CN107160048A - A kind of inline diagnosis method of the laser welding defect based on spectral information - Google Patents

Abstract

The present invention proposes an online diagnostic method for laser welding defects based on spectral information, which includes the following steps: adjusting the position of the optical fiber probe; collecting photoplasma information to determine characteristic spectral lines; temperature to obtain the electronic temperature time-domain diagram; according to the material and size of the workpiece to be welded, use different welding parameters for welding, and select the parameter that minimizes the standard deviation of the electronic temperature time-domain change as the optimal welding parameter; under the above parameters, N Weld the pieces to be welded to obtain N electronic temperature time-domain diagrams, and then obtain the SPC control diagram; adjust the welding parameters according to the actual situation, and weld the parts to be welded to obtain the electronic temperature time-domain diagrams and record them as test time-domain diagrams; The domain diagram is drawn into the SPC control diagram in real time, and whether the welding defects exist is judged by judging whether each point in the test time domain diagram exceeds the upper and lower limits of the SPC control diagram. The above method can quickly and accurately judge whether there is a defect in the laser welding process.

Description

An Online Diagnosis Method of Laser Welding Defects Based on Spectral Information

technical field

The invention relates to the field of online diagnosis of laser welding quality, in particular to an online diagnosis method of laser welding defects based on spectral information.

Background technique

Laser welding uses a high-energy-density laser as a heat source to melt metals to form welded joints. It is an efficient and precise welding method. In recent years, with the rapid development of aerospace, automotive, microelectronics, light industry, medical and other industries, the shape of product parts has become more and more complex, and the requirements for the surface shape and deformation of welded components have become higher and higher. Traditional welding The method is difficult to meet its needs, so laser welding has received extensive attention. In order to improve the efficiency of laser welding and reduce the defects of welded joints, the online diagnosis method of welding defects is indispensable. The traditional offline detection method is time-consuming and laborious and requires professional equipment and related operators.

Contents of the invention

In order to solve the problems existing in the prior art, the present invention proposes an online diagnosis method of laser welding defects based on spectral information, so as to quickly and accurately determine whether there are welding defects in the laser welding process.

In order to achieve the above object, the present invention proposes an online diagnostic method for laser welding defects based on spectral information, wherein the online diagnostic device includes an optical fiber probe, and the optical fiber probe is connected to a fiber optic spectrometer via an optical fiber, and the fiber optic spectrometer is also connected to a computer The optical fiber probe is fixed on the laser head via a mechanical universal rod, and the method includes the following steps:

Step 1, adjusting the position of the optical fiber probe;

Step 2. When the laser welding process starts, the optical signal in the photoplasma generated during the laser welding process is collected by the optical fiber probe, and the spectral information formed in the optical fiber spectrometer is transmitted to the computer;

Step 3. Observing the fluctuation of the relative intensity of each spectral line in the frequency domain distribution diagram of the above-mentioned spectral information in real time from the display interface of the computer, so as to determine the characteristic spectral line as the analysis object;

Step 4. At any point in the collection process, calculate the electron temperature of the photoplasma by using the relative intensity of the selected characteristic spectral lines, so as to obtain the change curve of the electron temperature with time;

Step 5. According to the material and size of the parts to be welded, different welding parameters are used for welding, and the parameter that minimizes the standard deviation of the electronic temperature time domain change is selected as the optimal welding parameter;

Step 6. Welding N workpieces to be welded under optimal welding parameters, and recording the electronic temperature and time-domain diagram of the photoinduced plasma generated by each workpiece to be welded during the laser welding process through a computer;

Step 7, obtain the SPC control diagram according to the time-domain diagram of each electronic temperature in step 6;

Step 8. Adjust the welding parameters according to the actual situation, and perform laser welding on the parts to be welded;

Step 9, when the laser welding process starts, collect the optical signal in the photoplasma produced in the laser welding process through the optical fiber probe, transmit the spectral information formed to the computer through the optical fiber spectrometer, and select the characteristic spectral line in step 3 as Characteristic spectral lines, calculate the electron temperature of the photoinduced plasma through the selected characteristic spectral lines, and obtain the time-domain diagram of the electron temperature, which is called the test time-domain diagram;

Step 10, during the laser welding process, draw the test time domain diagram into the SPC control diagram in real time by the computer, and judge whether each point in the test time domain diagram exceeds the upper and lower limits of the SPC control diagram;

Step 11, if no, there is no welding defect;

Step 12, if yes, judge whether there are continuous a points in the continuous A points that exceed the upper and lower limits of the SPC control chart, wherein the values of A and a depend on the sensitivity of defect judgment;

Step 13, if no, there is no welding defect;

Step 14, if yes, there is a welding defect.

