“PROCEDURE FOR QUALITY CONTROL OF LASER WELDING”

Description

DESCRIPTION of the industrial invention entitled: “Procedure for quality control of laser welding” by: CRF Consortium Company, nationality Italian, Strada Torino 50, 10043 Orbassano (TO) Designated Inventors: PASQUETTAZ Giorgio Filed on: July 3, 2024 **** DESCRIPTION TEXT Field of invention The present invention relates to quality control of laser welding. In particular, the invention was developed with reference to a quality control by operate in real time. Known technique The execution of laser welding between two or more elements metals, in particular metal sheets in the context of an industrial assembly for mass production of car bodies, is susceptible to a result – in terms of quality – highly variable depending on the welding parameters and conditions (operating and environmental conditions) in which the welding itself develops. penetration depth of laser welding is, in this case sense, the parameter of the welded joint that is most susceptible of variations and oscillations. Currently a quality control on welded joints with laser technology it can be operated in a conventional way through sample analysis and downstream of the production of a batch of components welded with the same parameters, with the obvious consequence of a marked variability in the characteristics of the joint and / or a high quantity of components rejected because they are out of specification. This is not desirable neither in terms of quality nor efficiency of production, nor in terms of production costs. A more recently conceived alternative consists of systems quality monitoring based on complex algorithms and integrated with artificial intelligence solutions, which However, they require a learning phase (for the training of the algorithm) complex and articulated, with related costs and need for highly skilled personnel specialized. Purpose of the invention The purpose of the invention is to solve the problem previously described technical. In particular, the purpose of the invention is to provide a method for controlling of laser welding quality that can be implemented in real time during production so as to control and standardize the penetration depth of the weld and minimize the number of discarded welded components because out of specification and in general minimize variations in quality of the welded joint. Summary of the invention The purpose of the invention is achieved by a process having the characteristics forming the object of the claims that follow, which form part integral to the technical teaching administered here in relation to the invention. Brief description of the figures The invention will now be described with reference to the attached figures, provided purely as an example, do not limiting, where: - Figure 1 is a schematic view of a kit of devices for the implementation of the procedure according to the invention, and - Figure 2 illustrates a time diagram of signal processing according to the procedure in accordance with the invention. Detailed description With reference to figure 1, it illustrates schematically a set of devices and a relative laser welding equipment usable for the implementation of a process according to the invention. Laser welding equipment is identified globally from reference 1 and includes a head of welding 2 configured to operate a two or multiple elements P1, P2 at a welding area WA. The welding equipment 1 also includes, which measuring equipment for the implementation of the process according to the invention, a first photodetector 4 having a first operating band, a second photodetector 6 having a second operating band, and a third photodetector 8 having a third band of operation. The first photodetector 4, the second photodetector 6 and the third photodetector 8 are operationally associated with the welding head 2 and are configured to detect radiation electromagnetic coming from the WA welding area while performing laser welding for the making a welded joint. Each operating band, as is known, is definable with reference to a lower limit and a lower limit higher wavelength of radiation incident electromagnetic field with respect to which each photodetector is sensitive, and consequently gives rise to a voltage signal at the photodetector terminals same. In a preferred embodiment, the first photodetector is a silicon photodiode, the second photodetector is a germanium photodiode equipped with a filter optical for the acquisition of reflected laser radiation (cd back reflected light), and the third photodetector is a indium gallium arsenide photodiode, or InGaAs. The following description is valid for any photodetector 4, 6, 8 and for the specific selection including photodiodes to silicon (4), germanium (6) and InGaAs (8). Where some aspects of the invention are specific for silicon photodiodes (4), germanium (6) and InGaAs (8) this is expressly indicated. Still referring to figure 1, the photodetectors 4, 6, 8 are installed on an optical complex 10 which includes a semi-reflective mirror 12 – in turn including a deflector mirror 14, a first beam splitter 16 and a second beam splitter 18, a collimating lens 20 and a focusing lens 22. The radiation electromagnetic radiation emitted by welding – indicated by the EW reference in figure 1 - reaches the photodetector 8 / germanium photodiode for direct reflection by of the deflector mirror 14, as it reaches the photodetectors 4 and 6 / silicon and InGaAs photodiodes through, respectively, the beam splitter 14 and the beam splitter 16. The set of photodetectors 4, 6, 8 is