CA2054718C - Magnetic inspection apparatus for thin steel strip - Google Patents

Magnetic inspection apparatus for thin steel strip

Info

Publication number
CA2054718C
CA2054718C CA 2054718 CA2054718A CA2054718C CA 2054718 C CA2054718 C CA 2054718C CA 2054718 CA2054718 CA 2054718 CA 2054718 A CA2054718 A CA 2054718A CA 2054718 C CA2054718 C CA 2054718C
Authority
CA
Canada
Prior art keywords
steel strip
magnetic
thin steel
defect
inspection apparatus
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CA 2054718
Other languages
French (fr)
Inventor
Kenichi Iwanaga
Seigo Ando
Masaki Takenaka
Takato Furukawa
Atsuhisa Takekoshi
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
JFE Engineering Corp
Original Assignee
NKK Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from JP18948490A external-priority patent/JPH07109416B2/en
Priority claimed from JP19787190A external-priority patent/JPH07104329B2/en
Application filed by NKK Corp filed Critical NKK Corp
Application granted granted Critical
Publication of CA2054718C publication Critical patent/CA2054718C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Landscapes

  • Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)

Abstract

In a magnetic inspection apparatus for a thin steel strip of the invention, a pole-to-pole distance of a magnetizer, which is housed in a hollow roller contact-ing a traveling thin steel strip to oppose the thin steel strip, is set between twice and eight times a dis-tance between magnetic poles and the thin steel strip.
A position of a magnetic sensor in the travel direction of the thin steel strip is shifted from the central position of the magnetic poles to a travel direction side by a small distance. In this invention, a pair of hollow rollers are provided to sandwich upper and lower surfaces of a traveling thin steel strip, and magnetic sensors are respectively housed in the hollow rollers.
As a result, a position of a defect in the direction of the thickness of the thin steel strip, and a defect size can be accurately detected. In a magnetic inspection apparatus for a thin steel strip of the invention, a magnetic detection circuit for detecting magnetic flux leakage caused by a defect in the interior or on the surface of a thin steel strip uses an over-saturation type magnetic sensor constituted by applying a detection coil around a ferromagnetic core. As a result, a defect detection sensibility and a defect detection precision are greatly improved.

Description

S P E C I F I C A T I O N
"MAGNETIC INSPECTION APPARATUS FOR THIN STEEL STRIP"
[Technical Field]
The present invention relates to a magnetic inspec-tion apparatus for a thin steel strip which detects a defect present in the interior or on the surface of a thin steel strip in a traveling state and, more particularly, to a magnetic inspection apparatus for a thin steel strip which urges a hollow roller, rotatably supported on a stationary shaft perpendicular to a travel path of the thin steel strip, against the steel strip, which houses a magnetizer in the hollow roller, and which detects by a magnetic sensor magnetic flux leakage occurring due to a defect.
[Background Art]
A magnetic inspection apparatus detects a defect, e.g., a flaw and an inclusion, present in the interior or on the surface of a thin steel strip by utilizing magnetism. It is reported that a magnetic inspection apparatus for a thin steel strip cannot only examine a thin steel strip as an object to be examined in a still state but also continuously detect defects present in a thin steel strip traveling along, e.g., a manufacture line of a factory and the like (Published Unexamined Japanese Utility Model Application No. 63-107849).
Figs. 31 and 32 are sectional views taken along different directions, respectively, of - 2 - 205~718 an above-described magnetic inspection apparatus for a thin steel strip which continuously detects defects of a traveling thin steel strip.
A hollow roller 1 is made of a non-magnetic material. One end of a stationary shaft 2 extends along the axis of the hollow roller 1. The other end of the stationary shaft 2 is fixed on the frame of a base (not shown). The stationary shaft 2 is supported on the inner circumferential surfaces of the two ends of the hollow roller 1 by a pair of rolling bearings 3a and 3b so that it is located along the axis of the hollow roller 1. Hence, the hollow roller 1 freely rotates about the stationary shaft 2 as the central axis of rotation.
Inside the hollow roller 1, a magnetization core 4c having a substantially U-shaped section is fixed on the stationary shaft 2 through a support member 5 such that their magnetic poles 4a and 4b constituting a magnetic path are close to the inner circumferential surface of the hollow roller 1. A-magnetization coil 6 is wound around the magnetization core 4c. Hence, the magnetiza-tion core 4c with the magnetic poles 4a and 4b, and the magnetization coil 6 constitute a magnetizer 4. A plu-rality of magnetic sensors 7 are arranged between the magnetic poles 4a and 4b of the magnetization core 4c of the magnetizer 4 in the axial direction. Each magnetic sensor 7 is fixed on the stationary shaft 2.

20~4718 A power cable 8 for supplying an excitation current to the magnetization coil 6 and a signal cable 9 for deriving the respective detection signals output from the respective magnetic sensors 7 extend to the outside through the interior of the stationary shaft 2. Hence, the positions of the magnetization core 4 and the respective magnetic sensors 7 are fixed, and the hollow roller 1 rotates at a small gap to the outer surfaces of the magnetizer 4 and the respective magnetic lo sensors 7.
When the outer circumferential surface of the hol-low roller 1 of the magnetic inspection apparatus having such an arrangement is urged against one surface of a thin steel strip 10, traveling in the direction of an arrow A, at a predetermined pressure, the hollow roller 1 rotates in the direction of an arrow B as the station-ary shaft 2 is fixed on the frame of the base.
When an excitation current is supplied to the mag-netization coil 6, a closed magnetic path is constituted by the magnetization core 4c and the traveling thin steel strip 10. Therefore, if a defect described above is present in the interior or on the surface of the thin steel strip 10, the magnetic path within the thin steel strip 10 is disturbed, and magnetic flux leakage occurs.
The magnetic flux leakage is extracted by a magnetic sensor 7 at a corresponding position and is detected as a defect signal.