Preferably, the optical fiber probe is equipped with a COL-UV/VIS collimating lens.

Preferably, the resolution of the fiber optic spectrometer is not lower than 0.1 nm.

Preferably, in the step 1, the optical fiber probe is set at a position 15-20 cm away from the incident point of the laser beam on the workpiece to be welded, and at a position 7.5-10 cm higher than the upper surface of the workpiece to be welded, while ensuring that the The optical fiber probe mentioned above collects the photoinduced plasma generated in the laser welding process.

Preferably, in the step 4, the electron temperature of the photoplasma is calculated using the Boltzmann diagram method or the two-spectrum method.

Preferably, in said step 7, obtaining the SPC control chart includes the following steps:

Step 71. Dividing the data of the N electronic temperature time-domain diagrams in the step 6 into several subgroups, and the data of different welding passes at the same time is a subgroup;

Step 72, calculating the arithmetic mean value for the electronic temperature data in each subgroup, and using the arithmetic mean value as the value of each subgroup;

Step 73, calculating the arithmetic mean value for the values of each subgroup, which is the center line in the SPC control chart;

Step 74. Calculate the standard deviation of the electronic temperature data in each subgroup, which is called the subgroup standard deviation;

Step 75, calculating the arithmetic mean value for all subgroup standard deviations to obtain the overall standard deviation;

Step 76, by adding or subtracting the overall standard deviation of the corresponding multiple on the value of the central line to obtain the value of the upper control limit or the lower control limit in the SPC control chart, wherein the multiple checks the coefficient of the SPC control chart according to the number of samples in the subgroup Table available.

The beneficial effect of this solution of the present invention is that the above-mentioned online diagnosis method of laser welding defects based on spectral information establishes a specific quantitative standard to judge the existence of defects in the welding process. It can quickly and accurately judge whether there are defects in the laser welding process, so as to apply to the online diagnosis of defects in industrial mass production, and lay the foundation for the judgment and classification of defect types.

Description of drawings

Fig. 1 shows a schematic diagram of the principle of an online diagnostic device and a schematic diagram of a laser welding system involved in the present invention.

Fig. 2 shows a flowchart of the steps of the online diagnosis method involved in the present invention.

Fig. 3 (a) to Fig. 3 (d) respectively show the schematic diagram of the welded joint in the first embodiment of the present invention, the relative intensity of photoinduced plasma and time, the corresponding space diagram of wavelength, the temperature of photoinduced plasma electrons with Time domain diagram of time change, SPC control diagram for judging welding quality.

Figure 4(a) to Figure 4(d) respectively show the distribution of the test time domain diagram in the SPC control diagram under the conditions of ideal weld, misalignment, weld penetration and butt gap variation.

Fig. 5 shows a schematic diagram of a welded joint in a second embodiment of the present invention.

Fig. 6(a) and Fig. 6(b) respectively show the distribution of the test time-domain diagram in the SPC control diagram under the state of weld penetration and weld bead separation in the second embodiment of the present invention.

Reference signs: 1-fiber optic spectrometer, 2-computer, 3-fiber optic probe, 4-mechanical universal rod, 5-workbench, 6-piece to be welded, 7-laser beam, 8-Ar shielding gas, 9-welding seam, 10-photoplasma, 11-optical fiber, 12-laser head, A-misalignment area, B-weld-through area caused by the thinning of the center of the weldment, C-unwelded area caused by the change of butt gap, D - Spacer, E - galvanized steel sheet, F - area of weld penetration caused by slowing down of welding speed, G - area of weld bead separation.

detailed description

The specific embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

As shown in Figure 1, the online diagnostic device involved in the present invention includes a fiber optic probe 3, which is connected to a fiber optic spectrometer 1 via an optical fiber 11, and the fiber optic spectrometer 1 is also connected to a computer 2; wherein the fiber optic The probe 3 is fixed on the laser head 12 via a mechanical universal rod 4, and the angle and height of the fiber optic probe 3 can be adjusted according to requirements so as to collect information in alignment with the photoplasma generated during the laser welding process.