operational associated with a 24-bit electronic processor equipped with a signal acquisition card of each photodetector, and a communication interface input / output to a laser source 26 that powers the welding head 2 (the reference L indicates the radiation laser emitted by the welding head 2). According to the invention, the process comprises: - perform laser welding to create a first welded joint at the WA welding area, - acquire a first S1 signal from the first photodetector 4, a second signal S2 from the second photodetector 6, and a third signal S3 from the third photodetector 8, each of called first signal S1, second signal S2 and third signal S3 being a consequence of the detection, by the corresponding photodetector 4, 6, 8, of a radiation electromagnetic coming from the WA welding area when making a welded joint, - inspect the first welded joint and check its conformity to one or more reference specifications, and elect the first joint welded to the reference joint if it is confirmed compliance with said one or more specifications reference; - elect each of the said first signal S1, second S2 signal and third S3 signal acquired with respect to the realization of the first welded joint chosen as a joint reference to first reference signal RS1, second RS2 reference signal, and third reference signal RS3 respectively, - acquire the first S1 signal from the first photodetector 4, the second signal S2 from the second photodetector 6, and the third signal S3 from the third photodetector 8 during the creation of one or more welded joints after the realization of the reference welded joint, - make a comparison for each welded joint made after the reference welded joint, between the first signal S1, the second signal S2, and the third signal S3 and the corresponding first reference signal RS1, second reference signal RS2, and third reference signal RS3 reference, - determine a welding power correction and / or a judgment of conformity to one or more specifications of reference of each welded joint made subsequently to the reference welded joint in function of the outcome of the said comparison. Each of the above steps will now be described in detail. The realization of the first welded joint, the inspection and the possible election of the same as reference joint for the implementation of the process according to the invention, is a complex of operations that allows you to eliminate the use of systems based on artificial intelligence and the related training by operators with related qualification in artificial intelligence and related fields training: inspection on the first welded joint is performed by a qualified inspection operator of welding, so it is a type of inspection traditional performed by normally present staff on the welding line. The operation has the sole purpose of determine whether the first welded joint complies with the expected specifications and whether, consequently, it can be taken as a reference, meaning by this to evaluate whether the signals S1, S2, S3 can be taken as reference. If the first welded joint was found not to comply with the specifications, it would be rejected and a new “first” welded joint to be subjected to the same inspection, electing as a reference for the implementation of the process according to the invention the “first” welded joint which meets the specifications in this sense, the phrase “first” welded joint is to be understood as welded joint upstream of the automated inspection sequence implemented according to the process of the invention. Once the reference welded joint has been selected, for the purpose of making a comparison of one or more welded joints subsequently made with it, the invention provides for define a first confidence interval B1 with respect to the first reference signal RS1, a second interval of B2 confidence with respect to the second reference signal RS2, and a third confidence interval B3 with respect to the third RS3 reference signal. A qualitative visualization of the confidence intervals B1, B2, B3 and of the signals of reference RS1, RS2, RS3 is given in figure 2, where the The time diagram illustrated therein is applicable (as qualitative indication) regardless of RS1 signals, RS2 and RS3 and their respective confidence intervals. The definition of the confidence intervals B1, B2, B3 therefore allows the comparison foreseen by the process according to the invention in terms of a comparison between each of the said first signal S1, second signal S2, and third signal S3 with the first interval of confidence interval B1, the second confidence interval B2 and the third confidence interval B3 respectively (therefore S1 is compared with B1, S2 with B2, S3 with B3). In more detail, in one form of execution preferred the procedure provides for determining, for the reference joint and for each of the said first signal reference signal RS1, second reference signal RS2 and third reference signal RS3 (each, preferably, subjected to block media mining operations and oversampling in anti-aliasing function at frequency of acquisition of 2048 Hz, in order to avoid complications and increase in costs, and also in order to avoid slowdowns in processing times, which are almost incompatible with (real-time processing): - an average value A1, A2, A3 (figure 2), - a standard deviation σ1, σ2, σ3 - a signal length L1, L2, L3 (expressed in time terms [s] or, multiplied by a speed of welding [mm / s], in linear terms [mm]), - a total energy of the signal E1, E2, E3, preferably calculated according to Parseval's theorem (sum of the squares of the signal samples). Based