Improvements in and relating to welding and quality control

Abstract

The present invention provides a welding method and apparatus for carrying out the method, which employs inspection of a weld formed by the welding method. The inspection of the weld is performed by an inspection apparatus at an inspection location on one or more welded substrates, the inspection location being at an elevated temperature above ambient temperature, for example, the elevated temperature being at least 180° C. above ambient temperature, or at least 350° C. above ambient temperature. In this manner, the inspection may be performed without the need to cool the substrates, achieving faster inspection and faster weld repair times.

Description

[Technical field]

[0001] The present disclosure relates to improvements in welding, quality control in welding, and their use in non-destructive ultrasonic testing, particularly but not exclusively to their high temperature deployment.

[0002] Ultrasonic testing is used for non-destructive testing of a variety of objects. A transmitting transducer emits ultrasonic waves which enter the object, interact with the object and its sub-features, and return to the receiving transducer. Efficient transmission of the ultrasonic waves to the object is important to avoid high levels of reflection at interfaces. Liquid couplers are often used to improve transmission at interfaces.

[0003] When testing welds, it is desirable to be able to perform ultrasonic testing immediately after the weld or weld pass or weld layer is formed, either after the weld is completed or after each individual weld pass or weld layer. This minimizes the time it takes to perform, test, and make any corrections required for the weld or weld pass or weld layer. However, liquid couplers such as water have definite limitations on the temperature of the object they can operate at, so in such cases the object must be cooled to some degree before testing. If corrections are required, the object must be brought up to temperature again.

[0004] One potential objective of the present disclosure is to provide a method of welding and quality verification that performs the verification at an elevated temperature without the need for first cooling. One potential objective of the present disclosure is to provide a method of welding and verification of welding conditions that can be performed in real time and provide further control of the welding process.

[0005] According to a first aspect of the present disclosure, there is provided a welding method, the welding method including an inspection of a weld formed by the welding method, the welding method comprising: (i) providing a welding device; (ii) providing a weld inspection device; (iii) introducing one or more substrates to be welded into a welding apparatus; (iv) using heat to elevate the temperature of the one or more substrates above ambient temperature; (v) performing welding of the one or more substrates at an elevated temperature above ambient temperature using a welding device; (vi) inspecting the produced weld using a weld inspection device; Including, The inspection of the weld is performed at one or more test locations on the substrate with an inspection device at an elevated temperature above ambient temperature; A welding method is provided.

[0006] The method may be arc welding.

[0007] The method may include inspection of the entire weld. The method may include inspection of a multi-pass weld, for example, after each pass and before the next pass, before all passes are completed.

[0008] The welding device can be mounted on any autonomous self-deployment system. The welding device can be mounted on a guide rail system or a column and boom system. The welding device can be mounted on a robotic arm, such as a multi-axis arm. The welding device can be an arc welder.

[0009] The weld inspection device may use ultrasound. The weld inspection device may be mounted on any autonomous self-deployed system. The weld inspection device may be mounted on a guide rail system or a column and boom system. The weld inspection device may be mounted on a robotic arm such as a multi-axis arm. The weld inspection device may be a phased array ultrasonic transducer. The weld inspection device may include an ultrasonic emitter and a receiver. The weld inspection device may include a substrate contacting surface. The weld inspection device may be in physical contact with the inspection location. The substrate contacting surface may be in physical contact with the inspection location.

[0010] The method may include rolling a weld inspection device, e.g., a substrate contacting surface thereof, over a surface of the substrate, e.g., to one side, but optionally parallel to the weld.

[0011] The weld inspection apparatus may be provided with an internal cooling system and the method may include providing and / or removing coolant to and / or from the weld inspection apparatus.

[0012] The method may provide that the inspection of the weld occurs at an elevated temperature above ambient temperature, the elevated temperature being at least 80° C. above ambient temperature. The method may provide that the inspection of the weld occurs at an elevated temperature above ambient temperature, the elevated temperature being at least 180° C. above ambient temperature. The method may provide that the inspection of the weld occurs at an elevated temperature above ambient temperature, the elevated temperature being at least 250° C. above ambient temperature. The method may provide that the inspection of the weld occurs at an elevated temperature above ambient temperature, the elevated temperature being at least 300° C. above ambient temperature. The method may provide that the inspection of the weld occurs at an elevated temperature above ambient temperature, the elevated temperature being at least 350° C. above ambient temperature. The method may provide that the ambient temperature is 20° C.+ / -10° C.

[0013] The method may provide that inspection of the weld is performed without time for the one or more substrates to cool below an elevated temperature, and may provide that inspection of the weld is performed with heat (e.g., preheat) applied to the one or more substrates.

[0014] The method may provide that the duration of the inspection of the weld is performed for a period of time equivalent to the period during which the weld is performed. The inspection period of the weld may begin after the welding period. The inspection period of the weld may overlap with the welding period, but the overlap may be at least 50%, or at least 80%, or at least 90% of the inspection period of the weld.

[0015] The method may provide that welding is performed at the weld location when inspection is performed at the inspection location, the weld location and the inspection location are less than 2 m apart along the weld. The method may provide that welding is performed at the weld location when inspection is performed at the inspection location, the weld location and the inspection location are less than 1 meter apart along the weld. The method may provide that welding is performed at the weld location when inspection is performed at the inspection location, the weld location and the inspection location are less than 0.5 m apart along the weld. The method may provide that welding is performed at the weld location when inspection is performed at the inspection location, the weld location and the inspection location are less than 0.25 m apart along the weld. The method may provide that welding is performed at the weld location when inspection is performed at the inspection location, the weld location and the inspection location are less than 0.15 m apart, in some cases 0.1 m apart along the weld.

[0016] The method may provide that welding is performed at the welding location when inspection is performed at the inspection location, and the inspection location was the welding location less than 10 minutes ago. The method may provide that welding is performed at the welding location when inspection is performed at the inspection location, and the inspection location was the welding location less than 5 minutes ago. The method may provide that welding is performed at the welding location when inspection is performed at the inspection location, and the inspection location was the welding location less than 3 minutes ago.

[0017] The method may provide that welding is performed at a weld location and that the temperature at the weld location during welding is 800°C.

[0018] The method may provide that by raising the temperature of the one or more substrates above ambient temperature, the temperature of the one or more substrates is at least 110°C. The method may provide that by raising the temperature of the one or more substrates above ambient temperature, the temperature of the one or more substrates is at least 180°C. The method may provide that by raising the temperature of the one or more substrates above ambient temperature, the temperature of the one or more substrates is at least 250°C. The method may provide that by raising the temperature of the one or more substrates above ambient temperature, the temperature of the one or more substrates is at least 300°C. The method may provide that by raising the temperature of the one or more substrates above ambient temperature, the temperature of the one or more substrates is at least 350°C. The method may provide that by pre-heating the one or more substrates prior to initiating welding, the temperature of the one or more substrates is raised above ambient temperature.

[0019] The method may provide for detection of defect locations to within 15 mm, in some cases within 10 mm, in some cases within 5 mm, or even within 1 mm of the actual defect location.

[0020] The method may provide that the characteristics include one or more of a shape of the weld and / or a size relative to the length of the weld, a position, a type of defect, a shape of the defect, or a location of the defect.

[0021] The method may provide for performing one or more repair steps on the defective location of the weld if the inspection indicates a defect at the defective location of the weld. One of the repair steps may be removing a portion of the weld, e.g., excavating a portion of the weld. One of the repair steps may be performing a further inspection at the inspection location where the repair step was performed, e.g., after removing the portion of the weld. The further inspection may reveal that the defect has been removed or that a further repair step is required, such as removing a further portion of the weld. For example, one of the repair steps after it is determined that the defect has been removed may be rewelding. The method may include rewelding at the location of the defect and / or where the portion of the weld was removed.

[0022] The method may provide for providing a step of repairing when the defect location is at an elevated temperature above ambient temperature. The method may provide for said elevated temperature to be at least 80° C. above ambient temperature, optionally at least 180° C. above ambient temperature, optionally at least 250° C. above ambient temperature. The method may provide for said elevated temperature to be at least 300° C. above ambient temperature, optionally at least 325° C. above ambient temperature, optionally at least 350° C. above ambient temperature. The method may provide for all repair steps to be performed when the defect location is at an elevated temperature above ambient temperature.

[0023] According to a second aspect of the present disclosure, there is provided an apparatus for performing a welding method including inspection of a weld formed by the welding method on a substrate, the apparatus comprising: (i) a welding device; (ii) a weld inspection apparatus, the weld inspection apparatus including an ultrasonic transmitter and a receiver, the weld inspection apparatus being provided with a substrate contact surface, the substrate contact surface having a melting point greater than 250°C; An apparatus is provided comprising:

[0024] The weld inspection device may use ultrasonics. The weld inspection device may be mounted on any autonomous self-deployed system. The weld inspection device may be mounted on a guide rail system or a column and boom system. The weld inspection device may be mounted on a robotic arm, such as a multi-axis arm. The weld inspection device may be a phased array ultrasonic transducer. The weld inspection device may include an ultrasonic emitter and a receiver. The weld inspection device may include a substrate contact surface.

