputs the elastic wave into the substance, the elastic wave receiving sensor receives the received wave. A time-series filter section applies the above-mentioned time-series filter to this received wave to obtain a modified received wave. Since the modified received wave is composed of substantially reflected wave components, the frequency of the reflected wave in the received wave can be specified with high accuracy as described above.

When the superposed wave filtering processing device is provided with a filtering processing section, the modified receiving wave is subjected to frequency filtering processing in this filtering processing section. Thereby, the frequency of the reflected wave in the received wave can be specified with higher accuracy.

When the time-series filter section is formed by a computer or the like, the reception wave is time-series filtered in a short time to obtain a modified reception wave. As a result, the reflected wave is extracted and the frequency of the reflected wave (deformed received wave) is specified in a short time even for a complicated received wave. Further, when the filtering processing unit is formed by a computer or the like, the modified reception wave is subjected to high-speed filtering processing.

The frequency filtering process described above is not particularly limited, and various frequency filtering processes can be used. However, the frequency of the reflected wave can be specified in any frequency filtering process.
An example of such filtering processing is shown below.

For example, the fundamental frequency is f ₀ , m is a predetermined positive integer, and Δt = m / 2f ₀ is defined.
After setting b and b to predetermined real numbers respectively, the modified received wave f
For (t), F ₁ (t) = {f (t) −af (t + Δ
The filtering process F ₁ (t) represented by t)} / b is performed. Then, among the components of the wave constituting the _{f (t), f 0 /} m and f ₀ / odd multiple frequencies of the m (f ₀ / m,
_{3f 0 / m, 5f 0 /} m, the amplified component is ·····), f ₀ / even multiple frequency of _{m (2f 0 / m, 4f} 0 / m,
The component of 6f ₀ / m, ...) Is reduced or deleted. In this case, f ₀ / m and f ₀ / odd multiple frequencies of the _{m (f 0 / m, 3f} 0 / m, 5f 0 / m, ·····) the amplification factor of the X, f ₀ / Even multiple frequency of m (2f ₀ / m, 4
The attenuation rate of the component of f ₀ / m, 6f ₀ / m, ...
In this case, simultaneous equations represented by a + b = X and a−b = Y may be calculated and the values of a and b may be determined. However, a and b are not limited to this value. Also,
m may be determined according to the purpose, but when m is 1, f
Odd frequency of ₀ and f ₀ (f ₀ , 3f ₀ , 5f ₀ , ...
..) component is amplified, and an even multiple frequency of f ₀ (2f ₀ ,
4f ₀ , 6f ₀ , ...) Components are reduced or deleted. If a 2 m addition, f _0/2 odd multiple frequency _{(f 0 / 2,3f 0 / 2,5f} 0/2, ·····) the component of the amplified, f _0/2 Even frequency of (2f ₀ / 2,4
_{f 0 / 2,6f 0/2,} ····· components) are reduced or removed. That is, an integral multiple frequency of _{f 0 (f 0, 2f 0} , 3
The components of f ₀ , ...) Can be reduced or deleted.

Further, assuming that the fundamental frequency is f ₀ , m is a predetermined positive integer, and Δt = m / 2f ₀ is defined, and a and b are set to predetermined real numbers respectively, and then the modified reception is performed. Wave f
For (t), F ₂ (t) = {f (t) + af (t + Δ
The filtering process F ₂ (t) represented by t)} / b may be performed. Then, among the components of the waves that form f (t), f ₀ / m and the odd harmonic frequency (f _{0 of} f ₀ / m)
_{/ M, 3f 0 / m,} 5f 0 / m, the component removed or reduced in ·····), f ₀ / even multiple frequency of m (2f ₀ / m,
The components of 4f ₀ / m, 6f ₀ / m, ...) Are amplified. In this case, the values of a and b can be determined by calculating the simultaneous equations as described above. However, a and b are not limited to this value as in the case described above. Also may be determined depending on m also object, when m is 1, an odd multiple frequencies of f ₀ and f ₀ (f _0, 3f _0, 5
f _0, are components removed or reduced in .....), even multiple frequency of _{f 0 (2f 0, 4f 0} , 6f 0, ····· components) are removed or reduced. When m is 2, f
_0/2 of an odd multiple frequency _{(f 0 / 2,3f 0 / 2,5f} 0 /
2, ...) component is deleted or reduced, and f _0/2
Even multiple frequency _{(2f 0 / 2,4f 0 / 2,6f} 0/2,
...) is amplified. That is, an integral multiple frequency of _{f 0 (f 0, 2f 0} , 3f 0, ·····) can amplify a component of.

In any frequency filtering process, this filtering process may be carried out a plurality of times as necessary. In this case, the above-mentioned F ₁ (t) and / or F ₂ (t) may be subjected to a superimposed wave f (t), also F ₁ (t) and / or F ₂ a (t) F ₁ (T) and / or F ₂ (t) may be substituted a plurality of times to form a new filter, and this filter may be used to perform frequency filtering on the superimposed wave f (t). With this new filter, the same result as when the filtering process is performed a plurality of times is obtained by one filtering process.

[0028]

BEST MODE FOR CARRYING OUT THE INVENTION Next, an embodiment of the present invention will be described.
This will be specifically described with reference to the accompanying drawings. First, a superposed wave filtering processing method according to an embodiment of the present invention will be described.

First Embodiment Method FIG. 1 is a schematic view showing a reinforced concrete model in the first embodiment method of the present invention. FIG. 2 is a cross-sectional view of FIG.
It is sectional drawing which shows A'plane. 1 and 2, in FIG.
The same parts as 9 are designated by the same reference numerals, and detailed description thereof will be omitted. As shown in FIG. 1, a reinforcing bar 4 is embedded in the reinforced concrete 1 so as to be parallel to the upper surface 2. Further, as shown in FIG. 2, the oscillation sensor 6 and the reception sensor 7
Are arranged on the upper surface 2 immediately above the reinforcing bar 4 along the longitudinal direction of the reinforcing bar 4.

Reinforced concrete 1 constructed in this way
At, the elastic wave is oscillated from the oscillation sensor 6 and received by the reception sensor 7. Then, as indicated by an arrow 61, the elastic wave is reflected by the reinforcing bar 4, and the reflected wave is reflected by the reception sensor 7.
At the same time, the elastic wave is reflected by the bottom surface 8 and reaches the reception sensor 7, as indicated by an arrow 62. Further, as shown by the arrow 63, the forced excitation wave is excited by the elastic wave, and the forced excitation wave reaches the reception sensor 7. In this way, the reflected wave and the forced excitation wave are superimposed to form a superimposed wave,
This superimposed wave is received by the reception sensor 7.

FIG. 3 is a graph showing the relationship between the time t (seconds) on the horizontal axis and the amplitude on the vertical axis. In FIG. 3, the upper stage is a graph diagram schematically showing the oscillating wave oscillated by the oscillation sensor 6, and the lower stage is a graph diagram schematically showing the received wave received by the receiving sensor 7. Also, t _h in Figure 3 shows the arrival time of the received wave. A time series filter is applied to this received wave.

FIG. 4 is a graph showing the relationship between the time t (seconds) on the horizontal axis and the time series filter H (t) on the vertical axis, and is a diagram schematically showing the time series filter. is there.
The time series filter H (t) is 0 before the arrival time t _h, as indicated by reference numeral 51. Also, from the arrival time t _h to t
_{After a} second has elapsed, the time series filter H (t) becomes 1 as indicated by reference numeral 53. Forced excitation wave is caused by forced excitation, in consideration of the time required to decay the forced excitation wave, setting the elapsed time t _a in advance. Time t is t _h
In the range from to t _h + t _a , the time series filter H
(T) is a smooth curved line as indicated by reference numeral 52. As such a curve, for example, a sine wave may be used. Such curve, H (t _h) from _{= 0 H (t h + t} a) = to 1, so that H (t) smoothly varies continuously. Such a time series filter H
(T) is applied to the received wave shown in FIG.

FIG. 5 is a graph showing the relationship between the two, with the horizontal axis representing time t (seconds) and the vertical axis representing amplitude. The received wave is time-series filtered to obtain the modified received wave. It is a graph figure which shows typically. As shown in FIG. 5, in the range of t _h to t _h + t _a, the time-series order filter H (t) is 0 to 1, the deformation reception wave is attenuated in a range of t _h to t _h + t _a There is. Usually, the wave indicated by the arrow 63 in FIG. 1 is generated by being forcibly excited by the oscillating wave, has the same frequency as the oscillating wave, and attenuates and disappears in an extremely short time. On the other hand, the reflected waves indicated by arrows 61 and 62 in FIG. 1 are elastic waves generated by resonance, and thus maintain a high amplitude for a long time. Thus, by the time series filter H (t), if the attenuation of the received wave in the range of t _h to t _h + t _a, component of the force excitation wave (arrow 63) from the received wave is reduced, the reflected wave (arrow 61 , 62) is extracted and remains.
The transformed received wave thus obtained is subjected to Fourier transform. The time-series filter H (t) may be applied to the received wave a plurality of times. In this case, the component of the obtained modified received wave becomes closer to the component of the reflected wave.

FIG. 6 is a graph showing the relationship between the two by plotting the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and is a diagram schematically showing the obtained Fourier spectrum. While amplifying the spectral value of a specific frequency,
The modified received wave is frequency filtered to attenuate other specific spectral values.

FIG. 7 is a graph showing the relationship between the horizontal axis and the vertical axis representing the frequency (Hz) and the spectrum value, respectively, and showing the Fourier spectrum after frequency filtering. is there. As shown in FIG. 7, since the forced excitation wave has already been reduced, the spectrum of the specific reflected wave is amplified by the frequency filtering process, and its frequency f _n (f ₁ and f ₂ in the figure) Is specified. The relationship between the fundamental frequency f _n of the reflected wave and the reflector is shown in the following mathematical formula 1. If the spectrum of the reflected wave is clear and there is no particular need for amplification, the frequency filtering process may not be applied to the modified received wave.

