JP6883476B2 - Welding defect detection system and welding defect detection device equipped with this system - Google Patents

Description

The present invention relates to a welding defect detection system and a welding defect detecting device including the same.

In arc welding such as MIG welding and MAG welding, an arc is generated from the tip of the wire, causing a part of the wire to melt, and the thermal pinch effect of the magnetic field causes the wire to constrict, causing droplets to come from the wire. It leaves in a drop shape. In such welding, when the voltage is increased, the droplets become smaller and become spray-like. When the voltage is set so that the droplets are between the drop shape and the spray shape, spatter occurs frequently.

As welding to prevent such spatter, there are pulse mig welding and pulse mug welding in which the welding current has a pulse waveform. By making the welding current a pulse waveform, droplets are formed and separated from the wire in a timely manner, so that the occurrence of spatter can be suppressed.

By the way, in the conventional welding defect detecting device for detecting welding defects such as spatter, the welding current and the welding voltage are measured, and when these measured values deviate from the target range by a predetermined number of times, an alarm is output. It is configured (see, for example, Patent Document 1). Such a welding defect detecting device has an advantage that it can detect the presence or absence of welding defects while proceeding with welding.

Japanese Patent No. 3285203

However, since the welding defect detection device of the above patent document is intended for welding in which the welding current and welding voltage do not fluctuate significantly, sufficient accuracy can be obtained when applied to pulse mig welding or pulse mug welding in which the welding current has a pulse waveform. There wasn't.

Therefore, an object of the present invention is to provide a welding defect detection system capable of detecting welding defects, particularly blow holes, with high accuracy in pulse arc welding, and a welding defect detecting device including the same.

In order to solve the above problems, the welding defect detection system according to the first invention performs pulse arc welding, and at the same time, detects defects in the welding bead placed in this operation and the measured welding current of the pulse arc welding. And a welding defect detection system that detects based on the value of welding voltage.
A drawing unit that creates a frequency distribution map at a predetermined time with the above welding current and welding voltage values as axes of the Cartesian coordinate system, respectively.
And have you in the histogram, the maximum value of the frequency in the area smaller than the average of the peak of the set value of the base of the set value and the welding current of the welding current detected one of the frequencies in larger area than the average A first section that detects one maximum value and divides the frequency distribution map into short-circuited and non-short-circuited areas with a line parallel to the line connecting these two maximum values and having a lower welding voltage than the line.
In the second section, which divides the frequency distribution map in the area of the base and peak of the welding current on average,
A statistic calculation unit that calculates a summary statistic of each area partitioned by the first section and the second section, and a summary statistic based on the value of the welding current at the predetermined time, respectively.
It is provided with a defect determination unit for determining the presence or absence of defects in the welding bead depending on the degree to which the summary statistic calculated by the statistic calculation unit exceeds the corresponding reference value.

Further, in the welding defect detection system according to the second invention, the defect determination unit in the welding defect detection system according to the first invention
An individual judgment unit that determines whether the summary statistics of each area exceed the respective standard values,
It has a comprehensive judgment unit that determines the presence or absence of defects in the weld bead by a value that weights each judgment by the individual judgment unit.

Further, the welding defect detection system according to the third invention sets a reference value corresponding to the summary statistics of each area and / or sets a weighting coefficient in the welding defect detection system according to the second invention. , It is done automatically.

In order to solve the above problems, the welding defect detection system according to the fourth invention performs pulse arc welding, and at the same time, detects defects in the welding bead placed in this operation and the measured welding current of the pulse arc welding. And a welding defect detection system that detects based on the value of welding voltage.
A drawing unit that creates a frequency distribution map at a predetermined time with the above welding current and welding voltage values as axes of the Cartesian coordinate system, respectively.
In the above frequency distribution diagram, one maximum value of the frequency is detected in an area smaller than the average of the set value of the base of the welding current and the set value of the peak of the welding current, and the maximum value of the frequency is detected in the area larger than the average. The first section that divides the frequency distribution map in the short-circuited and non-short-circuited areas with a line parallel to the line connecting these two maximum values and having a lower welding voltage than the line.
The base area of the welding current, and a second partition part partitioning the area of a peak of the welding current, a frequency distribution diagram and a transition area between the base and the peak area of the welding current,
A statistic calculation unit that calculates a summary statistic of each area partitioned by the first section and the second section, and a summary statistic based on the value of the welding current at the predetermined time, respectively.
It is provided with a defect determination unit for determining the presence or absence of defects in the welding bead depending on the degree to which the summary statistic calculated by the statistic calculation unit exceeds the corresponding reference value.

