JP5055797B2 - Welding quality judgment device and welding quality judgment method - Google Patents

Description

  The present invention relates to a welding quality determination device and a welding quality determination method.

In spot welding of steel plates, there is a technique for determining the welding quality by measuring the expansion amount of the plate thickness during welding in real time in order to improve the welding quality. One such technique is to obtain an approximate linear equation for each unit time from the thermal expansion waveform of the base metal obtained during welding current energization. On the other hand, a quadratic approximate curve equation at the time of contraction of the base material is obtained from the thermal expansion waveform of the base material, the approximate linear equation having the maximum inclination, the minimum approximate linear equation, and the quadratic approximation Obtain the time at which each of the curve equations was obtained, and based on the approximate linear equation with the maximum slope, the approximate linear equation with the smallest slope, the quadratic approximate curve equation, and the time at which these approximate linear equations were obtained, the nugget There is a method of generating a characteristic parameter of the generation process, estimating a nugget diameter from the obtained characteristic parameter by at least one analysis of a stratified analysis unit or multiple regression analysis, and determining a welding condition (see Patent Document 1). ).
JP 2003-181649 A

  However, in the conventional method, when there is an abnormal occurrence of spatter or displacement of the hit point position, a discontinuous point is generated in the thermal expansion waveform of the base material. There is a problem that an appropriate judgment cannot be made when a point occurs.

  Accordingly, an object of the present invention is to provide a welding quality determination method and apparatus capable of accurately determining welding quality even when discontinuities occur in the thermal expansion waveform of a base material.

  In order to solve the above problems, the present invention provides an electrode interval measuring means for measuring an interval between welding electrodes sandwiching a member to be welded for each unit time, and storing the measured electrode interval for each unit time as electrode interval data. And a decrease start point from which the electrode interval has once increased after the start of energization and the electrode interval has started to decrease by a decrease amount greater than or equal to a predetermined amount from the electrode interval data stored in the storage unit, and an electrode after decrease It is determined that spatter has occurred when a re-increase start point at which the interval starts to increase by a predetermined amount or more is detected, and the maximum value of the electrode interval is equal to or less than a predetermined value and after the decrease start point is detected from the maximum value. And a determination means for determining that there is an abnormality in the spot position when the value up to the minimum value of the electrode interval is a predetermined amount or more.

  Further, the present invention for solving the above problems is based on the data of the distance between the welding electrodes sandwiching the member to be welded measured every unit time, and the electrode spacing is increased by a predetermined amount after the electrode spacing once increases after the start of energization. It is determined that spatter has occurred when the decrease start point that has started to decrease with the above decrease amount and the re-increase start point at which the post-reduction electrode interval has started to increase more than a predetermined amount is detected, and the maximum value of the electrode interval is the start of energization It is a welding quality determination method characterized by determining that there is an abnormality in the spot position when the value from the maximum value to the minimum value of the electrode interval after detection of the decrease start point is equal to or greater than a predetermined amount. .

  According to the present invention, when the decrease start point of the electrode interval, and the re-increase start point at which the electrode interval starts to increase again by a predetermined amount or more after the decrease start point is detected, it is determined that sputtering has occurred, and the maximum value of the electrode interval is Determining that there is an abnormality in the striking point position when it is not more than a predetermined value, for example, not more than the value of the electrode interval at the start of energization, and the minimum value of the electrode interval after detection of the decrease start point from the maximum value is not less than a predetermined amount. Therefore, even if there is a discontinuous point where the waveform of the electrode interval data drops sharply, by detecting the occurrence of such a discontinuous point, the occurrence of spatter and abnormal spot position can be detected. Cases can be sorted out accurately. Since the determination of the welding quality is performed for each of the occurrence of spatter that causes the defect and the abnormal spot position, the accuracy of the welding quality determination can be increased.

  Exemplary embodiments of a welding quality determination method and apparatus according to the present invention will be described below in detail with reference to the accompanying drawings. In the present embodiment, a case where a general mild steel sheet is spot welded using a single-phase AC spot welding apparatus is illustrated.

  In this embodiment, basically, the expansion state of the base material during energization of the welding current and the contraction state (nugget formation process) of the base material after the end of the energization of the welding current are generated from the amount of displacement between the electrodes of the welding machine. In addition, by classifying abnormalities in the spot position, etc., the quality of the welding quality is judged after the occurrence of these abnormalities is divided into each case.

  FIG. 1 is a functional block diagram of a welding quality determination apparatus according to the present invention.