- 4 _ 20S4718 The signal level of the detected defect signal cor-responds to the size of the defect in the interior or on the surface of the thin steel strip 10. Therefore, the presence and size of the defect of the thin steel strip 10 can be obtained in the form of the signal level of the defect signal.
However, the signal level of a defect signal is largely changed depending on the state of the magnetic path constituted by the thin steel strip 10 and the lo magnetizer 4 comprising the magnetization core 4c and the magnetization coil 6, a distance L between the magnetizer 4 and the thin steel strip 10, a distance ~
between the thin steel strip 10 and the respective mag-netic sensors 7, which is called a lift-off distance, and so on.
In order to eliminate these drawbacks, the distance L between the thin steel strip 10 and the magnetizer 4 and the distance ~ between the thin steel strip 10 and the respective sensors 7 are constantly kept to be pre-determined values by using the hollow roller 1 having apredetermined thickness t, as shown in Figs. 31 and 32.
If the hollow roller 1 is made of a magnetic material, formation of the magnetic path into the thin steel strip 10 is interfered. Therefore, the hollow roller 1 is made of a non-magnetic material.
Accordingly, the smaller the thickness t of the hollow steel strip 1, the smaller the distance L between - 20~4718 the magnetic poles 4a and 4b of the magnetizer 4 and the thin steel strip 10, and the larger the magnetic field formed within the thin steel strip 10, thereby obtaining stable magnetic fluxes. For this reason, it is prefera-ble that the thickness t of the hollow roller 1 be madesmall.
If the hollow roller 1 has a large thickness t, its moment of inertia becomes large. Then, when the travel speed of the thin steel strip 10 fluctuates, a sliding phenomenon may occur between the contact surfaces of the hollow roller 1 and the thin steel strip 10 to flaw the surface of the thin steel strip 10. Therefore, the moment of inertia must be decreased by decreasing the thickness t of the hollow roller 1. When only the moment of inertia is to be decreased, the outer diameter of the hollow roller 1 may be set small. The outer diameter is restricted by the size of the magnetizer 4 or the magnetic sensors 7 housed in the hollow roller 1.
As described above, in order to continuously detect a defect in the traveling thin steel strip 10 with a high precision, as described above, the surface of the thin steel strip 10 must be in constant contact with the outer circumferential surface of the hollow roller 1.
As a result, a downward force caused by the tension of the thin steel strip 10 and a downward force caused by the weight of the thin steel strip 10 itself are applied to the hollow roller 1. When a downward force is - 6 - 205~718 applied, the hollow roller 1 is deformed or damaged.
Then, the distance L between the thin steel strip 10 and the magnetizer 4 and the distance ~ between the thin steel strip 10 and the respective magnetic sensors 7, described above, cannot be controlled to be the prede-termined values. As a result, a defect detection precision may be degraded, or inspection may become impossible.
Hence, if the true circle state of the hollow roller 1 is to be kept over a long period of time, the thickness t of the hollow roller 1 should not be made smaller than a predetermined limit. For example, under the condition that the travel speed of the thin steel strip 10 is 100 m/min., the limit thickness t is about 2 mm.
The strength of the magnetic field generated by the magnetizer 4, housed in the hollow roller 1, comprising the magnetization core 4c and the magnetization coil 6 may be increased. However, if the size of the magneti-zation core 4c or the intensity of the current supplied to the magnetization coil 6 is increased over a prede-termined limit, the entire apparatus may be made large, or the manufacturing costs may be greatly increased.
[Disclosure of Invention]
It is the first object of the present invention to provide a magnetic inspection apparatus for a thin steel strip, in which the S/N ratio of a defect signal ~ 7 ~ 205~718 detected by a magnetic sensor can be increased without largely increasing the manufacturing costs, and a defect detection sensibility and a detection precision can be greatly increased.
s It is the second object of the present invention to provide a magnetic inspection apparatus for a thin steel strip, which can easily detect a location and a size of a defect of a thin steel strip in the direction of the thickness.
In order to achieve the first object of the present invention, according to the present invention, the pole-to-pole distance of a magnetizer arranged in a hollow roller rotated as it contacts the surface of a traveling thin steel strip is set between twice or more and eight times or less of the distance between each magnetic pole and the thin steel strip.
As is known, in a magnetizer having a pair of sepa-rated magnetic poles, the magnetic fluxes output from one magnetic pole are input to the other magnetic pole through the space (magnetic gap) between magnetic poles.
In this case, if a thin steel strip made of a magnetic material is present close to the magnetic gap, some of the magnetic fluxes output from one magnetic pole are input to the other magnetic pole through the thin steel strip without passing through the magnetic gap.
In this case, the ratio of the magnetic fluxes passing through the magnetic gap to those passing -through the thin steel strip is largely influenced by the size (pole-to-pole distance W) of the magnetic gap and the distance L between each magnetic pole and the thin steel strip. More specifically, when the pole-to-pole distance W is constant, more magnetic fluxes areconcentrated on a portion corresponding to one of the distances w and L of the magnetic circuit that has a smaller magnetic resistance. Therefore, when the dis-tance L between each magnetic pole and the thin steel strip becomes small, the density of the magnetic fluxes passing through the thin steel strip is increased accordingly; when the distance L becomes large, the density of the magnetic fluxes passing through the thin steel strip is decreased.
When the distance L is constant, the ratio of the magnetic fluxes passing through the thin steel strip is increased as the pole-to-pole distance W is increased.
However, when the pole-to-pole distance W is excessively increased, although the total number of the magnetic fluxes described above is increased, the amount of mag-netic fluxes passing through the thin steel strip is decreased.
Hence, the pole-to-pole distance W has a predeter-mined optimum range. The optimum range is influenced by the distance L between each magnetic pole and the thin steel strip. More specifically, when the distance L is large, the optimum range is largely influenced by the -- 9 - 2Q5~718 pole-to-pole distance W; when small, it is largely influenced by the distance L.
The present inventor has obtained the relationship between the pole-to-pole distance W and the distance L
through experiments and has confirmed that, if the pole-to-pole distance W falls within a range of twice to eight times the distance L (2L < W < 8L), magnetic fluxes passing through the thin steel strip have a high density of a level sufficiently used in practice.
Therefore, even if the distance L cannot be set small due to the limitations of the thickness of the hollow roller, as described above, the detection sensi-tivity of the magnetic sensor can be set maximum by setting the pole-to-pole distance W to satisfy the relationship described above, and a defect detection precision is improved.
According to the present invention, the position of the magnetic sensor in the travel direction of the thin steel strip is set at a position shifted from the central position of the magnetic poles to the travel direction side by a small distance determined by the residual magnetization characteristics of the thin steel strip.
Assume that a completely defectless thin steel strip is placed to oppose the magnetic poles of the mag-netizer and the magnetization coil is excited by a DC
power. In this state, when the position of the magnetic lO- 20~4718 sensor is changed in the travel direction of the thin steel strip, the strength of the magnetic field detected by the magnetic sensor is maximum at one magnetic pole and is m; n; ~um at the other magnetic pole, exhibiting a vertical magnetic field distribution characteristic curve that crosses the 0 level line at the central position of the pole-to-pole distance W due to the floating magnetic flux. Therefore, if the magnetic sensor is set at the central position of the pole-to-pole distance W where the vertical magnetic field dis-tribution characteristic curve crosses the 0 level line, the influences of the floating magnetic fluxes can be removed.
In the actual magnetic inspection apparatus, however, the thin steel strip travels in one direction at a constant speed. At this time, the thin steel strip is magnetized by the magnetizer, and the magnetic fluxes corresponding to the magnetization strength and the coercive force of the thin steel strip remain in the thin steel strip. As a result, a position where the vertical magnetic field distribution characteristic curve crosses the 0 level line is not always the central position of the pole-to-pole distance W but shifts in the travel direction side.
More specifically, when the thin steel strip travels, the central position of the pole-to-pole distance w does not correspond to the 0 level of the ll- 2054718 vertical magnetic field distribution characteristic curve. Rather, the 0 level position is shifted from the central position of the pole-to-pole distance W to the travel direction side. Hence, floating magnetic fluxes are present at the central position.
In fine, according to the present invention, the magnetic sensor is moved to the shifted 0 level position. As a result, the magnetic sensor does not detect floating magnetic fluxes. Therefore, the detec-tion sensibility of the magnetic sensor can be easily increased.
Furthermore, according to the present invention, the pair of hollow rollers housing either a magnetizer or a magnetic sensor are arranged to sandwich a thin steel strip between them.
Hence, for example, the thickness of the hollow roller of a side to which the weight and tension force of the thin steel strip are directly applied is set large, and the thickness of the hollow roller of a side to which the weight and tension force of the thin steel strip are not directly applied is set small. When the magnetic sensor is housed in the thin hollow roller, the lift-off distance can be set short, and the defect detection sensibility can be improved.
Furthermore, according to the present invention, the magnetic detection circuit for detecting magnetic flux leakage occurring due to a defect in the interior or on the surface of the thin steel strip comprises an over-saturation type magnetic sensor housed in the hollow roller and constituted by winding a detection coil around a ferromagnetic core, excitation power supply means for exciting the magnetic sensor to an over-saturation range by supplying an AC power to the detection coil of the magnetic sensor through a station-ary impedance, voltage detecting means for detecting positive and negative values of a voltage generated across two terminals of the detection coil, and arithme-tic means for adding the positive and negative values that are detected by the voltage detecting means, and determining the sum as a measured value corresponding to the magnetic flux leakage.
Generally, an over-saturation type magnetic sensor obtained by applying a detection coil around a ferromag-netic core has a considerably excellent detection sensi-bility and a temperature characteristic when compared to those of a magnetic sensor which uses a magnetic diode, a magnetoresistive element, or a Hall element.
In order to achieve the second ob;ect, according to the present invention, a pair of hollow rollers are arranged to respectively contact the upper and lower surfaces of a thin steel strip. A magnetizer is arranged in one hollow roller. A magnetic sensor for detecting magnetic flux leakage occurring due to a defect in the interior or on the surface of the thin - 13 - 2~5~71$

