JP2003098159A - Apparatus for inspecting surface of cylindrical body and method for processing data - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that a surface defect of a cylindrical body can be inspected accurately only when a reflected wave from the defect is discriminated from noise, and to provide an apparatus for inspecting the surface of a cylindrical body and a method for processing data in which the effect of noise is eliminated by setting a decision criterion of significant data depending on the scanning direction of the cylindrical body. SOLUTION: The apparatus for inspecting the surface of a cylindrical body comprises a plurality of probes arranged on the surface of the cylindrical body while having circumferential and oblique scanning directions, a first grouping number of scanning data within the circumference of the cylindrical body predetermined in correspondence with individual probes and a second grouping number having a predetermined number of pitches appearing in the scanning pitch of continuous scanning data, and a processing unit for evaluating the scanning data from grouped scanning data satisfying the first and second grouping numbers.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data processing method in a cylinder surface inspection device for inspecting defects on the surface and surface layer of a cylinder such as a rolling roll by ultrasonic flaw detection.

[0002]

2. Description of the Related Art Most of conventional devices are disclosed in Japanese Patent Laid-Open No. 07-2.
As described in Japanese Patent No. 80777, a single element type surface wave probe, a single-channel flaw detector, a data recording processing device for recording and processing an echo signal, or a recorder for direct recording is used. A device that evaluates the soundness of the cylinder surface and the surface layer by injecting ultrasonic waves in the circumferential direction of the cylinder by a single element type surface wave probe and performing automatic flaw detection with an ultrasonic beam in only one direction. is there.

Regarding data grouping, there is, for example, Japanese Patent Laid-Open No. 8-198818. It is disclosed that in an ultrasonic flaw detector, data obtained by sampling is grouped according to the identity of a reflection source, and related data is used to store the data. Regarding data grouping, Japanese Patent Laid-Open No. 8-248016
There is a gazette. This makes the inter-data linkage during grouping easy to reorganize. It has a judgment condition table for grouping it. Further, as a prior art, there is Japanese Patent Laid-Open No. 3216548. It is stated that the sampling cycle of the measurement data is determined from the number of data in the memory for storing the data, and the sampling cycle obtained by A / D conversion is further made 1/2 times shorter than this. ing.

[0004]

In the above-mentioned prior art, since the single probe is used, good probe performance is easily obtained, and the defect detectability itself is good. However, since the ultrasonic wave incident direction is only one direction, there is a problem that the detectability is lowered depending on the direction of the defect. In addition, depending on the material of the rolling roll, if the flaw detection sensitivity is not increased and flaw detection is not performed, it may be difficult to detect defects, and reflection noise other than defects of the cylindrical body may be affected by reflection from minute water droplets. However, this may have been an obstacle to defect determination. Further, there are the prior arts as described above, but the grouping is based on the identity of the reflection sources, and the inter-data linkage is structured to be easily reorganized. It also relates to how to determine the data collection cycle.

An object of the present invention is to detect a reflected wave from a defect as significant data by distinguishing a signal for each probe from the above-mentioned reflection noise, and to detect flaws on the surface of a cylindrical body. The object of the present invention is to provide a cylindrical body surface inspection device and a data processing method by grouping significant data according to a plurality of arranged probes.

[0006]

The above-mentioned problems can be solved by the following means. A plurality of probes for transmitting and receiving ultrasonic waves, and a cylindrical body surface inspection device for detecting flaws on the surface of a cylindrical body by ultrasonic waves, which has a switcher for performing switching synthesis of ultrasonic wave transmission and reception with the probe. The plurality of probes arranged on the surface of the cylindrical body with the flaw detection direction in the axial direction of the body, the circumferential direction, and the diagonal direction, and a predetermined one circle of the cylindrical body corresponding to the plurality of individual probes The first grouping number of flaw detection data and the second grouping number that predetermines the number of pitches in which flaw detection data appears in continuous flaw detection data, and the grouping flaw detection data satisfying the first and second grouping numbers An arithmetic unit that evaluates data,
It is characterized by having and. Further, the arithmetic unit arranges the second grouping number for a probe which is obliquely arranged with a flaw detection direction among the plurality of probes, and arranges the second grouping number with a flaw detection direction in an axial direction or a circumferential direction of the cylindrical body. The probe is characterized in that it is a computing unit that determines the first grouping number and evaluates flaw detection data.

