JP2002079386A - Method for detecting weld defect in laser beam welding - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for detecting a weld defect or a method for monitoring quality of a weld zone which method is carried out by using a measuring device having a comparatively low resolution in a laser beam welding and enables obtaining of a stable and excellent correlation with the presence or absence of blow holes or pit defects even when a secular contamination is generated at the light-receiving part of the measuring device. SOLUTION: The method for detecting the weld defect in the laser beam welding is characterized in that the intensity of a predetermined light, which is in the range of wave length between 741 and 748 nm including the light emission spectrum of nitrogen among the light emission of the laser beam- induced plasma formed at a point irradiated or pierced with the laser beam, and the intensity of a predetermined light, which includes the light emission spectrum of iron and excludes the light emission spectrum of argon, are each measured in a laser beam welding, a ratio of the light intensities derived by dividing the former intensity by the latter intensity is used as a decision signal, and the presence or absence of the weld defect is determined by comparing the decision signal with the decision value preliminarily decided by an experiment.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a welding defect in laser welding, in which the presence or absence of a welding defect such as a blowhole or a pit generated in a laser welding portion can be reliably known during welding.

[0002]

2. Description of the Related Art In laser welding, since the heat source is a laser beam, the amount of heat input can be easily controlled as compared with a welding method utilizing a discharge phenomenon (arc) in a gas such as TIG welding or MIG welding. Yes, if the welding conditions are set appropriately, a high quality welded joint can be obtained efficiently.

However, even in laser welding, it is difficult to completely eliminate fluctuations in welding conditions such as turbulence of the shield gas flow and welding speed. Since defects may occur, it is important to detect whether or not welding defects have occurred.

[0004] As a conventional method for monitoring a defect in a welded portion during laser welding, Japanese Patent Application Laid-Open No. 08-267241 discloses a method in which a part of a material to be welded or a shielding gas is ionized and becomes a plasma during laser welding. Of the A on the back side of the weld
There is disclosed a method of measuring the emission spectrum line intensity of r, He or Fe, and detecting the burn-through of the welded portion or the state of the back bead shape from the change in the intensity.

In Japanese Patent Application Laid-Open No. 08-281457, the intensity of light on the front side of the weld is measured from two directions with different depression angles, and the plasma light intensity inside the keyhole and the plasma light intensity above the keyhole are separately measured. And a method for accurately detecting the molten state of a welded portion and monitoring the occurrence of a defective bead shape.

However, Japanese Patent Application Laid-Open No. 08-267241 discloses
In the technology disclosed in Japanese Patent Application Laid-Open No. 08-281457 and Japanese Patent Application Laid-Open No. 08-281457, although it is possible to monitor the shape of a weld bead or burn-through of a welded portion, blow holes and pit defects have different generation mechanisms. There was a problem that it could not be detected.

[0007] The emission intensity of laser-induced plasma is also measured in order to know the occurrence of blowholes and pit defects, which are problems in carrying out the present invention.

In Japanese Patent Application Laid-Open No. 11-188489, when welding the back surface of a steel sheet by shielding it with He gas,
A method for detecting the presence or absence of a blow hole or a pit defect by determining whether the intensity of light emitted from a plasma plume generated on the back surface of a steel plate is twice or more a predetermined intensity, and a method for detecting the back surface of the steel plate with Ar or Ar and He. When shielding with a mixed gas, of the light emitted by the laser-induced plasma on the back of the steel sheet, the intensity of the specific wavelength light of nitrogen (emission spectrum line) is measured, and by monitoring whether the value abnormally fluctuates. A method for detecting the presence or absence of a welding defect is disclosed. As the specific wavelength light of nitrogen, 410.
It is stated that it is preferable to use emission lines of 3 nm, 441.0 nm, and 568.0 nm.

However, in the monitoring method disclosed in the above-mentioned Japanese Patent Application Laid-Open No. H11-188489, the measuring instrument used is highly accurate.
Unless the wavelength resolution is sufficiently high, the correspondence between the change in light intensity and the presence or absence of a borrow hole or a pit defect becomes extremely ambiguous, and there is a problem that detection of the defect becomes difficult. In addition, since the presence or absence of welding defects is determined only by the intensity of light emitted from the plasma plume or the intensity of light of a specific wavelength of nitrogen, the light receiving portion of the plasma light is contaminated by fumes generated by welding, and the signal intensity decreases over time. In such a case, there is a problem that the presence / absence of a welding defect is erroneously determined.

[0010]

SUMMARY OF THE INVENTION In view of the above-mentioned problems of the prior art, an object of the present invention is to use an inexpensive measuring instrument having relatively low resolution during laser welding, and to generate fumes generated by welding. Even if the light receiving ability is reduced due to contamination of the light receiving part due to the deterioration of the signal intensity and the signal strength is reduced, it is possible to stably detect the presence or absence of a welding defect over a long period of time. It is to provide a monitoring method.

[0011]

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and the gist thereof is as follows.