on these parameters the first interval of confidence interval B1, the second confidence interval B2 and the third confidence interval includes a limit upper UL1, UL2, UL3 corresponding to a sum (A1 + σ1; A2 + σ2, A3 + σ3) of the mean value A1, A2, A3 of the respective reference signal RS1, RS2, RS3 and deviation standard σ1, σ2, σ3 of the respective reference signal RS1, RS2, RS3, and a lower bound LL1, LL2, LL3 corresponding to a difference (A1 - σ1; σ2 – A2, σ3 – A3) between the average value of the respective reference signal RS1, RS2, RS3 and the standard deviation σ1, σ2, σ3 of the respective reference signal RS1, RS2, RS3. With this premise, and always on the subject of developing the signal, in the context of the aforementioned comparison the procedure according to the invention provides - for each welded joint including the reference joint and for each of the said first signal S1, second signal S2 and third signal S3 including the first reference signal RS1, the second RS2 reference signal and the third reference signal RS3: - split each signal S1, S2, S3 and signal reference RS1, RS2, RS3 in a number of intervals temporal TS (figure 2) determined as the product of a ratio between a laser welding length of the respective welded joint (expressed in [mm]) and a speed of welding (expressed in [mm] / [s]), and a frequency of acquisition of the corresponding photodetector 4, 6, 8 -1 (expressed in [s ]), - for each TS interval compare an average value of the first signal S1, of the second signal S2 and of the third S3 signal acquired on the joints made subsequently at the reference junction with the first confidence band B1, the second confidence band B2 and the third confidence band confidence B3, respectively; in other words, it is checked whether the average value of the signals coming from each photodetector 4, 6, 8 is or is not inside the band B1, B2, B3 of the reference signals RS1, RS2, RS3 (hence the role of the joint is even more evident reference), - assign a first indicator to the TS intervals for which the average value of the signal S1, S2, S3 is inside of the respective confidence band B1, B2, B3, of preference the value “1”, - assign a second indicator to the TS intervals for which the average value of the signal S1, S2, S3 is above outside the respective confidence bands B1, B2, B3, of preference the value “0” The subsequent deductions are based, substantially, on the occurrences of the values ​​“1” and “0” in relation to the total of occurrences. More specifically, determine a welding power correction and / or a judgment of compliance with one or more reference specifications such as provided for by the process according to the invention comprises determine a percentage of weld penetration laser in one or more welded joints made subsequently to the reference welded joint as a ratio between a number of occurrences of the first indicator (“1”) and a total number of occurrences (“1” and “0”) of the first and of the second indicator and multiplying the ratio by 100 between the number of occurrences of the first indicator (“1”) and the total number of occurrences (“1” and “0”) of the first and second indicator. This is because the intensity of the signals S1, S2, S3 (the same applies, of course – since it is the same type of signal – for signals RS1, RS2, RS3) is correlated with the penetration depth of the weld, therefore the number of samples for which the signals S1, S2, S3 are within the confidence band B1, B2, B3 (defined with reference to the signals RS1, RS2, RS3) is in direct correlation with the penetration rate of the welded joint. Processing of S1, S2, S3 signals and comparison with the RS1, RS2, RS3 signals according to the above-mentioned methods It also helps determine the need for corrections of welding power. In particular, the process according to the invention it provides for: - correct the welding power if at a range TS relating to each of the said first signal S1, second Signal S2 and third signal S3 are assigned simultaneously the second indicator (“0”), so if each of the signals in question has an average value outside the band of B1, B2, B3 confidence of the respective reference signal RS1, RS2, RS3, - maintain welding power if at an interval TS relating to only one of the said first signal S1, second signal S2 and third signal S3 is assigned the second indicator (“0”), therefore if only one of the signals in question has a mean value outside the confidence band B1, B2, B3 of the respective reference signal RS1, RS2, RS3. The the phrase "contemporaneously" is to be understood as referring at temporally aligned TS intervals, therefore at intervals TS with the same temporal position. This determination is repeated in real time for each time interval TS following the previous one. The power regulation is preferably performed in steps, in particular by commanding an increase or decrease of entity equal to 10% of the power level in force at the time where the increment or decrement is commanded. With the corresponding choice of photodetectors 4, 6, 8 to the devices listed above, therefore photodetector 4 = silicon photodiode, photodetector 6 = germanium photodiode and photodetector 8 = InGaAs photodiode, it is also possible to make further deductions on the subject of need for power regulation, for example: - correct the welding power if it is assigned at the same time the second indicator (“0”) to a TS interval related to the first S1 signal acquired by the silicon photodiode 4, and a TS interval relative to the third signal S3 acquired by the arsenide photodiode 8 indium-gallium