  The welding quality determination apparatus 1 includes an electrode interval measurement unit 10, a welding sequence unit 12, a layer analysis unit 13, a quality determination unit 14, a storage unit 15, and a clock generation unit 16.

  In actuality, the welding quality determination apparatus 1 is configured such that the function of each unit is implemented by executing a program created according to a layer-based determination procedure described later by a computer such as a personal computer. Therefore, in the actual apparatus configuration, it is not necessary that the function of each unit is divided for each function as illustrated.

  In addition, although not shown, the welding quality determination apparatus 1 is connected to an input device such as a keyboard, a touch panel, a pen tablet, and a mouse, an output device such as a display and a printer, and another computer that are usually provided in the computer. Network connection function.

  Moreover, although the welding sequence part 12 is built in here as the welding quality determination apparatus 1, this is made to function by the controller of a welding machine, and the welding quality determination apparatus 1 itself is classified into the stratification analysis part 13, the quality determination part 14, The storage unit 15 and the like may be included.

  Hereinafter, the function of the welding quality determination apparatus will be described in detail with reference to FIG.

  The electrode interval measuring unit 10 (electrode interval measuring means) is a part that measures how the electrodes move in the nugget formation process. When spot welding is performed on the base material, the base material is sandwiched between electrodes of the welding machine 50, and a welding current is supplied to the base material in this state. When the supply of the welding current starts, the base material of the portion sandwiched between the electrodes expands, and when the supply of the welding current ends, the base material of that portion contracts. The electrode moves up and down according to the expansion and contraction of the base material. The electrode interval measuring unit 10 detects the electrode interval by the sensor 11 attached to the welding machine 50. The sensor 11 is, for example, an encoder attached to a motor or a gear mechanism for sending out one or both of the electrodes of the welding machine 50.

  The electrode interval measurement unit 10 acquires the measured electrode interval value for each unit time based on the clock signal from the clock generation unit 16 with the spot welding start time as a reference, and stores it in the storage unit 15 together with the acquisition time. The electrode interval data acquired for each unit time is electrode interval data.

  The welding sequence unit 12 controls the energization amount at the time of welding based on the clock signal output from the clock generation unit 16.

  The stratification analysis unit 13 (stratification means) performs stratification in the welded state. In the present embodiment, the welding state stratification means that a spatter abnormality has occurred during welding of the workpiece, a striking point position abnormality (hereinafter referred to as a striking point position abnormality), or both have not occurred. It means to separate each.

  The quality judging unit 14 (quality judging means) judges the quality of the welding point for each welding state stratified by the stratified analyzing unit 13. The quality judgment here is performed by comparing the result of the actual judgment and the characteristic parameter calculated from the electrode interval data of the welding point with the characteristic parameter calculated from the electrode interval data in the current welding. .

  Here, simply comparing feature parameters makes it difficult to make a pass / fail decision based on various feature parameter values. However, it has been found that the value of the characteristic parameter shows a characteristic tendency depending on the quality of the welding quality (that is, the result of the negative judgment). Therefore, a discriminant is created by combining this tendency with the result of the determination of the chisel, and the discriminant is used for determination.

  The discriminant is based on the electrode interval data at the time of welding, where the target variable is the result of the determination of each of the welds when the spatter has occurred, the spot position is abnormal, and there is no spatter and no spot position abnormality. This is a formula for determining good or bad welding using the acquired characteristic parameters as explanatory variables (details will be described later). The discriminant is stored in the quality judgment unit 14.

  The memory | storage part 15 (memory | storage means) memorize | stores the value of the electrode space | interval measured in the electrode space | interval measurement part 10 with the time which acquired the value.

  The clock generation unit 16 is a part that supplies a clock signal to all the components constituting the welding quality determination device 1. All the components including the electrode interval measuring unit 10 described above operate based on this clock signal.

  Next, the welding quality determination method according to the present invention will be described.

  FIG. 2 is a flowchart showing the stratification procedure.

  First, the welding quality determination apparatus 1 is for the layer analysis unit 13 to start a welding sequence by the welding sequence unit 12 and to start welding by the welding machine 50, and at the same time to measure the time on the basis of the welding start time. The time is set to 0 (time clear S1).

  Subsequently, the electrode interval measuring unit 10 measures the electrode interval due to thermal expansion or contraction of the base material every unit time St by a signal from the sensor 11 provided on the electrode (S2). The electrode interval measuring unit 10 stores the measured electrode interval in the storage unit 15 as electrode interval data together with the time for each unit time St (S3). The sampling unit time St (for example, 10 msec) is set in advance.