(1) When laser welding is performed by shielding the welded portion on the front surface or the back surface of the material to be joined with He gas,
Among the light emitted by the laser-induced plasma generated near the laser processing point of the welded portion shielded by e-gas, a wavelength range 741 to 748n including an emission spectrum line of nitrogen.
m, and the intensity of the predetermined light including the emission spectrum line of iron was measured, and the intensity ratio of light obtained by dividing the former by the latter was determined as a determination signal. A method for detecting welding defects in laser welding, wherein the presence or absence of a welding defect is determined by comparing a determination value.

(2) When laser welding is performed by shielding the welded portion on the front or back surface of the material to be joined with Ar gas or a mixed gas containing Ar, the material is shielded with Ar gas or a shield gas containing Ar gas. Of the light emitted by the laser-induced plasma generated in the vicinity of the laser processing point of the welding portion, a wavelength range 741 including a light emission spectrum line of nitrogen is included.
A predetermined light intensity at 748 nm and a predetermined light intensity including an emission spectrum line of iron and not including an emission spectrum line of Ar are measured, and an intensity ratio of light obtained by dividing the former by the latter is used as a determination signal. A method for detecting welding defects in laser welding, wherein the presence or absence of a welding defect is determined by comparing the determination signal with a determination value determined in advance by an experiment.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below.

In general, when the shield gas flow rate at the welded portion fluctuates during laser welding or when the shield gas flow is disturbed due to a change in the geometrical shape near the processing point, the shielding performance is reduced and nitrogen molecules in the atmosphere are reduced. Penetrates into the vicinity of the processing point irradiated with the laser beam. The invading nitrogen molecules are dissociated or ionized into atoms in the laser-induced plasma, but these atomic nitrogens are much easier to dissolve in molten steel than nitrogen molecules. Therefore, a large amount of nitrogen is dissolved in the molten steel, and when the molten metal solidifies, the nitrogen dissolved in the molten steel forms blowholes and pit defects.

This is because the nitrogen solubility of the solidified phase is extremely small as compared with the molten steel. Therefore, when nitrogen is dissolved in excess of the solubility of the solidified primary crystal, nitrogen that cannot be dissolved in the primary crystal is added to the molten steel. The released nitrogen concentration in the molten steel increases with the progress of solidification. When the amount of nitrogen in the molten steel exceeds the solubility of the molten steel, bubbles are formed in the molten steel. Bubbles that cannot be removed from the molten steel and remain in the weld metal
The bubbles solidified in a state of being separated from the molten steel form "pits" opened on the bead surface.

Therefore, there is a correlation between the amount of atomic nitrogen dissociated in the laser-induced plasma and the occurrence of blowholes or pit defects.

The above-mentioned Japanese Patent Application Laid-Open No. 11-188489 discloses a method for detecting a welding defect utilizing this correlation. In this method, when the back surface of a steel plate is shielded with Ar or a mixed gas of Ar and He, the emission spectrum line intensity of nitrogen is measured in the light emitted by the laser-induced plasma on the back surface. As spectral lines, wavelengths 410.3 nm, 411.0 nm, 56
A light of 8.0 nm is preferred.

However, according to experiments by the present inventors, when a measuring instrument having a wavelength resolution of about 2 nm is used, the wavelengths of 410.3 nm, 411.0 nm, 411.0 nm and 411.0 nm disclosed in JP-A-11-188489 are used. The emission spectrum line at 568.0 nm was indistinguishable from the emission spectrum line of iron, making measurement difficult.

Therefore, in the method of detecting blowholes and pit defects of the weld bead by measuring only the intensity of the emission spectrum line of nitrogen, the emission intensity and the welding defect are required unless the accuracy (wavelength resolution) of the measuring instrument used is extremely high. It is not possible to clarify the correspondence between the occurrence and the occurrence of the problem.

Further, even if the wavelength resolution of the measuring instrument is sufficiently high, the signal intensity decreases with time due to contamination of the light receiving surface, and the correspondence between the intensity of the emission spectrum line and the occurrence of weld bead defects is long term. Stability was not obtained.

Therefore, the present inventors can use an inexpensive measuring instrument having a relatively low resolution (wavelength resolution of about 6 nm) and reduce the signal intensity of the plasma light due to the temporal contamination of the light receiving section. However, a method for measuring laser-induced plasma that can obtain a stable correlation with the presence or absence of laser welding defects in the long term was studied.

Hereinafter, selection of a nitrogen emission spectrum line which can be easily measured and a method of producing a determination signal which can detect the presence or absence of a blowhole or a pit defect even using a measuring device having relatively low resolution will be described in order.

First, selection of a nitrogen emission spectrum line that can be easily measured will be described.

Generally, what emission spectrum lines each element has is described in the literature. Such documents include, for example, M.S. I. t. WAVELEN
GTH TABLES, Volume2, Wave
lengths by Element, THE M
There is IT Press. However, since the intensity of each emission spectral line greatly changes depending on the plasma state such as temperature and density, the intensity of each emission spectral line in the specific situation of laser-induced plasma generated during laser welding should be estimated only from literature information. It is difficult.

Therefore, the present inventors prepared a cold-rolled steel sheet having a thickness of 1 mm using a laser output of 4 kW at a processing point and a welding speed of 6 m / m.
butt laser welding experiments were performed for the case where He was used as the shielding gas and the case where nitrogen was used as the shielding gas. In each case, the emission spectra were compared by spectrally measuring the light of the laser-induced plasma generated on the steel sheet surface. Then, emission spectrum lines of nitrogen, which are easy to measure in laser welding, were investigated.