InGaAs, - maintain welding power if assigned at the same time the second indicator (“0”) to a TS interval related to the second S2 signal acquired by the germanium 6 photodiode and a TS interval related to the first signal S1 acquired by silicon photodiode 4 or a TS interval related to the third S3 signal acquired by the Indium gallium arsenide (InGaAs) photodiode 8. Of new, the expression "contemporaneously" is to be understood as referred to temporally aligned TS intervals, therefore at TS intervals with the same temporal position. The difference compared to the previous eventuality (“0” for signals S1 and S3) lies in the fact that the photodiode at germanium 6 only acquires electromagnetic radiation at wavelength of light reflected from the weld (1064 nm – so-called “back reflected light”), therefore it has data density lower than the photodiodes 4, 8, which are configured to acquire electromagnetic radiation in a band of wider wavelengths, therefore they return higher data density information, and – lastly – more reliable. In this sense, the germanium photodiode 6 It is used essentially as a guide signal, but it is not sufficient in and of itself for a conclusive evaluation. With reference to figures 3A, 3B and 4A, 4B, in addition to the RS1, RS2, RS3 and S1, S2, S3 signal processing based on the temporal evolution of the signal amplitude acquired by each of the photodetectors 4, 6, 8, the procedure preferably involves further processing based on a temporal evolution of the frequency of the signal acquired by each of the photodetectors 4, 6, 8. The reference number 30 in figure 3A indicates in complex a timing diagram of the signal S1, S2, S3 (amplitude – time) acquired by one of the photodetectors 4, 6, 8 corresponding to a welded joint condition acceptable and in line with the specifications, while the number reference 32 indicates overall a time-diagram frequency of the same signal, therefore always in condition of acceptable weld joint and in line with the specifications (in this sense, diagrams 30, 32 are also indicative and especially of the RS1, RS2, RS3 signals). A single time-amplitude diagram and a single time- frequency only for illustrative reasons, but the implementation of the process according to the invention plans to carry out the same treatment described here follow-up on the signals of all photodetectors. In addition, the reference number 40 in figure 4A indicates overall a timing diagram of the signal S1, S2, S3 (amplitude – time) acquired by one of the photodetectors 4, 6, 8 corresponding to a non-welded joint condition acceptable and not in line with the specifications, while the reference number 42 indicates overall a diagram time-frequency of the same signal, therefore always in weld joint condition not acceptable and not in line with the specifications (in this sense, diagrams 40, 42 do not are indicative of the RS1, RS2, RS3 signals, as they refer to a condition opposite to that referred to by the signals RS1, RS2, RS3). On each of the two diagrams a trace of is visible lighter color associated with the reference fe corresponding, essentially, to a threshold frequency. The components of the signal acquired by each of the photodetectors 4, 6, 8 having a frequency above this threshold are indicative of the porosity of the weld, with entity substantially proportional (or in any case correlated) to the amplitude of the components in question. Experimentally it is observed that the threshold frequency f is equal to 300 Hz, and in general the frequencies of interest for an indication of weld porosity is in the range 300- 1000 Hz. The time-frequency distribution diagrams of the Figures 3B and 4B well describe the phenomenon in question: in case of figure 3B, with reference also to the scale of amplitude at the top right in the diagram of figure 3B, at above the frequency f the amplitudes are mainly concentrated in the area with negative values ​​on the scale width (almost uniform gray tint above the frequency f), while in the diagram of figure 4B there are many occurrences of amplitudes concentrated in the region with values positives of the amplitude scale (cf. the numerous bands white / whitish resulting in a haze effect “fading” of the diagram). For this reason, for each welded joint including the reference joint and for each of the said first signal S1, second signal S2 and third signal S3 including the first RS1 reference signal, the second reference signal RS2 and the third reference signal RS3 the procedure according to the invention preferably comprises: - obtain a time-frequency distribution of the signal RS1, RS2, RS3, S1, S2, S3 - divide the time-frequency distribution of each of the signals RS1, RS2, RS3, S1, S2, S3 in a number of time intervals TS determined as the product of a ratio between a laser welding length of the respective welded joint and a welding speed, and a acquisition frequency of the corresponding photodetector, - assign a first indicator (preferably “1”) to the TS intervals for which the amplitude of the signals (all) S1, S2, S3 in a frequency range above a threshold frequency, preferably equal to 300 Hz, and even more preferably in the range 300 – 1000 Hz, is less than an amplitude of the corresponding signal reference RS1, RS2, RS3, - assign a second indicator (preferably “0”) to the TS intervals for which the amplitude of the signals (all) S1, S2, S3 in a frequency range above a threshold frequency, preferably equal to 300 Hz, and even more preferably in the range 300 – 1000 Hz, is greater than an amplitude of the corresponding signal reference RS1, RS2, RS3, - determine