steel strip is arranged in each of the hollow rollers.
A data processing unit calculates a defect occurring position in the direction of the thickness of the thin steel strip and a defect size from each magnetic flux leakage detected by each of the pair of magnetic sensors.
As the pair of hollow rollers contact the upper and lower surfaces of the steel strip, the distance between the upper surface of the steel strip and the correspond-ing magnetic sensor and that between the lower surfaceof the steel strip and the corresponding magnetic sensor are kept constant-. A magnetic field is generated in the steel strip by the magnetizer. Therefore, if a defect is present, each magnetic sensor detects magnetic flux leakage corresponding to the defect. The magnetic flux leakage detected by each magnetic sensor can be expressed as a defect size and a distance to the defect, i.e., a function of the defect depth from the surface of the given magnetic sensor. As a result, the defect size and the defect position can be calculated by resolving the two functions as simultaneous equations.
[srief Description of Drawings]
Fig. 1 is a sectional view of an embodiment of a magnetic inspection apparatus for a thin steel strip according to the present invention, along a plane paral-lel to the travel direction of the thin steel strip;
Fig. 2 is a sectional view of the apparatus of the - 14 - 2~ 8 same along a plane perpendicular to the travel direction of the thin steel strip;
Fig. 3 is a sectional view of a test system for confirming the effect of the apparatus of the same;
Fig. 4 is a graph of detection characteristics obtained by the test system;
Fig. 5 is a graph of the detection characteristics of Fig. 4 but indicated by using a different parameter;
Fig. 6 is a waveform chart of a defect signal obtained by the test system;
Fig. 7 is a sectional view of another embodiment of a magnetic inspection apparatus for a thin steel strip according to the present invention along a plane paral-lel to the travel direction of the thin steel strip;
Fig. 8 is a schematic diagram of the main part of the same apparatus for confirming its effect;
Fig. 9 shows the relationship among the respective magnetic poles, the horizontal magnetic field distribution, and the vertical magnetic field distribution;
Fig. 10 is a graph of detection characteristics showing the relationship between an excitation current and a detection voltage of the magnetic sensor;
Fig. 11 is a graph of detection characteristics of the magnetic sensor when the magnetization current of the test system is changed;
Fig. 12 is a graph of detection characteristics of -- 15 - 205~71~

the magnetic sensor when the position of the magnetic sensor in the test system is changed;
Fig. 13 is a sectional view of still another embodiment of a magnetic inspection apparatus for a thin steel strip according to the present invention along a plane parallel to the travel direction of the thin steel strip;
Fig. 14 is a sectional view of the same along a plane perpendicular to the travel direction of the thin steel strip;
Fig. 15 is a sectional view of still another embod-iment of the magnetic inspection apparatus for a thin steel strip according to the present invention along a plane parallel to the travel direction of the thin steel strip;
Fig. 16 is a sectional view of the same along a plane perpendicular to the travel direction of the thin steel strip;
Fig. 17 is a schematic diagram showing the system of the entire apparatus of the same;
Fig. 18 is a graph of detect signal characteristics of the same;
Fig. 19 is a graph showing the steps of calculating a defect position and a defect size from the defect sig-nal characteristics;
Fig. 20 is a graph showing the relationship between a measured value and a human visual evaluation;

-- 16 - ~0~17~8 Fig. 21 is a table showing the relationship among the respective measured values, the defect positions, and the defect sizes of a still another embodiment of a magnetic inspection apparatus;
Fig. 22 is a block diagram showing a magnetic detection circuit of still another embodiment of a magnetic inspection apparatus for a thin steel strip according to the present invention;
Fig. 23 is a timing chart showing the operation of the magnetic detection circuit;
Fig. 24 is a waveform chart of a voltage applied to the detection coil of the magnetic detection circuit;
Fig. 25 is a waveform chart of a voltage output from a coil of the magnetic detection circuit;
Fig. 26 is also a waveform chart of a voltage out-put from the coil of the magnetic detection circuit;
Fig. 27 is a graph of magnetization characteristics of a ferromagnetic core;
Fig. 28 is a graph of an output voltage with respect to the magnetic flux density of the magnetic detection circuit;
Fig. 29 is a block diagram showing a magnetic detection circuit of a still another embodiment of the present invention;
Fig. 30 is a graph of an output voltage with respect to the magnetization current of the magnetic detection circuit of the same;

-- 17 - 2rDS~ 7 Fig. 31 is a sectional view of a general magnetic inspection apparatus for a thin steel strip along a plane parallel to the travel direction of the thin steel strip;
Fig. 32 is a sectional view of the conventional apparatus along a plane perpendicular to the travel direction of the thin steel strip;
Fig. 33 is a view showing the relationship between the arrangement directions of the magnetic sensors and the detection output waveforms of the respective mag-netic sensors; and Fig. 34 is a view showing the relationship between the vertical and horizontal magnetic fields detected by the respective magnetic sensors of Fig. 33.
[Best Mode of Carrying Out the Invention]
An embodiment of the present invention will be described with reference to the accompanying drawings.
Figs. 1 and 2 are sectional views showing a sche-matic arrangement of an embodiment of a magnetic inspec-tion apparatus. Note that the same portions as in the conventional apparatus shown in Figs. 31 and 32 are denoted by the same reference numerals. Accordingly, a detailed description of overlapping portions is omitted.
One end of a stationary shaft 2 extends along the axis of a hollow roller la made of a non-magnetic material. The inner circumferential surfaces of the two ends of the hollow roller la are rotatably supported on 205~18 the stationary shaft 2 by a pair of roll bearings 3a and 3b. Thus, the hollow roller la freely rotates about the stationary shaft 2 as the central axis of rotation.
Of the hollow roller la, a thickness to f each end of it on which the roll bearing 3a or 3b is mounted is set large, and a thickness tl its central portion to which a thin steel strip 10 contacts is set small, as shown in Fig. 2. In this embodiment, the thickness to of each end is set to 6 to 10 mm, and the thickness t of the central portion is set to be 1 to 4 mm.
In the hollow roller la, a magnetization core 4c having a substantially U-shaped section is fixed to the stationary shaft 2 through a support member 5 such that its magnetic poles 4a and 4b are close to the inner circumferential surface of the hollow roller la. The distal end of each of the magnetic poles 4a and 4b is curved to correspond to the radius of curvature of the inner circumferential surface of the hollow roller la.
A magnetization coil 6 is wound around the magnetization core 4c. A plurality of magnetic sensors 7 are arranged between the magnetic poles 4a and 4b of the magnetiza-tion core 4c in the axial direction. Each magnetic sensor 7 is fixed to the stationary shaft 2. The magnetization core 4c and the magnetization coil 6 con-stitute a magnetizer 4 for generating a magnetic fieldin the thin steel strip 10 through the hollow roller la.
Each magnetic sensor 7 uses an over-saturation type 19- 20~4718 magnetic sensor described in Published Unexamined Japanese Patent Application No. 1-308982.
A power cable 8 for supplying an excitation current to the magnetization coil 6 and a signal cable 9 for deriving the detection signals output from the respec-tive magnetic sensors 7 extend to the outside through the interior of the stationary shaft 2. Hence, the positions of the magnetization core 4c and the respec-tive magnetic sensors 7 are fixed, and the hollow roller la rotates at a small gap to the outer surfaces of the magnetization core 4c and the respective magnetic sen-sors 7.
A pole-to-pole distance (magnetic gap size) W
expressed as the distance between the magnetic poles 4a and 4b of the magnetizer 4 is set to be a value between twice or more and eight times or less of a distance L
between the thin steel strip 10 and each of the magnetic poles 4a and 4b (2L < W < 8L).
The position of each magnetic sensor 7 in the travel direction of the thin steel strip 10 is set at substantially an intermediate position of the magnetic poles 4a and 4b. A lift-off distance ~ between each magnetic sensor 7 and the thin steel strip 10 is set at 3 mm in this embodiment.
When the outer circumferential surface of the hollow roller la of the magnetic inspection apparatus having such an arrangement is urged against one surface -- 20 - 2~5~7~1~