[0036]

## EQU1 ## f _n = nV / (2d) where V is the elastic wave velocity of the elastic body, d is the distance between the receiving sensor and the reflector, and n is a positive integer.

By substituting the value of f _n into Equation 1 above, the distance d between the receiving sensor and the reflector can be calculated. For example, by substituting f ₁ having a low frequency among the frequencies of the reflected wave shown in FIG. 7 into the above-mentioned mathematical formula 1, the distance between the distant reflector and the receiving sensor 7, that is, the concrete shown in FIG. The distance between the bottom surface of 1 and the receiving sensor 7 can be calculated. In addition, the high frequency f ₂ is calculated by the above formula 1
By substituting into the distance, it is possible to calculate the distance between the nearby reflector and the receiving sensor 7, that is, the distance between the reinforcing bar 4 and the receiving sensor 7.

As described above, in the present embodiment, the forced excitation wave component in the received wave is reduced and the reflected wave component is extracted. Therefore, the frequency of each reflected wave can be specified with high accuracy, and the concrete 1 Between the bottom of the base and the receiving sensor 7 and the rebar 4
The distance between the sensor and the receiving sensor 7 can be accurately measured.

There are various types of frequency filtering processing described above, and a specific example thereof will be described later.

Second Embodiment Method In the second embodiment method, the superimposed wave filtering method is applied to a concrete structure in which a plurality of reinforcing bars are arranged to specify the positions of the reinforcing bars. Before starting the description of the method of this embodiment, an outline of the concrete structure will be described.

Typical concrete structures include, for example, columns, beams and walls. FIG. 8 is a schematic view showing a column in which reinforcing bars are arranged, (a) is a plan sectional view, and (b) is a side sectional view showing an XX 'plane.
As shown in FIG. 8, the large diameter muscles 3 are arranged in the pillar 101 along the longitudinal direction (vertical direction) of the pillars, and the small diameter muscles 5 are arranged so as to be perpendicular to the large diameter muscles 3. Have been streaked. The large diameter muscles 3 are parallel to each other and are equally spaced. Similarly, the small diameter muscles 5 are parallel to each other and are equally spaced.

FIG. 9 is a schematic view showing a beam in which reinforcing bars are arranged, (a) is a front sectional view, and (b) is YY '.
It is a side surface sectional view showing a field. 10 is a schematic view showing a wall in which reinforcing bars are arranged, (a) is a front sectional view, and (b) is a side sectional view showing a ZZ 'plane. As shown in FIG. 9, large-diameter muscles 3 are horizontally arranged along the beam in the beam, and small-diameter muscles 5 are arranged so as to be perpendicular to the large-diameter muscle 3. Further, as shown in FIG. 10, a large diameter muscle 3 is arranged in the wall so as to be perpendicular to the ground, and a small diameter muscle 5 is arranged so as to be perpendicular to the large diameter muscle 3. ing. In each of the beam and the wall, the small diameter streaks 5 are parallel to each other and are equally spaced.

As described above, in concrete structures such as columns, beams and walls, the reinforcing bars are arranged at substantially equal intervals and parallel to each other. Accurately know the position of the reinforcing bar in the concrete by utilizing the structure of the concrete structure and oscillating and receiving the elastic wave in the concrete, and subjecting the received wave to the superimposed wave filtering process. be able to. That is, it is possible to accurately see through the concrete.

FIG. 11A is a schematic diagram showing concrete walls and measuring points in the second embodiment method, and FIG.
FIG. 6 is a sectional view showing a BB ′ surface. As shown in FIG. 11, reinforcing bars 11a are arranged in the concrete 1 so as to be parallel to the upper surface 2 of the concrete 1, and
The reinforcing bars 11b are arranged so as to be parallel to the bottom surface 8. The reinforcing bar 11a and the reinforcing bar 11b are parallel to each other. A reinforcing bar 12 is arranged between the reinforcing bars 11a and 11b, and the reinforcing bars 12 are perpendicular to the reinforcing bars 11a and 11b. Further, the respective reinforcing bars 12 are parallel to each other and arranged at equal intervals. A plurality of measurement points 9 are predetermined on the upper surface 2 of the concrete 1, and the measurement points 9 are installed at equal intervals L along the longitudinal direction of the concrete wall. As shown in FIG. 11B, the measurement point 9
An oscillating sensor 6 and a receiving sensor 7 are disposed in the concrete, and the oscillating sensor 6 oscillates an elastic wave in the concrete, and the receiving sensor 7 receives the elastic wave.

After placing the oscillation sensor 6 and the reception sensor 7 at the measurement point 9 at the concrete wall and the measurement point thus configured, an elastic wave is oscillated from the oscillation sensor 6 into the concrete 1, and the reception sensor 7 Receive elastic waves. Next, the received wave obtained at each measurement point 9 is subjected to a time series example filter to reduce the components other than the reflected wave and extract the reflected wave component. In this way, the modified received wave is obtained.

FIG. 12 is a graph showing the relationship between the two, with the horizontal axis representing the frequency (Hz) and the vertical axis representing the spectrum value, showing the Fourier spectrum of the modified received wave obtained at the measurement point 9. It is a graph figure. FIG. 12A shows a measurement point 9a.
Shows the Fourier spectrum at the measurement point 9 (b).
The Fourier spectrum in b is shown. As shown in Formula 1, the frequency of the reflected wave is inversely proportional to the distance (shortest distance) between the reception sensor 7 and the reinforced concrete. Therefore, the shortest distance L ₁ between the measurement point 9a and the reinforcing bar 11a is determined depending on whether the reflected wave from the reinforcing bar 11a is received at the measuring point 9a and the reflected wave from the reinforcing bar 11a is received at the measuring point 9b. Is the same as the shortest distance L ₂ between the measuring point 9b and the reinforcing bar 11a, the frequency g ₁ of the reflected wave from the reinforcing bar 11a at the measuring point 9a is equal to the measuring point 9 as shown in FIG.
It becomes equal to the frequency h ₁ of the reflected wave from the reinforcing bar 11a at b. Similarly, the reflected wave from the reinforcing bar 11b is measured at the measurement point 9a.
And 9b, the frequency g ₂ of the reflected wave from the reinforcing bar 11b at the measurement point 9a is the same as that at the measurement point 9b.
Is equal to the frequency h ₂ of the reflected wave from the reinforcing bar 11b. On the other hand, since the shortest distance L ₃ between the measuring point 9a and the reinforcing bar 12a is different from the shortest distance L ₄ between the measuring point 9b and the reinforcing bar 12a, it is different from the case where the reflected wave from the reinforcing bar 12a is received at the measuring point 9a. The frequency of the reflected wave is different from the case where the reflected wave from the reinforcing bar 12a is received at the measurement point 9b, as indicated by the frequencies g ₃ and h _{3 in} FIG.

From the above, the frequency of the reflected waves from the reinforcing bars 11a and 11b is constant regardless of the measuring points, while the frequency of the reflected waves from the reinforcing bars 12a and 12b is different at each measuring point. Therefore, the reception wave is received at each measurement point 9, the time-series filter is applied to each reception wave to reduce the forced excitation wave, the reflected wave is extracted as a modified reception wave, and the modified reception waves are added to each other. In the case where the combined received wave is subjected to Fourier transform, as shown in FIG. 13, among the spectrum values of the obtained Fourier spectrum, the reinforcing bar 11a is included.
The spectral value of the reflected wave from is high because the frequency g ₁ is constant. Similarly, in the spectrum value of the reflected wave from the reinforcing bar 11b, the frequency g ₂ is constant,
Its strength increases. That is, these spectral values are amplified, and in the ideal case, if the number of measurement points is s,
s times. On the other hand, since the frequency of the reflected wave from the reinforcing bar 12 differs depending on the measurement point, the spectral value of the reflected wave from the reinforcing bar 12 stands for each frequency, but its intensity does not increase and is substantially constant. Thereby, the reinforcing bar 11a
When a large number of measurement points are set along with, the received waves obtained at each measurement point are time-series filtered, and the received waves are added to each other, the reflected wave components of the reinforcing bars 11a and 11b are remarkable Fourier spectra. And the spectral value of these reflected waves becomes high. Therefore, when the received wave is time-series filtered and then frequency filtered, the reinforcing bar 11
Since the spectrum value of the reflected wave from a and 11b becomes high, the frequency of the reflected wave component of the reinforcing bars 11a and 11b can be easily specified, and the positions (depth) of the reinforcing bars 11a and 11b can be detected with high accuracy. it can.

Further, the fact that the distance between the rebars 11a and 11b and the upper surface 2 of the concrete 1 is constant is used to increase the spectral value of the reflected wave from the rebars 11a and 11b. If the wave is not buried in other reflected waves and the spectral value of the reflected wave is clear, the modified received wave may not be subjected to frequency filtering processing.

The method of this embodiment can be applied to other than the position measurement of the reinforcing bar as long as the distance between the upper surface of concrete and the reflector is constant. Further, in the present embodiment, each received wave is time-series filtered, a modified received wave is obtained for each received wave, the modified received waves are added together, and a combined received wave is obtained. Although the Fourier transform is performed, the order of the time series filter and the addition may be exchanged. Further, the order of addition and Fourier transform may be exchanged.

FIG. 14 (a) is a schematic view showing a concrete wall and measuring points, and FIG. 14 (b) is a sectional view showing a CC 'plane. 14, the same parts as those in FIG. 11 are designated by the same reference numerals, and detailed description thereof will be omitted. As shown in FIG. 14, the distance between the top surface 1 and the bottom surface 8 of the concrete 1 is constant, and each measurement point 9 is similar to the measurement point 9 shown in FIG.
Are installed at equal intervals L along the longitudinal direction of the concrete wall.