The laser welding system belongs to the prior art, and only a schematic diagram is given here for a brief description. As shown in FIG. Both sides are clamped on the workbench 5 by clamps (not shown in the figure), and the laser beam 7 is irradiated vertically on the surface of the workpiece 6 to be welded, and the base material metal is melted to form a joint to realize the laser welding process. During the welding process, the Ar The shielding gas 8 is delivered at a certain flow rate to prevent the weld seam 9 and its adjacent areas from being oxidized during laser welding.

In this embodiment, the laser beam 7 is produced by a YLS-6000 fiber laser produced by IPG Photonics, Germany. The maximum output power of the laser is 6000W, the wavelength is 1060-1070nm, and the beam quality is BPP≥4.0. The welding process The motion execution system adopts the KRC-60HA six-axis linkage manipulator produced by KUKA Company, and the protective gas is laterally supplied during welding. The fiber optic spectrometer 1 adopts AvaSpec-ULS2048-8-USB2 multi-channel fiber optic spectrometer, the resolution of the fiber optic spectrometer is 0.052±0.001nm, and the measurable wavelength range is 200-1100nm. The fiber optic probe 3 is equipped with a COL-UV/VIS collimating lens.

The step process of the online diagnosis method of laser welding defects based on spectral information involved in the present invention is shown in Figure 2, and the method includes the following steps:

Step 1. Adjust the position of the optical fiber probe 3 . The optical fiber probe 3 is set at a position 15-20 cm away from the incident point of the laser beam 7 on the workpiece 6 to be welded, and at a position 7.5-10 cm higher than the upper surface of the workpiece 6 to be welded, while ensuring that the optical fiber probe 3 The location is capable of collecting the photoplasma 10 generated during the laser welding process; as shown in step S101 in FIG. 2 .

Step 2, when the laser welding process starts, the optical signal in the photoplasma 10 generated during the laser welding process is collected by the optical fiber probe 3, and the spectral information formed in the optical fiber spectrometer 1 is transmitted to the computer 2, as shown in Figure 2 Shown in step S102. In a specific collection process, the sampling frequency of the online diagnostic device is determined according to the material of the workpiece 6 to be welded and the welding speed of the material.

Step 3, observe in real time from the display interface of the computer 2 the fluctuation of the relative intensity of each spectral line in the frequency domain distribution diagram of the above-mentioned spectral information over time, so as to determine the characteristic spectral line as the analysis object, as shown in step S103 in Figure 2 Show. The selection of the characteristic spectral lines has the following requirements: sufficient resolution, high signal-to-noise ratio, and obvious features.

Step 4. At any point in the collection process, calculate the electron temperature of the photoplasma 10 using the relative intensity of the selected characteristic lines to obtain a time-varying curve of the electron temperature, as shown in step S104 in FIG. 2 .

In order to calculate the electron temperature of the photoplasma 10, the Boltzmann diagram method or the two-line method can be used. Among them, the Boltzmann diagram method is a method to calculate the electron temperature through multiple characteristic spectral lines of selected wavelengths, such as the formula , where I _mn is the relative intensity of the characteristic spectral line (assuming that the characteristic spectral line is produced by the transition from m energy level to n energy level), A _mn is the transition probability, g _mn is the statistical weight, and λ _mn is the characteristic spectral line Wavelength, Z is the partition function, h is Planck's constant, N is the total density of particles, c is the speed of light, k is the Boltzmann constant, E _m is the energy of energy level m (based on the ground state energy level being zero). It can be seen that the energy level E _m is taken as the abscissa, and As the ordinate, if each parameter of each selected characteristic spectral line is substituted into the above-mentioned abscissa and ordinate, the horizontal and vertical coordinates of the corresponding groups (that is, a characteristic spectral line corresponds to a group of horizontal and vertical coordinates) will be obtained. The square method fits these points into a straight line, and the slope of the fitted line is In this way, the electron temperature T can be calculated from the slope.