of the thin steel strip 10, which is traveling in, e.g., the direction of an arrow A, at a predetermined pressure, as the stationary shaft 2 is fixed to the frame of the base, the hollow roller la rotates in the direction of an arrow B.
When an excitation current is supplied to the mag-netization coil 6 from an external magnetization power unit (not shown), a closed magnetic path is constituted by the magnetic poles 4a and 4b of the magnetization core 4c and the traveling thin steel strip 10. If a defect is present in the interior or on the surface of the thin steel strip 10, the magnetic path inside the thin steel strip 10 is disturbed, and magnetic flux leakage occurs. The magnetic flux leakage is detected by a magnetic sensor 7 at a given position as a defect signal.
The signal level of the detected defect signal corresponds to the size of the defect in the interior or on the surface of the thin steel strip 10. Hence, the presence and size of a defect in the interior or on the surface of the thin steel strip 10 can be determined from a change in a signal level of a defect signal.
An experimental result which explains why the pole-to-pole distance W of the magnetic poles 4a and 4b of the magnetizer 4 is set to be a value between twice or more and eight times or less of the distance L to the thin steel strip 10, as described above, will be - 21 - 205~718 described.
Referring to Fig. 3, a magnetization coil 33 is applied around a magnetization core 32, having separated magnetic poles 32a and 32b, to constitute a magnetizer 31. A thin steel strip 10 is arranged to be separated from the magnetizer 31 by a distance L. A magnetic sen-sor 7a is arranged on the other side of the thin steel strip 10 to be separated by a distance d. The position of the magnetic sensor 7a corresponds to the central position of a pole-to-pole distance w. The magnetic sensor 7a indirectly detects, of the magnetic fluxes of the magnetic field generated by the magnetizer 31, the density of the magnetic fluxes passing through the thin steel strip 10. A plurality of magnetizers 31 having different pole-to-pole distances W are prepared. The distance L between the thin steel strip 10 and the mag-netizer 31 can also be arbitrarily changed.
In such a test system, the magnetic sensor 7a was arranged such that its axis was perpendicular to the thin steel strip 10, and the vertical component of the magnetic flux leakage caused by each of four types of calibration flaws having outer diameters 0.2 mm to 0.9 mm formed in the thin steel strip 10 was measured.
Similarly, the magnetic sensor 7a was arranged such that its axis was parallel to the thin steel strip 10, and the horizontal components of the magnetic flux leakages were measured under the same conditions. The - 22 - ~Y~71 measurement results are shown in Fig. 33. A signal wav-eform a is the horizontal component of the magnetic field, and a signal waveform b is the vertical component of the magnetic field.
Fig. 9 shows the positional relationship among the magnetic poles of the magnetizer, a horizontal magnetic field distribution characteristic F, and a vertical mag-netic field distribution characteristic D. As shown in Fig. 9, the horizontal magnetic field distribution char-acteristic F has a substantially inverted U shape, and the vertical magnetic field distribution characteristic D has a wavefom of substantially a sine curve.
A distance L between the magnetizer 31 and the thin steel strip lOa is 3.5 mm, the pole-to-pole distance W
is 20 mm, and the distance d between the magnetic sensor 7a and the thin steel strip lOa is 3 mm.
Fig. 34 shows the relationship between the relative outputs of the vertical and horizontal components of the magnetic field generated by each magnetic sensor 7a. As is understood from the graph of Fig. 34, the horizontal and vertical components of the magnetic field have a positive correlation.
Based on this finding, the following embodiments exemplify a case in which a vertical component type mag-netic sensor is used if not otherwise specified.
As shown in Fig. 33, the detection sensibility of a horizontal component detection type magnetic sensor is -- 23 - 2~7I~

higher than that of a vertical component detection type magnetic sensor. When a horizontal component detection type magnetic sensor is used, however, a separate high-pass filter must be provided in order to extract a defect signal from magnetic noise of the ob;ect to be examined, e.g., a thin steel strip lOa, leading to a complicated circuit configuration.
An output voltage of the magnetic sensor 7a was measured while changing a pole-to-pole distance W of the magnetizer 31 from, e.g., 5 mm to 25 mm and fixing the distance L at a predetermined value, e.g., 3 mm. If no defect is present inside the thin steel strip lOa, the magnetic flux density inside it is measured by the magnetic sensor 7a as the magnetic flux leakage is pro-portional to the magnetic flux density inside the thin steel strip lOa. Fig. 4 shows the measurement result.
In the experiment, the magnetization current supplied to the magnetization coil 33 is gradually increased from 0 A to rated 5 A.
It is understood that, when the magnetization cur-rent is increased, the density of the magnetic fluxes passing through the thin steel strip 10 is changed depending on the pole-to-pole distance W, as shown in Fig. 4. More specifically, in an area where the pole-- 25 to-pole distance W is excessively smaller than the dis-tance L, like under a condition w = 5 mm, the magnetic flux density is small. The magnetic flux density is - 24 - 2~5~718 also small in an area where the pole-to-pole distance W
is excessively larger than the distance I like under a condition W = 25 mm. This tendency was observed within an area of 0.5 mm < L ~ 8.0 mm which was measured actually.
Characteristics shown in Fig. 5 are obtained by plotting a ratio (W/L) of the pole-to-pole distance W
to the distance L along the axis of abscissa and a rela-tive output of the magnetic sensor 7a along the axis of ordinate. More specifically, in Fig. 5, the pole-to-pole distance W is adjusted within an actually measured range of the distance L (0.5 mm < L < 8.0 mm), and a maximum output for each distance L is indicated with respect to each pole-to-pole distance W normalized by the distance L.
Generally, the characteristics of a measuring device is evaluated with reference to [-3 dB]. In Fig. 5, a relative output of 70% or more is considered to be sufficiently used in practice. Hence, a range of W/L not less than 2 and not more than 8 is the optimum range.
Fig. 6 is a waveform chart of defect signals detected by the respective magnetic sensors 7 when the distance L between the magnetizer 4 and the thin steel strip 10 is changed from 1 mm to 5 mm in the actual apparatus shown in Figs. 1 and 2 while the above condition (2L < W < 8L) is satisfied. The signal -- 25 ~ 7~8 waveforms of this chart are obtained by differentiating the vertical components of the magnetic fields. The experiment was conducted by using a thin steel strip 10 having three types of defect samples having predeter-mined pinhole outer diameters, i.e., 0.9 mm, 0.6 mm, and0.3 mm.
When the distance L is increased, the signal level of an entire detected defect signal is decreased accordingly. However, the S/N ratio of the obtained detect signal is increased. Therefore, if the gain is increased by using an amplifier, even a small defect of, e.g., 0.3 mm, can be detected with a high precision.
As shown in Fig. 2, the thickness t of the hollow roller la is set large at two ends thereof on which the roll bearings 3a and 3b are mounted and small at the central portion thereof which contacts the thin steel strip 10. As described above, the thickness t of the hollow roller lb is preferably small. However, if it is excessively small, the strength of the hollow roller la may be degraded. In order to compensate for the degradation in strength, the thickness to f the hollow roller la at two ends thereon on which the roll bearings 3a and 3b are mounted is set larger than the thickness tl at the central portion thereof which the thin steel strip 10 contacts, thereby compensating for the degrada-tion in strength caused by setting the overall thickness t of the hollow roller la to a certain degree.