After placing the oscillation sensor 6 and the reception sensor 7 at each measurement point 9 on the concrete wall and the measurement point thus configured, an elastic wave is oscillated from the oscillation sensor 6, and this elastic wave is applied to the bottom surface 8. The received wave that is reflected and received by the reception sensor 7 is time-series filtered to reduce components other than the reflected wave to form a modified received wave. Then, among the frequencies of the components of the modified received wave, the frequency of the reflected wave from the bottom surface 8 is always constant. Therefore, when the received waves obtained at each measurement point 9 are time-series filtered and then Fourier-transformed and then the received waves are added to each other, the Fourier spectrum includes many reflected wave components of the bottom surface 8, The spectrum value of this reflected wave becomes large. Thereby, the position of the bottom surface 8 of the concrete 1 can be detected with high accuracy as in the case described above.

FIG. 15A is a schematic view showing a cracked concrete wall and a measuring point, and FIG. 15B is a sectional view showing a DD 'plane. 15, the same parts as those in FIG. 11 are designated by the same reference numerals, and detailed description thereof will be omitted. FIG.
As shown in, the concrete 1 is formed with a crack 10 having a constant depth, and the crack extends in the longitudinal direction of the concrete 10. Measurement points 9 are provided before and after the crack 10, and like the measurement points 9 shown in FIG. 11, the measurement points 9 are installed at equal intervals L along the longitudinal direction of the concrete wall.

In the cracked concrete wall and the measuring points thus constructed, the oscillation sensor 6 is attached to each measuring point 9.
After the reception sensor 7 is arranged, the oscillation sensor 6 oscillates an elastic wave. The elastic wave is reflected by the bottom surface 10 a of the crack 10 and is received by the reception sensor 7. A reflected wave is extracted by applying a time series filter to the obtained received wave and used as a modified received wave. Then, among the frequencies of the respective components of the modified received wave, the frequency of the reflected wave from the reflecting surface 10a is always constant. Therefore, as in the case described above, the reflecting surface 10a of the concrete 1, that is, the depth of the crack can be detected with high accuracy.

Next, the superposed wave filtering processing device will be described. FIG. 16 is a schematic diagram showing a superimposed wave filtering processing device according to an embodiment of the present invention. As shown in FIG. 16, the reception sensor 71 is connected to the A / D converter 72. The elastic wave received by the reception sensor 71 is converted into an electric signal, and this electric signal is converted into a digital signal by the A / D converter 72. The A / D converter 72 is connected to a computer 73 such as a personal computer or an EWS (Engineering Workstation), and the computer 73 has a CR.
T (display device) 74 and DISK (auxiliary storage device) 75
Is connected. As a result, the digital signal is taken into the computer 73, displayed on the CRT 74, and stored in the DISK 75. The computer 73 applies the above-mentioned time series filter to the received wave to calculate the modified received wave. The computer 73 is connected to a filtering device 76, and the filtering device 76 includes a filtering CPU 77 and a memory 78. The filtering device 76 is a DM
Computer 7 by A (Direct Memory Access) transfer
The data of the modified received wave is fetched from No. 3, and the filtering CPU 77 performs frequency filtering processing on this data at an extremely high speed. An elastic wave oscillation sensor (not shown) is provided at a predetermined place.

In the superposed wave filtering processing device thus constructed, an elastic wave is oscillated from the oscillation sensor, the elastic wave is reflected by a reflector or the like to generate a reflected wave, and the elastic wave forces the elastic body. A forced excitation wave is generated by the excitation. The forced excitation wave and the reflected wave are superposed to form a superposed wave, and the superposed wave is received by the receiving sensor 71 as a received wave 79. The reception sensor 71 converts the received wave 79 into an electric signal, and the electric signal is converted into a digital signal by an A / D converter and input to the computer 73. The computer 73 processes the digital signal, displays the received wave 79 on the CRT 75, and stores the received wave 79 data in the DISK 74.
Under the control of the computer 73, a time series filter is applied to the data, and only the reflected wave component of the received wave component is extracted and left. The data of the obtained modified reception wave is sent to the filtering device 76, and this data is written in the memory 78. The data in the memory 78 is filtered for C
PU77 performs ultra high-speed frequency filtering,
The obtained result is returned to the computer 73. in this way,
After the computer 73 performs the time-series filter on the received wave to calculate the modified received wave, the filtering CPU 77 performs the frequency filtering process on the modified received wave. Therefore, the frequency of the reflected wave can be accurately determined in a short time. You can decide.

Next, a filter used when the above frequency filtering processing is carried out will be described. Even when a large number of reflected waves are superposed to form a superposed wave and the superposed wave has an extremely complicated shape, if the superposed wave f (t) has a repeating cycle T (second). , Can be represented by a simple superposition of sinusoidal waves, as shown in Equation 2 below.

[0057]

F (t) = a ₀ + {Σ (n = 1, k) a _n cos (n
ωt)} + {Σ (n = 1, k) b _n sin (nωt)} where t: time (second) ω: angular frequency a ₀ , a _n , b _n : coefficient k, n: positive number.

It should be noted that Σ (n = 1, k) is defined as changing n from 1 to k and taking the sum. For example, {Σ
(N = 1, k) X n} represents _{X 1 + X 2 + ····· +} X k.

The filter in this embodiment will be described using the above equation 2. First, in order to simplify the discussion, the repetition period T is 1 second (the angular frequency ω of the above formula 2 is 2π), and k = 9, a ₀ = a ₁ = ... = a ₉
= 0 and b ₁ = b ₂ = ... = b ₉ = 1 the superimposed wave f (t) and the period f of this superimposed wave f (t) are shifted by Δt = 0.5 seconds. The property of (t + Δt) will be described.

In FIG. 17, the horizontal axis represents time t (seconds), the vertical axis represents superposed wave f (t), and a graph showing the relationship between the two is shown. The horizontal axis represents time t (seconds) and the vertical axis represents time t (seconds). Superposed wave f (t +
It is a graph which shows (DELTA) t and shows the relationship of both. That is, the upper part of FIG. 17 is a diagram showing Σ (n = 1, k) sin (2nπt). In the lower part of FIG. 17, Σ (n = 1, k) sin {(2nπ (t +
It is a figure which shows (DELTA) t).

FIG. 18 is a graph showing the relationship between the horizontal axis and the vertical axis representing time t (seconds) and amplitude. The solid line represents each sin (2nπt) constituting f (t). And the broken line represents each sin {(2
nπ (t + Δt)} is shown. FIG. 19 shows the wave G ₁ (t) obtained by substituting each sine wave shown in FIG.

[0062]

## EQU3 ## G ₁ (t) = Σ (n = 1, k) sin (2nπt) −
Σ (n = 1, k) sin {(2nπ (t + Δt)}

Further, each n component of G ₁ (t) (frequency n)
Is shown in FIG. The numbers on the right side in FIG. 20 indicate the maximum amplitude of each n component. Similarly, by substituting each sine wave shown in FIG. 18, the equation shown in the following equation 4 is calculated, and the obtained wave G ₂ (t) is shown in FIG.

[0064]

[Number 4] _{G 2 (t) = Σ (} n = 1, k) sin (2nπt) +
Σ (n = 1, k) sin {(2nπ (t + Δt)}

Further, each n component of G ₂ (t) (frequency n)
Is shown in FIG. The numbers on the right side in FIG. 22 indicate the maximum amplitude of each n component.

From FIGS. 19 and 20, the following can be said. That is, the superimposed wave f (t) and the cycle T (T
= 1.0 second) is shifted by a half cycle (ΔT = T / 2) from the superposed wave f (t + ΔT), the fundamental frequency f ₀ (= 1 / T) and this fundamental frequency The amplitude of the sin wave having an odd multiple (3 / T, 5 / T, 7 / T, and 9 / T) is twice that of an even multiple (2 / T, 4 / T, 6).
/ T and 8 / T) sin waves have an amplitude of 0. As a result, the above mathematical formula 3 is calculated from the superimposed wave f (t) to the fundamental frequency f
Removing the ₀ even multiple frequency components of an odd-numbered multiple frequency components of the fundamental frequency f ₀ can be said that a is a filter to amplify doubled. Then, FIG. 19 can be interpreted as a diagram showing the result of applying this filter to f (t).

On the other hand, as shown in FIG. 22, in the equation 4, the amplitude of the odd harmonic frequency of the fundamental frequency f ₀ is set to 0, and the amplitude of the even harmonic frequency of the fundamental frequency f ₀ is doubled. As a result, the above-mentioned mathematical formula 3 is obtained from the superimposed wave f (t) to the fundamental frequency f ₀
It can be said that it is a filter that removes the component of the odd-numbered harmonic frequency and doubles the component of the even-numbered harmonic frequency. Then, FIG. 21 can be interpreted as a diagram showing the result of applying this filter to f (t).

Next, the fundamental frequency f _{0 is set} to 4 Hz, the cycle increment Δt in this case is determined to be 1 / (2f ₀ ) = 0.125 seconds, and further k = 9, a ₀ = a ₁ =.・ ・ ・ ・ ＝ a ₉ ＝
0 and b ₁ = b ₂ = ... = b ₉ = 1 the superimposed waves f (t) and f (t + Δ
The property of t) will be described. In FIG. 23, the horizontal axis represents time t.
(Sec), the vertical axis is the superimposed wave f (t), and the horizontal axis is the time t (seconds), and the vertical axis is the superimposed wave f (t + Δt). It is a graph which shows the relationship of. FIG. 24 is a diagram showing the relationship between the time t (seconds) on the horizontal axis and the n components (frequency n) of the superimposed waves f (t) and f (t + Δt) on the vertical axis, showing the relationship between the two. Is f
(T) shows each n component sin (2nπt), and the broken line is f
Each n component of {t + Δt) sin {(2nπ (t + Δ
t)} is shown. The wave G ₁ (t) obtained by calculating the formula shown in Formula 3 is shown in FIG. Also, each n of G ₁ (t)
The component (frequency n) is shown in FIG. The number on the right side in FIG. 26 indicates the maximum amplitude of each n component. As shown in FIG. 34, when the fundamental frequency f ₀ is 4 Hz, the above formula 3 doubles the amplitude of the odd-numbered harmonic frequency of the fundamental frequency f _{0 and} sets the amplitude of the even-numbered harmonic frequency to 0. To do. Further, for a wave whose frequency is close to an odd-numbered frequency, the amplitude of the wave is amplified, while for a wave whose frequency is close to an even-numbered frequency, the amplitude of the wave is attenuated. FIG. 25 is a diagram showing a wave obtained by applying such a filter.