[0025] The weld inspection device can be configured to be in physical contact with the inspection location. The substrate contact surface can be configured to be in physical contact with the inspection location.

[0026] The weld inspection device, e.g., its substrate contacting surface, can be adapted to roll over a surface of the substrate, e.g., to one side of the weld, and possibly parallel to the weld.

[0027] The weld inspection apparatus may be provided with an internal cooling system. The ultrasonic transmitter and receiver may be provided within the weld inspection apparatus with the internal cooling system. The weld inspection apparatus may be provided with an inlet for coolant to flow into the weld inspection apparatus and / or an outlet for coolant to flow out of the weld inspection apparatus.

[0028] This first aspect of the disclosure may include any of the features, possibilities, or additional matters described elsewhere in the document, including other aspects of the disclosure.

[0029] According to a third aspect of the present disclosure, there is provided a welding method, comprising: (a) providing a welding device; (b) providing a plurality of sensor types; (c) defining a first set of welding conditions for the welding method; (d) introducing one or more substrates to be welded into a welding apparatus; (e) performing welding of one or more substrates; (f) acquiring data from a plurality of sensor types during welding; (g) comparing the acquired data from the plurality of sensor types with reference data for one or more sensor types; (h) determining whether the weld is of acceptable or unacceptable quality based on the one or more comparisons; Including, (i) A method of welding is provided that includes performing one or more actions if the quality of the weld is unacceptable.

[0030] A third aspect of the present disclosure may include any of the features, capabilities, or additions described in the first and / or second aspects of the present disclosure relating to inspection of welds, including other aspects of the disclosure.

[0031] The welding method may further include inspection of a weld formed by the welding method. (i) providing a weld inspection device; (ii) using heat to elevate the temperature of one or more substrates above ambient temperature; (iii) performing welding of the one or more substrates at an elevated temperature above ambient temperature using a welding device; (iv) inspecting the produced weld using a weld inspection device; It may include, (v) Inspection of the weld is performed at an inspection location on one or more substrates with an inspection device at an elevated temperature above ambient temperature.

[0032] The method may be arc welding.

[0033] The method may include inspection of the entire weld. The method may include inspection of a multi-pass weld, for example, after each pass and before the next pass, before all passes are completed.

[0034] The welding device can be mounted on any autonomous self-deployment system. The welding device can be mounted on a guide rail system or a column and boom system. The welding device can be mounted on a robotic arm, such as a multi-axis arm. The welding device can be an arc welder.

[0035] The weld inspection device may use ultrasound. The weld inspection device may be mounted on any autonomous self-deployed system. The weld inspection device may be mounted on a guide rail system or a column and boom system. The weld inspection device may be mounted on a robotic arm such as a multi-axis arm. The weld inspection device may be a phased array ultrasonic transducer. The weld inspection device may include an ultrasonic emitter and a receiver. The weld inspection device may include a substrate contacting surface. The weld inspection device may be in physical contact with the inspection location. The substrate contacting surface may be in physical contact with the inspection location.

[0036] The method may include rolling a weld inspection device, e.g., a substrate-contacting surface thereof, over a surface of the substrate, e.g., against one side, but optionally parallel to the weld.

[0037] The weld inspection apparatus may be provided with an internal cooling system and the method may include providing and / or removing coolant to and / or from the weld inspection apparatus.

[0038] The method may further specify that at least one sensor type of the plurality of sensor types is part of a weld inspection device, the method including inspecting the weld using the weld inspection device. The method may include inspecting the weld using the weld inspection device to determine one or more characteristics of the defect. The method may specify that the characteristics include one or more of a shape of the weld and / or a size relative to a length of the weld, a location, a type of defect, a shape of the defect, or a location of the defect.

[0039] The method may further include comparing the one or more characteristics to one or more criteria, and may further include determining whether the defective weld meets or does not meet the welding criteria.

[0040] The method may provide that if the weld meets the welding criteria, a record of the weld is created and stored and may include the location of the defect relative to the geometry of the weld and / or the location of the defect relative to the length of the weld. The method may further provide that the record includes data from one or more of a plurality of sensor types. The method may provide that if the weld does not meet the welding criteria, one or more repair steps are applied to the weld.

[0041] The method may provide that at least two sensor types of the plurality of sensor types are weld condition sensors, the method including inspecting the weld condition using the weld condition sensors.

[0042] The method may provide for including inspecting the weld as it is formed to determine one or more parameters of the weld.

[0043] The method may provide for including a comparison of the one or more parameters to one or more control parameters, and further including a determination of whether a risk level for a weld defect has been exceeded.

[0044] The method may provide for including one or more actions of changing the welding conditions from a first set of welding conditions of the welding method. The method may provide for changing the welding conditions from the first set of welding conditions to include stopping welding and / or alerting an operator. The method may provide for changing the welding conditions from the first set of welding conditions to include changing the welding conditions back to the first set of welding conditions and / or changing the welding conditions to a second set of welding conditions.

[0045] The method may provide that at least two of the sensor types include being selected from a voltage sensor, a current sensor, a welding arc sound emission sensor, a welding topology sensor, a welding image sensor, and an ultrasonic image sensor.

[0046] This third aspect of the disclosure may include any of the features, capabilities, or options described elsewhere in the document, including other aspects of the disclosure.

[0047] According to a fourth aspect of the present disclosure, 1. An apparatus for monitoring welding, comprising: (a) multiple sensor types; (b) a control unit for receiving a first set of welding conditions for the welding method; (c) a comparator that receives and compares the data from the plurality of sensor types with reference data for one or more sensor types; An apparatus is provided for monitoring a weld, wherein the comparator outputs a determination of whether the weld is of acceptable or unacceptable quality based on the compared data, and if the determination results in the weld being of unacceptable quality, the apparatus further provides a control signal for triggering one or more actions by the apparatus.

[0048] The apparatus may provide for including a weld inspection device for determining one or more characteristics of the defect. The apparatus may further provide a comparator including a first comparator for receiving and comparing the one or more characteristics to one or more criteria, the first comparator outputting a first determination as to whether the weld having the defect meets or does not meet the welding criteria. The apparatus may further provide a comparator including a second comparator for receiving and comparing one or more parameters of the weld as the weld is formed to one or more control parameters, the second comparator outputting a second determination as to whether a risk level of the weld defect has been exceeded. The apparatus may provide such that if a risk level of the weld defect is exceeded, a control signal provided by the apparatus is sent to a controller to trigger one or more actions, the one or more actions being to change the welding conditions from a first set of welding conditions of the welding method, e.g., to stop welding, and / or to alert an operator.

[0049] This fourth aspect of the disclosure may include any of the features, capabilities, or options described elsewhere in the document, including other aspects of the disclosure.

[0050] Various embodiments of the present disclosure will now be described, by way of example only, with reference to the accompanying drawings, in which: [Brief description of the drawings]

[0051] [Figure 1] FIG. 1 is a schematic diagram of a process sequence according to an example of the prior art. [Diagram 2] FIG. 1 is a perspective view of a preferred environment for use of the present disclosure. [Diagram 3] FIG. 2 is a schematic diagram of a process sequence according to the present disclosure. [Figure 4] FIG. 1 is a schematic diagram of an adaptive control function for a welding process provided by the present disclosure. [Diagram 5] FIG. 2 is a perspective view of a probe according to a first embodiment of the present disclosure. [Figure 6] FIG. 6 is a plan view of the probe of FIG. 5. [Figure 7a] FIG. 7 is a cross-sectional side view of the probe of FIG. 6 taken along plane BB. [Figure 7b] FIG. 7b is the same view as FIG. 7a, but with further features shown. [Figure 8a] FIG. 2 is a perspective view of a probe and a transmission block. [Figure 8b] FIG. 8b is a top view of the probe and transmission block of FIG. 8a. [Figure 8c] FIG. 8b is a side view of the probe and transmission block of FIG. 8a. [Figure 8d] FIG. 13 is a detail of a side view of the top of the interface between the probe and the transmission block. [Figure 8e] FIG. 8d is a top view detail of the same location as in FIG. 8d. [Figure 9] FIG. 13 is a view of the underside of the transmission block facing the compliant element. [Figure 10] FIG. 1 is a schematic diagram of a cooling circuit for use in a probe according to the present disclosure. [Figure 11] 11 is a comparative plot of temperature of an actively cooled probe versus a fixed cooling volume probe in accordance with the present disclosure. [Figure 12a] 1 is an image of a defect detected by an ultrasonic probe. [Figure 12b] FIG. 12b is an image of the ultrasound probe used for detection in FIG. 12a. [Figure 13] 1 is a plot of outlier scores obtained from acoustic signal v data points for various welding conditions. [Figure 14a] FIG. 2 is a perspective view of a profile sensing device relative to a substrate and a weld. [Figure 14b] FIG. 1 is a schematic diagram of the overall weld formation sequence and shape through a series of weld passes. [Figure 15] 1 is a sequence of camera images of a weld site during welding. [Figure 16] 1 is a plot of arc voltage and a plot of Gaussian amplitude x Gaussian center v time of a welding process. [Figure 17] FIG. 13 is a diagram of combined data types displayed to a user. [Figure 18]FIG. 13 is a diagram of a second level of processing applied to data from multiple sensor types.