a percentage of porosity of the laser welding in one or more additional welded joints in function of a ratio between a number of occurrences of the first indicator and a number of total occurrences of said first indicator and second indicator (so like the ratio in question multiplied by 100). Once the penetration percentage has been determined and the percentage of porosity, it is possible to operate further deductions, in particular it is possible to determine a non conformity of the welded joint if at least one of the following occurs: following conditions: - a time extension of each signal S1, S2, S3 is less than 90% of a time extension of the corresponding reference signal RS1, RS2, RS3, - total energy of each signal S1, S2, S3 is lower total energy of the corresponding signal reference RS1, RS2, RS3 when the time extension of each signal S1, S2, S3 is greater than 90% of the extension time of the corresponding reference signal RS1, RS2, RS3 - the percentage of porosity is greater than 2% when the time span of each signal S1, S2, S3 is greater than 90% of the time span of the corresponding reference signal RS1, RS2, RS3, and the total energy of each signal S1, S2, S3 is greater than to the total energy of the corresponding reference signal RS1, RS2, RS3, - the penetration percentage is less than 80% when the time span of each signal S1, S2, S3 is greater than 90% of the time span of the corresponding reference signal RS1, RS2, RS3, and the total energy of each signal S1, S2, S3 is greater than to the total energy of the corresponding reference signal RS1, RS2, RS3, the process further comprising determining a need for further verification if at least one occurs of the following conditions: - the penetration percentage is between 80% and 90% when the time span of each signal S1, S2, S3 is greater than 90% of the time span of the corresponding reference signal RS1, RS2, RS3, and the total energy of each signal S1, S2, S3 is greater than to the total energy of the corresponding reference signal RS1, RS2, RS3. Thanks to the process according to the invention it is therefore quality control of laser welding is possible real time (all operations described are performed in real time thanks to simplicity and efficiency of the development of the process according to the invention) during production by controlling and standardizing it weld penetration depth (via welding power regulation), minimizing the number of welded joints rejected because they were out of specification and in general minimizing the variations in quality of the welded joint itself without the limitations of the traditional manual random checks, and without the costs and hassles incumbent on control procedures based on artificial intelligence. Of course, the manufacturing details and shapes of execution may be widely varied with respect to as described and illustrated without going beyond from the scope of this invention as defined from the attached claims.

Claims

CLAIMS 1. A method for quality control of laser welding, comprising: - providing laser welding equipment (1) including a welding head (2) configured to perform a welding of two or more elements (P1, P2) at a welding area (WA), - providing a first photodetector (4) having a first operating band, a second photodetector (6) having a second operating band, and a third photodetector (8) having a third operating band, the first photodetector (4), the second photodetector (6) and the third photodetector (8) being operatively associated with said welding head (2) and being configured to detect electromagnetic radiation coming from said welding area (WA) during the execution of a laser welding for the realization of a welded joint,the process comprising: - performing laser welding to produce a first welded joint at the welding area (WA), - acquiring a first signal (S1) from said first photodetector (4), a second signal (S2) from said second photodetector (6), and a third signal (S3) from said third photodetector (8), each of said first signal (S1), second signal (S2) and third signal (S3) being a consequence of the detection, by the corresponding photodetector (4, 6, 8), of electromagnetic radiation coming from said welding area (WA) during the production of the welded joint, - inspecting the first welded joint and verifying its conformity to one or more reference specifications, and electing the first welded joint as the reference joint if conformity to said one or more reference specifications is confirmed, - electing each of said first signal (S1), second signal (S2) and third signal (S3),acquired with respect to the realization of the first welded joint elected as reference welded joint, first reference signal (RS1), second reference signal (RS2), and third reference signal (RS3) respectively, - acquire the first signal (S1) from said first photodetector (4), the second signal (S2) from said second photodetector (6), and the third signal (S3) from said third photodetector during the realization of one or more welded joints subsequent to the realization of the reference welded joint, - make a comparison, for each welded joint realized subsequent to the reference welded joint, between the first signal (S1), the second signal (S2), and the third signal (S3) and the corresponding first reference signal (RS1), second reference signal (RS2), and the third reference signal (RS3),- determine a welding power correction and / or a judgment of conformity to one or more reference specifications of each welded joint made after the reference welded joint based on the outcome of said comparison.