2 and 3 show the results of the experiment.

FIG. 2 shows the emission spectrum of the laser-induced plasma in the nitrogen shield and the ratio of the intensity (spectral intensity) at each wavelength of the plasma light in the nitrogen shield to the spectral intensity of the plasma light in the He shield. Range 300 ~
The result measured at 600 nm is shown. Here, the wavelength resolution of the measurement device is about 2 nm in full width at half maximum, and the measurement wavelength interval is 0.5 nm.

In normal laser welding, He hardly emits light. Therefore, in FIG. 2, if light emission specific to nitrogen exists, the intensity ratio should increase. However, there is no wavelength at which the intensity ratio is remarkably increased, including 410.3 nm, 411.0 nm, and 568.0 nm described in JP-A-11-188489, and the emission spectrum line specific to nitrogen is I could not confirm.

In FIG. 2, the wavelength range is 450 nm to 450 nm.
Although the light intensity ratio is large at 470 nm, 570 nm to 600 nm, etc., in this wavelength range, the intensity of the plasma light is extremely weak regardless of the shield gas type, and the noise level and the signal level are almost the same. It is a meaningless change.

According to this experiment, when an inexpensive measuring instrument having a relatively low wavelength resolution is used, the wavelength range is 300 nm to 600 nm.
In nm, no emission line specific to nitrogen could be found.

FIG. 3 shows a wavelength range 60 on the longer wavelength side than FIG.
At 0 to 800 nm, the emission spectrum of the laser-induced plasma in the nitrogen shield and the ratio of the spectral intensity of the plasma light in the nitrogen shield to the spectral intensity of the plasma light in the He shield are shown as in FIG.

Referring to FIG. 3, the wavelengths are 742 nm and 747 n.
It is observed that the intensity ratio is higher at m.
Observing the emission spectrum closely together with the literature information,
Where these intensity ratios are high, the wavelength 742.
364 nm, 744.230 nm, 746.831 nm
The emission spectral lines of 742.364 nm and 744.230 nm overlap without being able to completely separate the spectrum lines because of the problem of the performance of the spectroscope used for the measurement. It was found that two peaks were observed as the intensity ratio together with the emission spectrum line.

From the above findings, even if a measurement device having a relatively low resolution (about 2 nm in full width at half maximum in the case of the above experiment) is used, the emission spectrum line of nitrogen is 742.364n.
It was found that light of three wavelengths of m, 744.230 nm and 746.831 nm could be used.

Next, the above-mentioned 742.364 nm and 744.230 were measured using a measuring device having a relatively low resolution.
The method of detecting the presence or absence of blowholes and pit defects using the emission spectral line intensities of nitrogen having three wavelengths of nm and 746.831 nm was further examined by experiments. Therefore, first, the problem of the method for determining the presence or absence of a defect using the plasma plume or the intensity of the emission spectral line of nitrogen described in Japanese Patent Application Laid-Open No. 11-188489 was examined.

FIG. 4 shows that the back surface of a cold rolled steel sheet having a thickness of 1.4 mm was Ar at a processing point output of 4 kW and a welding speed of 2 m / min.
Shielded and butt welded, A
FIG. 9 compares emission spectra of laser-induced plasma generated on the back surface of a steel plate when butt welding is performed with shielding with r-25% N ₂ . At this time, the shielding gas on the surface of the steel sheet is both Ar.

When an X-ray transmission test of the weld bead was performed under the above two types of shield conditions, the back surface of the steel plate was Ar
The blowholes in the weld bead when shield was not observed, when the steel sheet back surface and Ar-25% N ₂ shield, blowholes were observed in the weld bead.

As shown in FIG. 4, the intensity of the plasma light is lower as a whole when the atmosphere of the back surface of the steel sheet is Ar-25% N _{2 than in} the case where the atmosphere of the back surface of the steel sheet is pure Ar. 7 above
42.364 nm, 744.230 nm, 746.83
Looking at the spectral intensities corresponding to the emission spectral lines of nitrogen having three wavelengths of 1 nm, these spectral intensities should be high as a result of nitrogen being mixed into the plasma.
Although the intensity of light at 6.831 nm is not so reduced, the other two wavelengths, 742.364 nm and 744.230, are used.
The intensity of the light of nm has been reduced.

Therefore, it was determined that it was impossible to detect the contamination of the shielding gas with nitrogen by increasing the intensity of the emission spectrum line of nitrogen.

As a result of the investigation by the inventors, it was found that the emission spectrum lines of Fe (744.098 nm, 741.867 nm, 740.
169 nm) and the resolution of the measurement system is low (in this case, about 2 nm in full width at half maximum), the adjacent F
It was found that the emission spectrum line of e could not be completely separated and was measured as noise.

Since N ₂ and O ₂ are harder to be turned into plasma than Ar, the shield becomes defective and entrainment of nitrogen or oxygen lowers the purity of Ar, which lowers the plasma temperature and is the main element constituting the plasma. The emission of iron is weakened, and the emission of plasma is weakened as a whole.