[0052] Ultrasonic testing is used for non-destructive testing of a variety of test pieces. A transmitting transducer emits ultrasonic waves which enter the test piece, interact with the test piece and its sub-features, and return to the receiving transducer. It is important to effectively transmit the ultrasonic waves to the test piece to avoid high levels of reflection at interfaces. Liquid couplers are often used to improve transmission at interfaces.

[0053] When testing a weld, it is desirable to be able to perform ultrasonic testing immediately after the weld is made in order to minimize the time it takes to perform, test, and correct any corrections required to the weld once the weld is complete. However, liquid couplers such as water have definite limitations on the temperature at which the test piece can be operated, so in such cases it may be necessary to cool the test piece to or near ambient temperature before testing.

[0054] FIG. 1 provides a schematic of a typical conventional process sequence. This schematic shows the process sequence beginning at the point where the weld is completed. Typically, this sequence is applied after all passes have occurred and the weld is complete, as this is when the total cooling time occurs. For multi-pass welds, this may be performed after each weld pass, but is very time consuming. Allowing the object to cool between weld passes for inspection is also undesirable, as repeated cycles of heating, cooling, heating, cooling may have undesirable effects on the object's microstructure.

[0055] In the process sequence on the left, the weld is completed and the preheat that is applied to the substrate to be welded in the first step is turned off. This preheat was used to bring the substrate up to an optimal temperature for welding and to maintain that temperature during welding. Once the preheat is turned off, the second step is for the substrate to cool to the environmental temperature surrounding the substrate (usually room temperature). For substrates that are particularly susceptible to hydrogen cracking, a further 24 hour delay may be built in before testing can be performed. Once the substrate has cooled to a temperature low enough to perform non-destructive testing, NDT, the third step may be performed to inspect the entire weld. In this case, no defects are detected, the weld passes inspection, and the process is complete. As can be seen from the time required for each step of the process, the substrate needs to be cooled to a relatively low temperature in order to perform NDT. Performing NDT before it has cooled sufficiently may cause irreparable damage to the ultrasonic-based NDT device used.

[0056] In the process on the right side of Figure 1, the first three steps are performed as before and take the same amount of time to perform. In this case, the third step identifies a defect in the weld, so a repair action is necessary.

[0057] During the repair phase, the first repair step is to drill the weld to reach the defect. The time required depends on the depth and size of the defect. For example, if the defect is in the initial weld pass, a longer period will be required and multiple subsequent passes will need to be drilled. After drilling, a further NDT cross-check is performed as a second repair step to ensure the defect has been found and completely drilled. The base material is now ready for rewelding.

[0058] During the rewelding phase, the first rewelding step involves turning on preheat to gradually raise the temperature of the substrate to the optimum welding temperature, and then the second rewelding step involves further welding to repair the weld.

[0059] The process sequence on the right side may then return to the NDT stage and that step, turn off the preheat, cool to ambient temperature, wait another 24 hours, and perform the NDT. If no defects are present, the weld proceeds to the next stage. If new defects are found, the repair step must be repeated.

[0060] As can be seen from the process step times on the right, defect correction can add significant time to the process, making it less efficient.

[0061] The ability of ultrasonic probes to operate properly at high temperatures means that the probes can be used closer in time and distance to the weld location, including between individual weld passes, reducing the time delay between when a weld is formed at one location and when a weld problem is detected at that location.

[0062] This improvement allows for faster inspection to detect defects and cracks at the point of production, allowing repair work to begin sooner. For example, defects may be detected without completing all passes or waiting for the object to cool. Defects may be corrected sooner because there is less delay than bringing the object down to measurement temperature and then back up to welding temperature. Additionally, much less material needs to be excavated to reach the defect, greatly reducing the scope and time of the excavation process. These steps increase throughput, minimize risk, and reduce costs.

[0063] This improvement may facilitate greater closed loop control and automation to minimize the occurrence of defects and cracks in the welded components. By monitoring conditions and adapting the control of the welding equipment, the system may maintain optimal conditions.

[0064] The ability of ultrasonic probes to operate normally at high temperatures allows for significantly different welding and NDT sequences to be performed. The ability of ultrasonic probes to operate normally at high temperatures allows for significantly different welding, NDT and subsequent repair sequences to be performed. Details of suitable embodiments of the probes are provided at the end of this document.

[0065] With reference to Figure 2, it is shown how an object 1 made of substrate 2 is being welded at a welding location 3 by a welding device 5. The multi-axis robot 7 is also provided with an arm 9 having a probe 13, shown diagrammatically at its distal end 11. The probe 13 is in contact with a test location 15, which was previously the welding location. The physical distance D and the temporal distance are likewise very short and both act simultaneously. The weld and substrate cooling takes place from the welding temperature to or towards the preheat temperature of the substrate 2.

[0066] Figure 3 shows a schematic diagram of the process sequence according to the present disclosure. With the substrate 2 preheated to an optimum welding temperature, welding is performed at weld location 3 while NDT is performed at test location 15. The NDT is performed without any cooling time for the substrate 2, so no testing time is added to the overall process time, except for a small time increment from when the weld reaches its final location until the test reaches its final location shortly thereafter. As shown in Figure 2, this is a relatively short distance and therefore short time.

[0067] The top part of Figure 3 shows the process sequence when a defect is detected. In the sequence, repair steps are applied. In the repair steps, the first repair step is to drill the weld to reach the defect. The time required depends on the depth of the defect and the size of the defect. After drilling, a further NDT cross check is performed. This is the second repair step to ensure that the defect has been found and completely drilled out. The base material is now ready for rewelding.

[0068] Because rewelding does not require power down for preheating and no cooling is required, rewelding may begin immediately. Therefore, a second rewelding step to repair the weld with an additional weld may begin without delay. An additional NDE may be performed on the rewelded location to verify that no new defects exist and that the old defects have been completely repaired.

[0069] Once all welds have been completed, subjected to NDT, and defects repaired, the weld is complete and the substrate may simply enter a cooling sequence. In the first cooling step, the pre-heat applied to the substrate to be welded is turned off. Once the pre-heat is turned off, the substrate is cooled in a second cooling step to the temperature of the environment surrounding the substrate (usually room temperature). This completes the process, but a final NDT step may be performed.

[0070] The step times and total times of the process in Figure 3 clearly show that significant time savings and therefore increased efficiency have been achieved compared to Figure 1 and its process sequence. Not only are the need for various steps avoided, but some steps, such as the weld drilling step where only the last weld pass needs to be drilled, are themselves shorter if they occur at all.

[0071] Since both the multi-axis robot 7 with the arm 9 and the probe 13 and the multi-axis robot for the welding device 5 are controlled autonomously, a very wide range of welding and NDT situations can be accurately performed and tested without user intervention.

[0072] While the above disclosure relates to shortening the time between welding and NDT so that defects can be repaired more quickly, this disclosure also reduces the risk of defect formation by improving initial weld quality, by expanding the scope of data sets about the weld that are collected, analyzed, and used while the weld is in progress.

[0073] In the illustrated disclosure, a combination of four different sensor types are used to collect data and provide adaptive control of the weld to mitigate the possibility of defect formation.

[0074] Referring to FIG. 4, a schematic diagram of the adaptive control function is shown.

[0075] In the welding process, on the left side of Figure 4, the NDT step described above [Post-weld high temperature NDT] is performed using a high temperature capable roller type ultrasonic probe [described in more detail below], which provides ultrasonic data [NDT display] that is fed into an overall assessment of weld qualification marking the end of the process.

[0076] In the welding process, a series of steps are performed in parallel, starting with process monitoring, which aims to prevent or minimize the occurrence of defects as well as to monitor the weld produced and prompt repair actions thereon. In the welding process, a series of variables are defined and controlled that influence and control the weld. These may include wire feed speed, voltage, current, welding speed, separation of welding device and welding location, shielding gas, shielding gas flow rate, substrate shape / profile / configuration / dimensions, substrate preheat temperature, weld groove shape / profile / configuration / dimensions, weld or weld pass shape / profile / configuration / dimensions.

[0077] In the process monitoring step, data regarding the time variations of the voltage and current applied to the welding device 5 is measured and collected.

[0078] At this point, data from the five types of sensors are fed into data streams to form sensor data. These are current, voltage, laser, vision, and acoustic data streams from appropriate sensors as described in more detail below. In this embodiment, sensing of the weld profile using laser scanning, visual assessment of the weld, and acoustic sensing of sounds emanating from the weld are also used to address weld formation corrections or issues in weld formation corrections. Voltage and current sensing are used in embodiments where weld formation corrections or issues in weld formation corrections are addressed.