- 26 - 205~718 As a result, the lift-off distance ~ expressed as the distance between the magnetic sensor 7 and the thin steel strip 10 can be set short, and thus the detection sensibility of the magnetic sensor 7 can be increased.
Fig. 7 is a sectional view schematically showing the arrangement of another embodiment of the magnetic inspection apparatus for a thin steel strip according to the present invention. The same portions as those of the magnetic inspection apparatuses shown in Figs. 31 and 1 are denoted by the same reference numerals. Thus, the detailed description of the overlapping portions is omitted.
In the magnetic inspection apparatus of this embodiment, the position of each magnetic sensor 11 mounted to a stationary shaft 2 in a hollow roller 1 in the travel direction of a thin steel strip 10 is set at a position shifted from a central position P of mag-netic poles 4a and 4b by a small distance ~X0 in the travel direction of the thin steel strip 10. In this embodiment, the small distance ~X0 is set at 1 mm.
A pole-to-pole distance W of a magnetizer 4 is set to be 56 mm, and a lift-off distance ~ between each mag-netic sensor 11 and the thin steel strip 10 is set to be 3 mm.
Subsequently, the reason why the magnetic sensors 11 are shifted from the central position P of the mag-netic poles 4a and 4b to the travel direction side of -- 27 - 205~718 the thin steel strip 10 by the small distance ~X0, as described above, and its effect will be described with reference to Figs. 8 to 12.
Fig. 9 shows a vertical magnetic field distribution characteristic curve D of floating magnetic fluxes obtained when a completely defectless thin steel strip 10 is set still with respect to the magnetic poles 4a and 4b. As the thin steel strip 10 travels in one direction at a constant speed, magnetic fluxes corre-sponding to the coercive force of the thin steel strip 10 remain in the thin steel strip 10. As a result, the position where the vertical magnetic field distribution characteristic curve D crosses the 0 level does not necessarily correspond to the central position of the pole-to-pole distance W but is shifted to the travel direction side.
Fig. 10 shows actually measured values that repre-sent the relationship between the excitation current and the detection voltage of the magnetic sensor 7 when the completely defectless thin steel strip 10 is caused to travel at a constant speed through a position opposing the magnetic poles 4a and 4b. It can be understood from Fig. 10 that when the excitation current is increased, the detected floating magnetic fluxes are increased.
Fig. 8 is a schematic view of the main part of Fig. 7. A predetermined shift occurs between a vertical magnetic field distribution characteristic curve G, -- 28 - 20~47 18 which is obtained when the completely defectless thin steel strip 10 is arranged to oppose the magnetic poles 4a and 4b, and a vertical magnetic field distribution characteristic curve E, which is obtained when the same thin steel strip 10 is caused to travel in the direction of an arrow A. This shift amount is substantially determined by the residual magnetization characteristics of the thin steel strip 10. Each magnetic sensor 11 is mounted at such a position in the travel direction where the vertical magnetic field distribution characteristic curve E becomes 0 level. In other words, this shift amount corresponds to the small distance ~X0 described above. Hence, no floating magnetic flux is generated at a position where a magnetic sensor 11 is mounted.
If a floating magnetic flux component is not mixed in the detection voltage of a magnetic sensor 11, even if its detection sensibility is increased, the magnetic sensor 11 is not saturated. Thus, magnetic flux leakage caused by a defect, if small, can be detected at a high precision.
Fig. 11 is a graph of actually measured values that indicates the relationship between a magnetization current applied to the magnetization coil 6 and the detection voltage of the magnetic sensors 11 when each magnetic sensor 11 is mounted at a position shifted from the central position P to travel and reverse directions of a completely defectless thin steel strip 10 by 1 mm, respectively, in the same manner as in this embodiment.
The characteristic curve indicated as a solid line represents actually measured values obtained when the magnetic sensors ll are shifted to the travel direction.
The characteristic curve indicated as a broken line represents actually measured values obtained when the magnetic sensors ll are shifted to the reverse direction. The pole-to-pole distance W is 56 mm.
As apparent from this experimental result, the vertical magnetic field generated by the floating mag-netic fluxes detected by the magnetic sensors ll when the magnetic sensors ll are shifted in the travel direc-tion is considerably smaller than that obtained when the magnetic sensors ll are shifted in the reverse direction. Also, this vertical magnetic field is much smaller than the vertical magnetic field of the case of Fig. 10 which is detected when the magnetic sensors 7 are set at the central position P.
In other words, when the magnetic sensors 11 are shifted in the travel direction of the thin steel strip lO, the floating magnetic field detected by the magnetic sensors ll is greatly reduced.
Fig. 12 shows the result of defect detection of a thin steel strip 10 in which a defect having a through hole having a diameter of 0.6 mm was artificially formed. This experiment shows a case indicated by a solid curve in which the positions of the magnetic 2~ 71~

sensors 11 were shifted in the travel direction, and a case indicated by a broken curve in which the posi-tions of the magnetic sensors 11 were shifted in the reverse direction.
As indicated in Fig. 12, an optimum detection sensibility was obtained when the magnetic sensors 11 were shifted in the travel direction by a predetermined distance.
Even if the mount position of a magnetic sensor 11 in the travel direction (X direction) of the thin steel strip 10 is deviated a little, a fluctuation in detec-tion sensibility against an artificial defect is small.
Therefore, the magnetic sensors 11 can be easily mounted. For example, in the apparatus of this embodiment, an allowable range X of a mount position is X = 1 + 0.5 mm.
Figs. 13 and 14 are sectional views of still another embodiment of the magnetic inspection apparatus for a thin steel strip according to the present invention. Same portions as those of the magnetic inspection apparatuses shown in Figs. 31, 32 and 1 are denoted by the same reference numerals. Hence, a detailed description of overlapping portions is omitted.
In the magnetic inspection apparatus of this embodiment, a pair of hollow rollers 1 and lb are verti-cally arranged to sandwich a thin steel strip 10 between them. Each of the hollow rollers 1 and lb is made of a non-magnetic material. The outer diameters of the hollow rollers 1 and lb are set equal. However, a thickness t3 of the upper hollow roller lb is set smaller than a thickness tl of the lower hollow roller 1. One end of a hollow stationary shaft 2 extends along the axis of the hollow roller 1, and one end of a hollow stationary shaft 2a extends along the axis of the hollow roller lb. The other end of the stationary shaft 2 of the lower hollow roller 1 is fixed on the frame of a base (not shown). The other end of the stationary shaft 2a of the upper hollow roller lb is biased by a weak spring (not shown) toward the stationary shaft 2 of the lower hollow roller 1. The stationary shafts 2 and 2a are supported by the inner circumferential surfaces of the two ends of the hollow rollers 1 and lb, respectively, through a pair of roll bearings 3a and 3b, such that they are located along the axes of the hollow rollers 1 and lb, respectively. As a result, the hollow rollers 1 ad lb respectively rotate freely about the stationary shafts 2 and 2a as the central axes of rotation. When the thin steel strip 10 travels in the direction of an arrow A, the hollow rollers 1 and lb are rotated in the directions of arrows B and C, respectively.
In the upper hollow roller lb, a plurality of mag-netic sensors 7b are fixed to the stationary shaft 2a through a support member such that they face downward.