2. A method according to claim 1, wherein said first photodetector (4) is a silicon (Si) photodiode, said second photodetector (6) is a germanium (Ge) photodiode, and said third photodetector (8) is an indium gallium arsenide (InGaAs) photodiode.

3. A method according to claim 1 or claim 2, comprising defining a first confidence interval (B1) with respect to said first reference signal (RS1), a second confidence interval (B2) with respect to said second reference signal (RS2), and a third confidence interval (B3) with respect to said third reference signal (RS3), and wherein said performing a comparison comprises comparing each of said first signal (S1), second signal (S2), and third signal (S3), with the first confidence interval (B1), the second confidence interval (B2), and the third confidence interval (B3) respectively.

4. A method according to any of the preceding claims, comprising determining, for the reference weld joint and for each of said first reference signal (S1), second reference signal (S2) and third reference signal (S3): - an average value (A1, A2, A3), - a standard deviation (σ1, σ2, σ3), - a signal length (L1, L2, L3), - a total signal energy (E1, E2, E3).

5. Method according to claim 4, wherein said first confidence interval (B1), second confidence interval (B2) and third confidence interval (B3) comprise an upper limit (UL1, UL2, UL3) corresponding to a sum of the mean value of the respective signal and the standard deviation of the respective signal (A1 + σ1, A2 + σ2, A3 + σ3), and a lower limit (LL1, LL2, LL3) corresponding to a difference between the mean value of the respective signal and the standard deviation of the respective signal (A1 - σ1, A2 - σ2, A3 - σ3).

6. A method according to claim 5, - 19 comprising, for each welded joint including the reference welded joint and for each of said first signal (S1), second signal (S2) and third signal (S3) including the first reference signal (RS1), the second reference signal (RS2) and the third reference signal (RS3): - dividing each signal (S1, S2, S3) and reference signal (RS1, RS2, RS3) into a number of time intervals (TS) determined as the product of a ratio between a laser welding length of the respective welded joint and a welding speed, and an acquisition frequency of the corresponding photodetector (4, 6, 8), - for each time interval (TS) comparing an average value of the first signal (S1), the second signal (S2) and the third signal (S3) with the first confidence band (B1), the second confidence band (B2) and the third confidence band (B3), respectively,- assign a first indicator (“1”) to the time intervals (TS) for which the mean value (A1, A2, A3) of the signal (S1, S2, S3) is within the confidence band (B1, B2, B3) of the corresponding reference signal (RS1, RS2, RS3), - assign a second indicator (“0”) to the intervals for which the mean value (A1, A2, A3) of the signal (S1, S2, S3) is outside the confidence band (B1, B2, B3) of the corresponding reference signal (RS1, RS2, RS3), wherein said determination of a welding power correction and / or a judgement of conformity to one or more reference specifications comprises determining as a function of a ratio between a number of occurrences of said first indicator (“1”) and a total number of occurrences of said first indicator (“1”) and second indicator (“0”).

7. A method according to claim 6, - 20 comprising determining a percentage of laser welding penetration in said one or more further welded joints as said ratio between a number of occurrences of said first indicator (1) and a number of total occurrences of said first indicator (1) and second indicator (0), multiplied by 100.