[0079] Once data is collected from all sensors and the necessary data is obtained, the data is sent to a data reconciliation step. Data reconciliation is based on input provided through the user interface and / or historical data obtained from storage. The storage may contain initial data for this welding run, data for many previous welds performed by the system, and / or data from other systems (e.g., calibration processes, etc.), all of which may contribute to the historical data and therefore to the data reconciliation.

[0080] By applying data reduction and / or machine learning, live analysis of the data may be obtained and a copy may be sent to storage for future use or may contribute to the knowledge of the system and / or similar systems.

[0081] An important step in the analysis results is to compare the location of the data to one or more thresholds set for that type of data (threshold setting). The location relative to the threshold is considered to be indicative of a defect (defect indicator) and can be an important part of the weld quality results. The weld quality results can be displayed to a user via a user display, which also receives and displays the received sensor data. If the weld quality results step detects an unacceptable weld, it can stop the welding process and / or issue a warning to the operator, such as via a user display.

[0082] Based on the position of the user display, the operator or the system itself may adjust one or more variable parameters used to control and perform the welding process in real time.

[0083] The weld quality results are reflected in a Weld Qualification which is provided for the entire weld. If a defect is identified by NDT, the size and location of the defect are checked against appropriate criteria to ensure that the level of the defect is within tolerance. If not acceptable, remedial action is taken. If acceptable, 3D life record data is stored within the Weld Qualification and made available for the life of the weld and for any necessary decommissioning support.

[0084] (Information on Ultrasonic Probes for High Temperature Substrates) 5 is a perspective view of an embodiment of the probe 13. An axis of rotation R-R extends through the probe 13. A first mounting location 20 is provided on the axis, together with a second mounting location 22 on the other side of the probe 13. The first mounting location 20 and second mounting location 22 provide attachment of the probe 13 to the robot 7 such that the probe 13 can rotate as it moves over the surface of the object 1.

[0085] The probe 13 has rigid end structures 24a, 24b at each end with bolts 25 connecting them and annular mounting 26 for a generally cylindrical coupling element 28.

[0086] A manifold element 30 is connected to and extends axially from the first mounting location 20. The manifold element 30 has a coolant inlet 32 ​​which, in use, is connected to a coolant supply conduit [not shown], and a coolant outlet 34 which, in use, is connected to a coolant outlet conduit [not shown].

[0087] FIG. 6 is a cross-sectional plan view of the probe of FIG. 5, which also shows many of the features described above.

[0088] 7a, which is a cross-sectional side view of the probe of FIG. 5, manifold element 30 is fluidly connected to first mounting location 20. A first fluid connection is made by first inlet bore section 36 connecting to second inlet bore section 38 leading via conduit 40 to an interior volume 42 of probe 13. A second fluid connection is made by first outlet bore section 44 connecting to a second outlet bore section (not shown) leading from interior volume 42 of probe 13.

[0089] The interior volume 42 extends between opposing wall sections 46a, 46b of the coupling element 28 and also extends between opposing interior surfaces 48a, 48b of the rigid end structures 24a, 24b. The interior volume 42 is filled with coolant to at least a level above a maximum vertical extent 50 of the transducer 52.

[0090] The joining element 28 is a unitary piece of compatible material, as will be further described below. The joining element 28 has a generally right cylindrical body portion 54 with inwardly turned rims 56a, 56b at either end. The rims 56a, 56b are each compressed between the outer element 24a and the inner element 24b to form a rigid end structure 24a. The outer element 24a and the opposing inner element 24b are connected to one another by a series of releasable fasteners, in this case bolts 25.

[0091] Thus, as the probe 13 rolls over the surface of an object, different portions of the body portion 54 of the coupling element 28 contact the object, causing the outer element 24a and inner element 24b, which form the rigid end structure, to also rotate.

[0092] The rigid end structures 24a, 24b are free to rotate relative to the shaft element 64, with its extension providing the first mounting location 20. A first shaft-type seal 66 and a second shaft-type seal 68 allow rotation while sealing against coolant leakage between the shaft element 64 and the rigid end structures 24.

[0093] The second shaft element 70 is connected to the shaft element 64 by a series of releasable fasteners 72. The second shaft element 70 serves as a mounting for the transducer 52, an anti-echo block 74, and an ultrasound transmission block 76.

[0094] Thus, the transducer 52, anti-echo block 74, and transmission block 76, together with the second axis element 70, axis element 64, and first mounting location 20, do not rotate as the probe 13 rolls over the surface of the object. Thus, the transducer 70 and associated components always maintain the same sensing orientation facing the object.

[0095] As a result of the above arrangement, there is relative movement between the inner surface 78 of the coupling element 28 and the radial surface 80 of the transmission block 76 as the probe 13 rolls over the surface of the object.

[0096] Also attached to the second shaft element 70 is a mounting element 82 carrying a thermistor 84 for sensing the temperature of the interior roller probe area at a location 86 proximate the portion of the interior surface 78 adjacent the weld location.

[0097] As shown in Figures 8a to 8e, alternative embodiments of the probe 13, transducer 52, and transmission block 76 may be provided. As shown in Figure 8a, the transducer 52 is attached to an inclined axially facing surface 88 of the transmission block 76 by fasteners 92. The transmission block 76 is provided with a coolant inlet 93 and a coolant outlet 94. As shown in Figure 8c, the coolant inlet 93 leads to a serpentine passageway 95 leading to a coolant supply outlet 96 in fluid communication with the interior volume 42 of the probe 13. A similar structure on the other side of the transmission block 76 extracts the coolant from the interior volume and directs it through a coolant removal inlet into a second serpentine passageway and to the coolant outlet 94. These passageways and the configuration of these passageways aid in cooling the transmission block 76.

[0098] The axial surface 88 of the transfer block 76 and / or the radial surface 90 of the transducer 52 may have gaps, slots, or grooves to aid in the flow of coolant between the two surfaces.

[0099] For example, referring to Figures 8c and 8d, the inclined axial surface 88 has an intersection with a second axial surface 97 that is generally parallel to the axis of rotation. The transducer 52 terminates near the intersection. As shown in detail in Figure 8d, a channel 98, such as a groove, is provided in the transmission block 76. The channel 98 improves coolant access to the gap 92 between the axially facing surface 88 of the transmission block 76 and the radially facing surface 90 of the transducer 52. The channel 98 is in fluid communication with a series of further grooves 99 in the axially facing surface 88 of the transmission block 76, further facilitating the flow of coolant to the gap 92. The channel 98 and / or the further channels 99 may be provided in the transducer 52 and / or the transmission block 76.

[0100] As shown in the detailed views of Figures 8b and 8e, the channel 98 extends across the width of the conductor block 76 from one side to the other. Other channels 99 are spaced apart along the channel 98 and extend away from the channel 98 along the inclined axially facing surface 88.

[0101] Another potential detail is that the leading edge 100 and trailing edge 102 of the transfer block 76, considered in relation to rotation as the probe 13 moves in the direction A, are each provided with a chamfer. Thus, as the transfer block 76 effectively moves through the coolant during rotation, the coolant is encouraged by the chamfer of the leading edge 100 towards the gap 108 between the radially facing surface 80 of the transfer block 76 and the inner surface 78 of the coupling element 28. This encourages a continuous presence of coolant between the radial surface 80 and the inner surface 78, which is highly beneficial to the passage of ultrasonic waves across the interface between the transfer block 76 and the coupling element 28. The continuous presence of coolant also aids in cooling the radial surface 80.

[0102] The separation between the rotation axis RR and the object 1 to be probed is such that, in use, the object 1 pushes the coupling element 28 towards the axis, so that the coupling element 28 is compressed between the object 1 and the transmission block 76 and makes good contact with the transmission block 76.

[0103] In another potential detail, as seen in Figure 8a, the axial face 88 of the transfer block 76 has a greater extent along the axis and perpendicular to the axis than the radial face 90 of the transducer 52. As the transfer block 76 and transducer 52 move through the coolant due to rotation of the probe 13, this shape encourages the coolant to flow to the junction of the transfer block 76 and transducer 52 and into the gap 92 between them. The flow direction of the coolant entering the device along the conduit 42 also encourages flow toward the gap 92.

[0104] In yet another potential detail shown in Figure 9, the radially facing surface 80 of the transmission block 76 is provided with a series of interface channels 300. These are recessed in the curved radially facing surface 80 which, in use, faces the inner surface 78 of the coupling element 28. The interface channels 300 extend the entire length of the transmission block 76 and run parallel to the axis, although other extents and profiles may be provided.

[0105] The continued presence of coolant in the gap 92 between the axial face 88 of the transmission block 76 and the radial face 90 of the transducer 52 is highly beneficial to the passage of ultrasonic waves across the interface therebetween.