  The layer-by-layer analysis unit 13 calculates necessary characteristic parameters for each layer based on the electrode interval data stored in the storage unit 15 after the end of welding of one welding point (S4). Details of the feature parameter will be described later.

  After calculating the feature parameter, the stratified analysis unit 13 first determines whether the hit point position is abnormal from the obtained feature parameter (S5).

  As a result of the determination, if there is a hit point position abnormality (S6: Yes), the quality determination unit 14 performs the hit point position abnormality process (the process after S100).

  There are two types of hitting point anomalies and a worsening protrusion. These dot position anomalies will be described later.

  First, the quality determination unit 14 determines whether it is edged or protruding (S100).

  As a result of the determination, if it is determined that the protrusion is present (S101: Yes), the welding point defect is recorded as it is (S110). For example, the defective weld point is stored in the storage unit 15 together with the welding position (or an identification number indicating the weld point). At the same time, it may be displayed on a display or the like, or may be notified to another computer or the like (hereinafter, the same applies to other determination results).

  On the other hand, in the case of a non-extrusion that does not lead to overhanging (S101: No), the quality determination unit 14 then determines the chisel data for the end (S102). Although details on the seed data discrimination will be described later, this process is performed even when there is a punch. As a result of the judgment, there are some welds that have sufficient strength. It is judged from the interval data and taken out as a non-defective product.

  If it is determined that the product is a non-defective product (S103: Yes), that effect is stored in the storage unit 15 (S104), and if it is determined to be defective, it is stored in the storage unit 15 (S105).

  If it is determined in step S6 that there is no hit point position abnormality, the layer-by-layer analysis unit 13 determines the presence or absence of sputtering from the feature parameter (S7).

  As a result of the determination, if there is spatter (S8: Yes), the quality determining unit 14 performs processing when spatter occurs (processing after S200).

  When the spatter has occurred, the quality determining unit 14 determines the chisel data from the electrode interval data (S200). This is because even if spatter occurs, the welded part is joined with sufficient strength according to the result of the judgment, so that it is not a defective weld due to spatter alone. It is a process for taking out.

  If it is determined that the product is a non-defective product (S201: Yes), that effect is stored in the storage unit 15 (S202), and if it is determined to be defective, it is stored in the storage unit 15 (S203).

  On the other hand, when it is determined that there is no spatter (S8: No), the quality determination unit 14 continues to determine the defect data of the hit point without abnormality from the characteristic parameter (S9). As a result of the judgment data determination, if it is determined to be a non-defective product (S10: Yes), that effect is stored in the storage unit 15 (S11), and if it is determined to be defective, it is stored in the storage unit 15 (S12). ). Then, all processing is finished.

  FIG. 3 is a drawing for explaining the edge punching.

  As shown in FIG. 3A, the striking points for the members to be welded are A1 and A2 when the spot nugget 100 is completely inside the plate 101 to be welded, and B when a part is protruding from the end of the plate 101. is there. Of these, A1 and A2 do not end. On the other hand, “B” is called “end hit”. Of these B cases, the degree is poor, and the case where most of the electrodes protrude is referred to as protrusion.

  As shown in FIG. 3 (B), the edge punching and protruding judgment criteria are such that the formed nugget 100 is regarded as a circle, the diameter in the direction protruding from the plate 101 is a, and the diameter in the direction not protruding is b. The case of a + b) / 2 <reference nugget diameter is projected, and the case of (a + b) / 2 ≧ reference nugget diameter is assumed to be an edge (not a protrusion). In the present embodiment, it is determined from the feature parameter whether it is edged or protruding.

  Next, a characteristic parameter and a method of layering using the characteristic parameter will be described with reference to a specific electrode interval waveform (thermal expansion waveform).

  FIG. 4 is a graph showing the electrode interval data along the time axis. This graph is a waveform graph without a discontinuous portion that causes defects such as spattering and edge cutting.

  As shown in the figure, the electrode interval waveform during welding changes according to the welding sequence. The welding sequence is in the order of squeeze, first energization, second energization, third energization, and hold.

  In the electrode interval waveform, the electrode interval at the time of squeezing before starting energization is 0, and in normal welding, as shown in FIG. In this case, the nugget continues to expand and increases as it is, passes through the second energization, and decreases from the second half of the third energization to the hold. That is, there is no discontinuity in the waveform of the electrode interval data.

  FIG. 5 is a graph showing an electrode interval waveform when sputtering occurs.