Similarly, using each n-component wave shown in FIG. 24, the equation shown in the above equation 4 is calculated, and the obtained wave G is obtained.
₂ (t) is shown in FIG. 28 shows each n component (frequency n) of G ₂ (t). The numbers on the right side of FIG. 28 indicate the maximum amplitude of each n component. As shown in FIG. 28, in Equation 4, the amplitude of the odd-numbered frequency of the fundamental frequency f ₀ is set to 0, and the amplitude of the even-numbered frequency is doubled. For a wave whose frequency is close to an odd-numbered frequency, the amplitude of the wave is attenuated, while for a wave whose frequency is close to an even-numbered frequency, the amplitude of the wave is amplified. FIG. 27 is a diagram showing a wave obtained by applying such a filter.

The above consideration is based on the above equation 2 with k = 9, a ₀ = a ₁ = ... = a ₉ = 0 and b ₁ =
When b ₂ = ... = b ₉ = 1, that is, si
As for n waves, a ₁ = ... = a ₉ =
1 and a ₀ = b ₁ = b ₂ = ... = b ₉ = 0, co
The above considerations can be applied to the s wave.

The properties of the above filter will be described using a Fourier spectrum. FIG. 29 shows Σ (n = 1, k) sin
For (2nπt), the horizontal axis is the frequency (Hz),
It is a graph which shows a spectrum value on a vertical axis and shows a relation between both. This spectrum has a constant spectrum value of 1.0. Then, FIG.
FIG. 4 is a graph showing a spectrum value obtained by applying the above equation, in which a horizontal axis represents a frequency (Hz) and a vertical axis represents a spectrum value, showing a relationship between the two. Equation 3 above is Σ
The waveform obtained by moving this waveform by the period Δt is subtracted from the waveform represented by (n = 1, k) sin (2nπt). The action of Equation 3 will be examined from the relationship between the frequency and the spectrum value. As shown in FIG. 29, Δt = 1 / (2f
₀ ), that is, when f ₀ = 1 / (2Δt), as shown in FIG. 30, the frequencies f ₀ , 3f ₀ , 5f ₀ , ...
The spectrum value of is doubled and the frequency is 2f ₀ , 4
The spectral values of f ₀ , 6f ₀ , ... Are set to 0. Further, the frequencies are f ₀ , 3f ₀ , 5f ₀ , ... Amplifying the spectrum values in the vicinity, and the frequencies are 2f ₀ , 4f ₀ , 6f ₀ ,
...... Attenuates nearby spectrum values. Formula 3 above performs such frequency filtering processing,
This means that a spectrum with an odd multiple of the fundamental frequency is extracted.

FIG. 31 is obtained by applying the equation 4 to FIG.
It is a figure which shows the obtained spectrum value, is a graph which shows a frequency (Hz) on a horizontal axis and a spectrum value on a vertical axis, and shows the relationship between both. Equation 4 above is Σ (n = 1, k)
The waveform represented by sin (2nπt) and the waveform obtained by moving this waveform by the period Δt are added. Formula 4 above
From the relationship between the frequency and the spectrum value, the effect of is as shown in FIG. 31, the frequencies f ₀ , 3f ₀ , 5
The frequency is 2 with the spectrum value of f ₀ , ...
The spectral values of f ₀ , 4f ₀ , 6f ₀ , ... Are doubled. The frequencies are f ₀ , 3f ₀ , 5f ₀ , ...
... Attenuation of nearby spectrum values, and frequency is 2
f ₀ , 4f ₀ , 6f ₀ , ... Amplifies nearby spectrum values. Formula 4 above means that such frequency filtering processing is performed to extract a spectrum of an even multiple of the fundamental frequency.

An example in which the above-described frequency filtering process is performed on an actual waveform will be shown. FIG. 32 is a graph showing the relationship between the two with the horizontal axis representing time t (seconds) and the vertical axis representing amplitude. Δt = 30 μsec, f ₀ = 1 / (2Δ
t) ≈16.6 kHz, and as shown in FIG. 32, when waveforms are added and subtracted, F ₁ (t) = f (t) −
f (t + Δt), F ₂ (t) = f (t) + f (t + Δ
A time series wave represented by t) is obtained. Fourier transform is performed on each of F ₁ (t) and F ₂ (t).

FIGS. 33 (a) and 33 (b) respectively show f
(T) is a Fourier spectrum of F ₁ (t), in which the frequency (Hz) is plotted on the horizontal axis and the spectrum value is plotted on the vertical axis, showing the relationship between the two. The broken line in FIG. 33A shows the shape of the filter. The spectrum of f (t) in the vicinity of f ₀ ≈16.6 kHz is amplified by the filter shown by the broken line, and the spectrum of the frequency extremely different from f ₀ is attenuated.

34 (a) and 34 (b) respectively show f
(T) is a Fourier spectrum of F ₂ (t), where the horizontal axis is the frequency (Hz) and the vertical axis is the spectrum value, showing the relationship between the two. The broken line in FIG. 34 (a) shows the shape of the filter. f (t) is attenuated by the filter indicated by the broken line in the spectrum near f ₀ ≈16.6 kHz, and the spectrum having a frequency extremely different from f ₀ is amplified.

Let us consider what mathematical expression the filters in FIGS. 33 (a) and 26 (a) can represent. Let us consider a component in which nω is R in the above formula 1 by taking it out from f (t). f _R (t) and F for this f _R (t)
₁ (t) and F ₂ (t) are shown in Equations 5 to 7 below.

[0077]

F _R (t) = a _R cos (Rt) + b _R sin (Rt)

[0078]

[Equation 6] F _1R (t) = f _R (t) −f _R (t + Δt)

[0079]

## _EQU7 ## F _2R (t) = f _R (t) + f _R (t + Δt) The above equation 5 is substituted into equation 6 for rearranging. The obtained result is shown in the following formula 8.

[0080]

## _EQU8 ## F _1R (t) = C ₁ {a _R cos (Rt + R · Δt
/ 2) + b _R sin (Rt + R · Δt / 2)} where C ₁ is defined by the following formula 9.

[0081]

## EQU9 ## C ₁ = 2cos (RΔt / 2) Similarly, Equation 5 is substituted into Equation 7 for rearranging. The obtained result is shown in the following mathematical formula 10.

[0082]

[Formula 10] F _2R (t) = C ₂ {a _R sin (Rt + R · Δ
t / 2) + b _R cos (Rt + R · Δt / 2)} where C ₂ is defined by the following mathematical formula 11.

[0083]

[Equation 11] C ₂ = ₂ sin (R · Δt / 2)

The filter shown by the broken line in FIG. 33 (a) is the one expressed by the above-mentioned mathematical expression 9, and is shown in FIG. 34 (a).
The filter shown by the broken line in the above is the filter shown by the above-mentioned mathematical expression 11.

In Equation 6, the coefficient of f _R (t) was 1, and the coefficient of f _R (t + Δt) was −1. Further, in the above formula 7, the coefficients of f _R (t) and f _R (t + Δt) were both 1. The above equations 6 and 7 are generalized to the following equations 12 and 13, respectively, so that other coefficients can be taken.

[0086]

F ₁ (t) = {f (t) −af (t + Δt)} / b

[0087]

F ₂ (t) = {f (t) + af (t + Δt)} / b where a and b are positive real numbers.

Further, Δt is defined by the following formula 14.

[0089]

Δt = m / 2f ₀ where m is a positive integer.

The coefficients "a" and "b" in the equations 12 and 13 and the coefficient "m" in the equation 14 are set to appropriate values, and the superimposed wave is subjected to the frequency filtering process shown in the equation 12 or 13. In the above formula 12, f (t)
Among the components of the wave which constitutes a odd multiple frequency components of f ₀ / m and f ₀ / m is amplified, even multiple frequency components of f ₀ / m is reduced or removed. In this case, when the maximum amplification rate is X and the maximum attenuation rate is Y, simultaneous equations represented by a + b = X and a−b = Y may be calculated and the values of a and b may be determined. However, a and b are not limited to this value. Also, m may be determined according to the purpose, but when m is 1, the components of the odd harmonic frequencies of f ₀ and f ₀ are amplified, and the components of the even harmonic frequency of f ₀ are reduced or deleted. It In the case where the 2 m, odd multiple frequency components of f _0/2 is amplified, even multiple frequency components of f _0/2 is reduced or removed. That is, it is possible to reduce or eliminate the component of the integral multiple frequency of f ₀ .

[0091] Also, among the components of the wave constituting the f (t) in the equation 13, an odd multiple frequency components of f ₀ / m and f ₀ / m is removed or reduced, the f ₀ / m The components of even frequency are amplified. In this case, the values of a and b can be determined by calculating the simultaneous equations as described above. However, a and b are not limited to this value as in the case described above. Also, m may be determined according to the purpose. When m is 1, the components of f ₀ and the odd harmonic frequency of f ₀ are deleted or reduced, and the components of the even harmonic frequency of f ₀ are deleted or reduced. To be done. When m is 2, f ₀
/ ₀ odd harmonic frequency components are deleted or reduced, and f ₀ /
A component with an even frequency of 2 is amplified. That is, it is possible to amplify the component of the integral multiple frequency of f ₀ .