The two-spectrum method is a method of solving electron temperature with slightly lower precision but simple calculation, such as the formula , where I is the relative intensity of the characteristic spectral lines (where 1 represents the selected first characteristic spectral line, and 2 represents the selected second characteristic spectral line), A is the transition probability, g _m is the statistical weight, λ is the wavelength of the characteristic spectral line, k is the Boltzmann constant, and E _m is the energy of the energy level m (based on the ground state energy level being zero). It can be seen that only any two characteristic spectral lines of the same particle are required by the two-spectral method The ratio, energy range and some of its parameters can be used to obtain the electron temperature T.

Step 5. According to the material and size of the workpiece 6 to be welded, different welding parameters are used for welding, and the parameter that minimizes the standard deviation of the electronic temperature time domain change is selected as the optimal welding parameter, as shown in step S105 in FIG. 2 . The laser welding performed under the optimal welding parameters can form a good weld 9, so the laser welding under the optimal welding parameters is in a stable and controlled state.

Step 6. Weld the N pieces to be welded 6 under the optimal welding parameters, and record the electronic temperature time domain diagram of the photoinduced plasma 10 generated by each piece to be welded 6 in the laser welding process through the computer 2, as shown in Figure 2 Shown in step S106.

Step 7. Obtain the SPC control diagram according to the time-domain diagram of each electron temperature in step 6, as shown in step S107 in FIG. 2 .

The specific process of obtaining the SPC (Statistical Process Control, Statistical Process Control) control chart is as follows: Assuming that five parts to be welded 6 are welded under the optimal welding parameters, the data of the electronic temperature time-domain diagram of the five welds collected will be It is divided into several subgroups, and the data of different welding passes at the same time is listed as a subgroup, so that the data of each welding pass at different times can be divided into different subgroups, so that the data of the electronic temperature time domain map obtained by five weldings can be divided into several subgroups. For example, the first point collected in the first welding to the first point collected in the fifth welding are classified into a subgroup, and so on, in the case of five weldings, each subgroup has five points.

After that, calculate the arithmetic mean value of the electronic temperature data in each subgroup, use the arithmetic mean value as the value of each subgroup, and then calculate the arithmetic mean value of the values of each subgroup, and the arithmetic mean value is the SPC control chart The center line in the center line; calculate the standard deviation for the data in each subgroup, which is called the subgroup standard deviation, and then calculate the arithmetic mean for all the subgroup standard deviations to obtain the overall standard deviation. The value of the upper control limit or the lower control limit in the SPC control chart can be obtained by adding or subtracting a certain multiple of the overall standard deviation to the value of the center line, and the multiple is related to the size of the subgroup (the size of the subgroup is the weld within the subgroup The quantity of track, in the present embodiment, the size of subgroup is five), according to the number of samples in the subgroup (that is, the number of welding passes), check the SPC control chart coefficient table to get the multiple, when the size of the subgroup is adjusted , this multiple must be adjusted accordingly, and thus the three lines of the SPC control chart can be obtained.

Step 8. Adjust the welding parameters according to the actual situation, and perform laser welding on the workpiece 6 to be welded, as shown in step S108 in FIG. 2 .

Step 9, when the laser welding process starts, collect the optical signal in the photoplasma 10 produced in the laser welding process through the optical fiber probe 3, transmit the formed spectral information to the computer 2 through the optical fiber spectrometer 1, and select the optical signal in step 3 The characteristic spectral lines are used as the characteristic spectral lines, and the electron temperature of the photoplasma 10 is calculated through the selected characteristic spectral lines to obtain a time-domain diagram of the electron temperature, which is called a test time-domain diagram, as shown in step S109 in FIG. 2 . Wherein the sampling frequency of the online diagnostic device is consistent with the sampling frequency in step 2.

Step 10, in the process of laser welding, by computer 2, the test time domain diagram is drawn into the SPC control diagram in real time; whether each point in the test time domain diagram exceeds the upper and lower limits of the SPC control diagram, as shown in step S110 in Figure 2 shown.