  As shown in the figure, when spatter occurs and a part of the welded member melts and jumps out, the electrode spacing increases slightly at first, but rapidly decreases as spatter occurs. The decrease start point is the feature parameter DownStartFlg, and the minimum value that has dropped is the feature parameter MaxDown. The maximum value of the electrode spacing is the feature parameter Hmax. Hmax is only once before DownStartFlg when sputtering occurs. The re-increase starting point at which the electrode interval starts to increase again after MaxDown is the characteristic parameter DownEndFlg. A feature parameter indicating the maximum increase rate after MaxDown is SpaSlopeMax. The maximum value after MaxDown is set as a feature parameter Hmax2. In addition, the final value of the electrode interval after the end of welding (after the end of the hold period) is used as the feature parameter Hend.

  From this waveform, when sputtering occurs, the electrode interval slightly increases but decreases rapidly to the minimum value, but after that the electrode interval starts to increase, but it does not return to the maximum value once recorded. I understand that.

  In order to detect this characteristic pattern, the present embodiment uses three characteristic parameters. First, when energization is started, the electrode interval once becomes larger than 0, so that Hmax> 0. Then, a decrease start point DownStartFlg at which the electrode interval starts to decrease with a decrease amount greater than or equal to a predetermined amount, and a re-increase start point DownEndFlg at which the electrode interval starts increasing again by a predetermined amount or more after passing through MaxDown. When these three characteristic parameters are detected, it is determined that there is sputtering. In other words, after the electrode interval once increased after the start of energization, the electrode start interval was recorded with the decrease start point at which the electrode interval began to decrease by a decrease amount greater than or equal to the predetermined amount, and the minimum electrode interval after the decrease start point. When the re-increase start point where the increase is started more than the fixed amount is detected, it is determined that the spatter is generated.

  As a specific method for detecting these characteristic parameters, first, it may be determined whether or not Hmax is greater than zero.

  The decrease start point DownStartFlg is 5 when the waveform of the electrode interval continues to drop by 5 points or more between the start of the first energization and the third energization, and when the decrease amount of the 5 points becomes a predetermined amount or more. A point at which a decrease of more than the point has started is defined as a decrease start point DownStartFlg. Here, the “point” of 5 points corresponds to one unit time because the electrode interval data itself is measured every unit time. In addition, the predetermined amount here is a value obtained in advance by statistical processing, for example, from a plurality of actual welding results, the amount of depression for five points at the time of occurrence of spatter. For example, a minimum value or a value of 3σ may be used.

  Subsequently, when the points when DownStartFlg is set up (Point P1 5 points ahead of DownStartFly), the subsequent points are compared one after another, and there is a point that has increased by a predetermined amount or more at the points after P1 The point of the increase start is defined as DownEndFlg. The predetermined amount here may be set from the data when the actual spattering occurs as described above. Therefore, if only this sputtering is detected, it is not necessary to detect the minimum value MaxDown.

  FIG. 6 is a graph showing an electrode interval waveform in the case of edge strike.

  As shown in the figure, when there is an edge or protrusion, the electrode interval hardly increases from 0 at the time of squeeze, and rapidly decreases with 0 being the maximum value Hmax. After the minimum value is recorded, it increases slightly, but Hmax2 is recorded without reaching Hmax, and the process ends with a gradual decrease.

  Therefore, the edge strike and the protrusion are from the maximum value Hmax ≦ 0 of the electrode interval and from the maximum value Hmax (in this case, Hmax is substantially 0) to the minimum value MaxDown of the electrode interval after detection of the decrease start point DownStartFlg. When the amount of the drop is greater than a predetermined amount, it can be determined that there is a slash or sticking out (that is, an abnormal spot position).

  Note that even in the rare cases where there is a limit, since Hmax may slightly exceed 0, x may be obtained in advance through experiments or the like as Hmax ≦ x and set to a value greater than 0.

  In this case, the predetermined amount may be set from data when the edge strike and the protrusion actually occur.

  FIG. 7 is a graph showing an electrode interval waveform in the case of protrusion.

  Although there is no significant difference between the edge pattern and the waveform shape, the size from Hmax to MaxDown is larger than the edge pattern (FIG. 6). In other words, the amount of drop of the waveform, that is, the amount of decrease from 0 of the electrode interval is larger when protruding from the edge.