- 32 - 2~71~

The distal end of each magnetic sensor 7b opposes the inner circumferenti-al surface of the upper hollow roller lb at a small gap. An output signal from each magnetic sensor 7b is supplied to the outside through a signal cable 9a extending through the interior of the station-ary shaft 2a.
In the lower hollow roller 1, a magnetizer 4 is fixed to the stationary shaft 2 such that magnetic poles 4a and 4b of its magnetization core 4c face upward. The excitation current to a magnetization coil 6 is supplied through a power cable 8 extending through the interior of the stationary shaft 2.
A pole-to-pole distance W of the magnetizer 4 in the hollow roller 1 and a distance L between the magnet-izer 4 and the thin steel strip 10 satisfy a predeter-mined relationship (2L < W < 8L) in the same manner as the embodiments described previously.
In the magnetic inspection apparatus of this type, since the weight or tension force of the thin steel strip 10 is not directly applied to the upper hollow roller lb, the thickness t3 of the upper hollow roller lb can be set smaller than the thickness tl of the lower hollow roller 1. Thus, the detection sensibility of the magnetic sensors 7b can be further improved by setting the lift-off distance ~ between the magnetic sensors 7b and the thin steel strip 10.
As the thin steel strip 10 is sandwiched between _ 33 - 205~718 the upper and lower hollow rollers lb and 1, vibration accompanying travel is suppressed. As a result, a fluc-tuation in a lift-off distance ~ is decreased, leading to an increase in defect detection precision.
Figs. 15 and 16 are sectional views schematically showing the arrangement of still another embodiment of the magnetic inspection apparatus for a thin steel strip according to the present invention. The same portions as those of the magnetic inspection apparatus shown in Figs. 13 and 14 are denoted by the same reference numerals. Accordingly, a detailed description of overlapping portions is omitted.
In this embodiment, magnetic sensors 7 having the same arrangement as that of the magnetic sensors 7b housed in an upper hollow roller lb are housed in a lower hollow roller 1.
Fig. 17 shows the entire system of this magnetic inspection apparatus. The thin steel strip 10 fed from a supply reel 12 is guided to the pair of hollow rollers 1 and lb through front press rollers 13a and 13b and is taken up by a take-up reel 15 through rear press rollers 14a and 14b at a constant speed. The lower hollow roller 1 is connected to a magnetization power unit 16 through the power cable 8. The hollow rollers 1 and lb are connected to signal processing circuits 17a and 17 through the signal cables 9 and 9a, respectively.
Defect signals Yl and Y2 output from the signal -- 34 - ~5~7~

processing circuits 17 and 17a are input to a data proc-essing unit. The data processing unit 18 calculates a defect size a and a defect occurring position Xl by using the input defect signals yl and Y2. The calcu-lated defect size a and defect occurring position Xl aredisplayed on a dlsplay unit 19 using, e,g., a CRT
display.
The steps of calculating the defect occurring position Xl in the thickness-wise direction of the thin steel strip 10 and the defect size a by using this magnetic inspection apparatus will be described.
The pair of hollow rollers lb and 1 contact upper and lower surfaces, respectively, of the thin steel strip 10. Therefore, the distance between the magnetic sensors 7b and the upper surface of the thin steel strip 10 and the distance between the magnetic sensors 7 and the lower surface of the thin steel strip 10 are kept constant. Thus, a magnetic flux leakage value Yl detected by the magnetic sensors 7b if a defect is pre-sent can be expressed as a function of the defect size aand a distance to the defect, i.e., a depth Xl from one surface of the thin steel strip 10 on the magnetic sen-sors 7b side, and a magnetic flux leakage value Y2 detected by the magnetic sensors 7 can be expressed as a function of the defect size a and a distance to the defect, i.e., a depth X2 from the other surface of the thin steel strip 10 on the magnetic sensors 7 side:

- 35 ~ 205~7i~

Yl = Fl(Xl, a) . . . (1) Y2 = Fl(X2~ a) ...(2) These functions Fl and F2 can be approximated by, e.g., exponent attenuation curves, as shown in Fig. 18.
In this embodiment, equations (1) and (2) are approxi-mated as exponential functions as expressed in the fol-lowing equations (3) and (4):

Yl = Cll eXP[C12Xl + a] (3) Y2 = C11 eXP[C22X2 + a~ (4) where Cll, C12~ C21, and C22 are constants that are experimentally obtained in advance.
As a thickness T of the steel strip is predetermined, T = Xl + X2 The defect size a and the defect position Xl can be obtained by resolving simultaneous equations (3), (4), and (5).
Fig. 19 schematically indicates the steps of calcu-lating the defect position Xl and the defect size a.
When the defect slze a is changed as a = a3, a = a2, and a = al, the characteristic curve of a defect signal Yl is translated toward the right. When the defect size A
is changed as a = a3, a = a2, and a = al, the character-istic curve of a defect signal Y2 is translated to the - 36 ~ B

left.
Hence, points on the characteristic curves corre-sponding to the measured signal value Yl are bl, b2,and b3. Similarly, points on the characteristic curves cor-responding to the measured signal value Y2 are cl, c2,and C3. Accordingly, points where the defect sizes a are equal and simultaneously satisfying the signal val-ues Yl and Y2 are the points b2 and c2. As a result, the position Xl corresponding to these points b2 and c2 is the defect occurred position, and the given detect size a2 is the defect size a of the corresponding defect.
with the magnetic inspection apparatus for the thin steel strip having the above arrangement, the position Xl and the size a of a defect present in the interior or on the surface of the thin steel strip 10 can be accu-rately obtained by simple calculation equations from the defect signals Yl and Y2 f the thin steel strip 10 respectively detected by the magnetic sensors 7b and 7 arranged in the pair of hollow rollers lb and 1 that oppose each other to sandwich the thin steel strip 10 as the inspection target.
Since the defect occurred position Xl and the defect size A can be accurately obtained, the type of the defect can substantially be accurately obtained.
The types of defects that can be obtained with this magnetic inspection apparatus include a visible gauge, _ 37 - 2~718 an invisible gauge, a weld, a blow hole, a hole, an edge tongue flaw, an ear crack flaw, a trimmer flaw, and the like.
Since the defect occurring position, the detect size, the defect type, and so on can be accurately obtained, in this manner, when this magnetic inspection apparatus is incorporated in the testing line of a factory, the test data obtained by it can be used as significant lnformation in improving the product quality.
The magnetic sensors 7b and 7 are housed in the hollow rollers lb and 1, respectively. The hollow rollers lb and 1 are constantly urged against the upper and lower surfaces, respectively, of the thin steel strip 10 by a predetermined biasing force. Therefore, the distance between the magnetic sensors 7b and the upper surface of the thin steel strip 10 and the dis-tance between the magnetic sensors 7 and the lower sur-face of the thin steel strip 10 can be constantly kept to be constant values. Hence, even if the thin steel strip 10 is vertically vibrated during traveling, the distances to it are kept to be the constant values, and thus the defect measuring precision is further improved.
Fig. 20 shows a correspondence between defect sizes of actually measured defects and their evaluation results obtained by actually revealing the measured - 38 - ~l71iB