When the atmosphere of the back surface of the steel sheet is Ar-25% N ₂ , the intensity of light having a wavelength corresponding to the emission spectrum line of nitrogen is lower than that in the case of pure Ar because nitrogen in the plasma is low. This is considered to be the result of the fact that the emission of iron, which is measured as noise, became weaker than the emission of nitrogen due to contamination.

FIG. 5 shows that when a cold-rolled steel sheet having a thickness of 1.4 mm was butted by laser welding at a processing point output of 4.5 kW and a welding speed of 4.5 m / min, He was used as a shielding gas on the steel sheet surface. FIG. 6 compares the emission spectra of laser-induced plasma generated on the steel sheet surface in the case of using He + 50% N ₂ . In addition, although Ar was used for the shielding gas on the back surface of the steel plate, no emission spectrum line of Ar was confirmed in the plasma light on the front surface of the steel plate.

An X-ray transmission test of the weld bead was carried out for the case where He was used as the shielding gas on the surface of the steel sheet and the case where He + 50% N ₂ was used. Numerous blowholes were observed in the bead.

In the measurement results shown in FIG.
When% N ₂ is used as the shielding gas, the spectral intensity at the wavelength corresponding to the emission spectrum line of nitrogen is higher than when He is used as the shielding gas.

However, even when He was originally used as the shielding gas, the spectral intensity of the wavelength corresponding to the emission spectrum line of nitrogen was smaller than that of the adjacent Fe (744.098 nm,
(E.g., 741.867 nm, 740.169 nm) is a significant value because it was measured as noise without being able to fully separate the spectrum, and the emission of nitrogen due to the fact that 50% of nitrogen was mixed into the shielding gas. The increase in the spectral intensity corresponding to the spectral line is reduced.

As shown in FIG. 5, the plasma light intensity when welding is performed over the entire wavelength range when He + 50% N ₂ is used as a shielding gas, and is higher than the plasma light intensity when welding is performed using He as a shielding gas. ing. This is because He is less likely to be turned into plasma than Fe or N ₂ , so He has the effect of cooling the laser-induced plasma in which Fe is the main constituent element.
This is because the purity is reduced and the cooling effect of the plasma is weakened, and as a result, the temperature of the plasma is increased.

Such a phenomenon is described in JP-A-11-18848.
It is pointed out in Japanese Patent Publication No. 9 and the method for detecting blowholes and pit defects disclosed in this publication utilizes this phenomenon. However, according to the experiments performed by the inventors, such an increase in the overall light emission intensity is a phenomenon that occurs even when the flow rate of He is slightly reduced without involving nitrogen in the shielding gas, and the occurrence of blowholes and pit defects occurs. We do not consider it to be the optimal method for detecting the presence or absence.

FIG. 6 shows that when a 1.4 mm-thick cold-rolled steel plate was butt-laser-welded at a working point output of 4 kW and a welding speed of 4 m / min, nitrogen was mixed with Ar at various ratios as a shielding gas on the back surface of the steel plate. In this case, the results of measuring the intensity of light corresponding to the emission spectrum line specific to nitrogen in the light emitted from the laser-induced plasma generated on the back surface of the steel sheet are shown.

In this experiment, the wavelength resolution was 2 at full width at half maximum.
A case where a light of 746.83 nm is measured using a spectrometer of nm (reference signal 1) and a case where a light of 747 nm is measured using a spectroscope having a resolution of about 6 nm (reference signal 2) were performed.

In FIG. 6, a relative intensity obtained as a ratio with respect to the light intensity when the shielding gas is 100% Ar is used as the light intensity. Also shown are the results of the transmission test.

As shown in FIG. 6, the measured light intensity is
The nitrogen concentration in the shielding gas on the back surface of the steel sheet does not change in the range of 0 to 25% (reference signal 1) or gradually decreases (reference signal 2), and rapidly decreases in the range of 25 to 35%. After that, it became almost constant. on the other hand,
Weld defects include nitrogen concentration in the range of 5-30 and 95-100%
Occurred in the range. In these two regions of the nitrogen concentration where welding defects occurred, the measured light intensity was 1:10
There is a difference of about 0, and it is understood that there is no correlation between the light intensity and the presence or absence of a welding defect.

As described above, when nitrogen, which is harder to be turned into plasma than Ar, is mixed into the shielding gas, the plasma temperature is lowered, and the emission of iron, which is a main element constituting the plasma, is weakened. As shown in FIG. 6, in the range of nitrogen concentration of 95 to 100% where a welding defect occurs, the measured value of the light intensity (reference signal 1 and reference signal 2) corresponding to the emission spectrum line of nitrogen is nitrogen. The reason why the concentration is about the same as the range of 35 to 95% is that the measured value is significantly affected by the emission of iron which is measured as noise more than the intensity of the emission spectrum line of nitrogen. is there.

