Friction stir welding device monitoring system

Abstract

To provide a friction stir welding device monitoring system that detects and analyzes an AE signal from a rotating tool in real time in a friction stir welding device, and monitors a condition of the rotating tool during welding, the condition being detectable only by the AE signal.SOLUTION: A friction stir welding device monitoring system detects, by an AE sensor disposed in the vicinity of a member to be joined, an AE signal generated from a rotating tool during welding of the member to be joined in a plunge start stage and a plunge end stage in friction stir welding, and quantitatively evaluates in real time an abnormality in the friction stir welding by an external device that performs bidirectional communication with CNC control means which executes operation control of a friction stir welding device while receiving the AE signal.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to a friction stir welding device monitoring system that detects and analyzes AE signals from a rotating tool in a friction stir welding device in real time, and monitors the status of the rotating tool during welding that can be detected only by the AE signal, using a high-speed, low-latency, multi-connection capable wireless communication method. [Background technology]

[0002] . In machining equipment controlled by CNC (Computer Numerical Control) using general cutting tools, etc., there are methods for predicting the occurrence of abnormalities such as chatter caused by tool wear by detecting the vibration, acceleration, audible sound, and motor control current value of the machine tool. In addition to these, there is also a known method of detecting AE signals using an acoustic emission (AE) sensor to monitor the machining status (see Patent Document 1, etc.).

[0003] Acoustic emission (AE) is a phenomenon in which strain energy stored inside a material is released as elastic waves when the material deforms or cracks. When predicting tool abnormalities, the elastic waves generated when an abnormality such as tool breakage occurs are detected by an AE sensor, and the tool abnormality is evaluated by performing signal processing.

[0004] Meanwhile, in CNC-controlled friction stir welding machines (FSW machines), there have conventionally been devices that monitor the welding status, which detect changes in physical quantities such as temperature, acceleration, and vibration in the depth direction from the probe tip of the rotary tool (FSW tool) in real time, and transmit this detected data wirelessly to the outside for monitoring, detecting and diagnosing abnormal signals to detect welding abnormalities and predict tool life (see Patent Documents 2 to 4, etc.).In these cases, a tool holder that can be attached to the spindle of the machine tool is used as a measuring device for changes in the physical quantities of the rotary tool (FSW tool), and various sensors are equipped on the tool holder, and data from the sensors is transmitted wirelessly to the outside, and this data is monitored by an external PC or the like.

[0005] However, conventionally, FSW machines have not specifically detected or analyzed the AE signals from the rotating tool (FSW tool). However, the inventors have detected and evaluated the AE signals from the rotating tool in FSW machines and found that it is possible to detect and evaluate the joining conditions in friction stir welding (FSW) that could not be detected using only conventional methods such as temperature and vibration.

[0006] Specifically, simply monitoring the temperature, acceleration, and vibration of the rotary tool (FSW tool) using a conventional tool holder, as in Patent Documents 2 to 4, did not provide a sufficient evaluation of whether the workpiece being machined was undergoing stable plastic flow in order to perform a good FSW. However, it was discovered that evaluation using AE signals allows for a detailed evaluation of whether stable plastic flow is occurring, particularly at the start and end of the plunge of the rotary tool (FSW tool) into the workpiece. Furthermore, when evaluating using AE signals, the AE signal (elastic wave signal) emitted externally by the FSW tool itself is observed, and a sensor (AE sensor) can be placed at a position on the machine tool other than the tool holder to receive the signal, which can then be wirelessly transmitted externally for monitoring. In particular, with the recent development of high-speed wireless communication technology (Local 5G), this technology is expected to attract particular attention as a technology for effectively monitoring the machining status from a position remote from the workpiece. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Application Publication No. 2018-94686 [Patent Document 2] International Publication No. WO2016 / 111336 [Patent Document 3] Japanese Patent Application Publication No. 2019-104061 [Patent Document 4] Japanese Patent Publication No. 2022-162847 Summary of the Invention [Problem to be solved by the invention]