Step 11. If not, there is no welding defect, as shown in step S111 in FIG. 2 .

Step 12. If yes, judge whether there are consecutive a points among the continuous A points that exceed the upper and lower limits of the SPC control chart, as shown in step S112 in FIG. 2 . Among them, the values of A and a are related to the sensitivity of defect judgment.

Step 13. If not, there is no welding defect, as shown in step S113 in FIG. 2 .

Step 14, if yes, there is a welding defect, as shown in step S114 in FIG. 2 . The position of the defect is the same as the position of the weld corresponding to the point crossing the limit.

Example 1

In this embodiment, the part 6 to be welded is made of 304 stainless steel, and the size of the part 6 to be welded is 100 × 50 × 3 mm. The form of the welded joint and the schematic diagram of the weld seam formed by laser welding are shown in Figure 3 (a) , Protected with Ar shielding gas 8 during the welding process.

Adjust the position of the fiber optic probe 3 so that during the laser welding process, the fiber optic probe 3 collects optical signals in the photoplasma with a wavelength of 200-1100 nm, and transmits the spectral information formed in the fiber optic spectrometer 1 to the computer 2 , by comparing the above spectral information, it is determined to use the 318-420nm spectral segment for analysis, so as to collect a spatial map reflecting the relationship between wavelength, time and intensity in the selected spectral segment, as shown in Figure 3(b). For the collected airspace map, based on the principles of high resolution, good profile, high sensitivity, moderate relative intensity, and close spectral lines, the characteristic spectral lines FeI375.8nm, FeI384.0nm, FeI385.9nm, FeI387.2nm, FeI387.8nm, FeI388.6nm, obtain the time domain diagram of the relative intensity of the spectrum changing with time during the welding process. According to the selected characteristic spectral lines, at any time of spectrum collection, the electron temperature is calculated by the Boltzmann diagram method, so as to establish the electron temperature change curve with time, as shown in Fig. 3(c).

The welding speed v=2m/min is fixed, and different welding parameters are used for welding. The time-domain diagram of the electronic temperature changing with time under different welding parameters is obtained by the above method, and the standard deviation of the electronic temperature changing with time is calculated. At this time, the welding parameters when the surface of the weld seam is well formed and the standard deviation of the time-domain change of the electron temperature is at the minimum value are considered as the optimal welding parameters under stable and controlled state. At this time, the laser power P=2600W, and the defocus The volume is +0mm, and the flow rate of Ar shielding gas is 20L/min. Weld the five parts 6 to be welded under the optimal welding parameters to ensure that there is no accidental disturbance during the welding process and the weld seam is well formed. The computer 2 records the photoplasma generated by each part 6 to be welded during the laser welding process. 10 electron temperature time-domain diagram, and according to each electron temperature time-domain diagram, obtain the SPC control diagram, as shown in Fig. 3(d).

After the SPC control chart is obtained, welding defects are intentionally created by changing the welding parameters or changing the clamping state of the workpiece 6 to be welded, so as to test the discrimination ability of the SPC control chart for welding defects.

As shown in Figure 4(a), the first weld was welded under the above-mentioned optimal welding parameters. It can be seen that its surface is well formed. It can be seen from the SPC control chart that the electron temperature during the welding process has no continuous point exceeding The upper and lower limits of the SPC control chart show that there are no welding defects, so the SPC control chart is consistent with the actual judgment of the actual welding forming.

Change the clamping state of the piece to be welded 6 so that the right end of the piece to be welded 6 is staggered to form a staggered side area A, and then laser welding is performed to simulate the situation that the wrong side occurs due to improper clamping in actual production, and the obtained weld and The corresponding SPC control chart is shown in Figure 4(b). It can be seen from the figure that the electronic temperature corresponding to the weld in the wrong side area A at the right end changes significantly, exceeding the upper and lower limits of the SPC control chart.