From the findings shown in FIGS. 4 to 6 and the like, when using a measuring device having a relatively low resolution, the inventors measure only the emission spectral line intensity of nitrogen to reduce the blowhole or the like in the laser weld. Judging that it is difficult to detect the presence or absence of pit defects, it is not affected by the emission spectrum of iron adjacent to the emission spectrum of nitrogen, and has a good correlation with the presence or absence of blowholes and pit defects in the laser weld. The method of processing the signal indicating

As a result, the intensity of the predetermined light including the emission spectrum line of nitrogen is used as a reference signal, and the intensity of the predetermined light including the emission spectrum line of iron without emission of Ar is measured for reference. If the intensity ratio obtained by dividing the former (reference signal) by the latter (reference signal) is used as the determination signal, it was found that this signal shows a good correlation with the presence or absence of a defect.

Hereinafter, a specific description will be given.

It suffices that the effect of the emission spectrum of the adjacent iron, which has been measured simultaneously, can be removed from the measured light intensities, but it is impossible to measure only the emission spectrum of nitrogen. Similarly, if only the emission spectrum of iron is measured, the emission spectrum of adjacent nitrogen is measured at the same time.

However, as a result of the experiments by the inventors, the intensity ratio between the emission spectrum lines of the individual irons did not change much depending on the welding conditions, and the intensity of the emission spectrum line of the iron was separated from the emission spectrum line of nitrogen. It was found that the intensity of the emission spectrum line of iron adjacent to the emission spectrum line of nitrogen can be estimated by using.

Therefore, the intensity of the predetermined light including the emission spectrum line of iron is used as the reference signal together with the reference signal which is the intensity of the predetermined light including the emission spectrum line of nitrogen. However, when the emission spectrum line of Ar is included in the measurement wavelength region of the reference signal, a determination signal showing a good correlation with the presence or absence of blow holes or pits cannot be obtained. This is Ar
This is because, while the intensity of the emission spectrum line of (1) significantly changes depending on the welding conditions, in the measurement of predetermined light including the emission spectrum line of nitrogen, the emission spectrum line of Ar is not directly measured as noise.

The method for determining the measurement wavelength of such a reference signal is as follows.

That is, spectroscopic measurement of laser-induced plasma light when a steel sheet is welded with shielding with He gas, and A
A laser-induced plasma is formed in the r gas and spectroscopic measurement of the light emitted by the Ar plasma is performed, and among the emission spectrum lines of iron, those that can be sufficiently distinguished from the emission spectrum lines of Ar are selected.

Here, in normal laser welding, He does not turn into plasma and does not emit clear light, so that the laser-induced plasma emission in laser welding of a steel plate with a He shield is due to iron.

FIG. 8 shows a wavelength range from 300 nm to 600 n.
It shows the spectroscopic measurement results in the range of m. The black one is the emission spectrum of iron, and the white one is the emission spectrum of Ar.

FIG. 9 shows a wavelength range from 600 nm to 800 n.
The emission spectrum of iron in the range of m (black fill)
And Ar emission spectra (shown in white).

FIGS. 8 and 9 show that the wavelength range usable as a reference signal is, for example, 300 to 340 nm, 520 to 520 nm.
570 nm and 620 nm to 690 nm. Also, the emission spectrum lines of iron 718.734 nm and 72
710 to 730 nm including 0.741 nm, and 780 to 790 n including 783.22 nm of iron emission spectrum line
It can be seen that a wavelength range such as m can be used as a reference signal.

Further, when Ar is not used as the shielding gas, the entire emission intensity, for example, light in a wavelength range of 300 nm to 800 nm can be measured and used as a reference signal.

Using the reference signal determined as described above, a blowhole or a pit is obtained from a reference signal that includes the emission spectrum of nitrogen and the intensity of light that has not completely separated the emission spectrum of an adjacent iron. In order to generate a determination signal that can detect the presence or absence of a defect, two types of signal processing methods are conceivable.

One is a method in which a value obtained by dividing a reference signal by a reference signal is used as a determination signal. Another method is to generate a determination signal by subtracting a constant coefficient times the reference signal from the reference signal. Is the way. According to the study results of the present inventors, the former signal processing method, that is, the determination signal obtained as a ratio of the reference signal to the determination signal is more constant coefficient of the reference signal than the latter signal processing method, that is, the reference signal. A good correlation with blowholes and pit defects was obtained over a wide range of welding conditions from the judgment signal obtained by subtracting.

Table 1 shows that the processing point output was 4 kW and the welding speed was 4
The figure shows the results of spectroscopic measurement of laser-induced plasma generated on the rear surface of a cold-rolled steel plate having a m / min of 6 m / min and a plate thickness of 1.4 mm when butt laser welding was performed. Here, Ar is used as the shielding gas on the steel sheet surface,
As a shielding gas on the back surface of the steel sheet, an Ar + N ₂ mixed gas is used in various mixing ratios in order to reproduce a shielding failure. The focal length of the focusing optical system is 10 inches, and the beam diameter (86% intensity) at the focusing point is about 0.45.
mm.

Table 1 shows the case where a light of 747 nm was measured using a spectroscope having a wavelength resolution of 2 nm at full width at half maximum (reference signal 1) as a reference signal indicating the intensity of light including the emission spectrum line of nitrogen. The case of measuring 747 nm light using a spectroscope having a resolution of about 6 nm (reference signal 2) was studied.

As the reference signal, light containing an emission spectrum line 783.2 nm of Fe, a center wavelength 783 nm,
The values measured using a spectroscope having a resolution of about 6 nm are shown.