Further, in the welding defect detection system according to the fifth invention, the summary statistics of each area in the welding defect detection system according to the fourth invention are dispersed when the base and transition areas are non-short-circuited, and the non-short-circuited. The peak area at is the variance, frequency and mode, and the short circuit area is the frequency.
A summary statistic of welding current at a given time is the standard deviation in the frequency domain.

Further, the welding defect detecting device according to the sixth invention is a welding defect detecting device including the welding defect detecting system according to any one of the first to fifth inventions.
An ammeter and a voltmeter that measure the welding current and welding voltage of pulse arc welding, respectively.
A recorder connected to the welding defect detection system and a recorder connected to the welding defect detection system while recording the welding current and welding voltage values measured by the ammeter and the voltmeter, respectively.
It is provided with a warning device that warns when a defect in the welding bead is detected by the welding defect detection system.

According to the welding defect detection system and the welding defect detection device provided with the welding defect detection system, the welding defect is detected by the summary statistics based on the frequency distribution map and the value of the welding current, so that the welding defect is detected with high accuracy in pulse arc welding. be able to.

It is a schematic block diagram which shows the welding defect detection apparatus which concerns on embodiment of this invention. It is a block diagram of the welding defect detection system included in the welding defect detection apparatus. It is a graph of the welding current (solid line) and welding voltage (broken line) at a predetermined time transferred to the welding defect detection system when there is no welding defect. It is a graph of the welding current (solid line) and welding voltage (broken line) at a predetermined time transferred to the welding defect detection system when there is a welding defect. It is a frequency distribution diagram when there is no welding defect created by the drawing part of the welding defect detection system. It is a frequency distribution map when there is a welding defect created by the drawing part of the welding defect detection system. It is a figure explaining the function of the 1st section diagram of the welding defect detection system, (a) shows the line connecting the maximum value of a frequency, and (b) is parallel to this line, and the frequency distribution is short-circuited and non-short-circuited. The lines that divide the figure are shown. It is a figure explaining the function of the 2nd section drawing of the welding defect detection system, and also explaining the 6 sections which were divided. It is a figure which shows the operation of the welding defect detection apparatus by a line chart. It is a figure which shows the display of the display provided in the welding defect detection apparatus.

Hereinafter, a welding defect detection system according to an embodiment of the present invention and a welding defect detection device including the same will be described. The welding defect detection system and the welding defect detection device are connected to a welding device that performs pulse arc welding (for example, pulse mig welding or pulse mug welding), and defects (particularly blow holes) in the weld bead placed in this work. Is detected based on the measured welding current and welding voltage values of the pulse arc welding. In the following, the blow hole in the welding bead is simply referred to as a welding defect.

First, the welding defect detection device will be described with reference to FIG.

As shown in FIG. 1, the welding defect detection device 1 includes an ammeter 2 for measuring the current (welding current I) of the circuit from the base material M to the welding power supply device A, and the wire supply device W and the base material M. It includes a voltmeter 3 that measures a voltage (welding voltage V), and a data recorder 4 that can record and output the values of the welding current I and the welding voltage V measured by the ammeter 2 and the voltmeter 3. Further, the welding defect detection device 1 includes a data processor 5 (for example, a personal computer) that processes the values of the welding current I and the welding voltage V output from the data recorder 4. The data processor 5 stores the welding defect detection system that detects welding defects based on the values of the welding current I and the welding voltage V from the data recorder 4, which will be described in detail later. Further, the welding defect detection device 1 is connected to a display 6 for displaying the result of processing by the data processor 5, a relay 7 for passing instructions between the data processor 5, and the relay 7. A welding robot control panel 9 for passing instructions between the warning light 8 and the relay 7 and controlling a welding robot (not shown) is provided.

Known equipment is used for the ammeter 2, the voltmeter 3, and the data recorder 4. For example, the ammeter 2 is a clamp meter, a shunt resistor, or the like. In FIG. 1, the ammeter 2, the voltmeter 3, and the data recorder 4 are shown as one set each, but may be two sets. That is, while the data processor 5 reads the recorded data from one set of data recorders 4 and performs welding defect detection processing, the data processor 5 records the data so that the other set of data recorders 4 record the data. It is clear that by alternating the operations of, welding and detection of welding defects can be performed without interruption. Each output of the voltmeter and the ammeter may be shared by two data recorders 4.