Further, a plurality of probes for transmitting and receiving ultrasonic waves are arranged on the surface of the cylindrical body, and the trigger signal from the flaw detector is controlled to be switched according to the probes, so that the trigger signals from the plurality of probes are transmitted. Cylindrical surface inspection that controls switching of signals for receiving and combining, performs probe signal transmission processing on the probe and signal reception processing from the probe, and performs flaw detection on the cylindrical surface with ultrasonic waves. In the data processing method, a probe having a flaw detection direction in the axial direction, the circumferential direction and an oblique direction is arranged on the surface of the cylindrical body, and the cylindrical body is predetermined in correspondence with the flaw detection directions of the plurality of probes. The flaw detection data satisfying the first and second grouping numbers by providing a first grouping number of flaw detecting data within one round and a second grouping number in which the number of pitches in which flaw detecting data appear consecutively are determined. Evaluation of flaw detection data using It is characterized in the data processing method of performing.

Further, the second grouping number is applied to the probes arranged so as to have a flaw detection direction in an oblique direction among the plurality of probes, and a columnar body is used among the plurality of probes. The probe arranged in the axial direction or the circumferential direction with the flaw detection direction is provided with the first grouping number, and the flaw detection data is evaluated using the flaw detection data satisfying the first and second grouping numbers. There is a feature in the inspection data processing method.

Further, a gate section is provided in advance for flaw detection signals from the plurality of probes, a threshold level is set for the data in the gate section, and the first grouping number and the data for the data exceeding the threshold level are set. A feature is the inspection data processing method provided with the second grouping number.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to FIGS. First, the overall configuration of the cylindrical body surface inspection apparatus according to the present embodiment will be described with reference to FIG. Trigger pulse signal 2 from flaw detector 1
Are switched and selected by the signal switch 4 in the multiplexer 3, and the pulsar 5 (ch1, ch2, ...
ch4, which is an example of 4ch in this case), and the corresponding channel of the multi-channel probe 7 having the ultrasonic beams 6 in four directions is excited. Then, the ultrasonic pulse signal is incident on the surface of the rolling roll 8 to be inspected.

At this time, the channel selection of the switching selector 4 is performed by the designation signal 4a from the computer 9. The ultrasonic signal reflected by the surface defect 10 of the rolling roll is received by the corresponding channel of the multi-channel probe 7, and is amplified by the preamplifier 11 (preamplifier ch1 to ch4) of the corresponding channel. The amplified received ultrasonic wave signal 12 is sent to the flaw detector 1 as a signal 12a via the signal switch 4 in the multiplexer 3 and amplified,
Signal processing such as detection is performed. The peak value of the received ultrasonic signal 12a subjected to signal processing is extracted by a flaw detection gate described later, peak-held, and then output as a signal 13 to the computer 9.

The peak hold output 13 is A / D converted and taken into the computer 9, and is held as flaw detection data 14 of the designated channel until the flaw detection cycle of the next 4 channels, and D / A converted to the corresponding channel of the recorder 15. It is output and recorded. For the above operation, ch1
It is possible to perform flaw detection with the ultrasonic beams 6 in four directions by sequentially repeating the process up to ch4. In FIG. 1, reference numerals 16 to 18 show a schematic configuration of the probe pressing mechanism portion, which are integrally configured. With the roller 16 in contact with the rolling roll 8 as the center, the probe holding head 1 is provided before and after the roller 16.
The probe pressing mechanism is configured by arranging 7 and the draining wiper 18. The probe holding head 17 is in the first quadrant of the cylinder, the drain wiper 18 is in the fourth quadrant, and the roller 16 is
Are placed between them (along the circumference).

The ultrasonic beam 6 of the multi-channel probe 7 is incident on the drain wiper 18 side. Therefore, the roller 16 is arranged at a position outside the propagation path of the ultrasonic beam 6. It is necessary to inject ultrasonic waves through a contact medium such as water, and the contact medium is supplied from the contact medium supply device 19 from the multi-channel probe unit.