defects by cutting or the like and visually evaluating the defect sizes in terms of 5 levels of A to E by an observer. It is understood that the measured defect sizes highly correspond to their visual evaluations.
The present invention is not limited to the embodi-ments described above. In the data processing unit 18 of the embodiment, in order to calculate the defect occurring position Xl and the defect size a from the signal values Yl and Y2, the defect signals Yl and Y2 are approximated by exponential functions indicated in equations (3) and (4). However, if approximation to a simple function is impossible, a large number of cali-bration flaw samples having known defect positions and defect sizes are used for measurement. The relationship among the signal values Yl and Y2 f the magnetic sen-sors 7b and 7, respectively, the defect position Xl, and the defect size a can be stored in the form of a table, as shown in Fig. 21. The table of Fig. 21 may be retrieved by using the measured values Yl and Y2 obtained by actually measuring the thin steel strip 10, and a pair of values Yl and Y2 closest to the actually measured values Yl and Y2 may be found. Then, a defect position and a defect size corresponding to the pair of the sample values Yl and Y2 may be read out and deter-mined as the measurement result.
A magnetic detection circuit which is used for detecting, by using an over-saturation type magnetic -sensor, magnetic flux leakage caused in the interior or on the surface of the thin steel strip 10 magnetized by the magnetizer 4 housed in the hollow roller will be described.
Fig. 22 is a block diagram showing the schematical arrangement of the magnetic detection circuit.
An over-saturation type magnetic sensor 7 comprises a rod-like ferromagnetic core 21 and a detection coil 22 applied around the ferromagnetic core 21. A pulse voltage generator 23 outputs positive and negative pulse voltages at an equal interval, as shown in Fig. 23.
The output terminal of the pulse voltage generator 23 is connected to one terminal of the detection coil 22 of the over-saturation type magnetic sensor 7 through a resistor 24 as a stationary impedance. The other terminal of the detection coil 22 is grounded. A pulse voltage output from the pulse voltage generator 23 is applied to the detection coil 22. As a result, the fer-romagnetic core 21 is magnetized to an over-saturation range.
The one end of the detection coil 22 is connected to the input terminals of a positive voltage peak detec-tor 25 and a negative voltage peak detector 26. The peak detectors 25 and 26 detect positive and negative peak values vl and -V2, respectively, of an input signal. The peak values Vl and -V2 obtained by the peak detectors 25 and 26 are input to an adder 27. The -205~

adder 27 adds the peak value vl with the peak value -V2 and outputs an output voltage VO.
The operation principle will be described with reference to Figs. 24 to 28.
An AC power having an AC voltage waveform as shown in Fig. 24 is applied to the detection coil 22 of the magnetic sensor 7 through the resistor 24. Then, a Ovoltage eO generated across the two terminals of the detection coil 22 is determined in accordance with a resistance R of the resistor 24 and an impedance zs of the detection coil 22. Namely, it is expressed as:

eO = e~Zs/(R + zs) ... (6) where e is a voltage to be applied.
As the detection coil 22 is applied around the fer-romagnetic core 21, the impedance zs changes in propor-tion to the magnetic permeability of the ferromagnetic core 21.
Assume that an AC current is supplied to the detec-tion coil 22 without applying an external magnetic field to the magnetic sensor 7. As shown in Fig. 27, the mag-netic permeability of the ferromagnetic core 21 changes in accordance with the hysteresis characteristic of the core 21. Note that reference symbol _ denotes the num-ber of coil turns; and i, a coil current.
As a result, an output voltage generated across the two terminals of the detection coil 22 has a waveform as - 41 - ~5~71~

shown in Fig. 25. When the external magnetic field is not applied, the positive and negative waveforms are symmetrical to each other, and the positive voltage V
is equal to the negative voltage V2.
In this state, when an external magnetic field is applied, the magnetic fluxes intersecting the ferromag-netic core 21 become the composite magnetic flux of the magnetic field generated by the detection coil 22 and the external magnetic field. Thus, the waveform gener-ated across the two terminals of the detection coil 22 is Vl > v2, as shown in Fig. 26.
Accordingly, the external magnetic field can be indirectly measured by comparing the positive and nega-tive voltages Vl and v2 of the output voltage generated across the two terminals of the detection coil 22 and obtaining their difference. In the magnetic inspection apparatus, an external magnetic field corresponds to the strength of the magnetic flux leakage caused by a defect.
When the over-saturation type magnetic sensor 7 having such an arrangement is used, an output voltage V0 of 0 to 500 mv can be obtained for a weak magnetic flux density of 0 to 10 gauss, as shown in Fig. 28.
In the magnetic detection circuit of the arrange-ment shown in Fig. 22, the AC power applied to the mag-netic sensor 7 has positive and negative pulse voltage waveforms as shown in Fig. 23.

~ 20547 1 8 Since the pulse voltage is supplied to the detec-tion coil 22 of the magnetic sensor 7 in this manner, the power consumption is decreased compared to a case in which the ordinary AC power shown in Fig. 24 is supplied, leading to energy saving. For example, if the ratio of the pulse width to the pulse period of the pulse voltage is set to 10 to 100, the average power supplied to the magnetic sensor 7 can be suppressed to about 1/10 to 1/100. As a result, a battery can be suf-ficiently used as the power source of the magnetic inspection apparatus.
As the peak values of the voltage generated across the two terminals of the detection coil 22 are detected, the relative detection sensibility of the weak magnetic flux is not substantially changed even if the ratio of the pulse width to the pulse period described above is changed over a wide range of 2 to 100.
As the power consumption is low, even if a large number of magnetic sensors 7 are arranged in the mag-netic inspection apparatus in the widthwise direction of the thin steel strip 10, the power consumption is not greatly increased.
Fig. 29 is a block diagram schematically showing another arrangement of a magnetic detection circuit.
In the magnetic detection circuit of this arrangement, a bias circuit for applying a DC bias volt-age to the pulse voltage to be applied to the magnetic -_ 43 - 2054718 sensor 7 is added to the circuit of Fig. 22.
An output voltage V0 from an adder 2 7 is converted into a DC voltage by a low-pass filter 41 and is input to a differential amplifier 42. The differential ampli-fier 42 outputs a differential voltage ~v between the input voltage vO and a reference voltage Vs output by a reference voltage generator 43. The differential volt-age ~v output from the differential amplifier 42 is input to a DC power source 44. The DC power source 44 applies, through an adder 45, a DC bias voltage vB pro-portional to the differential voltage ~V to the pulse voltage to be applied to the magnetic sensor 7.
The effect obtained by applying the DC bias voltage VB to the pulse voltage will be described.
Fig. 30 is a graph showing the result obtained by measuring the relationship between the magnetization current and the output voltage of the detection coil 22 while changing the DC bias current from the DC power source 44 to 0 mA, 50 mA, 100 mA, 160 mA, and 200 mA.
For example, when a DC vias current of 0 mA is supplied, the linear characteristic range of the output voltage with respect to the magnetization current is 0 to 2.5 A. When a DC bias current of 100 mA is supplied to the detection coil 22, the linear characteristic range is widened to 0 to 4.5 A. In this manner, the measurement span can be increased by changing the DC
bias current, thus improving the defect detection -~ 44 ~ æ~ ~ 7 precision.
When the DC bias current is further increased over 100 mA, although the measurement span is not changed, the measurement range of a magnetic flux leakage shifts.
Inversely, the DC bias current may be controlled such that, when a thin steel strip 10 completely free from a defect is magnetically examined, the output voltage V0 is constantly 0 v.
In the magnetic detection circuit shown in Fig. 29, the bias voltage VB proportional to the differential voltage ~v between the output voltage vO and the refer-ence voltage Vs is applied to the pulse voltage.
Therefore, if no defect is present, the output voltage V0 is automatically controlled to 0 V. The frequency response of the control 1QP is low whereas the defect of a traveling thin steel strip 10 is expressed as a high frequency component. Therefore, the defect is reliably detected.
In this manner, the operation point is automati-cally corrected to the central portion of the measure-ment range of the magnetic detection circuit. Even if the measurement conditions change, a good measurement range can be constantly obtained, and the defect detection performance can be further improved.

