A method for quality control of a spot welding

Abstract

There is described a method for monitoring the quality of resistance spot welding, the resistance spot welding being carried out by means of welding equipment (WF) comprising a first electrode ( T1 ) and a second electrode ( T2 ) between which a welding current (I) flows during welding and between which, during welding, a welding voltage (V) exists, the method including: - calculating, for each resistance welding spot, a welding energy value (E) as a function of the welding voltage (V ) and of the welding current (I) during the making of the welding spot, - calculating, for each resistance welding spot, an expected decay (E j ) of welding energy as a function of a wear indicator (X j ) of said first and / or second electrodes (T1, T2), - determining a value of a welding quality indicator (WPQE) as a function of said welding energy value (E) and of said expected decay (E j ) of welding energy, and - comparing said welding quality indicator (WPQE) with a threshold value (MaxDevPerc).

Description

[0001] " A method for quality control of a spot welding" ★ ★ ★ ★

[0002] TEXT OF THE DESCRIPTION

[0003] Field of the Invention

[0004] The present invention relates to the monitoring of the quality of resistance spot welding. The invention was developed with specific reference to quality control in the automotive sector, specifically with reference to welding lines of motor vehicle bodies.

[0005] Prior Art

[0006] In the field of quality control of resistance spot welding, there are currently available systems based on Al algorithms, which in practice exhibit poor reliance and poor flexibility. A further problem of such control systems is the high computational burden required in the execution of said algorithms.

[0007] Moreover, there is a further problem which is common to any Al algorithm, which regards the estimate of the welding quality in any wear condition of the welding electrodes; such a problem is not solved in the known art.

[0008] Obj ect of the Invention

[0009] The present invention aims at solving the technical problems outlined in the foregoing. Specifically, the obj ect of the present invention consists in providing a method for quality control of resistance spot welding which is reliable and accurate irrespective of the state of wear of the welding electrodes being monitored.

[0010] Summary of the Invention

[0011] The obj ect of the invention is achieved by means of a method having the features set forth in the claims that follow, which form an integral part of the technical disclosure provided herein in relation to the invention.

[0012] Brief Description of the Figures

[0013] The invention will now be described with reference to the annexed Figures, which are provided by way of non-limiting example only and wherein:

[0014] - Figure 1 schematically shows a welding kit in which it is possible to implement the method according to the invention,

[0015] - Figure 2 shows a diagram of welding current / welding voltage during a resistance spot welding,

[0016] - Figure 3 shows a model of welding energy decay implemented in a preferred embodiment of the method according to the invention,

[0017] - Figure 4 is a flow diagram representative of aspects of a preferred embodiment of the invention, and Figure 5 is a matrix representative of an implementation aspect of the diagram of Figure 4.

[0018] Detailed Description

[0019] With reference to Figure 1 and Figure 2, the invention defines a method for monitoring the quality of resistance spot welding, which is carried out by means of welding equipment WF (which in turn is part of a welding kit WK the details whereof will be described further below) comprising a first electrode T1 and a second electrode T2 between which, during welding, a welding current I flows and between which, during welding, a welding voltage V exists. The method according to the invention was developed with particular reference to resistance spot welding for assembling motor vehicle bodies, but it may be extended to any other operating context wherein a spot welding is implemented, i. e. wherein a certain number (e. g. a few hundreds) of welding spots are present the quality whereof must be monitored.

[0020] Figure 2 shows a diagram of evolution in time of the welding current I, which specifically highlights four reference time intervals including:

[0021] - a first interval tAduring which a rapid increase takes place of the value of the current which flows between the electrodes T1 and T2, up to a pre-heating (or tacking) value

[0022] - a second interval tBduring which the pre-heating (or tacking) occurs of the welding area, the welding current remaining at the pre-heating value, which has been reached at the end of the interval tA, until a rapid decrease of the current from the previously reached value a third interval

[0023]

[0024] during which the rapid decrease of the current from the pre-heating value has reached and maintains a zero value for a given time period, followed by a new increase up to the nominal value of the welding current I,

[0025] - a fourth interval tDduring which the resistance welding is completed, with the current I at the nominal value. Preferably, the fourth interval tDis the reference welding time interval, as regards the method according to the invention.