Hereinafter, the welding defect detection system, which is the gist of the present invention, will be described in detail with reference to FIG.

As shown in FIG. 2, the welding defect detection system 50 stored in the data processor 5 creates a frequency distribution map based on the values of the welding current I and the welding voltage V from the data recorder 4. And a first section 51 and a second section 52 for partitioning the frequency distribution map created by the drawing section 58, respectively. Further, the welding defect detection system 50 has a summary statistic in the area partitioned by the first section 51 and the second section 52, and a summary statistic based on the value of the welding current I from the data recorder 4. The statistic calculation unit 53 for calculating each of the above is provided. Further, the welding defect detection system 50 includes a reference value setting unit 57 for setting a reference value for comparison with the summary statistic calculated by the statistic calculation unit 53. In addition, the welding defect detection system 50 determines the presence or absence of the welding defect according to the degree to which the summary statistic calculated by the statistic calculation unit 53 exceeds the corresponding reference value. 54 is provided.

The drawing unit 58 creates a frequency distribution map of the values of the welding current I and the welding voltage V at a predetermined time with the values of the welding current I and the welding voltage V as axes of the Cartesian coordinate system, respectively. The procedure for creating this frequency distribution map is as follows. The values of the welding current I and the welding voltage V from the data recorder 4 are continuous values for a certain period of time (for example, the time required for one pass of welding). As shown in FIG. 3, the values at these constant times are filtered and extracted every predetermined time (for example, 1 second). In FIG. 3 and FIG. 4 described later, the continuous value of the welding current I is shown by a solid line, and the continuous value of the welding voltage V is shown by a broken line. The continuous values of the welding current I and the welding voltage V at the predetermined time are extracted as individual values in a predetermined processing cycle (for example, 1 kHz), and these extracted values are plotted in the Cartesian coordinate system. For example, when the predetermined time is 1 second and the processing cycle is 1 kHz, the points plotted in the Cartesian coordinate system are 1 second (predetermined time) / 1 kHz (processing cycle) = 1000 points. The Cartesian coordinate system in which the values are plotted in this way becomes a frequency distribution map, and an example thereof is shown in FIG.

Here, FIGS. 3 and 5 described above show a case where there is no welding defect, and FIGS. 4 and 6 show a case where there is a welding defect. As is clear from the comparison between FIG. 5 (when there is no welding defect) and FIG. 6 (when there is a welding defect), in the above frequency distribution diagram, when there is a welding defect, the frequency is large in the area corresponding to the short circuit. In the area corresponding to the non-short circuit, the welding voltage V varies greatly in the high and low directions (vertical direction in FIGS. 5 and 6), and the frequency is large in the area corresponding to the peak. This is to quantitatively grasp such features that appear in the frequency distribution map when there are welding defects as numerical values, compare the captured numerical values with the reference values, and use the results of the comparison to detect welding defects. It is the technical idea of the invention.

The first section 51 detects two maximum values of the frequency in the frequency distribution map, and divides the frequency distribution map into short-circuited and non-short-circuited areas with a line parallel to the line connecting the two maximum values. Is. The procedure for partitioning the frequency distribution map into short-circuited and non-short-circuited areas is as follows. As shown in FIG. 7A, in the frequency distribution map, for example, two portions H having the maximum frequency are detected in the histogram distribution, and these two portions H, that is, the two maximum values H are penetrated by the line L1. Let me. This line L1 is translated in the direction in which the welding voltage V becomes low (downward in FIG. 7), and as shown in FIG. 7 (b), the point where the frequency distribution diagram shows the large welding voltage V by the line L2. It is divided into a group (non-short circuit) and a small point group (short circuit).

The second section 52 partitions the frequency distribution map in the base, peak, and transition areas of the welding current I. As shown in FIGS. 3 and 4, in the pulsed welding current I (solid line) and welding voltage V (broken line), the base is the valley portion B and the peak is the peak portion P. , The transition is the rising and falling part rf. That is, as shown in FIG. 8, the second section 52 divides the frequency distribution map into a point group having a small welding current I, a point group having a large welding current I, and a point group in between. The point group having a small welding current I corresponds to the base of the welding current I, the point group having a large welding current I corresponds to the peak of the welding current I, and the point group between them corresponds to the transition (that is, rising) of the welding current I. And falling). In the following, in the frequency distribution map, as shown in FIG. 8, among the six areas partitioned by the first section 51 and the second section 52, the non-short-circuited area starts with the smaller welding current I. It is referred to as a non-short-circuit base area PB, a non-short-circuit transition area Prf, and a non-short-circuit peak area PP, and is referred to as a short-circuit base area sB, a short-circuit transition area srf, and a short-circuit peak area sP from the smallest welding current I in the short-circuit area. The non-short-circuited base area PB, the non-short-circuited transition area Prf, and the non-short-circuited peak area PP are point groups having a large welding voltage V. On the other hand, the short-circuit base area sB, the short-circuit transition area srf, and the short-circuit peak area sP, which are short-circuit areas, are point clouds with a small welding voltage V.