Claims (8)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:
1. A magnetic inspection apparatus for a thin steel strip, comprising a pair of hollow rollers, which are rotatably supported by stationary shafts perpendicular to a travel path of a thin steel strip, which sandwich the travel path, and which are rotated by being contacted by upper and lower surfaces of the thin steel strip that travels along the travel path, a magnetizer, arranged in one of said pair of hollow rollers, for generating a magnetic field in the thin steel strip, and a magnetic sensor, arranged in the other of said pair of hollow rollers, for detecting a magnetic flux leakage caused by a defect in an interior or on a surface of the thin steel strip.
2. A magnetic inspection apparatus for a thin steel strip according to claim 1, wherein one of said pair of hollow rollers housing said magnetic sensor is located above the other of said pair of hollow rollers housing said magnetizer.
3. A magnetic inspection apparatus for a thin steel strip according to claim 2, wherein the one of said pair of hollow rollers housing said magnetic sensor has a thickness larger than that of the other of said pair of hollow rollers housing said magnetizer.
4. A magnetic inspection apparatus for a thin steel strip, comprising a pair of hollow rollers, which are rotatably supported by stationary shafts perpendicular to a travel path of a thin steel strip, which sandwich the travel path, and which are rotated by being contacted by an upper and lower surface of the thin steel strip traveling along the travel path, a magnetizer, arranged on one of said pair of hollow rollers, for generating a magnetic field in the thin steel strip, a pair of magnetic sensors, respectively arranged in said hollow rollers, for detecting magnetic flux leakage caused by a defect in an interior or on a surface of the thin steel strip, and a data processing unit for calculating a defect occurred position in a direction of a thickness of the thin steel strip and a defect size, of the defect, from magnetic flux leakage detected by each of said pair of magnetic sensors.
5. A magnetic inspection apparatus for a thin steel strip according to claim 4, wherein said magnetizer is arranged on one of said pair of hollow rollers that contacts the lower surface of the thin steel strip.
6. A magnetic inspection apparatus for a thin steel strip, comprising:
a hollow roller, rotatably supported by a stationary shaft perpendicular to a travel path of a thin steel strip and rotated by being contacted by a surface of the thin steel strip traveling along the travel path;
a magnetizer, arranged in said hollow roller, for generating a magnetic field in the thin steel strip; and a magnetic detection circuit for detecting magnetic flux leakage occurring due to a defect in an interior or on the surface of the thin steel strip;

wherein said magnetic detection circuit comprises:
an oversaturation type magnetic sensor housed in said hollow roller, said magnetic sensor including a detection coil applied around a ferromagnetic core, said detection coil having two terminals;
excitation power supply means for exciting said magnetic sensor to an oversaturation range by supplying an AC power to said detection coil of said magnetic sensor through a stationary impedance, said excitation power supply means comprising a pulse voltage generator for supplying positive and negative pulse voltages to said detection coil so as to generate a voltage across said two terminals of said detection coil;
voltage detecting means coupled to said detection coil for detecting positive and negative values of said voltage generated across said two terminals of said detection coil, said voltage detecting means comprising a pair of peak value detection circuits connected to one of said two terminals of said detection coil for respectively detecting said positive and negative peak values of said voltage generated across said two terminals of said detection coil; and arithmetic means for adding the positive and negative values that are detected by said voltage detecting means, and for determining the sum as a measured value corresponding to the magnetic flux leakage.
7. A magnetic inspection apparatus for a thin steel strip according to claim 6, wherein pulse widths of said positive and negative pulse voltages are set to be 0.01 to 0.1 times pulse cycles thereof.
8. A magnetic inspection apparatus for a thin steel strip according to claim 6, further comprising direct current bias adding means for adding a DC bias voltage to the pulse voltages supplied by said pulse voltage generator to said detection coil, and bias control means for variably controlling said DC bias voltage in accordance with the measured value determined by said arithmetic means.
CA 2054718 1990-02-22 1991-02-22 Magnetic inspection apparatus for thin steel strip Expired - Fee Related CA2054718C (en)

Applications Claiming Priority (8)

Application Number Priority Date Filing Date Title
JP4184090 1990-02-22
JP2-41840 1990-02-22
JP7737990 1990-03-27
JP2-77379 1990-03-27
JP2-189484 1990-07-19
JP18948490A JPH07109416B2 (en) 1990-07-19 1990-07-19 Magnetic flaw detector for thin steel strip
JP2-197871 1990-07-27
JP19787190A JPH07104329B2 (en) 1990-07-27 1990-07-27 Magnetic flaw detection method and apparatus for steel strip

Publications (1)

Publication Number Publication Date
CA2054718C true CA2054718C (en) 1996-08-20

Family

ID=27461116

Family Applications (1)

Application Number Title Priority Date Filing Date
CA 2054718 Expired - Fee Related CA2054718C (en) 1990-02-22 1991-02-22 Magnetic inspection apparatus for thin steel strip

Country Status (1)

Country Link
CA (1) CA2054718C (en)

Similar Documents

Publication Publication Date Title
US5235275A (en) Magnetic inspection apparatus for thin steel strip having magnetizer and detection coil within a hollow roller rotated by the steel strip
KR100246244B1 (en) Magnetic sensor
CA2088918C (en) Magnetic detecting method and apparatus therefor
US5502382A (en) Apparatus for magnetic inspection using magnetic shield, with specific relation of distance E between magnetic sensors and distance L between each magnetic sensor and the object to be inspected, to obtain optimum S/N
JPH0572180A (en) Method and device for magnetic-field inspection
JP2848081B2 (en) Magnetic flaw detector
JPS6352345B2 (en)
CA2054718C (en) Magnetic inspection apparatus for thin steel strip
JP2697435B2 (en) Calibration method and apparatus for magnetic flaw detector
JP2002257789A (en) Leakage flux detecting device
JPH1183808A (en) Leakage flux flaw detecting method
EP0544911B1 (en) Device for detecting magnetic flux
KR960009762B1 (en) Magnetic flaw detector for thin steel belt
JP2617605B2 (en) Magnetic measuring device and diagnostic method for magnetic flaw detector
JPH11108900A (en) Method and apparatus for calibration of sensitivity of magnetic flaw-detecting device
JPH03239960A (en) Detecting device for flaw of steel plate
JPH09166582A (en) Electromagnetic flaw detection method
JP2861581B2 (en) Magnetic detection method and device
JP3530472B2 (en) Bar detection system
RU2032898C1 (en) Method of magnetic inspection of defects of long-length ferromagnetic articles
JPH07104329B2 (en) Magnetic flaw detection method and apparatus for steel strip
JPH03246463A (en) Flaw detecting device for steel plate
Janez Longitudinal and Circumferential Cracks in a Ferromagnetic Bar Detected Simultaneously
JP2000329746A (en) Apparatus and method for eddy current flaw detection
SU1293620A1 (en) Method of electromagnetic flaw detection of ferromagnetic objects

Legal Events

Date Code Title Description
EEER Examination request
MKLA Lapsed