  Therefore, the predetermined amount is set in two stages in order to determine the edge strike and the protrusion. In other words, if the predetermined amount (first predetermined amount) in the first step is exceeded and the predetermined amount (second predetermined amount) in the second step is not exceeded, it is determined as end-up. In this case, in this embodiment, as described above, even when it is determined to be edge-cutting, the edge data is judged to be acceptable, and the edge-cutting to the extent that no problem occurs in the welding itself is accepted. On the other hand, when it decreases beyond the predetermined amount (second predetermined amount) in the second stage, it is determined that the protrusion is present. In this case, in this embodiment, since it is unlikely to pass by the chisel determination, it is considered as poor welding without performing the chisel data determination.

  In the case shown in FIGS. 6 and 7, the first predetermined amount is set to −100 μm and the second predetermined amount is set to −2000 μm. The first and second predetermined amounts differ depending on the plate thickness and material to be welded. Therefore, the first and second predetermined amounts should be determined in advance according to the experiment, or set from electrode distance data of welding actually performed so far. become.

  Next, the chisel data determination will be described.

  In the welding data judgment, at the welding point after the actual welding, a judgment is made on the sputtered product and the end-punched product, and the characteristic parameter is calculated from the electrode interval data of each of the judgment-accepting product and the rejecting product, Judgment.

  Note that the chisel determination is performed for a long time in normal welding work, and is a method for confirming whether or not the chisel can be driven into an actually welded work.

  In this negative data determination, a determination formula is introduced in which the pass / fail judgment (OK or NG) is an objective variable and the characteristic parameter obtained from the electrode interval data is an explanatory variable.

  The judgment data when spattering occurs, the judgment data judgment when both ends occur, and the judgment data judgment when both do not occur are basically the same process. The only difference is that the created discriminant is used.

  FIG. 8 is a diagram for explaining the characteristic parameters.

  As the characteristic parameters, the maximum gradient θmax when rising in the direction in which the electrode interval of the electrode interval waveform increases, the value θmaxa where the straight line of the maximum gradient θmax when rising rises across the Y axis, and the maximum gradient θmax when rising are recorded. The time when the maximum time HmaxT, the maximum value Hmax, the maximum slope Hmax intersects the maximum value Hmax and the time VerT, the time VerT, the maximum slope inclination θmin, and the minimum slope θmin θminT, the maximum inclination d_θmax when descent in the direction in which the electrode interval of the electrode interval waveform increases, and the value d_θmaxa and the maximum inclination d_θmax when descent that the straight line of this maximum inclination d_θmax crosses the Y axis are recorded Time Peak_T when recording time d_θmaxT, minimum descent value d_θmin during descent, and maximum value Hmax . Further, a second-order approximation of the contracted portion after Hmax shown in FIG. 8 is made Y = ax2 + bx + c, a is used as X0, b is used as X1, and c is used as X2. Further, the above-described minimum value MaxDown, the maximum increase rate SpaSlopeMax after MaxDown (the maximum slope value when the waveform rises after MaxDown), and the final electrode interval value Hend are used.

  The discriminant is as follows.

Y = αθmaxa + β (d_θmaxa / 100) 2 + γθmaxa + δ (θmaxa / 100) 2 + εθmax 2 + ζθmaxa 2 + ηθmaxT + θ (θmaxT / 100) 2 + ιd_θmax + κ (d_θmax) 2 + λd_θmaxT + μ (d_θmaxT / 100) 2 + νθmin + ξ (d_θmin × 100) 2 + οθminT + π (θminT / ^{100) 2 + ρHend + σHend 2} + τ (Hmax-Hend) + υ (Hmax-Hend) 2 + φHmax + χHmax 2 + ψ P eakT + ωPeakT 2 + α1VerT 2 + β1X2 + γ1 (X2 × 100) 2 + δ1X1 + ε1X1 2 + ζ1X0 + η1X0 2 + θ1 (T1 + T2 + T3 + T4) + ι1Hmax2 + κ1 (Hmax2-Hend) + λ1MaxDown + μ1SpaSl pemax
In the formula, α, β, γ, δ, ε, ζ, η, θ, ι, κ, λ, μ, ν, ξ, ο, π, ρ, σ, τ are attached before each term. , Υ, φ, χ, ψ, ω, α1, β1, γ1, δ1, ε1, ζ1, η1, θ1, ι1, κ1, λ1, and μ1 are adjustment coefficients for using this discriminant as a discriminant score. . This adjustment coefficient is set so that the Y value is 0 or more when the result of the judgment is negative, and is less than 0 when the result is unacceptable. Specifically, it is set using a computer. As software used at that time, for example, JUSE-StatWorks (registered trademark) manufactured by Japan Science and Technology Institute can be used. In addition, the number of samples is required to be at least as large as the explanatory variable, but is preferably about 100 to 200, for example.