[0072]

[Table 1]

The judgment signal 1 and the judgment signal 2 are respectively formed by dividing the reference signal 1 and the reference signal 2 by the reference signal.

FIG. 7 shows the judgment signals 1 and 2
6 shows how the intensity of the reference signal has changed with respect to the nitrogen concentration of the Ar + N ₂ mixed gas on the back surface of the steel sheet, and the reference signal 1 and the reference signal 2 in FIG. Equivalent to.

In FIG. 7, as in the case of FIG. 6, the relative intensity which is the ratio to the determination signal intensity when the shielding gas is 100% Ar is used as the determination signal intensity. Is also shown with the result of the investigation by the X-ray transmission test.

As can be seen from FIG. 7, the judgment signal is high in the range of 5 to 30% and 95 to 100% of the nitrogen concentration where the welding defect occurs, while the judgment signal is high in the other nitrogen concentrations. It can be seen that the determination signal and the presence or absence of the occurrence of a welding defect show a good correlation.

The reference signal shown in FIG.
While the determination signal shown in FIG. 7 showed a tendency to hardly change or decrease in the range of 30%, whereas the determination signal shown in FIG. It depends on the reason. That is, when N _{2, which} is harder to be turned into plasma than Ar, is mixed into the Ar shielding gas, the temperature of the plasma gradually decreases, but does not significantly change when the nitrogen concentration is less than 30%. As a result, the rate of atomic dissociation of nitrogen molecules in the plasma is kept almost constant, and as the nitrogen concentration increases, the number of nitrogen atoms that dissociate atomically and emit light increases, corresponding to the emission spectrum of nitrogen. The intensity of the emitted light increases. However, since the reference signal has simultaneously measured the emission spectrum lines of the adjacent iron, it hardly changes or tends to decrease as the plasma temperature decreases. On the other hand, the determination signal has a tendency to increase without being affected by the gradually decreasing plasma temperature because the influence of the emission spectrum line of the adjacent iron is removed. The nitrogen concentration is 5 to 30%
In the range, since a large amount of atomic nitrogen is generated in the laser-induced plasma, a large amount of nitrogen dissolves in molten steel and forms blowholes and pits upon solidification.

When the nitrogen concentration is in the range of 35 to 95%,
Although both the reference signal and the judgment signal decrease, the temperature of the plasma is greatly reduced in this nitrogen concentration range as compared with the range of the nitrogen concentration of 0 to 30%, and the dissociation ratio of nitrogen molecules is small. Despite the large number of nitrogen molecules, the number of nitrogen atoms contributing to emission is
The number is smaller than the range of 0%, and as a result, both the reference signal and the determination signal become smaller. Further, since the number of nitrogen atoms is small, the amount of nitrogen dissolved in the molten steel is small, and no welding defects occur.

In the range of nitrogen concentration of 95 to 100%, as compared with the nitrogen concentration of 35 to 95%, the plasma temperature and the dissociation ratio of nitrogen molecules to atoms are almost the same, but the original nitrogen molecular weight is large. As a result, the number of nitrogen atoms in the plasma increases, and as a result, the light intensity corresponding to the emission spectrum line of nitrogen increases. Since the reference signal simultaneously measures the emission spectrum of the adjacent iron, it hardly changes, mainly due to the low plasma temperature, whereas the judgment signal removes the influence of the emission spectrum of the adjacent iron. Therefore, it becomes higher in response to the increase in the amount of nitrogen atoms in the plasma. At this time, since a large amount of atomic nitrogen is generated in the laser-induced plasma, the amount of nitrogen dissolved in the molten steel increases, and blow holes and pits are formed during solidification.

When the determination signal 2 shown in Table 1 is used, from the relationship between the intensity of the determination signal 2 and the presence / absence of a defect in the laser weld, when the welding speed is 4 m / min, the determination value of the presence / absence of a weld defect is 2 And welding speed of 6 m / mi
In the case of n, it can be seen that if 3.9 is used as the determination value, the presence or absence of a blowhole or a pit defect can be detected.

The judgment value is, for example, the lowest value of the judgment signal when there is a blowhole or a pit defect, and the judgment value when there is no blowhole or pit defect. The value can be selected by setting the value to an intermediate value from the maximum value of the signal.

Further, when the welding speed changes, the judgment value of the occurrence of blowholes and pit defects changes. Therefore, it is necessary to change the judgment value in accordance with the setting of the welding speed. By using the value divided by the speed as a new determination signal (determination signal 3), it is possible to determine a welding defect with a constant determination value (0.65 in Table 1) regardless of the welding speed.

As described above, in the present invention, when laser welding is performed by shielding the welded portion on the front surface or the back surface of the material to be joined with He gas, the welded portion is shielded by He gas near the laser processing point of the welded portion. Of the light emitted from the generated laser-induced plasma, the intensity of the predetermined light in the wavelength range of 741 to 748 nm including the emission spectrum line of nitrogen and the intensity of the predetermined light including the emission spectrum line of iron are measured. In the case of using a measuring device having a relatively low resolution (about 6 nm at full width at half maximum) by comparing this determination signal with a determination value determined in advance by experiment, using the intensity ratio of light obtained by dividing the above by the latter as a determination signal, Also, it is possible to accurately detect the occurrence of welding defects.