[0106] With respect to the passage of ultrasonic waves, the transmission block 76 is fabricated from polyetherimide, which provides the necessary heat resistance and capacity to handle repeated cycles of temperature change, and the material provides the necessary acoustic properties to balance with the other components.

[0107] For monitoring purposes, the thermistor 84 is provided in a separate block of polyetherimide offset to the side of the transmission block 76 so as not to interfere with the transmission block's role in transmitting ultrasound. At the same time, the location of the thermistor 84 is still effective to ensure that temperature limits of the components are not approached, in which case the probe 13 may be removed from the object to prevent damage to the components. In other embodiments, the block providing the thermistor 84 may be attached to the transmission block 76, and in other embodiments the thermistor 84 may be incorporated into the transmission block 76.

[0108] With regard to the passage of ultrasound, the coupling element 28 is silicone rubber, which is compatible with higher temperatures. The material selected is capable of withstanding temperatures in excess of 350° C. for extended periods of time. Such a material may have an attenuation of 0.87 dB / mm at 5 Mhz and an acoustic impedance of 1.12 MRayls, providing a good match with other materials used.

[0109] The thickness of the coupling element 28 is a balance between the increased thermal insulation from the contents of the probe that increases with thickness and the unfavorable increase in attenuation that increases with thickness. For the operating conditions under consideration, a thickness of 4 mm to 8 mm is suitable for such material.

[0110] The material selected for the coupling element 28 also provides sufficient conformability to conform to the surface of the object under moderate applied force levels. High force levels are undesirable in view of the equipment required to generate them and move the device over the test specimen. The surfaces of objects encountered in the real world are not highly finished or smooth, so a conforming material is necessary to obtain good contact for transmitting ultrasound without excessive loss.

[0111] With respect to the passage of ultrasound, the anti-echo block 74 plays an important role in preventing ultrasound from bouncing around within the probe and causing noise or other adverse effects on the probe. Hydrogenated nitrile rubber (HNBR), especially its N-filler form, was found to be a suitable material, as it provides an attenuation of 6.4 dB / mm at 5 MHz.

[0112] All these features aid in successful acoustic coupling of the probe to the object through the surfaces encountered in reality.

[0113] A cooling circuit for the probe 13 is used to withstand the rise in surface temperature of the object.

[0114] A schematic diagram of the coolant circuit of one embodiment is shown in Figure 10. A coolant reservoir 100 delivers coolant via a conduit 102 to a pump 104, which delivers coolant via a second conduit 106 to a coolant inlet 32 ​​in the probe 13.

[0115] Within the internal volume 42, the coolant is allowed to circulate freely within the entire volume of that internal volume, including around the transducer 52, around the transmission block 76 and the gap therebetween, around at least the lower periphery of the coupling element 28, and in the gap 92 between the radially facing surface 80 of the coupling element 28 and the transmission block 76.

[0116] Returning to Figure 10, from the interior volume 42, the coolant is discharged through the cooling fluid outlet 34 into a third conduit 108. Temperature sensors in the third conduit and / or in the interior volume 40 can be used to ensure that the cooling is as desired, and in some cases the pump speed can be controlled to increase or decrease the cooling. The third conduit 108 leads to a heat exchanger 110 which cools the coolant so that it can be reused. A fourth conduit 112 removes the cooled coolant from the heat exchanger 110 and returns it to the reservoir 100 for reuse.

[0117] The use of active cooling for the probe 13 and its components allows the probe 13 to be used for extended periods on hot object surfaces.

[0118] 11, a first temperature plot 200 and a second temperature plot 202 are shown versus time that the probe is in contact with a hot object. The first temperature plot 200 is for a probe 13 with a circulating coolant in accordance with the present invention, and shows that this approach successfully maintains the internal temperature of the probe 13 within operational limits. The active cooling ensures that the transducer remains below the maximum operating temperature of 55-60°C that applies to most transducers of the required type.

[0119] The second temperature plot 202 is for a probe with similar internal components, but with a fixed volume of coolant that is restricted to an amount sealed within the internal volume of the probe. As heat is transferred to and builds up within the probe, the temperature obviously increases over time, and after a relatively short time the temperature exceeds a reliable operating threshold of 50°C. In practice, such a probe would need to be removed from the object before it reached the 50°C threshold, and monitoring could not take place until the probe itself had cooled.

[0120] In terms of coolants, air has low heat capacity and thermal conductivity for active cooling. Water is also suboptimal as its acoustic impedance of 1.5 MRayls does not match with the other components. Providing the coolant in the form of a water-soluble oil, for example, with an acoustic impedance of 1.1 MRayls, would better match the acoustic impedance of the transmission block [around 1.1 MRayls].

[0121] The transducer 52 provides a 64 element phased array at 5 Mhz and is mounted to generate ultrasonic waves at 55° to the object. A 0.5 mm pitch and 10 mm elevation angle may be used. The angled beam is useful in that it allows the weld to be fully inspected from laterally spaced locations. Often the laterally spaced locations provide better contact between the probe and the object than the location where the weld is being made. For example, a multi-pass weld creates a large recess that prevents good contact and ultrasonic propagation to the object until the weld is complete. This is problematic with a 0° or low angle based approach.

[0122] This type of transducer and transmission block configuration can be used to provide a fan-shaped scan beam defined by an emitted top beam (angled away from normal to the transducer face) and a bottom beam (near normal to the transducer face).

[0123] Regarding the high temperature performance required of the probe, the present disclosure provides a probe capable of inspecting objects at temperatures of about 300° C. for an extended period of time.

[0124] The coupling is dry but achieves the level required to propagate ultrasound through the interface into and out of the object.

[0125] The high temperature polymers used in the coupling components are capable of withstanding prolonged contact with objects at such temperatures and allowing successful propagation of ultrasonic waves to and from the interface.

[0126] The coolant and the gap between the coolant-filled sensor-type elements may also allow for efficient propagation of ultrasonic waves to and from the transmission block.

[0127] Since the transmission block is only exposed to temperatures close to the ambient temperature, there is no need to select a high temperature resistant material with poor ultrasonic propagation characteristics, providing optimal propagation characteristics for the transport block.

[0128] (Information on using multiple sensor types)

[0129] (Conditions and effects of welding location) In arc welding, a power source is used to create a sufficient voltage difference between the electrode of the welding device and the substrates to be welded to create an arc, which results in an electric current. The arc heats the substrates to a molten state [potentially consuming the electrode]. Upon cooling, the molten metal solidifies and joins the two substrates.

[0130] The speed at which the welding device moves relative to the substrates affects the degree of melting, the shape of the weld pool, etc., which affects the quality of the weld.

[0131] The welding site is usually protected by a shielding gas to prevent atmospheric oxygen, water, or water vapor from reaching the weld site. Shielding gases are usually inert or semi-inert gases such as argon or helium. The flow rate of the shielding gas affects its ability to perform its function.

[0132] To ensure high weld quality, many operating variables must be carefully controlled within the welding process. These variables may be taken into account indirectly, as illustrated in the sections below. Although not directly sensed in the illustrated embodiment, other types of sensors may measure the speed of travel of the welding device, the flow of the shielding gas, and the flow rate of the shielding gas, and these sensor types and their data sets may be added to the processing.

[0133] (acoustic sensor type) The acoustic sensor type collects high frequency audio signals generated during welding. These signals arise from the arc forming the weld, the arc-substrate interaction, and the arc-shielding gas interaction. The detected audio signals were found to be sensitive to several key variables within the welding process. Figure 13 shows example values ​​of outlier scores for various weld characteristics.

[0134] The outlier score is obtained by a mathematical approach that considers how far a particular data value is from a set of known data values ​​that have been classified as acceptable values ​​for the characteristic and / or sensor type data being evaluated.

[0135] One such approach used in this disclosure is the use of a Mahalambis distance novelty detection model by comparing the input audio signal values ​​with previously developed principal component analysis (PCA) models of sensor types and their signal and / or data values.

[0136] Signal processing involves taking the audio signal and applying a denoising algorithm to the raw data. The audio signal is further processed using a short-time Fourier transform to convert the raw time series data into the frequency domain. Statistical features are then extracted from each of a series of bandwidths spanning the frequency range of interest. In this example, the bandwidth used is 39.1 kHz, generating 312 features that describe the acoustic signal of the particular signal instance. The entire feature set [312 features per signal instance] is then optimised by removing redundant features and standardised to improve robustness of features with small standard deviations.

[0137] In the initial setup of the PCA model, the signal resulting from the above processing, and therefore the remaining feature set, is set to be the acceptable performance of the weld with respect to that variable (in this case the acoustic signal). Thus, the remaining feature set may finally be reduced using PCA to obtain the principal components and model that define that model with respect to the acceptable performance of the weld.

[0138] The PCA model of acceptable performance may be used to consider subsequent signals. These subsequent signals are subject to the same noise cancellation and other steps defined above. For each feature, the location of the feature value relative to the distribution of feature values ​​of acceptable weld performance identified, i.e., the feature value, is determined. The mean of the distribution is calculated and the distance of the feature value relative to the mean is measured. The distance provides a quantification of the outliers, shown in FIG. 13, revealing distance variations consistent with acceptable weld conditions versus distance variations that are more exceptional and indicative of poor weld performance.