By applying a frequency filtering process to the superposed wave using such a filter, it is possible to amplify the necessary component in the superposed wave and to eliminate or reduce the unnecessary component. , It is possible to know the frequency of the reflected wave that constitutes the superimposed wave. The result of actual frequency filtering processing by setting a, b and m will be described in the embodiment.

[0093]

EXAMPLE Next, the result of actually extracting the reflected wave from the received wave and calculating the distance from the receiving sensor to the reflecting object by the above-mentioned superimposed wave filtering method will be described.

First Example In the reinforced concrete model shown in FIG. 1, the distance between the bottom surface 8 of concrete and the receiving sensor 7 and the distance between the reinforcing bar 4 and the receiving sensor 7 were calculated. The concrete used has a thickness of 175 cm and the diameter of the reinforcing bars in this concrete is 12 mm. Also, from the upper surface 2 to the rebar 4
Is 4.5 cm, and the distance between the oscillation sensor 6 and the reception sensor 7 is 3.5 cm.

Reinforced concrete 1 constructed in this way
At, the elastic wave is oscillated from the oscillation sensor 6 and received by the reception sensor 7. FIG. 35 is a graph showing the relationship between the oscillation frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, showing the Fourier spectrum of the oscillating wave (broken line) 91 and the received wave (solid line) 92. It is a figure. FIG.
As shown in FIG.
It is distributed around 751 kHz. Such an oscillating wave 91 was oscillated in the reinforced concrete 1 to obtain a received wave 92 having a plurality of peaks.

First, the result of performing only the frequency filtering process without using the time series filter in this embodiment will be described. The received wave 92 was frequency-filtered in order to amplify the necessary spectral component from the received wave 92 and reduce the unnecessary spectral component. This filtering process was performed after removing the spectra unrelated to the perspective of concrete. That is, the spectrum with a frequency of 1 kHz or less and the frequency of 9
After deleting the spectrum of 0 kHz or more, frequency filtering processing was performed. This filtering process was performed by correlating the oscillating wave 91 and the received wave 92, that is, after the received wave 92 was delayed by the oscillating wave 91, the basic frequency f ₀ of the filter was set to 62.419 kHz. . A specific filter of the frequency filtering process in this embodiment will be described later.

FIG. 36 shows the frequency (kHz) on the horizontal axis,
It is a graph which shows a Fourier spectrum on a vertical axis and shows a relation of both, and is a figure which shows a spectrum obtained by frequency-filtering the spectrum of the received wave 92.
Since the reflected waves indicated by the arrows 61 and 62 in FIG. 1 are buried in the forced excitation wave indicated by the arrow 63, the spectrum of the forced excitation wave is predominant and the spectrum of the reflected wave is buried as shown in FIG. Will end up. Thus, the received wave 9
If only 2 is subjected to frequency filtering, the received wave 9
It is difficult to specify the frequency of the reflected wave in 2. Therefore, the time series filter of the present embodiment is applied to the received wave 92 with t _h = 0 and t _a = 500 μsec, and then frequency filtering processing is performed.

37 to 39, the horizontal axis shows the frequency (kHz).
And a Fourier spectrum on the vertical axis to show the relationship between the two, and is a diagram showing a spectrum obtained by frequency-filtering the spectrum of the received wave 92 after time-series filtering. . In FIG.
The spectrum 31s shows the spectrum shown in FIG. 36, that is, the spectrum that has not been subjected to the time series filter. The spectra 31a, 31b, 31c and 31e are
Each shows a spectrum subjected to time series filters 1, 2, 3, and 5 times. The spectrum obtained by applying the time-series filter 4 times is very similar to that obtained by applying the time series filter 5 times, and thus is not shown in the graph. Further, in the time series filter H (t) shown in FIG.
A sine wave was applied to the portion indicated by 2.

As shown in FIG. 37, each peak becomes clear as the number of times the time series filter is applied increases. In particular, focusing on each spectrum having a frequency of 40 kHz to 45 kHz, there is only one large peak in the spectrum 31s, but the spectrum 31e is time-series filtered 5 times, Peaks have been found. The peak with the highest frequency exists at the position of 47.204 kHz, as shown by the broken line in FIG. 38, but this peak is caused by the vibration characteristics of the reception sensor 7, so this peak is excluded. . When this peak is excluded, the peak with the highest frequency exists at the position of 43.376 kHz as shown by the broken line in FIG. Substituting the frequency into f _n , n into 1, and sound velocity of concrete into 4000 m / sec and substituting it into the above formula 1, d becomes 46.1 mm. That is, the depth of the reinforcing bar 4 is 4.
It is approximately the same as 5 cm (45 mm). The lowest frequency peak is 1.13 as shown by the broken line in FIG.
It is located at 9 kHz. Substituting this frequency into f _n and substituting it in the above formula 1, d is calculated, and d is 175.6.
cm. That is, the distance between the receiving sensor 7 and the bottom surface 8 of the concrete is approximately 175 cm.

In FIG. 37, the frequency is 43.3.
The peak at 76 kHz does not seem clear. This matches the peak of the frequency of 47.204 kHz,
This is because each peak is smaller because the vertical axis of the graph is determined. For example, as shown in FIG. 40, the peak at the frequency of 43.376 kHz becomes clear by changing the scale of the vertical axis of the graph.

In FIG. 41, the horizontal axis represents frequency (kHz),
FIG. 7 is a graph showing the relationship between the two by taking the Fourier spectrum on the vertical axis, setting t _h = 0 and t _a = 300 μsec, applying a time series filter to the spectrum of the received wave 92, and then performing frequency filtering processing. It is a figure which shows the obtained spectrum. In FIG. 41, the spectrum 31s is shown in FIG.
6 shows the spectrum without the time series filter shown in FIG. Further, the spectra 32a, 32b, 32c, 3
2d and 32e respectively show spectra subjected to the time series filter 1 to 5 times. Similarly, FIG. 42, t _h
= 0, and a t _a = 200 [mu] seconds, after applying a time series filter spectrum of the received wave 92 is a diagram showing a spectrum obtained by frequency filtering. FIG.
In, the spectra 33a, 33b, 33c, 33d
And 33e respectively show spectra subjected to the time series filter 1 to 5 times.

In any of the spectra shown in FIGS. 41 and 42, as in the spectrum shown in FIG. 38, the highest frequency peak exists at the position of 47.204 kHz, but this peak is the vibration of the receiving sensor 7. This peak is excluded because it is caused by the characteristics. As shown in FIGS. 41 and 42, the highest frequency peak is 43.3.
It exists at the position of 76 kHz, and the lowest frequency peak exists at the position of 1.139 kHz. In other words, the maximum frequency of the peak and the lowest frequency of the peaks in Figure 41 and 42 are consistent with those in FIG. 37, even if there are some differences in the setting of t _a, the vibration of each peak The number is accurately derived, which allows the distance between the receiving sensor 7 and each reflector to be accurately measured.

The following filters were used in the frequency filtering process. FIG. 43 is a graph showing the relationship between the horizontal axis and the vertical axis representing the frequency (Hz) and the vertical axis, respectively. As shown in FIG. 43, each of the filters 81e, 82e, and 83e has a fundamental frequency f.
The spectral component of ₀ and the odd multiple of this frequency (3
The spectral components of f ₀ , 5f ₀ , ...) Are deleted or reduced. Moreover, this filter amplifies the spectral components of even-numbered frequencies (2f ₀ , 4f ₀ , ...). This filter is made on the basis of C ₂ in Expression 11 above. That is, assuming that m is 1, Δt = m / 2f
_The filter is based on the calculation shown in the above-mentioned formula 13 with ₀ = 1 / 2f ₀ .

Of the filters shown in FIG. 43, the filter 8
1e is obtained by multiplying C ₂ by 1/2. That is, F ₂ (t) where a is 1 and b is 2 in the above formula 13
Is calculated, and the obtained F ₂ (t) is a filter 81e for f (t).
Multiplied by The filter 82e is a filter that applies the filter 81e twice, and is a filter 83e.
Is a filter that applies the filter 81e three times. In the above embodiment, first the received wave is time-series filtered,
After using the modified received wave, and setting the fundamental frequency f ₀ of the filter shown in FIG. 43 to 62.419 kHz, this filter was applied once to the modified received wave to identify the frequency of each reflected wave.

Second Example A second example is one in which the position and diameter of the reinforcing bar in concrete are actually measured by the method of the second example in the embodiment of the present invention. However, in the second example, the concrete 1 in which one reinforcing bar 4 was reinforced was used.

FIG. 44 is a schematic diagram showing reinforced concrete and measuring points. As shown in FIG. 44, concrete 1
Reinforcing bars 4 are arranged inside the concrete 1 so as to be parallel to the upper surface 2 of the concrete 1. The reinforcing bar 4 is round steel, and its diameter is 12 mm. The distance d between the upper end of the reinforcing bar 4 and the upper surface 2 is 4.5 cm. Top 2 of concrete 1
Has a plurality of measurement points 9 set in advance, and the measurement points 9 are installed at equal intervals L along the longitudinal direction of the concrete wall. L is 1.5 cm, 20 measurement points, 2
Receives 0 waves.

After the oscillation sensor 6 and the reception sensor 7 are arranged at the measurement point 9 at the reinforced concrete wall and the measurement point thus configured, the oscillation sensor 6 oscillates an elastic wave into the concrete 1, and the reception sensor 7 Receive elastic waves. 45 and 46 are graphs showing the relationship between the horizontal axis and the vertical axis representing the frequency (Hz) and the spectrum value, respectively, showing the Fourier spectrum of the received wave obtained at each measurement point 9. It is a graph figure. Each Fourier spectrum was obtained by processing the received wave as follows. First, the received wave is Fourier-transformed, and 6k is obtained from the obtained Fourier spectrum.
The spectra below Hz and above 100 kHz were deleted. Next, this Fourier spectrum was subjected to inverse Fourier transform, and the obtained time series example wave was subjected to frequency filtering processing with f ₀ of the filter 81n shown in FIG. 47 set to 123.535 kHz. The method of deriving this filter will be described later. Next, in the time series wave after frequency filtering processing,
As 4095μs the elapsed time t _a of FIG. 4, over twice the time series filter far off, further over twice the time series filter short-range cut-off, was modified received wave.