The center of the workpiece 6 to be welded is thinned, and laser welding is performed without changing the welding parameters to simulate the uneven thickness of the workpiece in the actual production process. As shown in Figure 4(c), it can be seen from the weld image that the central area is due to welding There is an obvious weld-through phenomenon, that is, the weld-through area B caused by the thinning of the center of the weldment is formed, and the electron temperature at the corresponding position in the SPC control chart significantly exceeds the lower limit of control, indicating that welding defects occur here. This corresponds to the actual welding conditions.

Change the clamping state of the parts to be welded 6 again, so that the butt gap of the butt parts to be welded 6 changes, so as to simulate the change of the butt gap caused by improper clamping in actual production, and then perform laser welding, as shown in Figure 4(d) It can be seen from the weld seam image that when the butt gap is too large, it will cause unwelded defects, that is, the unwelded area C is formed due to the change of the butt gap, and the corresponding position of the SPC control chart has electron temperatures at multiple points in a row Below the lower control limit, it means that a welding defect has occurred, which is consistent with the actual weld formation.

Example 2

In this embodiment, the part 6 to be welded is DX51D+Z galvanized steel sheet E, and the form of the welded joint is a lap joint with a reserved gap, that is, pads D are placed on both ends of the welded joint, and the pads The size of sheet D is determined according to actual needs. The schematic diagram of the welded joint and the weld seam formed by laser welding is shown in Figure 5. Ar shielding gas is used for protection during the welding process.

The fixed welding speed v=2.5m/min is unchanged, and different welding parameters are used for welding to determine the optimal welding parameters. At this time, the laser power P=3900W, the gap value is 0.2mm, the defocus is +15mm, Ar protection The air flow is 20L/min. Weld the five parts 6 to be welded under the optimal welding parameters to ensure that there is no accidental disturbance during the welding process and the weld seam is well formed. The computer 2 records the photoplasma generated by each part 6 to be welded during the laser welding process. 10 electronic temperature time-domain diagrams, and obtain the SPC control diagram according to each electronic temperature time-domain diagram.

After the SPC control chart is obtained, welding defects are deliberately created by changing the welding parameters or changing the state of the clamping, and the ability of the SPC control chart to identify welding defects is tested.

Changing the welding parameters reduces the welding speed in the central area of the weld, as shown in Figure 6(a). From the image of the weld, it can be seen that the central area has a local burn-through phenomenon due to the speed reduction, that is, the welding speed slows down. The electron temperature in the corresponding area of the SPC control chart is significantly lower than the lower control limit, indicating that welding defects have occurred here, which is consistent with the actual welding situation.

Increase the gap value of the lap joint with the reserved gap and then carry out laser welding, as shown in Figure 6(b), it can be seen from the weld image that due to the large gap value, the weld bead separation has occurred obviously, that is, The weld bead separation area G is formed, and the electron temperature at the positions corresponding to multiple weld bead separations in the SPC control chart is significantly lower than the lower control limit, which means that welding defects have occurred, and the discrimination effect of the SPC control chart is good.

Through the above embodiment, the following conclusions can be drawn: when the test time-domain diagram is always between the upper and lower limits of the SPC control chart, no welding defects will occur; when the test time-domain diagram fluctuates greatly, but there are not enough continuous points beyond the SPC When the upper control limit or lower control limit of the control chart is reached, there is still no welding defect at this time; when a sufficient number of continuous points in the test time domain chart exceed the upper control limit or lower control limit of the SPC control chart, it means that some kind of welding defect has occurred here. form of welding defects.

The online diagnosis method of laser welding defects based on spectral information involved in the present invention establishes specific quantitative standards to judge the existence of defects in the welding process, and the online diagnosis method can obtain the spectral information of photoplasma very accurately, And quickly and accurately judge whether there are defects in the laser welding process, so as to be applied to the online diagnosis of defects in industrial mass production, and lay a foundation for realizing the judgment and classification of defect types.