When the contact medium supplied from the contact medium supply device 19 exists on the ultrasonic wave incident surface, it serves as a reflection source and hinders the detection of defects, but the rolling roll 8 rotates in the direction 20 shown in the drawing. , The contact medium flows on the surface in the direction opposite to the flaw detection surface, makes one revolution, and is removed by the drain wiper 18,
Does not affect the flaw detection surface. Further, it is more effective if this draining is performed by a nozzle using air pressure. The flaw scanning on the roll of the probe is spirally performed by the rotation of the rolling roll by the grinder and the axial feed.

FIG. 2 shows another embodiment. The computer 9 records and processes the flaw detection data, and outputs the flaw detection result to the display of the computer 9 and the printer 34 without using a recorder. The control of the flaw detector 1 and the process of fetching the peak hold data are performed by the computer 9 as in the embodiment of FIG. In this embodiment, in addition to the peak hold output from the flaw detector 1, beam path length data 35, and probe position data 36 from the probe pressing mechanism unit are taken in and recorded. By recording the beam path data 35 and the probe position data 36, the accurate position of the defect can be obtained, and the plan view, flaw detection chart, defect data list, etc. can be created by data processing, It can be evaluated by the chart output.

Next, the flaw detection signal will be described with reference to FIG. As shown in FIG. 3A, the trigger signal is sent from the computer 9 to the flaw detector at a constant cycle, and is further sent to the pulser 5 through the signal switch 4. The switch 4 is
Since the connection signal of the pulsar 5 is sequentially switched according to the designation from the computer 9, the pulsar of ch1 is driven at the first (cycle 1) and the pulsar of ch2 at the next (cycle 2) as shown in FIG. 3B. However, the pulser of ch4 is sequentially driven, and the pulser of ch1 is driven again in (cycle 4 + 1). Since each channel of the multi-channel probe 7 is connected to each channel of the pulsar 5 in a one-to-one manner, ultrasonic pulses are sequentially transmitted to the surface of the rolling roll 8 to be inspected at this timing.

The ultrasonic reflected wave reflected by the surface defect 10 of the rolling roll is received by the multi-channel probe 7, is sent to the preamplifier 11 and is amplified. The preamplifier 11 also has a channel (preamplifier ch1) like the pulsar 11. ~ C
h4) is selected. Therefore, the ultrasonic reflected waves are sequentially received and amplified as indicated by ch1 to ch4 at the timing shown in FIG. The signal sent to the flaw detector via the signal switch 4 is a signal obtained by superimposing all channels of the output signal of the preamplifier 11 as shown in (D) of FIG. Processed, FIG.
As shown in (E), the flaw detection gate 21 is applied to the reflected wave 22 from the defect. That is, the reflected wave 22 is detected as a flaw detection signal.

The reflection intensity of the reflected wave in the gate 21
A voltage proportional to the (echo height) is peak-held when the reflected wave 22 is generated, and is output to the computer 9 as a peak-hold signal as shown in FIG.
As shown in (F), the peak hold output is A / D converted after the flaw detection gate 21 is completed and before the next flaw detection cycle is started. The A / D converted data is
The signal 14 is again D / A converted at the timing shown in (H) of FIG. 3 and is output to the recorder 15. D / A
The converted data is separated for each channel and output to the corresponding channel of the recorder 15, and is held for only 4 cycles until updated in the next same cycle. By the signal processing described above, flaw detection can be performed with the ultrasonic beams 6 in four directions, and the flaw detection result can be output and recorded in the recorder 15.

Next, an example of flaw detection results in the present invention will be described. FIG. 4 shows a detection range of surface defects 10 (hereinafter abbreviated as defects) when using a multi-channel probe 7 (hereinafter abbreviated as probes) in which ultrasonic beams are incident in four directions at a pitch of 45 °. ing. FIG. 4A shows a state in which the defect 10 is detected in ch1. The ch1 ultrasonic beam 25 is incident at an angle of −45 ° with respect to the circumferential direction 23 of the rolling roll. When the flaw detection gate 21 is provided at the position of the distance W1 (29) to W2 (30) on the surface of the rolling roll, the probe position detected farthest is indicated by a solid line, and the probe position detected closest is detected. The position of the tentacles is indicated by the dotted line.