The short-distance cutoff time series filter is
A filter shown in FIG. 4, a zero magnitude at the arrival time t _h, refers to the size of the after t _a elapsed is 1. A specific shape of the short-distance cutoff time series filter H (t) used in the present embodiment is shown in the following formula 15.

[0109]

Equation 15] H (t) = sin (π / (2t a) × (t-t h)) _{where, t h ≦ t ≦ t h} + t a

On the other hand, the long-distance cutoff time series filter is
The size at the arrival time t _h is 1, and the size after the passage of t _a is 0. A specific shape of the long-distance cutoff time series filter H (t) used in this embodiment is shown in the following formula 16.

[0111]

Equation 16] H (t) = 1-sin (π / (2t a) × (t-t h)) _{where, t h ≦ t ≦ t h} + t a

[0112] When subjected to the time series filter near field cut off the superimposed wave, the amplitude in the vicinity of t _a is relatively amplified, the noise component in the vicinity of this is also amplified. Amplification of noise components can be suppressed by using a long-distance blocking time-series filter in addition to a short-distance blocking time-series filter.

The frequency f _{0 is set} to 3 in the obtained modified received wave.
The filter 82a shown in FIG. 48 is set to 0.762 kHz.
Therefore, frequency filtering processing was performed. Next, the frequency f ₀ is set to 59.570 kHz, the frequency filtering process is performed once by the filter 81e shown in FIG. 43, and the frequency f ₀ is set to 91.146 kHz, and the frequency filtering process shown in FIG. 43 is performed once. gave. The time-series wave after the frequency filtering process was Fourier-transformed to obtain a Fourier spectrum. The filter shown in FIG. 48 will be described later.

In each Fourier spectrum shown in FIGS. 45 and 46, the frequency indicated by the solid vertical line is the natural frequency of the receiving sensor. When this natural frequency peak is excluded and considered, the peak existing on the highest frequency side among the spectrum peaks is 3 as shown by the dotted line in FIG.
It exists near 7.445 kHz. Further, as shown by the dotted line in FIG. 46, there is a peak near 30.424 kHz.

FIGS. 49 and 50 are graphs showing the relationship between the horizontal axis and the vertical axis representing the frequency (Hz) and the spectrum value, respectively, showing the sum of the Fourier spectra at each measurement point. is there. However, in the Fourier spectra shown in FIGS. 49 and 50, the frequency f ₀ is 45.247 kHz.
The frequency filtering process shown in FIG.
After being applied 00 times, the frequency f ₀ is set to 49.642 kHz, and is obtained by applying the frequency filtering process shown in FIG. 51 100 times, and the spectrum by the receiving sensor is deleted. This deleted spectrum is shown by a dotted line. The filter shown in FIG. 51 will be described later.

As shown in FIG. 49, since the spectrum by the receiving sensor is deleted, the peaks on the high frequency side that can be clearly discriminated have a frequency of 37.435 k.
Hz. Further, as shown in FIG. 50, the next distinguishable peak is at 29.378 kHz.

In order to confirm whether or not these peaks are due to the reflected wave from the reinforcing bar 4, frequency filtering processing is sequentially performed, and the fundamental frequency f ₀ and its higher order vibrations are calculated from the low frequency side of the spectrum. Several components were deleted from the spectrum. The wave subjected to frequency filtering is G _A (t)
Then, the frequency filtering process shown in FIG. 51 is applied to G _A (t) N times in succession to obtain a wave G _B (t). next,
Calculate G _A (t) -G _B (t). For example, the filter 83i shown in FIG. 51 is a filter that applies the frequency filtering process 50 times, and only the fundamental frequency f ₀ and its higher-order frequency components remain, and the other components are substantially deleted. G _B subjected to such frequency filtering processing
By subtracting (t) from G _A (t), a wave in which the fundamental frequency f ₀ and its higher-order frequency components are substantially deleted can be obtained, and the vibration shown in FIG. 52 is shown in G _A (t). This means that the number filtering process (filter 81u) has been performed. This frequency filtering process was carried out for various basic frequencies f ₀ . The number of times N and the basic frequency f ₀ in each frequency filtering process are shown in Table 1 below.

[0118]

[Table 1]

53 and 54, the horizontal axis shows the frequency (Hz).
FIG. 4 is a graph showing the relationship between the two by taking the spectrum value on the vertical axis, and showing the Fourier spectrum after frequency filtering processing. As shown in FIG. 53, there is a clear spectrum peak at the position of the frequency of 37.435 kHz, and when this frequency is substituted into the above equation 1 with n being 1, the distance d of the above equation 1 The value is 5.3 cm. However, the sound velocity V in the concrete 1 was set to 4000 m / sec. Similarly, as shown in FIG.
There is a clear spectrum peak at a frequency of 29.378 kHz, and the frequency is set to 1 and V is set to 40.
When substituting 00 m / sec into Equation 1, the value of d in Equation 1 is 6.8 cm. Assuming 5.3 cm is the distance between the concrete 1 and the top of the rebar 4, 6.8
Assuming that cm is the distance between the concrete 1 and the lower end of the reinforcing bar 4, the diameter of the reinforcing bar is 6.8 cm-5.3 cm, that is, 1.5 cm (15 mm).

Generally, the speed of sound near the upper surface 2 of the concrete 1 is about 10 to 20% lower than the speed of sound in the concrete 1. Considering this point, the sound velocity is 4000m /
Table 2 below shows the result of deriving the diameter of the reinforcing bar 4 by calculating the obtained distance d on the assumption of seconds, 3600 m / sec and 3400 m / sec.

[0121]

[Table 2]

As shown in Table 2, the sound velocity V is 0.85.
When corrected by doubling, the distance d between the upper end of the reinforcing bar 4 and the upper surface 2 is calculated to be 45 mm (4.5 cm), which matches the actual distance and the diameter of the reinforcing bar is 12.8 mm.
Was calculated, and the actual diameter was approximately 12 mm. From the above, the sound velocity V near the upper surface 2 of concrete 1 is 3400 m
Assuming / sec, the distance and diameter between the upper end of the rebar 4 and the upper surface 2 are in agreement with the actual ones.

The filters shown in FIGS. 47, 48 and 51 will be described below. As shown in FIG. 47, each of the filters 81e, 82e, and 83e has a spectral component of the fundamental frequency f ₀ and an integral multiple frequency of this frequency (2f ₀ , 3f ₀ ,
......) is reduced. This filter is made on the basis of C ₁ in Expression 9 above. However, when m is 2, Δt = m / 2f ₀ = 1 / f _0, and the filter is based on the calculation shown in the above-mentioned formula 12.

Of the filters shown in FIG. 47, the filter 8
1n is obtained by multiplying C ₂ by 1/2. That is, F ₁ (t) where a is 1 and b is 2 in the above formula 12
Is calculated, and frequency filtering processing is performed. The obtained F ₁ (t) is the filter 81n in f (t).
Multiplied by The filter 82n is a filter that applies the filter 81n 20 times, and is the filter 83n.
n is a filter that applies the filter 81n 50 times.

As shown in FIG. 48, each filter 81a,
Reference numerals 82a and 83a denote spectral components of the fundamental frequency f ₀ and odd-number multiples (3f ₀ , 5f ₀ , ...) Of this frequency.
Amplify the spectral component of (). The filter also removes or reduces the spectral components of even harmonic frequencies (2f ₀ , 4f ₀ , ...). This filter is made on the basis of C ₁ in Expression 9 above. That is,
This is a filter based on the calculation shown in the above-mentioned formula 13 with m = 1 and Δt = m / 2f ₀ = 1 / 2f ₀ .

Of the filters shown in FIG. 48, the filter 8
1a is obtained by multiplying C ₂ by 1/2. That is, F ₁ (t) where a is 1 and b is 2 in the above formula 11
Is calculated and the frequency filtering process is performed. The obtained F ₁ (t) is converted into f (t) by the filter 81a.
Multiplied by The filter 82a is a filter that applies the filter 81a twice, and is a filter 83a.
Is a filter that applies the filter 81a three times.

As shown in FIG. 51, each filter 81i,
Reference numerals 82i and 83i denote spectral components of the fundamental frequency f ₀ and integral multiple frequencies (2f ₀ , 3f ₀ , ...) Of this frequency.
Amplify the spectral component of (). This filter is made on the basis of C ₂ in Expression 11 above. Where m
Is set to 2, and Δt = m / 2f ₀ = 1 / f _0, and the filter is based on the calculation shown in the above-mentioned formula 13.

Of the filters shown in FIG. 51, the filter 8
1i is obtained by multiplying C ₂ by 1/2. That is, F ₂ (t) where a is 1 and b is 2 in the above formula 13
Is calculated, and the obtained F ₂ (t) is a filter 81i for f (t).
Multiplied by The filter 82i is a filter that applies the filter 81i 20 times, and the filter 83i
i is a filter that applies the filter 81i 50 times.

Third Embodiment The third embodiment specifies the distance from the concrete top surface to the bottom surface and the distance from the concrete top surface to the crack tip (reflection surface) in the cracked concrete shown in FIG. The thickness of the concrete used,
That is, the distance from the top surface 2 to the bottom surface 8 is 30 cm, and the crack depth, that is, the distance from the top surface 2 to the reflecting surface 10a is 20 cm. Similar to the second embodiment, the measurement point 9 is set to 20
It is installed at one place, and the distance L between each measuring point 9 is 1.5 cm
A total of 20 waves were received at all measurement points 9.