  The slope θ in each characteristic parameter is measured every unit time St. Therefore, a recent linear equation for obtaining a slope by a time width Tw arbitrarily determined using the time unit Tn as a reference for the step size is calculated. To do. As a result, the recent linear equation is obtained by subtracting a solution obtained by dividing the time width Tw by the unit time St from n times (n = total measurement time Ts / unit time St) in which the electrode interval is measured (N = n− Only Tw / St) is obtained. Accordingly, for θmax, θmin, etc., the one having the largest coefficient indicating the inclination is selected from the obtained linear approximation equations.

  The time width Tw for calculating the approximate linear expression is not particularly determined and may be arbitrary. However, as the time width Tw is reduced, the accuracy of waveform analysis is improved, but the processing takes time. Become. Therefore, it is sufficient that the drop of the waveform due to the occurrence of spatter or spot position abnormality can be sufficiently confirmed.

  In the present embodiment, a discriminant in which Y ≧ 0 is obtained when a workpiece that has been actually welded is used for a weld point that does not generate spatter and does not have an abnormal spot position (referred to as a first discriminant). A discriminant in which Y ≧ 0 is obtained when the welding point at the time of occurrence of edge strike is Y ≧ 0 and a discriminant in which Y ≧ 0 is obtained when the weld point is sputtered. (Referred to as a third discriminant) is created and used for welding determination. In addition, the adjustment coefficient at the time of using for welding determination uses what was made when the 1st-3rd determination formula mentioned above was created.

  And the value of Y of a discriminant is calculated using each characteristic parameter obtained from electrode interval data after welding, and when the value of Y is 0 or more, it is judged as a pass product (good product).

  As described above, according to the present embodiment, first, the presence / absence of sputtering and the presence / absence of spot position abnormality are stratified (separated) from the electrode interval data, and it is determined that there is an edge finish at the time of spatter occurrence and spot position abnormality. Further, since it is determined whether or not there is a non-defective product in the welding point, it is possible to accurately determine the welding quality for various welding states.

  In particular, it is determined that sputtering has occurred when the re-increase start point at which the electrode interval starts to increase more than a predetermined amount is detected again after the electrode interval decrease start point and the minimum value of the electrode interval after the decrease start point. Since the maximum interval value is less than or equal to the interval at the start of energization and the minimum value of the electrode interval after detection of the decrease start point from the maximum value is greater than or equal to a predetermined amount, it is determined that there is a spot position abnormality. If there is a discontinuous point in which the waveform of A falls sharply, it is possible to easily stratify by detecting the occurrence of such a discontinuous point.

  In this embodiment, at the time of stratification, the presence / absence of a spot position abnormality is determined first, and the presence / absence of the next sputtering is determined. This is because, based on experience so far, if there is an abnormality in the dot position, there is a high possibility of a defective product, especially when there is an overhang, so most of the defects are defective. This is because the subsequent processing can be omitted. Thereby, when welding a some welding point continuously, a process can be accelerated as much as possible. Of course, the presence or absence of sputtering may be determined first.

  Although the embodiment to which the present invention is applied has been described above, the present invention is not limited to such an embodiment. For example, in the above-described embodiment, the chisel data determination is performed even in the case where both the sputter and the spot position abnormality are not present. Instead, in the case where neither the sputtering and the spot position abnormality are present, the conventional method (for example, patent document) Quality determination may be performed according to 1).

  In addition, only the layer-by-layer determination may be performed, and if there is any spatter or spot position abnormality, all may be determined as defective.

It is a functional block diagram of the welding quality judging device concerning the present invention. It is a flowchart which shows the stratification procedure. It is a figure for demonstrating edge punching. It is the graph which showed electrode space | interval data along the time axis. It is a graph which shows an electrode space | interval waveform when a sputter | spatter generate | occur | produces. It is a graph which shows an electrode space | interval waveform at the time of a slash. It is a graph which shows the electrode space | interval waveform at the time of a protrusion. It is drawing for demonstrating a characteristic parameter.

Explanation of symbols

1 ... Welding quality judgment device,
10: Electrode interval measuring unit,
11 ... sensor,
12 ... Welding sequence part,
13: Stratified analysis department,
14 ... Quality judgment part,
15 ... storage part,
16: Clock generator,
50 ... Welder.