8. Method according to claim 6 or claim 7, comprising: - correcting the welding power if the second indicator (0) is simultaneously assigned to a time interval (TS) relating to each of said first signal (S1), second signal (S2) and third signal (S3), - maintaining the welding power if the second indicator (0) is simultaneously assigned to a time interval (TS) relating to only one of said first signal (S1), second signal (S2) and third signal (S3), 9. Method according to claim 8, comprising: - correcting the welding power if the second indicator (0) is simultaneously assigned to a time interval (TS) relating to the first signal (S1) acquired by the silicon photodiode (4), and to a time interval (TS) relating to the third signal (S3) acquired by the indium gallium arsenide (InGaAs) photodiode (8), - maintaining the welding power if the second indicator (0) is simultaneously assigned to a time interval (TS) relating to the second signal (S2) acquired by the germanium photodiode (6), and to an interval (TS) relating to the first signal S1 acquired by the silicon photodiode (4) or to a time interval (TS) relating to the third signal (S3) acquired by the indium gallium arsenide (InGaAs) photodiode (8).

10. A method according to claim 6, comprising, for each welded joint including the reference joint and for each of said first signal, second signal and third signal including the first reference signal, the second reference signal and the third reference signal: - obtaining a time-frequency distribution of each of said first, second and third signals (S1, S2, S3) and first, second and third reference signals (RS1, RS2, RS3), - dividing the time-frequency distribution of each of said first, second and third signals (S1, S2, S3) and first, second and third reference signals (RS1, RS2, RS3) into a number of time intervals (TS) determined as the product of a ratio between a laser welding length of the respective welded joint and a welding speed, and an acquisition frequency of the corresponding photodetector,- assign a first indicator (“1”) to the time intervals (TS) for which the amplitude of the first, second and third signals (S1, S2, S3) in a frequency range above a threshold frequency, preferably equal to 300 Hz, and even more preferably in the range 300 - 1000 Hz, is less than an amplitude of the corresponding first, second and third reference signals (RS1, RS2, RS3), - assign a second indicator (preferably “0”) to the time intervals (TS) for which the amplitude of the first, second and third signals (S1, S2, S3) in a frequency range above a threshold frequency, preferably equal to 300 Hz, and even more preferably in the range 300 - 1000 Hz, is greater than an amplitude of the corresponding first, second and third reference signals (RS1, RS2, RS3),- determining a percentage of laser weld porosity in the one or more welded joints made subsequent to the reference welded joint as a function of a ratio between a number of occurrences of the first indicator (“1”) and a total number of occurrences of said first indicator (1) and second indicator (“0”), the process further comprising determining a non-conformity of the welded joint if at least one of the following conditions occurs: - a time extension of each of said first, second and third signals (S1, S2, S3) is less than 90% of a time extension of the corresponding first, second and third reference signals (RS1, RS2, RS3), - a total energy of each of said first, second and third signals (S1, S2, S3) is less than a total energy of the corresponding first, second and third reference signals (RS1, RS2, RS3) when the time extension of each of said first, second and third signals (S1, S2, S3) is less than 90% of a time extension of the corresponding first, second and third reference signals (RS1, RS2, RS3),S3) is greater than 90% of the temporal extension of the corresponding first, second and third reference signals (RS1, RS2, RS3), - the percentage of porosity is greater than 2% when the temporal extension of each of said first, second and third signals (S1, S2, S3) is greater than 90% of the temporal extension of the corresponding first, second and third reference signals (RS1, RS2, RS3), and the total energy of each of said first, second and third signals (S1, S2, S3) is less than the total energy of the corresponding first, second and third reference signals (RS1, RS2, RS3), - the percentage of penetration is less than 80% when the temporal extension of each of said first, second and third signals (S1, S2, S3) is greater than 90% of the temporal extension of the corresponding first, second and third reference signals (RS1, RS2, RS3), and the total energy of each of said first, second and third signals (S1, S2,S3) is less than the total energy of the corresponding first, second and third reference signals (RS1, RS2, RS3), the method further comprising determining a need for further verification if at least one of the following conditions occurs: - the penetration percentage is between 80% and 90% when the time span of each of said first, second and third signals (S1, S2, S3) is greater than 90% of the time span of the corresponding first, second and third reference signals (RS1, RS2, RS3), and the total energy of each of said first, second and third signals (S1, S2, S3) is less than the total energy of the corresponding first, second and third reference signals (RS1, RS2, RS3).,