The statistic calculation unit 53 calculates a summary statistic of the six areas partitioned by the first section 51 and the second section 52, and a summary statistic based on the value of the welding current I at the predetermined time. Is what you do. Specifically, as a summary statistic of the above six areas, the variance is calculated in the non-short-circuit base area PB, the variance is calculated in the non-short-circuit transition area Prf, and the variance, frequency and mode are calculated in the non-short-circuit peak area PP. Is calculated, and the frequency is calculated in each of the three short-circuit areas sB, srf, and sP. Further, as a summary statistic based on the value of the welding current I, the value of the welding current I at the predetermined time is Fourier transformed, the Fourier transformed value is extracted in a predetermined frequency region, and the standard deviation of the extracted value is obtained. calculate. That is, the summary statistics calculated by the statistic calculation unit 53 are the following nine (1) to (9). Specifically, (1) dispersion in the non-short-circuit base area PB, (1) 2) Variance in non-short-circuit transition area Prf, (3) Variance in non-short-circuit peak area PP, (4) Frequency in non-short-circuit peak area PP, (5) Mode in non-short-circuit peak area PP, ( 6) the frequency in the short-circuit base area sB, (7) the frequency in the short-circuit transition area srf, (8) the frequency in the short-circuit peak area sP, and (9) the standard deviation of the values of the Fourier transformed welding current I.

As shown in FIG. 2, the defect determination unit 54 weights the individual determination unit 55 that determines whether or not the summary statistic exceeds each reference value, and each determination by the individual determination unit 55. It has a comprehensive determination unit 56 that determines the presence or absence of welding defects based on the obtained values. The individual judgment unit 55 compares the summary statistic with each reference value, and sets the number 1 as the comprehensive judgment unit each time the summary statistic exceeds the reference value (or the summary statistic exceeds the reference value). Output to 56, otherwise the number 0 is output to the comprehensive judgment unit 56. The comprehensive judgment unit 56 multiplies the value (that is, 1 or 0) output by the individual judgment unit 55 by a weighting coefficient according to the importance of the summary statistics compared by the individual judgment unit 55, that is, Weight. These weighted values are summed up, and if the summed up values exceed the threshold value, it is considered that a welding defect has occurred, and the warning light 8 is instructed via the relay 7. Further, the comprehensive determination unit 56 stores the data generated in the process for the detection in the recording medium 59 (for example, a hard disk) regardless of whether or not the welding defect is detected, and also in this process. The progress is output to the display 6 as a table or a graph. The weighting coefficient may be set manually or automatically. When the weighting factor is set automatically, for example, a discriminant analysis based on a summary statistic of a welded sample whose presence or absence of defects is known in advance is used.

The reference value setting unit 57 sets nine reference values for comparison with the nine summary statistics based on the values of the welding current I and the welding voltage V from the data recorder 4. The reference value setting unit 57 calculates, for example, the average value and the mode value of the welding current I and the welding voltage V from the data recorder 4, and sets the above nine reference values from the average value and the mode value. .. The reference value may be set manually or automatically. When the reference value is set automatically, for example, a discriminant analysis based on a summary statistic of a welded sample whose presence or absence of defects is known in advance is used. In this case, the reference value may be one mathematical formula (multidimensional) instead of nine.

Hereinafter, the operation of the welding defect detection system 50 and the welding defect detection device 1 including the welding defect detection system 50 will be described.

When pulse arc welding is performed by the welding robot, the measured values of the welding current I and the welding voltage V are recorded in the data recorder 4.

Explaining these with reference to FIG. 9, when the welding signal is turned ON, the mode is during welding, and the welding defect detection device 1 starts data recording (welding mode).

The start of welding can be known by detecting the welding current I. Although not shown in FIG. 1, many welding power supplies A have a function of closing relay contacts while a welding current I is flowing, and can be used as a welding signal by extracting a contact signal to the outside. Welding is complete when the contacts switch from on to off. The same effect can be obtained by extracting the signal during welding from the welding robot to the outside.