55 and 56, the horizontal axis shows the frequency (Hz).
FIG. 4 is a graph showing the relationship between the two by taking the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point 9. Each Fourier spectrum was obtained by processing the received wave as follows. Each measurement point 9
For the received waves obtained by, as 4095μs the elapsed time t _a of FIG. 4, when multiplied by the sequence filter. Specifically, each reception wave is subjected to a long-distance cutoff time-series filter twice, and further, a short-distance cutoff time-series filter is applied twice to form a modified reception wave. After Fourier transforming the obtained modified received wave, spectra of 3 kHz or less and 100 kHz or more were deleted to obtain Fourier spectra shown in FIGS. 55 and 56. 55 where n is 1 in the above formula 1 and the sound velocity V in the concrete 1 is 4000 m / sec, the broken line in FIG. 55 shows the frequency at which the peak of the reflected wave by the bottom surface 8 appears and the integral multiple thereof. is there. Similarly, the broken line in FIG. 56 shows the position where a reflected wave is generated by the reflecting surface 10a of the crack 10 and a peak of this reflected wave appears.

As shown in FIG. 55, in some spectra, the peaks are buried and there is no clear peak at the position of the broken line. Similarly, as shown in FIG. 56, the peak is buried depending on the spectrum,
In some cases, there is no clear peak at the position of the broken line. Therefore, the received waves obtained at each measurement point are time-series filtered, and then Fourier-transformed, and then the received waves are added to each other.

57 and 58 are graphs showing the relationship between the horizontal axis and the vertical axis representing the frequency (Hz) and the spectrum value, respectively, showing the sum of Fourier spectra at each measurement point. is there. As shown in FIG. 57, there is a peak at a frequency of 6.669 kHz, and the frequency of this peak is substituted into the above mathematical formula 1 to calculate the distance d between the top surface 2 and the bottom surface 8 of the concrete 1. Then, d is 30
cm. In addition, n was set to 1 and V was set to 4000 m / sec.

Further, as shown in FIG. 58, the frequency is 1
There is a peak at a position of 0.011 kHz, and from the frequency of this peak, the crack 10 is cracked from the upper surface 2 of the concrete 1.
When the distance d to the reflecting surface 10a is calculated, d is 20c
m.

As described above, in the present embodiment, after the time series example sequence filter is applied to each received wave, the Fourier transform is performed and the obtained Fourier spectra are added to increase the distance between the reflecting object and the concrete upper surface. It was possible to calculate with accuracy. In this example, since the peak due to the reflected wave was clear, the frequency filtering process was not performed.

[0135]

As described above, according to the present invention,
Since the time-series filter is applied to the received wave, the frequency of the reflected wave in the received wave can be specified with high accuracy. As a result, it is possible to specify the position and the like of the reflector in the substance with high accuracy.

[Brief description of drawings]

FIG. 1 is a schematic view showing a reinforced concrete model in the first embodiment method of the present invention.

FIG. 2 is a cross-sectional view showing a plane AA ′ in FIG.

FIG. 3 is a graph showing the relationship between the two, with the horizontal axis representing time t (seconds) and the vertical axis representing amplitude.

FIG. 4 is a graph showing a relationship between the two, with the horizontal axis representing time t (seconds) and the vertical axis representing time series filter H (t), and a diagram schematically showing the time series filter.

FIG. 5 is a graph showing the relationship between the horizontal axis representing time t (seconds) and the vertical axis representing amplitude, showing a modified received wave obtained by applying a time series filter to the received wave. FIG.

FIG. 6 is a graph showing a relationship between the two, where the horizontal axis represents frequency (Hz) and the vertical axis represents spectrum value, and is a diagram schematically showing the obtained Fourier spectrum.

FIG. 7 is a graph showing the relationship between the two, where the horizontal axis represents the frequency (Hz) and the vertical axis represents the spectrum value, and is a diagram schematically showing the Fourier spectrum after frequency filtering processing.

FIG. 8 is a schematic view showing a column in which reinforcing bars are arranged.

FIG. 9 is a schematic view showing a beam in which reinforcing bars are arranged.

FIG. 10 is a schematic diagram showing a wall in which reinforcing bars are arranged.

FIG. 11 is a schematic diagram showing measurement of a concrete wall.

FIG. 12 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the modified received wave obtained at the measurement point. is there.

FIG. 13 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the modified received wave obtained at the measurement point. is there.

FIG. 14 is a schematic diagram showing a concrete wall and measurement points.

FIG. 15 is a schematic diagram showing a cracked concrete wall and measurement points.

FIG. 16 is a schematic diagram showing a superimposed wave filtering processing apparatus according to an embodiment of the present invention.

FIG. 17 shows time t (seconds) on the horizontal axis and superimposed wave f on the vertical axis.
(T) is a graph showing the relationship between the two, and the horizontal axis shows time t (seconds), and the vertical axis shows the superimposed wave f (t + Δt) showing the relationship between the two.

FIG. 18 is a graph showing the relationship between the two, where the horizontal axis represents time t (seconds) and the vertical axis represents amplitude, and the solid line represents f (t).
Showing sin (2nπt), and the broken line is f (t
Each sin {(2nπ (t + Δt)} that composes + Δt)
Is shown.

FIG. 19 is a graph showing G ₁ (t).

FIG. 20 is a graph showing each n component of G ₁ (t).

FIG. 21 is a graph showing G ₂ (t).

FIG. 22 is a graph showing each n component of G ₂ (t).

FIG. 23 (a) is a graph showing the relationship between the horizontal axis of time t (seconds) and the vertical axis of superimposed wave f (t), and FIG. 23 (b) is the horizontal axis of time t. FIG. 3 is a graph showing the relationship between (sec) and the superimposed wave f (t + Δt) on the vertical axis.

FIG. 24 shows time t (sec) on the horizontal axis and superimposed wave f on the vertical axis.
It is a figure which shows each relationship between both (t) and f (t + (DELTA) t) n component (frequency n) is taken, a solid line shows fundamental frequency wave sin (2n (pi) t) of f (t), and a broken line shows f. (T + Δ
The fundamental frequency wave sin {(2nπ (t + Δt)} of t) is shown.

FIG. 25 is a graph showing G ₁ (t).

FIG. 26 is a graph showing each n component of G ₁ (t).

FIG. 27 is a graph showing G ₂ (t).

FIG. 28 is a graph showing each n component of G ₂ (t).

FIG. 29 is a diagram of Σ (n = 1, k) sin (2nπt)
FIG. 7 is a graph showing the relationship between the horizontal axis and frequency (Hz), and the vertical axis with the spectrum value.

FIG. 30 is a graph showing a spectrum value obtained by applying the above-mentioned numerical formula 2 to FIG. 13, a graph showing a frequency (Hz) on the horizontal axis and a spectrum value on the vertical axis, showing a relationship between the two. It is a figure.

FIG. 31 is a graph showing a spectrum value obtained by applying the above-mentioned mathematical expression 3 to FIG. 13, a graph showing a frequency (Hz) on the horizontal axis and a spectrum value on the vertical axis, showing a relationship between the two. It is a figure.

FIG. 32 is a graph showing the relationship between the time t (seconds) on the horizontal axis and the amplitude on the vertical axis.

33 (a) and (b) are f (t) and F, respectively.
It is a Fourier spectrum of ₁ (t), the horizontal axis shows the frequency (Hz), and the vertical axis shows the spectrum value showing the relationship between the two.

34 (a) and 34 (b) are f (t) and F, respectively.
It is a Fourier spectrum of ₂ (t), and a horizontal axis shows a frequency (Hz) and a vertical axis shows a spectrum value, and is a graph showing the relationship between the two.

FIG. 35 is a graph showing a relationship between the oscillation frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, showing the Fourier spectrum of the oscillating wave (broken line) 91 and the received wave (solid line) 92. FIG.

FIG. 36 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and showing the spectrum obtained by frequency filtering the spectrum of the received wave 92. FIG.

FIG. 37 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the time-series filter is applied to the spectrum of the received wave 92. It is the spectrum obtained by processing.

FIG. 38 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the spectrum of the received wave 92 is time-series filtered. It is the spectrum obtained by processing.

FIG. 39 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the time series filter is applied to the spectrum of the received wave 92. It is the spectrum obtained by processing.

FIG. 40 is a graph showing a relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the time-series filter is applied to the spectrum of the received wave 92. It is the spectrum obtained by processing.

FIG. 41 is a graph showing the relationship between the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, where t _h = 0 and t _a = 300 μsec. It is a figure which shows the spectrum obtained by performing a frequency filtering process, after applying a time series filter to the spectrum.

42 is a diagram showing a spectrum obtained by performing frequency filtering processing after performing time series filtering on the spectrum of the received wave 92 with t _h = 0 and t _a = 200 μsec. FIG.

FIG. 43 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis and showing the filter.

FIG. 44 is a schematic diagram showing reinforced concrete and measuring points.

FIG. 45 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 46 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 47 is a graph showing a relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis and showing the filter.

FIG. 48 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the filter.

FIG. 49 is a graph showing a relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

FIG. 50 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

FIG. 51 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis and showing the filter.

FIG. 52 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the filter.

FIG. 53 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum after frequency filtering processing.

FIG. 54 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum after frequency filtering processing.

FIG. 55 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 56 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 57 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

FIG. 58 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

[Fig. 59] Fig. 59 is a schematic diagram showing a state where an oscillating sensor oscillates elastic waves in concrete and a receiving sensor receives the acoustic waves.

[Explanation of symbols]

1; Reinforced concrete 2; Top surface 3; Large diameter bar 4, 11, 11a, 11b, 12, 12a; Reinforcing bar 5; Small diameter bar 6; Oscillation sensor 7, 71; Reception sensor 8; Bottom surface 9, 9a, 9b; Measuring point 10 Cracks 10a; reflective surfaces 31s, 31a, 31b, 31c, 31e, 32a, 3
2b, 32c, 32d, 32e, 33a, 33b, 33
c, 33d, 33e; spectrum 51, 52, 53; time series filter 72; A / D converter 73; computer 74; CRT 75; DISK 76; filtering device 77; filtering CPU 78; memory 81a, 81e, 81n, 81i , 82a, 82e, 8
2n, 82i, 83a, 83e, 83n, 83i; Filter 91; Oscillating wave 79, 92; Received wave 101; Pillar 102; Beam 103; Wall

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] July 22, 1996

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Brief description of drawings

[Correction method] Change

[Correction contents]

[Brief description of drawings]

FIG. 1 is a schematic view showing a reinforced concrete model in the first embodiment method of the present invention.