[0026] In various embodiments, and specifically in the preferred embodiment shown in the Figures, the method according to the invention comprises:

[0027] - calculating, for each resistance welding spot, a welding energy value E as a function of the welding voltage V and of the welding current I during the making of the welding spot,

[0028] - calculating, for each resistance welding spot, an expected welding energy decay

[0029]

[0030] as a function of a wear indicator of said first and / or second electrodes Tl, T2 (in this regard, the wear indicator is generally representative of the wear of both electrodes, since the reconditioning takes place simultaneously. Therefore, it is generally a wear indicator of the first and of the second electrodes Tl, T2. However, with certain equipment - for example with single-sided welding clamps - the wear indicator may essentially refer to the welding electrode, since for geometric reasons the latter is subj ected to more frequent reconditionings (dressings) with respect to the ground electrode. In this case, the wear indicator only refers to one of the electrodes Tl, T2, specifically the welding electrode),

[0031] determining the value of a welding quality indicator WPQE as a function of the welding energy value E and of the expected welding energy decay Ej, and

[0032] - comparing the welding quality indicator WPQE with a threshold value.

[0033] According to the invention, the welding energy value E as a function of the welding voltage V and of the welding current I during the construction of the welding spot is calculated in accordance with the relation

[0034] N

[0035]

[0036] i=l

[0037] wherein:

[0038] E is the value of the welding energy

[0039] Ri is the value of the welding resistance at an i-th time instant of a reference interval of welding time, in particular the interval tD.

[0040] As a matter of fact, according to Ohm' s first law, the electrical resistance may be expressed as a ratio of the welding voltage V to the welding current I, and therefore it is possible to write – with reference to the i-th time instant of the reference time interval, in particular tD:

[0041]

[0042] wherein each i-th time instant of the reference interval tDof welding time is a sampled instant of the reference interval of welding time, wherein the sampling is carried out by means of a sampling device TC ( Figure 2 ) operatively associated with the welding equipment WF and configured to provide a time reference for all the determinations and the deductions that take place in the method.

[0043] The welding energy E considered in itself, however, is not indicative of the welding quality. This is due to the fact that the value of energy considered per se does not take into account a possible decay due to the wear of the welding equipment. Preferably, the method according to the invention considers, as a wear indicator, a number of reconditionings of the first and / or of the second electrodes Tl, T2. In this regard, among all the parameters related to the wear of the welding equipment, the inventors have ascertained that the reconditioning of the electrodes - i. e. the recovery of the electrode geometry with a consumption of the electrode itself - is the parameter which constitutes the main factor causing wear phenomena. As a consequence the calculation, for each resistance welding spot, of an expected decay as a function of an indicator of the wear of said first and / or second electrodes Tl, T2 comprises calculating a reduced energy value

[0044]

[0045] as a function of a number of reconditionings Xj of said first and / or second electrodes.

[0046] In more detail, the reduced energy value Eias a function of a number of reconditionings xjof the first and / or second electrodes T1, T2 is calculated according to the relation:

[0047] Ej = Po + Pi ‘xj

[0048] wherein:

[0049] Xj corresponds to a j -th occurrence of a reconditioning event

[0050] β0, β1are two numerical coefficients.

[0051] As it is evident, referring to Figure 3, said relation corresponds to a linear regression on a plurality of series of experimental points, which results into a regression line with intercept β0and angular coefficient β1. Basically, the angular coefficient β1is negative, since with the increase in the number of reconditionings of the first and / or second electrodes Tl, T2, the value of the energy developed during the welding decreases. In other words, the final deduction in the method according to the invention makes use of a mapping of the effects of wear on the welding quality.

[0052] Preferably, according to the invention, the welding quality indicator WPQE is calculated according to the following ratio:

[0053] WPQE = |E−Ej| / Ej· 100 = |E−(β0+β1·xj)| / (β0+β1·xj) · 100 =

[0054] wherein:

[0055] WPQE is the welding quality indicator

[0056] E is the welding energy

[0057] Ej is the expected decay of welding energy as a function of a wear indicator of said first and / or second electrodes.

[0058] The meaning of the formula for calculating WPQE is immediately evident: the numerator expresses a difference between the ideal value of energy developed during the welding (E) and the real (post-decay) value of the energy (Ej) developed during the welding as a consequence of wear, whereas the denominator expresses the real (post-decay) value of the energy (Ej) developed during the welding as a consequence of wear. If the numerator is small, then the energy Ej developed in practice is very close to the ideally developed energy E. In this case, the welding is compliant. On the contrary, if the numerator is great, then the energy

[0059]

[0060] developed in practice is in some way different from the ideally developed energy E. In this case, the welding is non-compliant.