[0139] This approach is useful for providing a unit-free, scale-invariant quantification that takes into account correlation of feature values ​​within a distribution. The mean can be recalculated each time an acceptable feature value is added to the distribution and / or PCA model, or it can be calculated based on a fixed set of existing acceptable feature values ​​used in the PCA model.

[0140] Referring to sample results obtained using this process and displayed in FIG. 13, the first set of data points indicate that good welding operation parameters are occurring. These parameters are verified individually as they are applied. As can be seen, they provide a nicely clustered data point set A against the log-scale outlier score axis.

[0141] The second set of data points indicates that the welding speed is too fast, i.e. the welding device and substrate are moving too fast relative to each other. Two different welding speed deviations are shown, which are well-spaced with few outliers in two sets of data points B. 1 and B. 2 Gives.

[0142] The third set of data points shows sidewall arcing during the weld, i.e. the arc is shorting to the sidewall instead of to the intended weld location in the weld groove. Again, these give a well-consolidated data point set C with few outliers.

[0143] The fourth set of data points shows sidewall melting, i.e. the arc melts the sidewall and causes melting at the sidewall and not in the weld groove. The data points are well clustered with few outliers as data point D.

[0144] The fifth set of data points indicates that the shielding gas flow rate is too high, which could potentially lead to undesirable porosity issues. Like the other data point sets, this set E is well defined with few outliers. A similar location would be found if the shielding gas flow rate was too low.

[0145] When the aforementioned undesirable welding conditions are present, the outliers or characteristic values ​​are much higher than when good welding conditions are present. As a result, the threshold value Th acoustic A Mahalanobis distance selection value can be set and used to distinguish between data obtained from an acoustic sensor type that indicates good welding conditions and data obtained from an acoustic sensor type that indicates defective welding conditions. Thus, the acoustic data type is selected based on the threshold Th acoustic The system has the ability to alert the operation or trigger a stop of welding if the defect is violated or remains violated for a specified number of data points. This provides real-time acoustic signal based identification of defect creation. This analysis can be provided continuously and for each pass of a multi-pass welding approach.

[0146] Importantly, the algorithm used is more complex than just using a single Principal Component Analysis (PCA) model previously developed for the sensor type and its signal and / or data values. As shown in Figure 14b, in multi-pass welding, the weld passes gradually fill the weld groove. This means that the weld groove depth, weld groove shape, and fill amount change from pass to pass. All these changes, and possibly other changes between passes, affect the acoustic signal emitted and detected. Therefore, this approach uses a separate model for each pass to most accurately evaluate the observed data against the predicted data for that pass.

[0147] The individual models can be obtained by a neural network approach that trains them based on existing paths, or they can learn to improve the model for a particular path in a sequence of paths as the paths are performed during operation and the number of paths increases.

[0148] The use of separate models for each pass also applies to the use of separate models for re-passes, for example as part of a weld repair operation: the geometry and therefore the acoustics of the weld groove in such a re-pass will be significantly different from a normal pass.

[0149] (Laser sensor type) The next sensor type used is visual and is intended to evaluate the geometry of the weld created.

[0150] FIG. 14a shows a cross section of a substrate 2 with a weld 20 formed thereon. In this example, the weld 20 is linear, but other weld paths are conceivable as well. The visual sensor type device 22 includes a casing 24 with a light source 26 therein that can illuminate the substrate 2 and the weld 20 over an illumination width 28. The device 22 has a working range 30 that allows for accurate imaging. The light returns to the device 22 where a receiver 32 focuses the light onto a sensor matrix 34 and a signal is generated. In this example, a 2D laser profile scanner is used, but other types could be substituted.

[0151] A laser as light source 26 is used to reveal various details of the weld and its surroundings, such as the weld profile, the remaining groove profile, the weld bead width, and the material deposited outside the weld groove. The inspection may use a plane perpendicular to the substrate surface beside the weld groove and perpendicular to the longitudinal axis of the weld groove. Additionally, the weld bead profile along the weld groove may also be considered.

[0152] Figure 14b is a diagram of a typical weld groove over the course of a series of weld passes. Each weld pass adds welds to the welds already in the groove in a predetermined order [numbered] to build up the overall weld. As can be seen, the weld passes contribute to a predictable shape of the weld pass itself and a predictable change in the shape of the weld groove if the weld is progressing correctly.

[0153] The signal from device 22 may be used to form a profile image of the entire weld track at each position along the weld track. The actual profile may be compared to an expected profile and deviations may be recorded. The deviations may be measured using a threshold Th. profile Thus, the profile data can be compared to a threshold value Th profile The ability to alert the operator and / or trigger a stop of welding if the profile signal is violated or remains violated for a specified number of data points. This can provide identification of defect creation based on real-time profile signals. This analysis can be provided continuously and for each pass of a multi-pass welding approach. Effective geometric validation is provided.

[0154] (visual sensor type) The next sensing area uses a high dynamic range camera to capture images of the weld location, including where the weld has not yet formed, where the weld is forming, and where the weld is solidifying and further cooling.

[0155] Figure 15 shows a series of images of this type collected from various welding locations. By processing these using a combination of artificial intelligence and traditional machine vision tools, the system can detect visual anomalies caused by changes in the images or defects associated with abnormal welding conditions. This can be accomplished by processing a single image or by combining and processing multiple images from previous data, both from recent images and from past parts and passes.

[0156] In this area, one or more variables may be considered. For example, the size of the weld pool [width, trailing length, leading length], the shape of the weld pool [oval, teardrop shaped, etc.], the temperature of the weld pool may be considered along with an analysis of the pattern, shape of the deposited material, and visible anomalies. A profile of known decision parameters for various types of defects may be used to compare the live output from each part of the algorithm to known good values. These values ​​may be a combination of the presence or absence of a particular visual feature, a numerical band or threshold, or a classification. If the image is determined to be beyond these parameters, the occurring situation for one or more or all may be compared to a desired position for one or more or all, and the deviation may again be used to trigger an alert or to stop the weld.

[0157] (Voltage and current sensor type) The applied voltage affects the formation of the arc and the current in the arc. This in turn affects the power and therefore the rate of dissolution of the base material [and electrode, if consumed]. These are important variables for the quality of the weld. These are the variables that affect, for example, the size of the weld pool.

[0158] The welding voltage also needs to be automatically and continuously adjusted to reflect the distance between the welding equipment and the substrate being welded, based on a known, fixed position of the substrate and a variable but known XYZ position of the robotic arm carrying the welding device, i.e. the separation of these two positions.

[0159] The sensing system for monitoring voltage and current is orders of magnitude faster than those built into existing automatic voltage control approaches. The power system is capable of handling 500A and is scalable up to 1000A and beyond, and also provides voltage and current monitoring with nanosecond levels between measurements. Thus, highly detailed information on voltage and current is obtained and short-term variations can be taken into account.

[0160] To avoid the sensing / detection itself interfering with the output power performance, there is a tendency for the input voltage and current, and therefore the input power, to be detected rather than the output power.

[0161] The data and the approach used to process it is similar to that for the acoustic sensor type described above.

[0162] 16, the arc voltage is plotted against time as plot V. In this example, the welding process is shut down after 30 seconds and the voltage returns to zero.

[0163] Also plotted in Figure 16 is the Gaussian Amplitude x Gaussian Center vs. Time. As can be seen, for an initial period of the first 2 or 3 seconds, the values ​​on this plot are above the acceptable threshold, indicating concern for the quality of the weld. After the first 2-3 seconds, and certainly after 8 seconds, the plot drops well below the threshold, indicating a high quality weld for this set of variables.

[0164] In addition to these variables, Figure 16 also includes an indication of when argon is scarce as a shielding gas. The indication shows the occurrence of argon deficiency with a plot point, and also the degree of argon deficiency with the vertical position of the plot point.

[0165] (Overall Process – continued) Returning to FIG. 4, now that the details of the operation of the sensor types and the consideration of their data have been made clear, the process and the data therein undergoes a data reconciliation step. Data reconciliation is provided based on input provided through the user interface and / or historical data obtained from storage. The storage may contain data from earlier in the performance of this weld and / or data from a number of previous welds performed by the system, and / or data from other systems [e.g., data from calibration processes, etc.], all of which may contribute to the historical data and therefore to the data reconciliation.

[0166] In the live analysis step, two different levels of processing may be implemented.

[0167] The first level of processing provides synchronization of data types from different sensor types, not only those illustrated above, but also any number of other sensor types that may be introduced and used to measure key characteristics directly and / or indirectly related to the welding system and the occurring weld.

[0168] A programmable logic controller (PLC) is used to temporally synchronize the various data types by providing a master timestamp at the start of data collection from the sensor types in the system. The PLC continuously checks that each sensor of each sensor type in the system is continuously providing data, and that the provided data is being collected at a constant rate. Additional master timestamps may be applied periodically while the data collection process is ongoing.