[0008] The present invention was created in light of the above circumstances, and has as its object to provide an FSW device monitoring system that detects and analyzes in real time AE signals from a rotating tool (FSW tool) in an FSW device, particularly at the start and end of pushing into the workpiece, to detect welding defects that can be detected in real time using AE signals alone, and further communicates those welding conditions to a CNC externally to correct the NC program. [Means for solving the problem]

[0009] In order to solve the above problems, the present invention provides: A friction stir welding monitoring system is provided in which an AE signal generated from a rotating tool during friction stir welding, including at the start and end of welding of the workpieces, is detected by an AE sensor arranged near the workpieces, and the friction stir welding status is quantitatively evaluated in real time using a CNC control means that receives the AE signal and controls the operation of the friction stir welding device, and an external device that communicates bidirectionally using a high-speed, low-latency, multi-connection capable wireless communication method.

[0010] In general-purpose friction stir welding equipment (hereinafter also referred to as "FSW equipment"), insufficient welding can occur due to insufficient rotation speed of the rotating tool relative to the workpieces or excessive plunge speed, or welding abnormalities can occur due to excessive burrs on the workpieces. As mentioned above, undesirable welding conditions can occur particularly during the plunge operation at the start of welding and at the end of the plunge operation at the end of the welding, when stable plastic flow is not occurring in the workpieces. Therefore, in order to detect the plastic flow conditions during plunge operation and at the end of the plunge operation in real time during welding, it has been found that detecting the plastic flow conditions during plunge operation and at the end of the plunge operation, as has been done conventionally, is insufficient to adequately capture the plastic flow conditions during plunge operation and at the end of the plunge operation. Therefore, it has been found that detecting and analyzing acoustic emissions (AE) emitted by the FSW tool itself during welding is effective.

[0011] In the present invention, an AE signal generated from the FSW tool during welding, particularly at the start and end of welding, is detected and transmitted externally by an AE sensor in the FSW device, and this signal is then analyzed by an external PC or the like, making it possible to quantitatively evaluate the occurrence of undesirable welding conditions from the specific AE signals generated during and at the end of the plunge.

[0012] In particular, because CNC communication is possible in response to the reception of AE signals, when specific AE signals are detected at the start and end of the plunge, the CNC can identify the joining position and joining conditions in real time, grasp the plastic flow status, and can advantageously stop the joining or modify the joining conditions. Furthermore, because the AE signal emitted from the FSW tool is actually captured, amplified, and wirelessly transmitted as a digital signal for precise external analysis, the recent development of high-speed wireless communication (Local 5G) technology is important for the practical application of this invention, and is also beneficial in that more precise analysis of the joining status is expected.

[0013] In addition, in the present preferred friction stir welding monitoring system, When a predetermined specific AE signal is detected from the AE signals received and monitored by the external device, it is determined that a specific friction stir welding situation has occurred, The external device includes external control means for informing the CNC control means and modifying the NC program of the friction stir welding device that controls the proper operation of the friction stir welding device and rotary tool to be executed.

[0014] In this friction stir welding monitoring system, specific AE signals known to correspond to undesirable welding conditions, such as insufficient plastic flow after the start of the plunge and the occurrence of burrs at the end of the plunge, are preset, and when those specific AE signals are detected at the start or end of the actual plunge, the NC program of the FSW device is corrected via CNC communication with the CNC control means of the FSW device from an external PC or the like, in order to make corrections to the proper welding operation. This has the advantage that when an AE signal is detected that can detect or predict undesirable welding conditions, such as insufficient plastic flow due to insufficient rotation speed of the rotary tool or the occurrence of excessive burrs due to an excessive plunge amount or excessive spindle rotation speed, the welding position and welding operation in the NC program at that time are identified in real time and a command is given to the CNC control means to make appropriate corrections to the welding operation, thereby guaranteeing welding quality and avoiding defective products.

[0015] In addition, in this friction stir welding monitoring system, It is preferable that the external device detects the received AE signals by excluding AE signals in a range other than a specific frequency range that is caused by the occurrence of a specific friction stir welding situation.