Claims (6)

An electrode interval measuring means for measuring an interval between welding electrodes sandwiching a member to be welded every unit time;
Storage means for storing the measured electrode interval per unit time as electrode interval data;
From the electrode interval data stored in the storage means, a decrease start point at which the electrode interval starts to decrease at a decrease amount of a predetermined amount or more after the electrode interval has once increased after the start of energization, and the electrode interval after the decrease is at least a predetermined amount It is determined that spatter has occurred when a re-increase start point that has started increasing is detected, the maximum value of the electrode interval is not more than a predetermined value, and the minimum value of the electrode interval after detecting the decrease start point from the maximum value Determining means for determining that there is an abnormality in the hit point position when the value up to a predetermined amount;
A welding quality determination device characterized by comprising:   The welding quality determination device according to claim 1, wherein the predetermined value is a value of an electrode interval at the start of energization. The characteristic parameters used in the following determination formula are calculated from the electrode spacing data for each of the cases where there is spatter stratified by the determination means, there is a hit point position abnormality, and there is no sputter occurrence and no hit point position abnormality,
The value of the feature parameters the calculated, and by putting the preset following discriminants, there spatters, there strike position abnormal, and determines the quality determining welding conditions about it in the absence of spatter nor strike position abnormality Further comprising means,
The discriminant formula uses as a target variable the result of each of the welding judgments in the case where the spatter has occurred, the spot position is abnormal, and there is no spatter occurrence and spot position abnormality, and the electrode interval data at the time of the welding in which the spot judgment was made. The characteristic parameter acquired from the above is used as an explanatory variable to determine whether welding is good when the Y value of the discriminant is 0 or more, and poor welding when the Y value of the discriminant is less than 0. The welding quality determination device according to claim 1 or 2, characterized in that:
Y = αθmaxa + β (d_θmaxa / 100) 2 + γθmaxa + δ (θmaxa / 100) 2 + εθmax 2 + ζθmaxa 2 + ηθmaxT + θ (θmaxT / 100) 2 + ιd_θmax + κ (d_θmax) 2 + λd_θmaxT + μ (d_θmaxT / 100) 2 + νθmin + ξ (d_θmin × 100) 2 + οθminT + π (θminT / ^{100) 2 + ρHend + σHend 2} + τ (Hmax-Hend) + υ (Hmax-Hend) 2 + φHmax + χHmax 2 + ψPeakT + ωPeakT 2 + α1VerT 2 + β1X2 + γ1 (X2 × 100) 2 + δ1X1 + ε1X1 2 + ζ1X0 + η1X0 2 + θ1 (T1 + T2 + T3 + T4) + ι1Hmax2 + κ1 (Hmax2-Hend) + λ1MaxDown + μ1SpaSlo pemax
(In the formula, the characteristic parameters are the maximum inclination θmax when rising in the direction in which the electrode interval of the electrode interval waveform increases, the value θmaxa that the straight line of this maximum inclination θmax rises across the Y axis, and the maximum inclination when rising. When the value θmax is recorded, the time θmaxT, the maximum value Hmax of the electrode interval, the straight line of the maximum slope of inclination θmax and the point where the maximum value Hmax intersects the time VerT, the minimum slope of rise θmin, the minimum slope The time θminT when the value θmin is recorded, the maximum inclination d_θmax when the electrode interval of the electrode interval waveform increases, d_θmax, the value d_θmaxa where the straight line of the maximum inclination d_θmax falls across the Y axis, and the inclination when descending When the time d_θmaxT when the maximum value d_θmax is recorded, the minimum inclination d_θmin when descending, and the maximum value Hmax are recorded In addition, when a contraction portion after the maximum value Hmax is second-order approximated to be Y = ax2 + bx + c, a is X0, b is X1, and c is X2. Further, when sputtering occurs, the electrode interval is used. The start point of suddenly decreasing is DownStartFlg, the minimum value dropped when sputtering occurs is MaxDown, the maximum increase rate SpaSlopeMax after the minimum value MaxDown (the maximum value of the slope when the waveform increases after MaxDown), The final electrode spacing value Hend is used, and α, β, γ, δ, ε, ζ, η, θ, ι, κ, λ, μ, ν, which precede each term. ξ, ο, π, ρ, σ, τ, υ, φ, χ, ψ, ω, α1, β1, γ1, δ1, ε1, ζ1, η1, θ1, ι1, κ1, λ1, μ1 Use as discriminant score The adjustment coefficient is a Y value of the discriminant that is greater than or equal to 0 when it passes as a result of the judgment, and the Y value of the discriminant is less than 0 when it fails. Set to be.)   