[0169] A master timestamp can mean that data collected from a microphone serving as a sensor type for acoustic analysis, an input power analysis for draw measurement during welding, an area scan camera providing a vision system, and a laser profile scanner providing a 3D profile sensor type are all coordinated to represent data from the same time, and therefore the same location within the weld.

[0170] Importantly, the PLC also receives data from an incremental encoder that provides position data. The encoder may be attached to the substrate being welded and / or the welding equipment, indicating its physical location at that time. The encoder triggers data collection and provides consistent correlation of the sensor data with the weld position of the component. Again, the same master timestamp is applied to the data from the encoder. This means that the actual location of the same revealed position is known, applicable at the same time in a synchronized data arrangement. The data is synchronized in time and space.

[0171] This processing allows sensor data collected at various locations on the component to be post-processed into one cohesive data structure, allowing the raw sensor data, post-processed information, and analytical output to be correlated with the physical location of the weld. This data can be displayed to the operator in a user display, allowing them to adjust the welding process based on the system's output.

[0172] The second level of processing is implemented by data reduction and / or the application of machine learning. This second level of processing takes the correlation data and feeds it to a second level of processing and analysis (using machine learning or other suitable analytical methods) to correlate any number of features from both the raw data and the output of the sensor type level analysis to weld defects. In practice, a decision engine supporting machine learning is employed.

[0173] In this method, where a defect is suspected, a general assessment is made as to whether the defect is within acceptable characteristics, such as size, or exceeds acceptable characteristics, and thus a defect requiring recording or repair action. When NDT-type sensing is performed, direct measurement of the defect is made by the imaging performed. This directly reports the size and possibly other characteristics of the defect. However, when the consideration of defects is based on the welding conditions occurring during welding, the potential defects are being considered indirectly, and the question is being considered: "Are these conditions likely to lead to defects?" A second level of processing may improve the determination of acceptable and / or unacceptable welds in this context.

[0174] 18, two different sensor types are considered: Type A on the left and Type B on the right. The two sensor types can be any of the sensor types described herein and / or other types of sensor types that provide data regarding the welding performance or results of the welding.

[0175] Referring to the Type A sensor type, within region 800 there is a series of data points that have been reliably established as indicative of an acceptable weld. These may be evident from other sensing and / or test runs verified by NDT or may be modeled cases.

[0176] There are also data points from a weld, such as data point 802, that are clearly outside the acceptable range and are well on the unacceptable side of threshold 804 that may be applied in initially determining whether a data point represents an acceptable or unacceptable weld. Other data points, such as data point 806, exceed the threshold and are considered unacceptable under the definition of threshold 804.

[0177] Instances that are difficult to interpret are data points that are outside of region 800 but below threshold 804, such as data points 808 and 810. A decision based on a single sensor would determine that these welds are acceptable because they are below the threshold.

[0178] Gains are gained from a second level of processing that considers data points [802, 806, 808, 810] across multiple sensor types to make a complete determination. Referring to the right side and to the Type B sensor type, data point 802 again lies within the identified acceptable weld region 800. Similarly, data point 806 is well above threshold 804, again indicating an unacceptable weld from this sensor type alone.

[0179] Looking at data points 808 and 810, both are outside of region 800 but below threshold 804. In a single sensor type approach, these edge cases would again be deemed acceptable welds, however, a second level of processing takes into account the location between multiple sensor types to gain additional information.

[0180] Using neural networks, two approaches to considering location across multiple sensor types can be employed: These two approaches can be used as alternatives to each other, one in parallel with the other, or even in series.

[0181] In the first approach, labeled data is provided and supervised learning takes place. The labeled data can come from either or both of two sources. First, especially in the early stages of processing, such as during calibration or early production weld runs, the labeled data can come from experimental results. So, continuing with the example of FIG. 18, the locations of both points 808 and 810 can be flagged to the operator to allow for the operator's evaluation of acceptable or unacceptable. The operator is not only provided with data from a single sensor type to review and make a decision, but with flagged data from multiple different sensor types on that data [a data point or sequence of data points] and can make a decision based on more nuanced criteria. The results are used to label the data and thus are utilized within the data pool on which the neural network learns. Human knowledge and interpretations are input into the neural network by user decisions to provide oversight.

[0182] A second method of providing labeled data is to utilize a library of existing data. This data is also labeled according to the decision and imparts the necessary operator knowledge. The library is fed into a neural network classifier to reveal the processing location of the library data set. The library can span data from multiple sensor types and determine a second level of processing type, which is the location of the data point across the tests and results of multiple sensor types. The classifier can then determine a classifier score for any data point associated with the library data, and then quantify the likely error of the data point [or set of data] that is being examined to make a decision. The classifier can use a Bayes' Theorem based approach for classification and error quantification.

[0183] The second method, a library, may be used as the starting point for a labeled data set. The first method, an operator invoked calibration or test run on the actual welding system may be used as an alternative from the start. However, the first method may be used to add to the data set of the second method to progress the training of the neural network from more general welding system positions to positions tuned for that particular welding system.

[0184] Returning to FIG. 18 , an operator may determine that data point 808 is acceptable because it is close enough to the tolerance region 800, but that data point 810 is unacceptable because it is too close to the threshold across multiple sensor types. This may cause the neural network to make small adjustments to the tolerance region 800 and / or threshold 804 [value or form]. Over time, repeated decisions of this type may lead to more pronounced and optimized modifications of the boundaries of the tolerance region 800 and / or threshold 804. For example, the tolerance region 800 may be expanded and / or the threshold may be tightened. The same results for data points 808 and 810 at a later, more advanced learning stage in the process may be called acceptable and unacceptable because in the case of data point 808 it is within the revised tolerance range, and in the case of data point 810 it is beyond the revised threshold.

[0185] An example of such an analysis in a real-world scenario might be where the left side relates to a visual image sensor type, and one or more images in succession suggest there is a problem with the sidewall proximity of the weld. Considering an acoustic sensor type on the right side, a sidewall proximity problem might be suggested, thereby confirming the overall determination that the weld is not acceptable.

[0186] While the above example references the position of one type of sensor relative to the other type of sensor to determine acceptability or unacceptability, the determination may be more detailed than that and may determine the nature of the problem with the welding conditions. Thus, where data from one type of sensor alone may only suggest a problem, data from two types of sensors may inform the nature of the problem.

[0187] An issue to note is that importing historical data from other welding situations and environments is not a strong starting point for a library for determining acceptable weld performance. This is because other welding operations and other welding environments have different variables that can affect the data for those welding operations. For example, in the case of a visual imaging sensor, the lighting and lighting angle of the welding environment, the nature of the substrate, the angle and spacing of the welding torch, etc. can all affect the data, which may not match data retrieved in another environment where the lighting is different, for example.

[0188] As an alternative or in addition to this first approach using labeled data and supervised learning, the ability of neural networks to perform clustering or grouping-based processing [looking for similarities and / or anomalies in the data] may be utilized to reduce or avoid the need for library-based data and / or supervised learning. There are various techniques, such as K-means clustering, that can establish the centroid of a cluster and reveal distance certainty around it. Other clustering techniques can also be applied. These can be used to reveal the location of acceptable regions 800 and / or thresholds 804, which can be refined with more data and learning. Large libraries of labeled data may not be required to get started.

[0189] As mentioned above, the first and second approaches may be used in combination, as well as alternatively. Thus, the first approach may be used to begin training the neural network, and the second approach may take over after the first approach has progressed. It is also possible for the neural network to simultaneously train from both the first and second approaches, in order to maximize the data fed to it, particularly if the library is being expanded from ongoing welds on other welding systems other than the welding system being considered. It is also possible to pool training from similarly configured and operated welding systems.

[0190] Over time, especially when the method is used in a production version with large volumes of welds, the amount of data, the accuracy of the evaluation, and the complexity of the various cases of data that can be successfully evaluated increases.

[0191] All results obtained from live analysis of data and / or data reduction and / or machine learning may have a copy sent to storage after acquisition for future use or to contribute to the knowledge of the system and / or similar systems.

[0192] An important step in the analysis results is to compare the data value or position to one or more thresholds established for the data type (thresholding). Examples of thresholding approaches are provided in various sections for the specific sensor types illustrated, but are broadly applicable to each sensor type and the data type it produces. The position relative to the threshold may be considered indicative of a defect (defect indicator) and may be an important part of the weld quality result, displayed to the user via the user display, which also receives and displays the received sensor data.

[0193] 17 is an example of a user display. The user display may provide a real-time output of the analysis of all sensor types so that the welding process can be reviewed as it is running and the process can be adjusted if a poor quality weld is found in the output of the analysis. Additionally, it is also possible to review the entire weld data after each pass and examine the data to identify areas that need repair or adjustment before or during the next weld.