When the welding for which the welding signal changes from ON to OFF (that is, one pass) is completed, the data processor 5 puts the welding robot on standby and outputs the values of the welding current I and the welding voltage V from the data recorder 4 (that is, the values of the welding current I and the welding voltage V from the data recorder 4 are output. Transfer) and perform defect detection processing (detection mode). The fixed time, which is the time when the welding signal changes from ON to OFF, is not limited to one pass, and is, for example, the time when one welding is started and finished.

When a welding defect is detected by the processing by the data processor 5, the warning light 8 is turned on to notify the operator. The warning light 8 is turned off by pressing a release button (not shown). By pressing the release button, the warning light 8 is turned off and the next welding is waited for.

The data processor 5 extracts values from the data recorder 4 at a fixed time (for example, the time required for one pass of welding) at shorter predetermined times (for example, 1 second), and weld defects at these predetermined times. Judge the presence or absence of. When a welding defect is detected, in order to make it easier to intuitively grasp where the welding defect is in the welding bead, as shown in FIG. 10, on the display 6, for example, the above-mentioned fixed time is displayed by a bar graph 61 or the like. Then, the portion 62 of the bar graph 61 corresponding to the predetermined time determined to have the welding defect d is emphasized.

As described above, according to the welding defect detecting system 50 and the welding defect detecting device 1 provided with the welding defect detecting system 50, the welding defect is detected by the summary statistic based on the frequency distribution map and the value of the welding current I. Welding defects can be detected with high accuracy.

Further, since the weighting is performed according to the importance of the summary statistic for detecting the welding defect, the welding defect can be detected with higher accuracy.

In addition, summary statistics have been adopted that are optimal for detecting welding defects in pulsed arc welding, ie pulsed mig or pulsed mug welding. That is, the summarized statistics adopted are variance for the non-short-circuit base area PB and non-short-circuit transition area Prf, variance, frequency and mode for the non-short-circuit peak area PP, and short-circuit area. Each is a frequency and is the standard deviation of the value of the Fourier transformed welding current I. Therefore, in pulse arc welding, welding defects can be detected with higher accuracy.

In addition, by automatically setting the reference value and / or the weighting coefficient, welding defects can be detected in a short time and with higher accuracy.

By the way, in the above-described embodiment, the first section 51 and the second section 52 have been described as dividing the frequency distribution map into six areas, but the non-short-circuit transition area Prf and the short-circuit transition area srf are omitted. It may be divided into two areas PB, PP, sB, and sP. With such a configuration, not only the number of divisions of the frequency distribution map is reduced, but also the number of summary statistics and reference values is the same as the number of areas, so that the processing by the data processor 5 is reduced. As a result, welding defects can be quickly detected in pulse arc welding.

Further, in the above embodiment, the order of partitioning the frequency distribution map is described with the first section 51 first and the second section 52 second, but the second section 52 comes first and the first section 51 May be later. Further, the first section 51 and the second section 52 may simultaneously partition the frequency distribution map.

In addition, in the above-described embodiment, it has been described that the frequency distribution map is divided into the six areas by the first division 51 and the second division 52 (that is, the two divisions). It may be a single compartment having the functions of the portion 51 and the second compartment 52.

Further, although not described in detail in the above embodiment, the data recorder 4 and the data processor 5 may be connected to the network of the factory where the welding apparatus is installed by a LAN or the like.

Further, in the above embodiment, as an example of the summary statistic, the non-short-circuit base area PB and the non-short-circuit transition area Prf are the variance, and the non-short-circuit peak area PP is the variance, the frequency and the mode, and the short circuit. In each of the areas, it is a frequency, and it is explained as a standard deviation of the value of the Fourier transformed welding current I, but these are only examples, and may be other summary statistics.

Further, in the above embodiment, as shown in FIG. 7B, it has been described that the line L2 parallel to the line L1 penetrating the two maximum values H is divided into a non-short-circuited area and a short-circuited area. However, the line L1 and the line L2 do not necessarily have to be completely parallel, and may be appropriately non-parallel depending on the distribution of the point group having a large welding voltage V and the point group having a small welding voltage V. Further, in the above-described embodiment, the area is partitioned as shown in FIG. 8, but the present invention is not limited to this, and may be partitioned by, for example, a welding current I and a welding voltage V in a certain range from the maximum value H. Also, it may be divided into 6 or more areas.