FIG. 2 is a cross-sectional view showing a plane AA ′ in FIG.

FIG. 3 is a graph showing the relationship between the two, with the horizontal axis representing time t (seconds) and the vertical axis representing amplitude.

FIG. 4 is a graph showing a relationship between the two, with the horizontal axis representing time t (seconds) and the vertical axis representing time series filter H (t), and a diagram schematically showing the time series filter.

FIG. 5 is a graph showing the relationship between the horizontal axis representing time t (seconds) and the vertical axis representing amplitude, showing a modified received wave obtained by applying a time series filter to the received wave. FIG.

FIG. 6 is a graph showing a relationship between the two, where the horizontal axis represents frequency (Hz) and the vertical axis represents spectrum value, and is a diagram schematically showing the obtained Fourier spectrum.

FIG. 7 is a graph showing the relationship between the two, where the horizontal axis represents the frequency (Hz) and the vertical axis represents the spectrum value, and is a diagram schematically showing the Fourier spectrum after frequency filtering processing.

FIG. 8 is a schematic view showing a column in which reinforcing bars are arranged.

FIG. 9 is a schematic view showing a beam in which reinforcing bars are arranged.

FIG. 10 is a schematic diagram showing a wall in which reinforcing bars are arranged.

FIG. 11 is a schematic diagram showing measurement of a concrete wall.

FIG. 12 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the modified received wave obtained at the measurement point. is there.

FIG. 13 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the modified received wave obtained at the measurement point. is there.

FIG. 14 is a schematic diagram showing a concrete wall and measurement points.

FIG. 15 is a schematic diagram showing a cracked concrete wall and measurement points.

FIG. 16 is a schematic diagram showing a superimposed wave filtering processing apparatus according to an embodiment of the present invention.

FIG. 17 shows time t (seconds) on the horizontal axis and superimposed wave f on the vertical axis.
FIG . 6 is a graph showing the relationship between (t) and f (t + Δt) .

FIG. 18 is a graph showing the relationship between the two, where the horizontal axis represents time t (seconds) and the vertical axis represents amplitude, and the solid line represents f (t).
Showing sin (2nπt), and the broken line is f (t
Each sin {(2nπ (t + Δt)} that composes + Δt)
Is shown.

FIG. 19 is a graph showing G ₁ (t).

FIG. 20 is a graph showing each n component of G ₁ (t).

FIG. 21 is a graph showing G ₂ (t).

FIG. 22 is a graph showing each n component of G ₂ (t).

FIG. 23 shows the time t (seconds) on the horizontal axis and the superimposed wave f on the vertical axis.
FIG . 6 is a graph showing the relationship between (t) and f (t + Δt) .

FIG. 24 shows time t (sec) on the horizontal axis and superimposed wave f on the vertical axis.
It is a figure which shows each relationship between both (t) and f (t + (DELTA) t) n component (frequency n) is taken, a solid line shows fundamental frequency wave sin (2n (pi) t) of f (t), and a broken line shows f. (T + Δ
The fundamental frequency wave sin {(2nπ (t + Δt)} of t) is shown.

FIG. 25 is a graph showing G ₁ (t).

FIG. 26 is a graph showing each n component of G ₁ (t).

FIG. 27 is a graph showing G ₂ (t).

FIG. 28 is a graph showing each n component of G ₂ (t).

FIG. 29 is a graph showing the relationship between Σ (n = 1, k) sin (2nπt), where the horizontal axis represents the frequency (Hz) and the vertical axis represents the spectrum value.

FIG. 30 is a graph showing a spectrum value obtained by applying the above-mentioned numerical formula 2 to FIG. 13, a graph showing a frequency (Hz) on the horizontal axis and a spectrum value on the vertical axis, showing a relationship between the two. It is a figure.

FIG. 31 is a graph showing a spectrum value obtained by applying the above-mentioned mathematical expression 3 to FIG. 13, a graph showing a frequency (Hz) on the horizontal axis and a spectrum value on the vertical axis, showing a relationship between the two. It is a figure.

FIG. 32 is a graph showing the relationship between the time t (seconds) on the horizontal axis and the amplitude on the vertical axis.

33 (a) and (b) are f (t) and F, respectively.
It is a Fourier spectrum of ₁ (t), a horizontal axis shows frequency (Hz), and a vertical axis | shaft takes a spectrum value, and is a graph figure which shows the relationship of both.

34 (a) and 34 (b) are f (t) and F, respectively.
It is a Fourier spectrum of ₂ (t), a horizontal axis shows a frequency (Hz), and a vertical axis | shaft takes a spectrum value, and is a graph figure which shows the relationship of both.

FIG. 35 is a graph showing a relationship between the oscillation frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, showing the Fourier spectrum of the oscillating wave (broken line) 91 and the received wave (solid line) 92. FIG.

FIG. 36 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and showing the spectrum obtained by frequency filtering the spectrum of the received wave 92. FIG.

FIG. 37 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the time-series filter is applied to the spectrum of the received wave 92. It is the spectrum obtained by processing.

FIG. 38 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the spectrum of the received wave 92 is time-series filtered. It is the spectrum obtained by processing.

FIG. 39 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the time series filter is applied to the spectrum of the received wave 92. It is the spectrum obtained by processing.

FIG. 40 is a graph showing a relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, and performing frequency filtering after the time-series filter is applied to the spectrum of the received wave 92. It is the spectrum obtained by processing.

FIG. 41 is a graph showing the relationship between the two by taking the frequency (kHz) on the horizontal axis and the Fourier spectrum on the vertical axis, where t _h = 0 and t _a = 300 μsec.
FIG. 5 is a diagram showing a spectrum obtained by performing frequency filtering processing after applying a time series filter to the spectrum of FIG.

[Figure 42] and t _{h =} 0, and t a ₌ 200 [mu] seconds, after applying a time series filter spectrum of the received wave 92 is a diagram showing a spectrum obtained by frequency filtering.

FIG. 43 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis and showing the filter.

FIG. 44 is a schematic diagram showing reinforced concrete and measuring points.

FIG. 45 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 46 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 47 is a graph showing a relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis and showing the filter.

FIG. 48 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the filter.

FIG. 49 is a graph showing a relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

FIG. 50 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

FIG. 51 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis and showing the filter.

FIG. 52 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the filter.

FIG. 53 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum after frequency filtering processing.

FIG. 54 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum after frequency filtering processing.

FIG. 55 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 56 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the Fourier spectrum of the received wave obtained at each measurement point. is there.

FIG. 57 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

FIG. 58 is a graph showing the relationship between the two by taking the frequency (Hz) on the horizontal axis and the spectrum value on the vertical axis, and showing the sum of Fourier spectra at each measurement point.

[Fig. 59] Fig. 59 is a schematic diagram showing a state where an oscillating sensor oscillates elastic waves in concrete and a receiving sensor receives the acoustic waves.

[Explanation of Codes] 1; Reinforced concrete 2; Top surface 3; Large diameter bar 4, 11, 11a, 11b, 12, 12a; Reinforcing bar 5; Small diameter bar 6; Oscillation sensor 7, 71; Reception sensor 8; Bottom surface 9, 9a, 9b; measurement point 10; crack 10a; reflection surface 31s, 31a, 31b, 31c, 31e, 32a, 3
2b, 32c, 32d, 32e, 33a, 33b, 33
c, 33d, 33e; spectrum 51, 52, 53; time series filter 72; A / D converter 73; computer 74; CRT 75; DISK 76; filtering device 77; filtering CPU 78; memory 81a, 81e, 81n, 81i , 82a, 82e, 8
2n, 82i, 83a, 83e, 83n, 83i; Filter 91; Oscillating wave 79, 92; Received wave 101; Pillar 102; Beam 103; Wall

Claims (7)

[Claims] 1. An elastic wave oscillating sensor and an elastic wave receiving sensor are arranged in a substance through which an elastic wave propagates, and the elastic wave oscillating sensor oscillates an elastic wave in the substance and then receives the elastic wave by the elastic wave receiving sensor. The obtained received wave is time-series filtered with a magnitude of 0 to 1 from the arrival time of the received wave to a predetermined time and a magnitude of 1 after the predetermined time. Superposed wave filtering method. 2. The superimposed wave filtering processing method according to claim 1, wherein the time series filter is applied a plurality of times. 3. The superposed wave filtering method according to claim 1, further comprising filtering by frequency. 4. A plurality of the elastic wave oscillating sensors and the elastic wave receiving sensor are arranged, and the received waves obtained by the respective elastic wave receiving sensors are mutually subjected to the time-series filter and then mutually. The addition is performed according to claim 1.
4. The superposed wave filtering method according to any one of items 1 to 3. 5. A plurality of the elastic wave oscillation sensors and the elastic wave receiving sensors are arranged, and after adding the received waves obtained by the respective elastic wave receiving sensors to each other,
2. The time series filter is applied.
4. The superposed wave filtering method according to any one of items 1 to 3. 6. An elastic wave oscillating sensor, an elastic wave receiving sensor, and a time-series filter section for applying a time-series filter to a received wave received by the elastic wave receiving sensor to obtain a modified received wave, The superposed wave filtering device, wherein the series filter has a size of 0 to 1 from the arrival time of the received wave to a predetermined time, and a size of 1 after the predetermined time. 7. The superimposed wave filtering processing apparatus according to claim 6, further comprising a filtering processing unit that performs filtering processing on the modified reception wave according to frequency.