From the data of the distance between the welding electrodes that sandwich the member to be welded measured per unit time, the electrode starting point decreases after the start of energization, and then the electrode starting point starts to decrease with a decrease amount of a predetermined amount or more, When the re-increase start point is detected when the electrode interval starts to increase by a predetermined amount or more after the decrease, it is determined that spatter has occurred, and the maximum value of the electrode interval is equal to or less than a predetermined value and starts decreasing from the maximum value. A welding quality determination method characterized by determining that there is an abnormality in the spot position when a value up to a minimum value of an electrode interval after point detection is a predetermined amount or more.   The welding quality determination method according to claim 4, wherein the predetermined value is a value of an electrode interval at the start of energization. The characteristic parameters used in the following determination formula are calculated from the electrode spacing data for each of the determined sputter occurrence, hitting point position abnormality, and when there is no sputter occurrence or hitting point position abnormality,
By putting the value of the calculated characteristic parameter in a discriminant set in advance, the welding state is determined with respect to the case where there is spatter occurrence, there is a spot position abnormality, and there is no sputter occurrence or spot position abnormality,
The discriminant formula uses as a target variable the result of each of the welding judgments in the case where the spatter has occurred, the spot position is abnormal, and there is no spatter occurrence and spot position abnormality, and the electrode interval data at the time of the welding in which the spot judgment was made. The characteristic parameter acquired from the above is used as an explanatory variable to determine whether welding is good when the Y value of the discriminant is 0 or more, and poor welding when the Y value of the discriminant is less than 0. The welding quality judgment method according to claim 4 or 5.
Y = αθmaxa + β (d_θmaxa / 100) 2 + γθmaxa + δ (θmaxa / 100) 2 + εθmax 2 + ζθmaxa 2 + ηθmaxT + θ (θmaxT / 100) 2 + ιd_θmax + κ (d_θmax) 2 + λd_θmaxT + μ (d_θmaxT / 100) 2 + νθmin + ξ (d_θmin × 100) 2 + οθminT + π (θminT / ^{100) 2 + ρHend + σHend 2} + τ (Hmax-Hend) + υ (Hmax-Hend) 2 + φHmax + χHmax 2 + ψPeakT + ωPeakT 2 + α1VerT 2 + β1X2 + γ1 (X2 × 100) 2 + δ1X1 + ε1X1 2 + ζ1X0 + η1X0 2 + θ1 (T1 + T2 + T3 + T4) + ι1Hmax2 + κ1 (Hmax2-Hend) + λ1MaxDown + μ1SpaSlo pemax
(In the formula, the characteristic parameters are the maximum inclination θmax when rising in the direction in which the electrode interval of the electrode interval waveform increases, the value θmaxa that the straight line of this maximum inclination θmax rises across the Y axis, and the maximum inclination when rising. When the value θmax is recorded, the time θmaxT, the maximum value Hmax of the electrode interval, the straight line of the maximum slope of inclination θmax and the point where the maximum value Hmax intersects the time VerT, the minimum slope of rise θmin, the minimum slope The time θminT when the value θmin is recorded, the maximum inclination d_θmax when the electrode interval of the electrode interval waveform increases, d_θmax, the value d_θmaxa where the straight line of the maximum inclination d_θmax falls across the Y axis, and the inclination when descending When the time d_θmaxT when the maximum value d_θmax is recorded, the minimum inclination d_θmin when descending, and the maximum value Hmax are recorded In addition, when a contraction portion after the maximum value Hmax is second-order approximated to be Y = ax2 + bx + c, a is X0, b is X1, and c is X2. Further, when sputtering occurs, the electrode interval is used. The start point of suddenly decreasing is DownStartFlg, the minimum value dropped when sputtering occurs is MaxDown, the maximum increase rate SpaSlopeMax after the minimum value MaxDown (the maximum value of the slope when the waveform increases after MaxDown), The final electrode spacing value Hend is used, and α, β, γ, δ, ε, ζ, η, θ, ι, κ, λ, μ, ν, which precede each term. ξ, ο, π, ρ, σ, τ, υ, φ, χ, ψ, ω, α1, β1, γ1, δ1, ε1, ζ1, η1, θ1, ι1, κ1, λ1, μ1 Use as discriminant score The adjustment coefficient is a Y value of the discriminant that is greater than or equal to 0 when it passes as a result of the judgment, and the Y value of the discriminant is less than 0 when it fails. Set to be.)

Images