Claims (6)

1. a kind of inline diagnosis method of the laser welding defect based on spectral information, wherein on-line diagnosing apparatus are visited including optical fiber Head, the fibre-optical probe is connected through optical fiber with fiber spectrometer, and the fiber spectrometer is also connected with computer, the light Fibre probe is fixed on laser head via mechanically gimbaled bar, it is characterised in that：It the described method comprises the following steps：

The position of step 1, the adjustment fibre-optical probe；

Step 2, when laser beam welding starts, pass through fibre-optical probe and gather the photic plasma that produces in laser beam welding Optical signal in body, and by the transfer spectral information formed in fiber spectrometer into computer；

Step 3, each spectral line is relative from the frequency domain distribution figure of the above-mentioned spectral information of Real Time Observation on the display interface of computer Intensity with the time fluctuation situation, be determined as analysis object characteristic spectral line；

Step 4, any time in gatherer process, the electricity of photo plasma is calculated using the relative intensity of selected characteristic spectral line Sub- temperature, to obtain electron temperature versus time curve；

Step 5, material and size according to part to be welded, are welded using different welding parameters, and selection makes electron temperature The parameter that the standard deviation of time domain change is minimum is used as optimal welding parameter；

Step 6, under optimal welding parameter N number of part to be welded is welded, by each part to be welded of computer recording in Laser Welding The electron temperature time-domain diagram of photo plasma is produced in termination process；

Step 7, according to each electron temperature time-domain diagram in step 6, obtain SPC control figures；

Step 8, according to actual conditions adjust welding parameter, to part to be welded carry out laser welding；

Step 9, when laser beam welding starts, pass through fibre-optical probe and gather the photic plasma that produces in laser beam welding Optical signal in body, the characteristic spectral line reached the spectral information of formation by fiber spectrometer in computer, selecting step 3 is made Spectral line is characterized, the electron temperature of photo plasma is calculated by selected characteristic spectral line, electron temperature time-domain diagram is obtained, by it Referred to as test time-domain diagram；

Step 10, during laser welding, by computer by test time-domain diagram be drawn into real time in SPC control figures, judge Whether each point tested in time-domain diagram exceeds the boundary up and down of SPC control figures；

Step 11, if it is not, then in the absence of weld defect；

Step 12, if it is, judging whether the bound for having continuous a point beyond SPC control figures in continuous A point Limit, wherein A and a value depend on the sensitivity of defect dipoles；

Step 13, if it is not, then in the absence of weld defect；

Step 14, if it is, there is weld defect.

2. the inline diagnosis method of the laser welding defect according to claim 1 based on spectral information, it is characterised in that： The fibre-optical probe is equipped with COL-UV/VIS collimation lenses.

3. the inline diagnosis method of the laser welding defect according to claim 1 or 2 based on spectral information, its feature exists In：The resolution ratio of the fiber spectrometer is not less than 0.1nm.

4. the inline diagnosis method of the laser welding defect according to claim 1 based on spectral information, it is characterised in that： In the step 1, the fibre-optical probe is arranged on 15 ~ 20cm of incidence point away from laser beam on part to be welded, and be higher by 7.5 ~ 10cm of part to be welded upper surface position, while ensure to produce in fibre-optical probe collection laser beam welding is photic etc. Gas ions.

5. the inline diagnosis method of the laser welding defect according to claim 1 based on spectral information, it is characterised in that： In the step 4, the electron temperature of photo plasma is calculated using Boltzmann figure method or doubl-line method.

6. the inline diagnosis method of the laser welding defect according to claim 1 based on spectral information, it is characterised in that： In the step 7, obtain SPC control figures and comprise the following steps：

Step 71, the data of N number of electron temperature time-domain diagram in the step 6 are divided into some subgroups, synchronization difference weldering The data in road are a subgroup；

Step 72, arithmetic mean of instantaneous value is asked for the electron temperature data in each subgroup, regard the arithmetic mean of instantaneous value as each subgroup Value；

Step 73, the value to each subgroup ask for arithmetic mean of instantaneous value, and the arithmetic mean of instantaneous value is the center line in SPC control figures；

Step 74, standard deviation, referred to as subgroup standard deviation are asked for the electron temperature data in each subgroup；

Step 75, arithmetic mean of instantaneous value is asked for all subgroup standard deviations, obtain overall standard deviation；

Step 76, by the value of center line plus or subtracting the overall standard deviation of corresponding multiple to obtain in SPC control figures The value of upper control limit or lower control limit, can be obtained wherein the multiple looks into SPC control figures coefficient table according to number of samples in subgroup.