The moving distance L1 in the axial direction of the rolling roll during this period
Is given by 24 is the axial direction of the rolling roll, 23
Indicates the circumferential direction of the rolling roll. L1 = (W2-W1) / √2 The flaw detection scanning is performed by feeding the roll 8 in the axial direction while rotating the rolling roll 8. Assuming that the axial feed pitch per rotation is P, m1 times (= 4 times) are detected almost every rotation. m1 = L1 / P

FIG. 4B shows a state in which the defect 10 is detected in ch2. The ch2 ultrasonic beam 26 is incident in the circumferential direction 23 of the rolling roll. In ch2, basically, at the position where the axial position of the probe matches the defect position, m
It is detected twice. m2 = (W2−W1) / (Ls × ΔT × n) where ΔT is the flaw detection repetition period and Ls is the rolling roll peripheral speed. However, it may be detected even before or after the ultrasonic beam due to its spread.

FIG. 4C shows a state in which a defect is detected on ch3, and the ultrasonic beam 27 on ch3 is incident at an angle of 45 ° with respect to the circumferential direction 23 of the rolling roll. ch1
Similarly to the above, m3 = m1 times (= 4 times) is detected almost every one rotation. FIG. 4D shows a state in which the defect 10 is detected on ch4, and the ultrasonic beam 28 on ch4 is incident 24 in the axial direction of the rolling roll. In ch4, the same idea as in ch1 is used, and m4 times are detected almost every one rotation. m4 = (W2-W1) / P

FIG. 5 shows an example of recording flaw detection data when the probe is fed from the right side to the left side of the rolling roll 8. The probe 7 approaches the defect 10 by the roll rotation feed operation,
First, ch1 is detected. This corresponds to (a) of FIG. In ch1, it is detected about m1 times (= 4 times),
Since the beam path length at the time of detection gradually becomes shorter, the intensity of the received reflected wave gradually becomes higher, and the beam path becomes the flaw detection gate start point 2
When it becomes shorter than 9, it will not be detected.

Next, at the position where the axial position of the probe matches the defect position, ch2 detects m2 times (= 3 times). This corresponds to (b) of FIG. Also in this case, since the probe gradually approaches the defect due to the roll rotation operation, the intensity of the received reflected wave gradually increases. ch2
In some cases, the beam divergence may be detected even at a position where it makes one more turn.

Next, the probe is detected by ch3 about m3 times (= 4 times) at the axial position beyond the defect position. This corresponds to (c) in FIG. Since ch3 is detected while moving away from the defect, the beam path length at the time of detection is gradually lengthened, the intensity of the received reflected wave is gradually lowered, and the beam path is not detected when the beam path length is longer than the flaw detection gate end point 30. The ch4 detected next (corresponding to (d) in FIG. 4) also shows the same mode as ch3.

In the flaw detection of the rolling roll, since water is used as the contact medium of ultrasonic waves, minute water droplets may remain and cause noise echo. However, since the water droplets are sequentially removed by the wiper (or compressed air), the adhesion is temporary and is not continuously detected like the defect 10, and the regularity of detection as described above is also present. Because I don't have
It can be easily identified as noise.

The above description is for the case where the reflection from the defect 10 is nearly omnidirectional, but the defect targeted for detection is a micro defect, and this condition is satisfied for the micro defect. On the other hand, a defect having a strong defect reflection directivity is generally a large defect and has a high ultrasonic reflection intensity, so that it can be detected in any channel and may be discriminated as a defect by detecting only one channel. As described above, by detecting defects with ultrasonic beams in a plurality of directions, it is possible to realize detection of surface and surface layer defects having high noise discrimination.

Next, the grouping process which is the data processing method of the present invention will be described. Table 1 shows an example of flaw detection data storage inside the computer 9. As shown in Table 1, only the crest value for each flaw detection cycle is stored, and the position and coordinate signals are not used as in the system of FIG. Also, the number of rotations and the feed rate are not taken into consideration. This data processing is suitable for the flaw detection method that focuses on the passage of time and the flaw detection cycle.
The time difference between the flaw detection times T1 and T2 is set to the time required for the roll to rotate once. A specific value is about 2 seconds per rotation. This is a so-called flaw detection pitch. The number of flaw detection cycles increases one by one at intervals of 125 (8 ms) when the transmission frequency is 500 Hz (2 ms) and the four-probe configuration is used. In the case of this numerical condition, 250 points of measurement data during one rotation,
It shows the case where it is detected and recorded for each channel.