[0194] The weld quality results step may stop the welding process and / or issue a warning to the operator, such as via a user display, if an unacceptable weld is encountered.

[0195] Based on the position of the user display, the operator or the system itself may adjust one or more variable parameters used to control and perform the welding process in real time.

[0196] The weld quality results are reflected in a Weld Qualification which is provided for the entire weld. If a defect is identified by NDT, the size and location of the defect are checked against appropriate criteria to ensure that the level of the defect is within tolerance. If not, remedial action is taken. If acceptable, 3D life record data is stored within the Weld Qualification and made available for the life of the weld and for any required decommissioning support.

[0197] (Data processing and storage) In addition to considering each data type individually, further benefits are gained by considering the data types as a combined data stream. To enable this, each data set is time-stamped, and synchronizing the timestamps synchronizes the time of all data sets.

[0198] This means that all data sets can be combined and saved as one data file, allowing for subsequent reuse.

Claims

1. 1. A welding method, the welding method including inspection of a weld formed by the welding method, the welding method comprising: (i) providing a welding device; (ii) providing a weld inspection device; (iii) introducing one or more substrates to be welded into said welding apparatus; (iv) using heat to elevate the temperature of the one or more substrates above ambient temperature; (v) performing welding of the one or more substrates at an elevated temperature above ambient temperature using the welding device; (vi) inspecting the produced weld using the weld inspection device; Including, The welding method, wherein the inspection of the weld is performed by the inspection device at an inspection location on the one or more substrates, the inspection location being at an elevated temperature above ambient temperature.

2. The method of claim 1 , wherein the welding inspection device uses ultrasound and the welding device is in physical contact with the inspection location.

3. 3. The method of claim 1, wherein the inspection of the weld is performed at an elevated temperature above ambient temperature, the elevated temperature being at least 80°C above ambient temperature.

4. 3. The method of claim 1, wherein the inspection of the weld is performed at an elevated temperature above ambient temperature, the elevated temperature being at least 180°C above ambient temperature.

5. 3. The method of claim 1, wherein the inspection of the weld is performed at an elevated temperature above ambient temperature, the elevated temperature being at least 250°C above ambient temperature.

6. 3. The method of claim 1, wherein the inspection of the weld is performed at an elevated temperature above ambient temperature, the elevated temperature being at least 350°C above ambient temperature.

7. 3. The method of claim 1, wherein the ambient temperature is 20°C + / - 10°C.

8. 3. The method of claim 1 or 2, wherein the welding is performed at a welding location when the inspection is performed at the inspection location, and the welding location and the inspection location are separated by less than 2 m along the weld.

9. 3. The method of claim 1, wherein the welding is performed at a welding location when the inspection is performed at the inspection location, and the welding location and the inspection location are separated by less than 1 meter along the weld.

10. 3. The method of claim 1 or 2, wherein the welding is performed at a welding location when the inspection is performed at the inspection location, and the welding location and the inspection location are separated by less than 0.2 m along the weld.

11. 3. The method of claim 1 or 2, wherein the welding is performed at a welding location when the inspection is performed at an inspection location, and the inspection location was the welding location less than 10 minutes prior.

12. 3. The method of claim 1 or 2, wherein the welding is performed at a welding location when the inspection is performed at an inspection location, and the inspection location was the welding location less than five minutes prior.

13. 3. The method of claim 1 or 2, wherein the welding is performed at a welding location when the inspection is performed at an inspection location, and the inspection location was the welding location less than three minutes prior.

14. 3. The method of claim 1 or 2, wherein the welding is performed at a welding location, and the temperature at the welding location is 800°C during welding.

15. 3. The method of claim 1 or 2, wherein the temperature of the one or more substrates is provided at a temperature of at least 110°C by increasing the temperature of the one or more substrates above ambient temperature.

16. 3. The method of claim 1 or 2, wherein the temperature of the one or more substrates is provided at a temperature of at least 180°C by increasing the temperature of the one or more substrates above ambient temperature.

17. 3. The method of claim 1 or 2, wherein the step of raising the temperature of the one or more substrates above ambient temperature is performed by preheating the one or more substrates before welding begins.

18. 3. The method of claim 1 or 2, wherein if the inspection indicates a defect at a defect location in the weld, one or more repair steps are performed on the defect location in the weld, the repair steps being performed at an elevated temperature at the defect location above ambient temperature, the elevated temperature being at least 180°C above ambient temperature.

19. 1. An apparatus for performing a welding process including inspection of a weld formed by the welding process on a substrate, comprising: (i) a welding device; (ii) a weld inspection device, the weld inspection device including an ultrasonic transmitter and receiver, the weld inspection device having a substrate contact surface, the substrate contact surface having a melting point greater than 250°C; 1. An apparatus for performing a welding method including inspecting a weld formed by the welding method on the substrate, comprising:

20. 20. The apparatus of claim 19, wherein the ultrasonic transmitter and receiver are disposed within the weld inspection device and the substrate contact surface is configured to roll on the substrate.

21. 21. The apparatus of claim 19 or 20, wherein the ultrasonic transmitter and receiver are provided within the weld inspection apparatus with internal cooling.

22. A welding method comprising: (a) providing a welding device; (b) providing a plurality of sensor types; (c) defining a first set of welding conditions for the welding method; (d) introducing one or more substrates to be welded into said welding apparatus; (e) performing welding of said one or more substrates; (f) acquiring data from the plurality of sensor types during welding; (g) comparing the data obtained from the plurality of sensor types with reference data for one or more sensor types; (h) determining whether the weld is of acceptable quality or unacceptable quality based on the one or more comparisons; (i) A method of welding, wherein if the quality of the weld is unacceptable, the method includes performing one or more actions.

23. 23. The method of claim 22, wherein at least one sensor type of the plurality of sensor types is part of a weld inspection device, the method including inspecting a weld using the weld inspection device.

24. 24. The method of claim 22 or claim 23, including inspecting the weld using a weld inspection device to determine one or more characteristics of a defect.

25. 25. The method of claim 24, wherein the characteristics include one or more of size, location, defect type, defect shape, or defect location with reference to the shape of the weld and / or relative to the length of the weld.

26. 25. The method of claim 24, wherein the method further comprises comparing one or more of the characteristics to one or more criteria, and further comprising determining whether the weld having the defect meets or fails to meet welding criteria.

27. 27. The method of claim 26, wherein if the weld meets the welding criteria, a record of the weld is created and stored.

28. 30. The method of claim 27, further comprising: the record including data from one or more of the plurality of sensor types.

29. The method of claim 26, wherein one or more repair steps are applied to the weld if the weld does not meet welding criteria.

30. 24. The method of claim 22 or 23, wherein at least two of the plurality of sensor types are weld condition sensors, and the method includes inspecting a weld condition using the weld condition sensors.

31. 31. The method of claim 30, including inspecting the weld as it is formed to determine one or more parameters of the weld.

32. 32. The method of claim 31, further comprising comparing the one or more parameters to one or more control parameters and determining whether a risk level for a weld defect has been exceeded.

33. The method of claim 1 or 2, further comprising one or more operations of changing the welding conditions from the first set of welding conditions of the welding method.

34. 34. The method of claim 33, wherein a change in welding conditions from the first set of welding conditions is to stop welding and / or alert an operator.

35. 34. The method of claim 33, wherein changing the welding conditions from the first set of welding conditions is changing the welding conditions back to the first set of welding conditions and / or changing the welding conditions to a second set of welding conditions.

36. 24. The method of claim 22 or 23, wherein at least two of the sensor types are selected from a voltage sensor, a current sensor, a welding arc sound emission sensor, a welding topology sensor, a welding image sensor, and an ultrasonic image sensor.

37. 1. An apparatus for monitoring welding, comprising: (a) a plurality of sensor types; (b) a control unit for receiving a first set of welding conditions for the welding method; (c) a comparator that receives and compares data from the plurality of sensor types with reference data for one or more sensor types; The comparator outputs a determination of whether the weld is of acceptable or unacceptable quality based on the compared data, and if the determination determines that the weld is of unacceptable quality, the apparatus further provides a control signal to trigger one or more actions by the apparatus. A device for monitoring welding.

38. 38. The apparatus of claim 37, wherein the apparatus comprises a weld inspection apparatus that determines one or more characteristics of a defect.

39. 39. The apparatus of claim 38, further comprising a first comparator for receiving and comparing one or more of the characteristics to one or more criteria, the first comparator outputting a first determination as to whether the weld having the defect meets or does not meet the welding criteria.

40. 38. The apparatus of claim 37, further comprising a second comparator for receiving and comparing one or more parameters of the weld as the weld is formed with one or more control parameters, the second comparator outputting a second determination of whether a risk level of a weld defect is exceeded.

41. 39. The apparatus of claim 38, wherein if a risk level of a weld defect is exceeded, a control signal provided by the apparatus is sent to a controller to trigger one or more actions, the one or more actions being to change the welding conditions from the first set of welding conditions for the welding method, for example, to stop welding and / or to alert an operator.

Images