[0016] For example, in the case of evaluating excessive burrs, a specific friction stir welding condition, as described below, the frequencies at which burrs occur are particularly high in the AE signal, particularly in the high-frequency range around 200 kHz. By filtering only this band, it is possible to detect only the AE intensity (voltage) caused by burrs. In other words, the occurrence point and amount of specific friction stir welding conditions can be detected based on CNC information, and undesirable welding conditions can be quantitatively evaluated, predicted, and corrected. As a result, it is possible to further guarantee welding quality and avoid defective products. [Effects of the Invention]

[0017] According to the present invention, specific AE signals emitted from the FSW tool due to insufficient plastic flow of the workpiece at the start and end of plunge in friction stir welding, or due to excessive plunge amount of the FSW tool or spindle rotation speed can be detected and analyzed in real time to detect undesirable welding conditions, and when such conditions occur, the NC program can be corrected in real time via external CNC communication to prevent the occurrence of such welding conditions in advance or to reduce their further occurrence. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a photograph illustrating an FSW device and its components used in the friction stir welding monitoring system of the present invention. FIG. [Figure 2] FIG. 1(a) shows a schematic perspective view of the exterior of a tool holder holding an FSW tool in the FSW device shown in FIG. 1, and FIG. 1(b) shows a schematic perspective view of the FSW tool held by the chuck portion of the tool holder shown in FIG. 1(a). [Figure 3] The measurement results of the AE signals generated during plunge welding using the FSW equipment shown in Figures 1 and 2 are shown. On the left side of each of (a) and (b) are photographs of the top surface of the workpieces after welding when the FSW tool was plunge-welded at a rotational speed of 750 rpm and 900 rpm, and on the right side of each of (a) and (b) are measurement results showing the frequency of the AE signal and the AE signal intensity at that frequency in time series. [Figure 4] The measurement results of the AE signals generated during plunge welding using the FSW equipment shown in Figures 1 and 2 are shown. On the left side of each of (a) and (b) are photographs of the top surface of the workpieces after welding when the FSW tool was plunge-welded at a rotational speed of 1,500 rpm and 3,000 rpm, and on the right side of each of (a) and (b) are measurement results showing the frequency of the AE signal and the AE signal intensity at that frequency in time series. [Figure 5]The AE signal intensities at each rotation speed, which are the measurement results of the AE signals shown in the right diagrams of Figures 3 and 4, are graphed over time. (a) is a graph showing the entire time region measured in Figures 3 and 4, and (b) is an enlarged graph of time region A in (a) at the beginning of indentation. [Figure 6] The measurement results of the AE signals generated during welding using the FSW equipment shown in Figures 1 and 2, from the plunge to travel and the end of plunge, are shown. On the left side of each of (a) and (b) are photographs of the top surface of the workpieces after welding when the FSW tool was joined at a rotational speed of 1,500 rpm and 3,000 rpm, and on the right side of each of (a) and (b) are measurement results showing the frequency of the AE signal and the AE signal intensity at that frequency in time series. [Figure 7] The measurement results of the AE signals generated during welding using the FSW device shown in Figures 1 and 2, from the plunge to travel and the end of plunge, are shown. On the left side is a photograph of the top surface of the workpiece after welding when the FSW tool was joined at a rotational speed of 6,000 rpm, and on the right side is the measurement results showing the frequency of the AE signal and the AE signal intensity at that frequency in chronological order. [Figure 8] FIG. 8 is a graph showing the intensity of the AE signal at each rotation speed in time series, which is the measurement result of the AE signal shown in the right diagrams of FIGS. 6 to 7. DETAILED DESCRIPTION OF THE INVENTION

[0019] 1 shows an example of a friction stir welding apparatus (FSW apparatus) and its components for detecting acoustic emission signals (AE signals) using the friction stir welding monitoring system of the present invention. In the FSW apparatus 10, the upper end of a tool holder 14 is attached to a spindle ("ROBODRILL-M" manufactured by FANUC Corporation) 12 that moves up and down in the Z direction and rotates around its axis, and a rotary tool (FSW tool) 20 is held in a chuck at the lower end of the tool holder 14. Below the rotary tool 20, a workpiece (base material to be welded) 22 is mounted on a table 24 that moves in the X and Y directions on a base 26, and the weld process is performed by pushing (in the Z direction) and moving (in the X and Y directions) the axially rotating rotary tool 20 against the welding point of the workpiece 22.