[0061] If this reasoning is applied to the comparison with the threshold value, it corresponds to declaring the welding compliant if WPQE is below the threshold value, and declaring the weld non-compliant if WPQE is equal to or higher than the threshold value.

[0062] Referring to Figure 2 and Figure 4, there will now be described a preferred embodiment of the invention -which has already been mentioned and described in the foregoing - and, specifically, the description will be set forth of the practical implementation thereof by using equipment and kits which are normally present in the welding lines for motor vehicle bodies.

[0063] Referring to Figure 1, the welding kit WK which is normally employed in a welding line for motor vehicles comprises the welding equipment WF, typically corresponding to an anthropomorphic robot having a welding clamp as end effector, wherein the welding clamp carries the electrodes T1 and T2.

[0064] The welding kit WK moreover comprises a PLC controller, identified by the reference CTR, which is operatively associated with the robot WF and which controls the movement and the operation of the same robot WF which carries the welding clamp. Another element which is operatively associated with the robot WF is the sampling device TC, which enables recording all the data coming from the welding kit WK, in order to store them in a mass memory device such as the hard disk of a computer dedicated to the welding kit or to the welding line as a whole.

[0065] In more detail, the data coming from the welding kit WK are preferably stored in the mass memory of a computer dedicated to the welding line as a whole, in a folder (named WeldLog) with. j son extension and with a sampling frequency equal to 1 ms. This frequency determines the number of i-th samples in the welding interval, in particular in the interval tD.

[0066] From the point of view of operation, the method according to the invention may be implemented in six operational steps, which are schematically described in the diagram of Figure 4. Said steps comprise:

[0067] - a step of acquiring data 2,

[0068] - a step of preparing data 4,

[0069] - a step of extracting data from the. j son file 6, - a step of calculating values from scalar data and curves 8

[0070] - a step of storing the threshold values for each welding spot 10

[0071] a step of calculating the welding quality indicator WPQE in real time 12.

[0072] Step 2: Acquiring data

[0073] It should be noted that not all files with. j son extension will have the same size, since the size thereof depends on the duration of constructing the single welding spot. If any fault is detected during the data recording of a welding spot in the. j son file, the software dedicated to recording is configured to repeat the execution of the faulty welding spot up to 3 times in a row. In this case, all the attempts carried out will be stored in the WeldLog folder, with a progressive file name for each of them. If it is not possible to carry out the welding spot in any way, the file which will be saved will be sent to another folder (containing faulty events) named FaultLog.

[0074] The data which do not exhibit faults (WeldLog folder) are further analysed to check whether the functions for calculating the welding current curves (Currentcurve) and the welding voltage curves (VoltageCurve ) have been filled in correctly.

[0075] Once this preliminary check is completed, a report is issued containing in detail the content of every. j son file which has been saved and given a time reference by means of the sampling device TC; said report includes both scalar parameters and curves.

[0076] The scalar parameters correspond to all the data of each file. j son which concern for example the name (or the unique identifier) of the welding spot, the date and the time when it was made, the number of electrodes which were used, the welding station (equipment WF) involved, the sampling device TC which supervised the making of that welding spot, the wear of the electrode, the possible occurrence of a material expulsion, and the like.

[0077] The curves correspond to all single samples (at the i-th instants) of acquisition of the values of the welding current Iiand the welding voltage Vi. From the parameters of welding current Iiand welding voltage Viit is possible, as described in the foregoing, to calculate the welding resistance Ri.

[0078] Preferably, the evaluation regards only the. j son files which do not contain faults, and in which there is consistency between the duration of the welding and the cumulative duration of the measurement samples of current and voltage. In other words, if the time duration of the acquisition of the current and voltage curves has the same duration of the making of the welding (preferably interval tD), then the file does not contain errors. Otherwise, the file will be discarded.

[0079] Step 4: Preparing data

[0080] Step 2 envisages, as described in the foregoing, a first step of filtering the acquired raw data, by discarding the. j son files which do not have the characteristics required for being analysed.

[0081] Once the. j son files have been selected, all the so-called VINs (Vehicle Identification Numbers), which are necessary for the trackability in production, are identified, in such a way as to be able to perform targeted correction interventions if the need arises.