[0029]

[Table 1]

In Table 1, the crest value of ch1 is 1 at the flaw detection cycle 1.
Is stored. As is clear from the concrete data in Table 2, the data of the crest value 1 is "5". Further, the specific data in the case of the flaw detection cycle 2 is “4”.

As already shown in FIG. 5, the flaw detection result of this apparatus is characterized in that defect signals are continuously detected in synchronization with the rotation cycle. Grouping processing is performed using this regularity. The procedure of the grouping process will be described below with reference to Tables 2 and 3. Table 2 focuses on the data of ch1 and describes the flaw detection times Tn to T (n + 3).

[0032]

[Table 2]

[0033]

[Table 3]

(1) The specified threshold value is compared with the crest value of flaw detection data, and data below the threshold level (threshold value) is replaced with a numerical value of zero. As the concrete value of the threshold level, the peak value of 5 or 10
Although a numerical value is set, the processing is performed as 10 in Table 2 (or the level may be normalized by the maximum peak value and the level may be set in%. In the case of the flaw detection time Tn and the flaw detection cycle 1, data is “ 4 ", and in this case the threshold level is" 10 ", so the noise removal data is" 0 ".
Becomes Also in the case of the flaw detection cycle (M-1), the data is "5" and the noise removal data is "0".

(2) In one rotation cycle, if the flaw detection data continues for more than a predetermined number of grouping constants within a predetermined cycle, it is determined that the data is valid, and if less than the constant number, it is regarded as noise. It is removed (in ch1, the ultrasonic beam is incident at an angle of 45 degrees, so it is basically only once, and in Table 2, processing is performed with the in-cycle grouping constant being 1). The data in the next rotation cycle is processed in the same manner, and the same processing is performed until the end of the rotation cycle.

(3) Further, it is determined that the data is valid when the flaw detection data is present at a continuous pitch (a flaw detection locus trace is drawn in a spiral shape but the number of feed pitches thereof) which is continuous for a grouping constant between pitches designated for each rotation cycle. Then, the same judgment is made up to the final data. As a concrete numerical value of the pitch grouping constant, a numerical value of 2 or more and m1 times (m1 = 4 as shown in FIG. 5A) or less is set. Table 2 is an example of the case where the process is performed as "4". This means that even if there is only one significant data in one cycle, the flaw detection time is T (n +
If the flaw detection can be continuously detected at 1), T (n + 2), and T (n + 3), it is determined as significant data.

(4) A management number (for example, AN1) is assigned to the number of flaw detection data grouped.

(5) The highest value among the grouped data groups is searched for and used as the characteristic value of that group.

Although the above description has been made focusing on ch1, the grouping processing can be performed on the other channels (ch3, 4) in the same way.

On the other hand, in ch2, since the ultrasonic beam is incident in the circumferential direction, the grouping constant is different from that of other ch. Table 3 shows an example of the flaw detection result in the case of ch2. c
In h2, the number of consecutive detections within one rotation cycle is theoretically m2 (= 3), so the intra-cycle grouping constant is set to a value of 2 to m2 (= “3”). . The reason why it may be set to m2 or less is to cope with the case where the difference between the threshold level and the height of the detected echo is small. On the other hand, in the case of ch2, a defect is basically detected in only one rotation cycle, so the number of pitch groupings is set to 1.

Table 3 shows a case where data is obtained at two consecutive flaw detection times, that is, the flaw detection time of Tn and the flaw detection time of T (n + 1). Noise removal data is L to (L + 4)
, And the grouping number within one cycle is set to “3”, these data are adopted as significant data. In this example, an example is shown in which the grouping number "3" within one cycle is satisfied even at the next flaw detection time T (n + 1).

FIG. 6 shows these processing flows in the computer 9. Step 62 is a comparison with a so-called threshold level, and data below a predetermined value is replaced with "0". In steps 63 and 64, the continuity of data is judged. In step 65, the data regarded as continuous data is left, grouping within one cycle, grouping between measurement pitches is performed, a management number is given (for example, AN1 to n), and the highest value of the wave height data is searched for. Characteristic value. In step 66, a chart or a plan view is developed based on the data obtained up to step 65 and the evaluation is performed.