Further, when laser welding is performed by shielding the welded portion on the front or back surface of the material to be joined with Ar gas or a mixed gas containing Ar, when the welded portion is shielded with a shield gas containing Ar gas. Among the light emitted by the laser-induced plasma generated in the vicinity of the laser processing point, the light includes a predetermined light intensity in a wavelength range of 741 to 748 nm including a nitrogen emission spectrum line and an emission spectrum line of iron, and A
r by measuring the intensity of predetermined light that does not include the emission spectrum line of r, and comparing the intensity of light obtained by dividing the former by the latter as a determination signal, and comparing the determination signal with a determination value determined in advance by experiment. Even when using a measuring device having a relatively low resolution (about 6 nm in full width at half maximum), it is possible to accurately detect the presence or absence of a welding defect.

Further, in the present invention, the ratio (relative intensity) of the reference signal obtained by measuring the intensity of light including the emission spectrum line of nitrogen to the reference signal obtained by measuring the intensity of light including the emission spectrum line of iron as described above. ) Was used as the determination signal, so that the light receiving portions of the respective lights were matched as shown in FIG. 1 and then spectrally separated. Even if the amount of light guided decreases,
A good correlation is maintained with time between the determination signal intensity and the presence or absence of a welding defect, which is very useful from the aspect of industrial use.

As the predetermined light whose intensity is measured as the reference signal, a wavelength range including the emission spectrum line of nitrogen is selected. As described above, the wavelength of the nitrogen emission spectrum line that can be measured is 742.364 nm and 744.23.
nm and 746.831 nm, and when the wavelength resolution can be secured to about 2 nm or less, 744.23 n
Measure the light at m or 746.83 nm. Also, 6
If only a resolution of about nm can be secured, 744.
It is assumed that light of 741 nm or more and 748 nm is measured around the light of 23 nm.

As the predetermined light including an emission spectrum line of iron as a reference signal for forming a determination signal, for example, a wavelength range of 300 to 340 nm, 520 to 570 n
m, light of 620 to 690 nm. In addition, an emission spectrum line of iron 718.834 nm,
m in the wavelength range of 710 to 730 nm, and 783.
Light having a wavelength range of 780 to 790 nm including 22 nm can be used as a reference signal.

Further, when Ar is not used as the shielding gas, the entire luminous intensity, for example, light having a wavelength range of 300 nm or more and 800 nm or less can be measured and used as a reference signal.

[0089]

FIG. 1 is a schematic diagram showing an example of an embodiment according to the present invention.

In FIG. 1, a material 1 to be welded, which is conveyed at a predetermined speed in the direction of the arrow in the figure, is welded through by a laser beam 2 irradiated from above.
Laser-induced plasma 3 is generated on the front and back surfaces of the weld. The front side of the weld is He or Ar gas,
Alternatively, it is shielded by a mixed gas of He and Ar, and the back side is shielded by Ar gas.

Here, monitoring of the back surface of the steel sheet will be described as an example, but the same applies to the front side. The same applies to the case where the shield gas on the back side of the steel plate becomes He.

The light of the plasma 3 is transmitted to the optical splitter 6 by an optical fiber 4 having one opening face directed toward the plasma. Here, in order to increase the light receiving area,
The plasma light is collected on the opening surface of the optical fiber through the light receiving lens 5.

The plasma light split by the light splitter 6 is converted into a photodetector, for example, by a spectroscope that transmits only specific light, for example, filters 7A and 7B that transmit only predetermined light. Photo diodes 8A and 8B
Thus, the light is sequentially converted into an electric signal corresponding to the light intensity.

Here, the light detectors 7A and 7B
Of the laser-induced plasma emission, the light intensity in the wavelength range including the emission spectrum line of nitrogen and the light intensity in the wavelength range including the emission spectrum line of iron are the reference signal and the reference signal, respectively.

The obtained electric signal is taken into the computer 10 via the voltage amplifiers 9A and 9B with the noise cut filter and the AD converter. The noise cut filter is necessary to kill a sudden and meaningless change in the light emission intensity, and a low-pass filter that passes a signal of 10 to 1 Hz or less is preferable. The computer calculates the ratio of the reference signal to the reference signal, calculates a determination signal obtained by dividing the ratio by the welding speed, and compares the calculated signal with a determination value that is a determination criterion for determining whether or not a welding defect has occurred. With C
The judgment result and the judgment signal are sequentially displayed on a display device such as an RT.

In addition to determining whether or not a welding defect has occurred,
An alarm may be issued by a buzzer or the like, or an abnormal portion may be marked on the front and back of the steel plate. Further, the output signal to the monitor may be recorded and stored in an appropriate recording device such as an optical disk.

In order to demonstrate the effect of the present invention, the emission intensity of laser-induced plasma generated on the back surface of the laser welded portion during actual laser welding was measured using a measuring device having the configuration shown in FIG.

The processing point output of the laser is 4.5 kW,
A 0.7 mm thick cold-rolled steel plate was butted and welded. The welding speed is 5 m / min. As shown in the table during welding, the relationship between the determination signal and the presence or absence of blow holes and pit defects was examined by intentionally deteriorating the butting condition and the Ar purity of the shielding gas on the back surface.