Further, in the above embodiment, the configuration for creating the frequency distribution map has been described, but instead of creating the frequency distribution map, an approximate model of the expected welding voltage based on the measured welding current I is created. Weld defects may be detected by comparing this approximate model with the measured welding voltage V. This approximate model utilizes the fact that the welding current I and the welding voltage V have a linear relationship. For example, the value obtained by multiplying the first coefficient by the measured welding current I, the second coefficient, and the third The expected welding voltage is the sum of the coefficient multiplied by the time change of the welding voltage V. Here, the first coefficient, the second coefficient, and the third coefficient are coefficients that are arbitrarily determined.

1 Welding defect detection device 2 Ammeter 3 Voltmeter 4 Data recorder 5 Data processor 6 Display 7 Relay 8 Warning light 9 Welding robot control panel 50 Welding defect detection system 58 Drawing section 51 First section 52 Second section 53 Statistics Quantity calculation unit 54 Defect judgment unit 55 Individual judgment unit 56 Comprehensive judgment unit 57 Reference value setting unit 59 Recording medium

Claims (6)

It is a welding defect detection system that detects defects in the welding bead placed in this construction based on the measured welding current and welding voltage values of the pulse arc welding while performing the pulse arc welding.
A drawing unit that creates a frequency distribution map at a predetermined time with the above welding current and welding voltage values as axes of the Cartesian coordinate system, respectively.
And have you in the histogram, the maximum value of the frequency in the area smaller than the average of the peak of the set value of the base of the set value and the welding current of the welding current detected one of the frequencies in larger area than the average A first section that detects one maximum value and divides the frequency distribution map into short-circuited and non-short-circuited areas with a line parallel to the line connecting these two maximum values and having a lower welding voltage than the line.
In the second section, which divides the frequency distribution map in the area of the base and peak of the welding current on average,
A statistic calculation unit that calculates a summary statistic of each area partitioned by the first section and the second section, and a summary statistic based on the value of the welding current at the predetermined time, respectively.
A welding defect is provided with a defect determination unit for determining the presence or absence of a defect in the welding bead depending on the degree to which the summary statistic calculated by the statistic calculation unit exceeds the corresponding reference value. Detection system. Defect judgment department
An individual judgment unit that determines whether the summary statistics of each area exceed the respective standard values,
The welding defect detection system according to claim 1, further comprising a comprehensive determination unit that determines the presence or absence of defects in the welding bead based on a value weighted for each determination by the individual determination unit. The welding defect detection system according to claim 2, wherein a reference value corresponding to a summary statistic of each area is set and / or a weighting coefficient is automatically set. It is a welding defect detection system that detects defects in the welding bead placed in this construction based on the measured welding current and welding voltage values of the pulse arc welding while performing the pulse arc welding.
A drawing unit that creates a frequency distribution map at a predetermined time with the above welding current and welding voltage values as axes of the Cartesian coordinate system, respectively.
In the above frequency distribution diagram, one maximum value of the frequency is detected in an area smaller than the average of the set value of the base of the welding current and the set value of the peak of the welding current, and the maximum value of the frequency is detected in the area larger than the average. The first section that divides the frequency distribution map in the short-circuited and non-short-circuited areas with a line parallel to the line connecting these two maximum values and having a lower welding voltage than the line.
The base area of the welding current, and a second partition part partitioning the area of a peak of the welding current, a frequency distribution diagram and a transition area between the base and the peak area of the welding current,
A statistic calculation unit that calculates a summary statistic of each area partitioned by the first section and the second section, and a summary statistic based on the value of the welding current at the predetermined time, respectively.
A welding defect is provided with a defect determination unit for determining the presence or absence of a defect in the welding bead depending on the degree to which the summary statistic calculated by the statistic calculation unit exceeds the corresponding reference value. Detection system. The summary statistics for each area are variance for non-short-circuit base and transition areas, variance, frequency and mode for non-short-circuit peak areas, and frequency for short-circuit areas. Yes,
The welding defect detection system according to claim 4, wherein the summary statistic of the welding current at a predetermined time is a standard deviation in the frequency domain. A welding defect detection device including the welding defect detection system according to any one of claims 1 to 5.
An ammeter and a voltmeter that measure the welding current and welding voltage of pulse arc welding, respectively.
A recorder connected to the welding defect detection system and a recorder connected to the welding defect detection system while recording the welding current and welding voltage values measured by the ammeter and the voltmeter, respectively.
A welding defect detecting device including a warning device that warns when a defect in a welding bead is detected by the welding defect detecting system.

Images