As described above, by performing the grouping process, even when a noise signal having no continuity is superimposed and detected, the noise data is removed and the number of defects and their Features such as peak value can be extracted. Finally, FIG. 7 shows an example of the overall flaw detection work flow when this data processing is incorporated.

First, at step 72, control conditions for driving the grinding machine are set. Next, in step 73, the flaw detection conditions for ultrasonic flaw detection are set, and as the flaw detection conditions, control condition settings for the grinder are set. Next, the flaw detection conditions are set. As the flaw detection conditions, in addition to the grinder, the inspection object management information and the flaw detection device conditions are set. Next step 7
Although the parameter for grouping is set in step 4, usually, an initial value is set so that it does not need to be changed except in a special case. Thus, the flaw detection preparation is completed, and subsequently, in step 75, flaw detection scanning and data recording are started to perform the flaw detection operation. After the flaw detection operation is completed, data analysis processing including grouping processing is performed in step 76, and evaluation / evaluation of the inspection object is performed from the data repair output diagram.

As described above, the hardware configuration is a multi-channel surface wave probe having two or more transducers for transmitting / receiving ultrasonic waves in at least two directions, and ultrasonic wave transmission / reception control of two or more channels. It is composed of a flaw detector and a multiplexer for performing the operation, a recorder and a computer for recording the echo signal from the flaw detector.

Reflected echoes from the defect are detected at a plurality of flaw detection positions. It is detected at a plurality of positions on one flaw detection scanning line, and is also detected at a plurality of positions on one flaw detection scanning line of the next feed pitch. The data captured by the computer will be stored in the form of a table. In the table, the data is continuous. It is possible to determine whether the data is continuous by determining whether there is data in the flaw detection cycle next to the flaw detection cycle in which the flaw detection data is stored or whether there is data in the flaw detection cycle next to the flaw detection cycle. it can.

It is also possible to carry out a comparison judgment at an interval twice or three times the flaw detection cycle, which is regarded as continuous, depending on whether or not there is data within the interval. Similarly, in the pitch feed direction, it is possible to determine whether there is data in the adjacent pitch or the next pitch. Continuity can be determined as an interval at twice or three times the pitch. There is an effect that data having no length or area such as data of only one point, data of only a few points, and data detected only on one scanning line can be omitted by these processes.

The number of defects can be obtained by assigning a management number to each grouped data group. The maximum peak value in the grouped data group is extracted and used as the representative peak value of this group. As a result of flaw detection, it is also possible to create an inspection table containing numerical values by outputting a group number and maximum peak value for each channel as a table. Then, it is possible to quantitatively evaluate and judge by numerical values.

The present invention differs from the conventional apparatus in that data processing is performed by using the flaw detection period signal and the peak value data of ultrasonic waves without using the position signal or the beam path signal. In addition, if a threshold value is set by a computer and data below the threshold value is removed before D / A conversion and only signals above the threshold value are recorded in the recorder in real time, flaw detection recording It is possible to remove unnecessary noise in the probe appearing in (3) and to identify defects easily.

As a pre-processing of the grouping process, the same effect can be obtained by performing the grouping process after leaving the data equal to or more than the threshold value. As described above, by providing the means for distinguishing the reflected wave from the defect from the reflected noise with a simple structure, it is possible to provide the cylindrical body surface inspection device having a good defect discrimination property.

[0051]

According to the present invention, noise can be removed by processing the data, and the accurate position or size of the surface defect can be obtained from the significant data. Moreover, there is an effect that it can be evaluated by various chart outputs.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a cylindrical body surface inspection apparatus according to a first embodiment of the present invention.

FIG. 2 is an overall configuration diagram of a cylindrical body surface inspection device according to another embodiment of the present invention.

FIG. 3 is a schematic time chart of flaw detection signals in the cylindrical body surface inspection apparatus.

FIG. 4 is an explanatory diagram of a detection range of surface defects by the multi-channel probe.

FIG. 5 is an example of detection of flaw detection data of rolling rolls by chart display.