FIG. 10 shows the measured judgment signal. Here, a light having a center wavelength of 745 nm is measured using a spectroscope having a wavelength resolution of 4 nm as a reference signal including a nitrogen emission spectrum line, and a light having a center wavelength of 785 nm is used as a reference light using a spectrometer having a wavelength resolution of 785 nm. And a 2 Hz low-pass filter is used.

As is evident from FIG. 10, in the sensitivity of the measuring system in the present embodiment, it is possible to determine whether or not a defect has occurred in the laser welded portion by using 2.2 as a quality determination criterion.

[0101]

According to the present invention, a measuring instrument having a relatively low resolution is used at the time of laser welding, and even if the signal intensity decreases due to the contamination of the light receiving surface of the measuring instrument over time, blowholes and pit defects can be prevented. It is possible to provide a method for detecting a welding defect or a method for monitoring the quality of a welded portion, which can obtain a good correlation stably with the presence or absence of occurrence.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing an example of an embodiment according to the present invention.

FIG. 2 is a schematic diagram showing an example of spectral measurement of laser-induced plasma emission on a steel sheet surface in nitrogen shield welding and He shield welding, and an example of a spectral emission intensity ratio in a wavelength range of 300 nm to 600 nm.

FIG. 3 is a schematic diagram showing an example of spectral measurement of light emission of laser-induced plasma on a steel sheet surface in nitrogen shield welding and He shield welding, and an example of a spectral emission intensity ratio in a wavelength range of 600 nm to 800 nm.

FIG. 4 shows a case where shielding is performed with a mixed gas of Ar and nitrogen;
It is a schematic diagram which shows an example of the spectrum measurement result of the laser induced plasma of the steel plate surface at the time of laser welding with an Ar shield.

FIG. 5 is a schematic diagram showing an example of spectral measurement results of laser-induced plasma on the back surface of a steel plate when the back surface of the steel plate is shielded with a mixed gas of He and nitrogen and when the back surface of the steel plate is shielded with He and laser-welded. .

FIG. 6 is a schematic diagram showing an example of a temporal change of a determination signal when a laser-induced plasma generated on a back surface of a steel sheet is monitored by a method according to the present invention.

FIG. 7 is a diagram showing how the intensities of the determination signal 1 and the determination signal 2 change with respect to the nitrogen concentration of the Ar + N ₂ mixed gas on the back surface of the steel sheet.

FIG. 8 is a diagram illustrating spectroscopic measurement results in a wavelength range of 300 nm to 600 nm. .

FIG. 9 is a diagram showing an emission spectrum of iron (filled black) and an emission spectrum of Ar (shown in white) in a wavelength range of 600 nm to 800 nm.

FIG. 10 is a diagram showing a determination signal obtained by measuring the emission intensity of laser-induced plasma generated on the back surface of a laser weld during laser welding.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 welding material 2 laser beam 3 laser induced plasma 4 optical fiber 5 light receiving lens 6 light splitter 7A, 7B spectroscope (filter) 8A, 8B photodetector (photodiode) 9A, 9B voltage amplifier with noise cut filter 10 AD Measurement computer with conversion board

Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat II (Reference) // B23K 31/00 B23K 31/00 L (72) Inventor Takashi Tanaka 20-1 Shintomi, Futtsu Nippon Steel Corporation Within the Technology Development Headquarters (72) Inventor Masahiro Ohara No. 1, Nishinoshu, Oita City Nippon Steel Corporation Oita Works F-term (Reference) 2G043 AA03 BA02 BA07 CA02 CA05 EA10 FA05 HA05 HA09 JA02 KA01 KA05 KA09 LA01 NA01 2G051 AA90 AB02 AB20 AC30 BA10 CC07 CC17 EA11 2G055 AA08 BA07 FA02 4E068 BA00 CA17 CC01 CJ01

Claims (2)

[Claims] When performing laser welding by shielding one of a front surface and a back surface of a material to be joined with He gas, a portion near a laser processing point of the weld portion shielded with He gas is provided. Of the light emitted by the laser-induced plasma, the wavelength range 741 to
The intensity of the predetermined light at 748 nm and the intensity of the predetermined light including the emission spectrum line of iron were measured, and the intensity ratio of light obtained by dividing the former by the latter was used as a judgment signal. A method for detecting a welding defect in laser welding, comprising determining whether or not a welding defect has occurred by comparing a determination value. 2. A laser welding method in which one of a front surface and a back surface of a material to be joined is shielded with Ar gas or a mixed gas containing Ar to perform laser welding.
Among the light emitted by the laser-induced plasma generated near the laser processing point of the welded portion shielded by the shielding gas containing gas, the intensity of the predetermined light in the wavelength range 741 to 748 nm including the emission spectrum line of nitrogen and The intensity of predetermined light that includes the emission spectrum line of iron and does not include the emission spectrum line of Ar is measured, and the intensity ratio of light obtained by dividing the former by the latter is used as a determination signal. A method for detecting a welding defect in laser welding, comprising determining whether or not a welding defect has occurred by comparing the detected value with a determined value.