[0020] In addition, the tool holder 14 is a measuring device ("MULTI INTELLIGENCE i-stir" manufactured by Yamamoto Metal Works) that measures the temperature, acceleration, etc. of the rotary tool 20 during the joining process in real time and transmits them wirelessly, as also shown in Patent Documents 2 to 4 mentioned above, and a dynamometer ("Multi-component dynamometer" manufactured by Kistler) 28 that measures the force acting on the rotary tool 20 is arranged on the base 26.

[0021] Figure 2(a) shows a schematic perspective view of the exterior of the tool holder 14 holding an FSW tool 20, and the tool holder 14 generally comprises a chuck portion 14a at the lower end that holds the FSW tool 20, a holding portion 14c at the upper end that is held by the spindle 12, and a storage portion 14b that encloses various sensors, electronic boards, and wireless devices between the chuck portion 14a and the holding portion 14c. Also, Figure 2(b) shows a schematic perspective view of the FSW tool 20 held by the chuck portion 14a of the tool holder 14 shown in Figure 2(a), and the FSW tool 20 has a plurality of channels that extend in the Z direction and have different lower end positions, and a thermocouple (not shown) is inserted and fixed in each channel.

[0022] In addition to the tool holder 14 and dynamometer 28 described above, the FSW apparatus 10 used in this friction stir welding monitoring system is equipped with an AE sensor 30 ("FAEN-S60I" manufactured by Shinwa Sangyo Co., Ltd.) mounted above the base 26 in the vicinity of the workpiece 22. This AE sensor is a so-called preamplifier-integrated AE sensor, a low-noise AE sensor with a 20 dB preamplifier and filter built into the housing. In this case, the sensor has a resonance point of 60 kHz and is equipped with measures to prevent EMI (electromagnetic interference) and RFI (radio frequency interference). Furthermore, with the recent development of high-speed, low-latency, multi-connection wireless communication (Local 5G), communication data speeds have increased and response times have not been delayed. Therefore, by mounting such an AE sensor 30 on the FSW apparatus 10, precise monitoring is possible, and conventional temperature and acceleration (vibration) measurements are also possible, enabling multifaceted evaluation.

[0023] <<Experimental Example 1 (Push-in Joining)>> Next, the acoustic emission signal (AE signal) generated during plunge welding using the FSW apparatus 10 shown in FIGS. 1 and 2 will be described with reference to FIGS. 3 and 4. FIG. 3(a) shows the FSW tool 20 plunge-welding in the Z direction at a spindle 12 rotational speed of 750 rpm. The left side shows a top view of the workpiece 22 after welding, and the right side shows a measurement result (spectrogram) displaying the frequency of the AE signal received by the AE sensor 30 and the AE signal intensity at that frequency in a time series. Specifically, the measurement result on the right side shows the elapsed time (s) from the start of rotation of the FSW tool 20, the vertical axis shows frequency (kHz), and the intensity of the AE signal corresponding to time and frequency is shown by the shade of color. Specifically, the shade of color represents the amplitude after Fourier transform processing of the AE voltage waveform (this also applies to FIGS. 3(b) to 4 and FIGS. 6 and 7 described below: the actual detected data is displayed in color, but here it is shown in shades of gray).