[0082] This parameter is extracted from the set of the scalar data (weldSpotCustDataP16_1 and weldSpotCustDataP16_2 ) and converted into decimal code, since it is formed by two hexadecimal elements.

[0083] It is therefore possible to calculate the energy associated with the welding resistance curve for each i-th instant and for each welding spot according to the relation described in the foregoing (R = Vt / lt), by association with the VIN of the vehicle to which the welding spot belongs.

[0084] Moreover, this step envisages the possibility of retrieving, wherever possible, the VIN code for the. j son files which might have been lost or which might be corrupted during saving, in such a way as to increase the reliability of the general data package.

[0085] Said operation is possible by considering the time of transit of a body from the instant when it enters until it leaves the last welding station to proceed to another line. The VIN code is not saved when the robot WF carrying out the welding is managed by a less recently designed controller CTR, for example a Comau® C4G controller; on the contrary, the VIN code is saved in any case when the robot WF is managed by a more recently designed controller CTR, for example a Comau® C5G controller. In this fashion it is possible to compensate for the variability of equipment in a welding line, which may not always rely on welding robots of one and the same kind or having the same design.

[0086] This step comprises calculating the linear regression line described in the foregoing for each saved .json file, which describes the wear of the electrodes T1 and / or T2 as a function of the number of electrode dressings / reconditionings before it is replaced. For each linear regression line an extraction is made of the three characteristics which define the same: angular coefficient (β1), intercept (β0), residues. The residues correspond to differences between the observed values and the estimated values in a linear regression. The observed values which are located above the regression curve have a positive residue value, and the observed values that are below the regression curve have a negative residue value.

[0087] Once this calculation has been completed, it is possible to collect all the spatial coordinates of the welding spots. This amount of information, together with the type of welding spot (tack or full penetration), is contained within a back-up file which will be used to supply the calculation model developed.

[0088] Once this step of data collection has been completed, it is preferable to divide the spatial coordinates of the welding spots into three main groups, according to the longitudinal coordinate of the vehicle.

[0089] Ideally, the vehicle body is divided into three macro areas, and each macro area encompasses a given number of welding spots. This also indirectly influences the basic conditions of the welding clamps because, since they are defined with a technological approach, some parts will be more easily accessible than others. In this regard, it should be considered that in the designing phase a number of welding spots is defined for each welding clamp to manage, as a function of predetermined criteria, i. a. the accessibility of the welding spot. One of the conditions which must be met regards the fact that the robot carrying the welding clamp must not operate in conditions of singleness (thus it must not interfere with the structure to be welded). As a consequence, each welding spot must be accessible along the best possible trajectory, and each welding clamp must be sized as a function of the respective cycle time.

[0090] When optimizing the cycle time, with a consequent increase in the number of elements to be produced, the welding spots may have to be reordered and the clamps may consequently be located in more disadvantageous working positions, thus jeopardizing the stability thereof in time (with the risk of increasing the rate of faults in the components which are more prone to wear, or with the risk of elements such as the motors burning out more rapidly).

[0091] Once this step is completed, it is possible to compose a matrix of " Risk-Probability" correlation, which assigns a rating for each welding spot (so-called Quality Risk Matrix). An example of said matrix is shown in Figure 5.

[0092] The welding spots which historically have incurred in more faults and have been characterized by a higher number of material expulsions will obtain a higher rating, since the probability of the same event taking place again is much higher for another welding spot having the same technological welding characteristics with respect to a spot that (historically) has not had the same recording of events. It is to be noted, therefore, that the information contained in the matrix as per Figure 5 is updated only consequent to a drift noticed over a certain number of cycles.

[0093] Step 6: Extracting data from .json files

[0094] In this step, all data which are relevant for implementing the method according to the invention are extracted, without specifically limiting them to the data which can be obtained from the .json files, but extending the extraction also to all the data which can be obtained (for example by means of a combination) from the .json files themselves.

[0095] The list of the data extracted in the preferred implementation of the method according to the invention is set forth in the following, including in the list the name assigned to each data item.