FIG. 6 is an overall flow diagram of data processing of the device.

FIG. 7 is a schematic overall flow chart of grouping processing.

[Explanation of symbols]

1 ... Flaw detector 2 ... Trigger pulse signal 3 ... Multiplexer 4 ... Signal switching 5 ... Pulser 6 ... Ultrasonic beam 7
... multi-channel probe 8 ... rolling roll 9 ... computer 10 ... surface defect 11 ... preamplifier 12 ... received ultrasonic signal 12a ... signal-processed received ultrasonic signal 13 ... peak hold output 14 ... flaw detection data 15 ... recorder 1
6 ... Roller 17 ... Probe holding head 18 ... Drain wiper 19 ... Contact medium supply device 20 ... Roll rotation direction 21 ... Flaw detection gate 22 ... Reflected wave from defect 23 ...
Circumferential direction of rolling roll 24 ... Axial direction of rolling roll 25
... ch1 ultrasonic beam 26 ... ch2 ultrasonic beam 27 ... ch3 ultrasonic beam 28 ... ch4 ultrasonic beam 29 ... flaw detection gate start point 30 ... flaw detection gate end point 34 ... printer 35 ... beam path data 3
6 ... Probe position data.

Claims (5)

[Claims] 1. A plurality of probes for transmitting and receiving ultrasonic waves, a flaw detector for controlling ultrasonic wave transmission and reception with the probe, and a switch for switching a trigger signal from the flaw detector according to the probe. Control and a switching device for synthesizing received signals from the plurality of probes, in a cylindrical body surface inspection device for flaw detection of the cylindrical body surface by ultrasonic waves, in the axial direction of the cylindrical body, the circumferential direction and A plurality of probes arranged on the surface of the cylindrical body with a flaw detection direction in an oblique direction; and a first grouping number of flaw detection data within one round of the cylindrical body, which is predetermined corresponding to the plurality of individual probes. A second grouping number in which the number of pitches in which flaw detection data appears in continuous flaw detection data is determined in advance, and an arithmetic unit for evaluating flaw detection data based on the grouped flaw detection data satisfying the first and second grouping numbers; Have Cylindrical surface inspection apparatus according to claim and. 2. The arithmetic unit according to claim 1, wherein the arithmetic unit has the second grouping number for a probe which is arranged obliquely with a flaw detection direction in the plurality of probes, A cylindrical body surface inspection device, wherein the probe arranged in the axial direction or the circumferential direction with a flaw detection direction is a computing unit that determines the first grouping number and evaluates flaw detection data. 3. A plurality of probes for transmitting and receiving ultrasonic waves are arranged on the surface of a cylindrical body, and a trigger signal from the flaw detector is controlled to be switched according to the probes, so that the trigger signals from the plurality of probes are transmitted. Cylindrical surface inspection that controls switching of signals for receiving and combining, performs probe signal transmission processing on the probe and signal reception processing from the probe, and performs flaw detection on the cylindrical surface with ultrasonic waves. In the data processing method, a probe having a flaw detection direction in the axial direction, the circumferential direction and an oblique direction is arranged on the surface of the cylindrical body, and the cylindrical body is predetermined in correspondence with the flaw detection directions of the plurality of probes. The flaw detection data satisfying the first and second grouping numbers by providing a first grouping number of flaw detecting data within one round and a second grouping number in which the number of pitches in which flaw detecting data appear consecutively are determined. Evaluate flaw detection data using Cylindrical surface test data processing method characterized by. 4. The probe according to claim 3, wherein the second grouping number is assigned to the plurality of probes arranged so as to have a flaw detection direction in an oblique direction. The probe arranged in the axial direction or the circumferential direction of the cylindrical body in the probe with the flaw detection direction is provided with the first grouping number, and flaw detection data satisfying the first and second grouping numbers is used. A cylindrical surface inspection data processing method, characterized in that flaw detection data is evaluated by means of the method. 5. The method according to claim 3, wherein a gate section is provided in advance for flaw detection signals from the plurality of probes, a threshold level is set for the data in the gate section, and a threshold level is set for data exceeding the threshold level. A cylindrical body surface inspection data processing method comprising providing a first grouping number and the second grouping number.