[0024] Furthermore, the three figures of Figures 3(b) to 4(a)(b) show the measurement results of the AE signal during push-in welding when the rotational speed of the FSW tool 20 is different from that of Figure 3(a). As with Figure 3(a), the left side shows a photograph of the top surface of the workpiece 22 after welding, and the right side shows the measurement results, which display the frequency of the AE signal received by the AE sensor 30 and the AE signal intensity at that frequency in chronological order. Figure 3(b) shows the measurement results at a rotational speed of 900 [rpm], Figure 4(a) shows the measurement results at a rotational speed of 1,500 [rpm], and Figure 4(b) shows the measurement results at a rotational speed of 3,000 [rpm].

[0025] Figure 5 is a graph showing the AE signal intensity at each rotation speed over time, based on the measurement results of the AE signals shown in the right diagrams of Figures 3 and 4. Figure 5(a) is a graph showing the entire time domain measured in Figures 3 and 4, and Figure 5(b) is an enlarged graph of time domain A (see the dotted-line box A in Figures 5(a) and 3(a)) at the beginning of indentation in Figure 5(a). Note that the AE signal intensity shows the result of calculating the RMS over the entire frequency domain (the same applies to Figure 8 described below).

[0026] Specifically, in the graphs of Figures 5(a) and (b), as in Figures 3 and 4, the horizontal axis represents the elapsed time [s] from the start of rotation of the FSW tool 20, and the vertical axis represents the intensity of the AE signal detected at each time in voltage RMS [V], with the broken line in circle 1 representing the voltage RMS [V] at a rotational speed of 750 [rpm] in Figure 3(a), the broken line in circle 2 representing the voltage RMS [V] at a rotational speed of 900 [rpm] in Figure 3(b), the broken line in circle 3 representing the voltage RMS [V] at a rotational speed of 1,500 [rpm] in Figure 4(a), and the broken line in circle 4 representing the voltage RMS [V] at a rotational speed of 3,000 [rpm] in Figure 4(b). It should be noted that a voltage of approximately 0.03 [V] is detected even when the probe 20a is not being pressed in and is separated from the workpiece 22, but this is a fixed AE signal that is originally generated due to a cause other than the FSW, and it can be determined that an AE signal due to the FSW has not been detected when the detected voltage is approximately 0.03 [V].

[0027] First, at the rotational speed of 750 rpm shown in Figure 3(a), as shown in the left diagram, the rotational speed of the spindle 12 is low, so frictional heat generated in the workpiece 22 is small, and the center, where the probe 20a of the FSW tool 20 contacts, is slightly depressed and the surrounding area is raised. It can be seen that elastic deformation is the main cause, with little plastic flow, and the weld mark 32 is small. Furthermore, in the right diagram, an AE signal of approximately 50 to 80 kHz was detected only in time region A (enclosed by dotted line) shortly after the FSW tool 20 began to plunge into the workpiece 22, and was no longer detected around time 7 s. This can also be seen from the fact that the AE signal drops sharply around time 7 s in Figure 5(b), and AE due to the FSW is no longer detected (see the broken line in circle 1). These results show that when the rotation speed is low, such as 750 rpm, the FSW tool 20 and the workpiece 22 rub against each other at low temperatures when the probe 20a is pressed in, and only low-frequency acoustic waves (AE) are generated continuously while increasing, and thereafter plastic flow does not occur, and the friction force becomes excessive, and the servo motor of the spindle 12 actually stops, showing measurement results that are in line with the fact.

[0028] Next, at a rotational speed of 900 rpm, as shown in Figure 3(b), the rotational speed of the spindle 12 is higher than that at 750 rpm in Figure 3(a), as shown in the left diagram. This causes a slight increase in frictional heat generated in the workpiece 22, pushing the workpiece into the shoulder 20b, resulting in elastic and plastic flow up to the contact point, and a larger weld mark 32. Also, in the right diagram, as with the rotational speed of 750 rpm in Figure 3(a), AE signals of approximately 50 to 80 kHz were detected only in time region A (enclosed by a dotted line) shortly after the FSW tool 20 began to plunge into the workpiece 22, indicating elastic deformation. Figure 5(b) also shows that the AE signal increases in time region A1, around 2 to 3 s, and then continues to be detected in time region A2, around 4 to 6 s (see the broken line in circle 2). Therefore, at a rotational speed of 900 rpm, the rotational speed is still low, just like at a rotational speed of 750 rpm, and the probe 20a and the workpiece 22 rub against each other at a low temperature, so low-frequency AE signals are continuously generated.Although the photograph on the left side of Figure 3(b) appears to show a good weld mark 32, the AE signal shows that sufficient plastic flow has not yet occurred.