[0096] i) Filename: name of the. j son file created during the making of the welding spot;

[0097] ii) progNo: number identifying the program executed by the sampling device TC during welding;

[0098] iii) spotName: identifier of the welding spot (name or number);

[0099] iv) timerName: code identifying the sampling device which has controlled the making of the welding spot; v) electrodeNo: number identifying the electrode which has been used;

[0100] vi) tipDressCounter: number of reconditioning events of the electrodes T1 and / or T2;

[0101] vii) uirExpulsionTime: i-th time instant at which a material expulsion is detected at the welding spot; viii) wear: it indicates the wear of the electrodes T1 and / or T2;

[0102] ix) weldSpotCustDataP16_1: hexadecimal code identifying the body whereon the welding is carried out;

[0103] x) weldSpotCustDataP16_2: hexadecimal code identifying the body whereon the welding is carried out;

[0104] xi) WeldTimeActualValue: actual duration of the welding (interval tD);

[0105] xii) weldTimeRefValue: reference duration of the welding.

[0106] Step 8: Calculating values from scalar data and curves

[0107] Among the further values which are extracted from the previous scalar values and by interpolating the information contained in the calculated welding resistance curve, it is possible to identify the following:

[0108] xiii) ΔWeldTime: difference WeldTimeActualValue - weldTimeRefValue;

[0109] xiv) WeldTimeError (WTE): coincidence between WeldTimeActualValue and WeldTimeRef erenceValue (either yes or no);

[0110] xv) R_max: maximum resistance detected at an i-th time instant after the intervals tA, tB, tC,

[0111] xvi) R_end: resistance value at the final instant (i = N, thus it may be the last value of the curve); xvii) TimeRmax: i-th time instant when Rmax is detected;

[0112] xviii) TimeRmax_after pause: i-th time instant when Rmax is detected after the intervals tA, tB, tC, and specifically after the end of the interval tCplus a following sampling instant;

[0113] xix) TimePrePulse: duration of the intervals tA, tB. xx) Pause: duration of the interval tC(zero welding resistance);

[0114] xxi) Leng (LEN): WeldTimeActualValue - TimePrePulse - Pause;

[0115]

[0116] Rend

[0117] xxv) Rexp: welding resistance value at the i-th time instant when a material expulsion occurs;

[0118] xxvi) Time_exp_after_Time_Rmax: i-th time instant when the material expulsion occurs after the instant Time_Rmax;

[0119] xxvii) Rmax after exp: welding resistance value R calculated after the expulsion (+ 20ms); xxviii) Time_Rmax_after_exp: i-th time instant when Time_Rmax is to be found after the material expulsion.

[0120] Step 10: Storing the threshold values of each welding spot

[0121] In this step it is important to store all the historical data which are necessary for a correct calculation of the threshold value for the indicator WPQE for each of the welding spots. Moreover, it is important to highlight that part of the historical information derives from reference values which are defined as "qualitatively acceptable" by an a posteriori control performed by a skilled line operator. In other words, the threshold values may derive from sample analyses performed with the purpose of statistically mapping the spot weldings.

[0122] The file created and stored contains all the information extracted from the single .json files, both directly (step 6) and indirectly (step 8 ).

[0123] Step 12: Calculating the welding quality indicator WPQE in real time

[0124] In this step the data which have been acquired, prepared and extracted in the previous steps converge into the calculation of the quality indicator WPQE based on the relation described in the foregoing, and in the comparison thereof with the threshold value (herein denoted with the reference MaxDevPerc, resulting from the product of a numerical coefficient, preferably equal to 8.4%, by the residual value determined by the previously calculated linear regression line) according to the fashion which will be better detailed in the following:

[0125] E - ( / ?o + Pi ‘xj) non compliant if > MaxDevPerc WPQE = ■ 100

[0126]

[0127] Po + Pi ■ Xj compliant if < MaxDevPerc At the end of this evaluation, a summarizing pie chart is prepared based on three indicators: consistency, conservativeness and error (which are all expressed in percentage).

[0128] Consistency: this condition is verified when the evaluation performed in the method according to the invention wholly satisfies the criteria defined by the operator, whether they be positive or negative. Specifically, consistency is evaluated by comparing the diagnosis resulting from the method according to the invention (either compliant or non-compliant ) with the diagnosis of an a posteriori ultrasonic quality control method, which is almost absolutely reliable. If the diagnoses coincide, the method according to the invention provides consistent results.

[0129] Conservativeness: this condition is verified when the evaluation performed by the method according to the invention deviates from the criteria which the operator considers acceptable. Specifically, this corresponds to a condition wherein the result of the method according to the invention is negative (non-compliant), whereas the result of the ultrasonic control is positive (compliant).