[0029] Next, at the rotational speed of 1,500 rpm shown in Figure 4(a), the rotational speed of the spindle 12 is higher than at the rotational speeds of 750 rpm and 900 rpm shown in Figures 3(a) and (b), and the frictional heat generated in the workpiece 22 also increases, so that the workpiece is pressed into the shoulder 20b, and sufficient plastic flow is likely occurring, as can be seen from the photograph of the welding mark 32 on the left. Also, in the right-hand image, it can be seen that the time region B (encircled by dotted line: frequency of approximately 50 to 80 kHz) in which the AE signal is detected from the start of pressing is shorter than at the rotational speeds of 750 rpm and 900 rpm shown in Figures 3(a) and (b). On the other hand, looking at Figure 5(b), in time region A1 around time 2 to 3 [s] within time region A, the AE signal initially increases, similar to when rotational speeds are 750 [rpm] and 900 [rpm], but begins to decrease in the latter half, and in time region A2, unlike when rotational speeds are 750 [rpm] and 900 [rpm], almost no AE signal due to FSW is detected (see the broken line in circle 3). Therefore, at a rotational speed of 1,500 [rpm], unlike rotational speeds of 750 [rpm] and 900 [rpm], the rotational speed is sufficiently high and the probe 20a and workpiece 22 rub against each other at a high temperature, so AE signals are no longer generated during plunge welding, indicating that sufficient plastic flow has occurred and good plunge welding has been achieved.

[0030] When the rotational speed increases further, as shown in Figure 4(b), the spindle 12 rotates at 3,000 rpm. The frictional heat generated in the workpiece 22 also increases, and visual inspection of the weld mark 32 on the left reveals sufficient plastic flow, with the exception of some burrs around the periphery. Furthermore, in the right-hand image, the AE signal is detected within time region C (encircled by dotted lines) even shorter than when the rotational speed was 1,500 rpm in Figure 4(a). Even if an AE signal is generated, it is quickly lost, indicating the disappearance of elastic energy generation. On the other hand, Figure 5(b) reveals that the AE signal increases sharply in time region A1, around 2–3 s, then almost disappears immediately after reaching its peak. It then rapidly increases again after time region A2 (see the broken line in circle 4). Therefore, at a rotational speed of 3,000 rpm, just like at a rotational speed of 1,500 rpm, the rotational speed is sufficient for plastic flow, but heat input extends to a large area outside the shoulder 20b, causing elastic deformation to spread to unnecessary areas. The burrs that appear around the periphery in the photograph in Figure 4(b) are also presumably the result of excessive temperature. As a result, it can be seen that if the rotational speed is too high, a push-joint that cannot be considered good will be formed.

[0031] As described above, from the measurement results using the AE signals shown in Figures 3 to 5, it was found that it is possible to detect good plastic flow of the workpiece 22 during push-joining in real time, which cannot be detected by simply sensing the weld marks 22 (or cross-sectional photographs), temperature, or acceleration, and to externally control the rotational speed of the FSW tool 20 in real time to maintain appropriate plastic flow.

[0032] <<Experimental Example 2 (Running Example After Push-In Welding)>> Next, as Example 2, acoustic emission signals (AE signals) generated when the FSW device 10 was moved in the X direction after being pressed in will be described with reference to Figures 6 to 8. Figure 6(a) shows a welding process in which the FSW tool 20 was pressed in the Z direction at a rotational speed of 1,500 rpm of the spindle 12 and then moved in the X direction (bottom of the left image). The left side shows a top view photograph of the workpiece 22 after welding, and the right side shows measurement results in which the frequency of the AE signal received by the AE sensor 30 and the AE signal intensity at that frequency are displayed in chronological order. As with Example 1, the measurement results on the right side show the elapsed time [s] from the start of rotation of the FSW tool 20, the vertical axis is frequency [kHz], and the intensity of the AE signal corresponding to time and frequency is shown by different shades of color.