[0130] Error: this condition is verified when the evaluation performed by the model deviates from the criteria which are considered erroneous by the operator. It is the opposite of the condition of conservativeness: the result of the method according to the invention is positive (compliant), whereas the result of the ultrasonic control is negative (non-compliant). It is the most dangerous condition, indicating a false conformity.

[0131] In order that the deductions of the method according to the invention may be constantly updated as a function of the production needs of the plant, without slowing down the production cycle, a nightly update is envisaged of the regression line for calculating Ej, taking into consideration the latest 500 welding spots carried out.

[0132] Therefore, the latest welding spot carried out during the day of production is taken into consideration, and the update will be determined backwards based on the 500 previous spots, in a time interval which depends on the production volume. This calculation may be considered valid only if the .json file under consideration correctly contains all the curves, and if the length of the curve (be it a welding voltage or a current voltage curve) has the same length of the scalar parameter weldTimeActualValue.

[0133] Once the calculation of the linear regression line for calculating Ejis completed, 10% of the samples, those which are farthest from the regression line, are automatically identified and discarded. The remaining samples ( 90%) are used for performing a new calculation of the linear regression line, which is necessary for a real-time evaluation, since it corresponds to a modelling / mapping of the effects of the wear.

[0134] Of course, the implementation details and the embodiments may amply vary with respect to what has been described and illustrated herein, without departing from the extent of the present invention, as defined in the annexed claims.

Claims

CLAIMS1. A method for monitoring the quality of resistance spot welding, the resistance spot welding being carried out by means of welding equipment (WF) comprising a first electrode (Tl ) and a second electrode (T2 ) between which a welding current ( / ) flows during welding and between which, during welding, a welding voltage exists (V), the procedure including:- calculating, for each resistance welding spot, a welding energy value (E) as a function of the welding voltage (V) and the welding current ( / ) during the making of the welding spot,- calculating, for each resistance welding spot, an expected decay (£j) of welding energy as a function of a wear indicator (%y) of said first and / or second electrodes (Tl, T2 ),- determining a value of a welding quality indicator (WPQE) as a function of said welding energy value (E) and of said expected decay (£y) of welding energy, and - comparing said welding quality indicator (WPQE) with a threshold value (MaxDevPerc').

2. The method of claim 1, wherein said calculating of a welding energy value (E) as a function of the welding voltage (V) and the welding current ( / ) during the construction of the welding point for each resistance welding point is included in calculating the welding energy value (E) in accordance with the relationNi=lwhere:E is the value of the welding energyRi is the value of the welding resistance at an i-th time instant of a reference interval (tD) of welding time.

3. The method of claim 2, wherein each i-th time instant of the reference interval (tD) of welding timeis a sampled instant of said reference interval (tD) of welding time.

4. The method of to claim 2 or claim 3, wherein the value of the welding strength ( / ? at the i-th time is calculated as the ratio of the welding voltage (V at the i-th time to the welding current ( / at the i-th time instant.

5. The method of any of the preceding claims, wherein the calculation of an expected weld energy decay (Ej) for each resistance weld spot as a function of the wear indicator (%y) of said first and / or second electrodes (Tl, T2 ) includes calculating a reduced energy value as a function of a number of reconditionings (Xj) of said first and / or second electrodes (Tl, T2 ).

6. The method of claim 5, wherein said reduced welding energy value (£y) as a function of a number of reconditionings of said first and / or second electrodes is calculated according to the relationEj = Po + Pi ‘xjwhere:Ej corresponds to the reduced welding energy value Xj corresponds to a j -th occurrence of a reconditioning eventβ0, β1are two numerical coefficients.

7. The method according to any of the above claims, wherein said welding quality indicator (WPQE) is calculated according to the following ratio:WPQE = ■ 100wherein:WPQE is said welding quality indicatorE is said welding energyEj is said the expected decay of welding energy as a function of a wear indicator of said first and / or secondelectrodes.

8. The method of claim 7, wherein said comparing the welding quality indicator (WPQE) with a threshold value (MaxDevPerc) includes:- declaring the weld compliant if the welding quality indicator (WPQE) is below said threshold value (MaxDevPerc')- declaring the weld non-compliant if the welding quality indicator (WPQE) is higher than said threshold value (MaxDevPerc).

Images