[0033] Furthermore, the two figures of Figures 6(b) to 7 show the measurement results of the AE signal when the FSW tool 20 is made to travel the same distance after being pressed in when the rotational speed is different from that of Figure 6(a). As with Figure 6(a), the left side shows a photograph of the top surface of the workpiece 22 after traveling, and the right side shows the measurement results in which the frequency of the AE signal received by the AE sensor 30 and the AE signal intensity at that frequency are displayed in chronological order. Figure 6(b) shows the measurement results at a rotational speed of 3,000 [rpm], and Figure 7 shows the measurement results at a rotational speed of 6,000 [rpm].

[0034] FIG. 8 is a graph showing, in time series, the entire time domain in which the intensity of the AE signal at each rotation speed was measured as a result of measuring the AE signal shown in the right diagrams of each of FIGS.

[0035] 6 and 7, the horizontal axis represents the elapsed time [s] from the start of rotation of the FSW tool 20, and the vertical axis represents the intensity of the AE signal detected at each time in terms of voltage RMS [V]. The broken line in circle 1 represents the voltage RMS [V] at a rotation speed of 1,500 [rpm] in FIG. 6(a), the broken line in circle 2 represents the voltage RMS [V] at a rotation speed of 3,000 [rpm] in FIG. 6(b), and the broken line in circle 3 represents the voltage RMS [V] at a rotation speed of 6,000 [rpm] in FIG. 7. Note that even when the probe 20a is not being pressed in and is separated from the workpiece 22, a voltage of approximately 0.03 [V] is detected. This detected voltage of approximately 0.03 [V] is a fixed AE signal that is originally generated due to a cause other than the FSW, and this voltage should be subtracted for judgment, as in FIG. 5 of the first embodiment.

[0036] First, at a rotation speed of 1,500 rpm as shown in Figure 6(a), sufficient frictional heat was generated in the workpiece 22, as in Figure 4(a), and visual inspection of the left-hand photograph from the weld mark 132a during the plunge to the weld mark 132b during travel, and then to the weld mark 132a at the end of the plunge, reveals that the workpiece was plunged all the way to the shoulder 20b, causing sufficient plastic flow, and that the weld continued to travel until the end of the plunge welding. Furthermore, the right-hand photograph reveals that a strong low-frequency AE signal of approximately 50 to 80 kHz was detected from the travel process after the plunge process to the end of the plunge process (time region D (enclosed by a dotted line)). On the other hand, looking at the graph for a rotational speed of 1,500 [rpm] in Figure 8 (see the broken line in circle 1), it can be seen that the strength of the AE signal (voltage RMS [V]) gradually increases from the start of the pressing through the travel process and up to the end of the pressing process (time region D (dotted line region)), and the fact that the AE signal is no longer detected after the pressing process ends shows that sufficient heat is input to the workpiece 22 throughout the entire time region of the joining process in Example 2, causing plastic flow and ensuring proper joining.

[0037] Next, at a rotational speed of 3,000 rpm as shown in Figure 6(b), the rotational speed is higher than that of 1,500 rpm, and as in Figure 4(b), the frictional heat generated in the workpiece 22 also increases, and it can be seen that sufficient plastic flow is occurring from the welding mark 132a at the time of pressing in the photograph on the left, to the welding mark 132b during travel, and to the welding mark 132a at the end of pressing, but it can be seen that some burrs have occurred around the welding mark 132a at the end of pressing.

[0038] In addition, in the right figure, the AE signal shows that strong low-frequency AE signals of around 50 to 80 kHz are detected from the travel process after the pressing process to the pressing end process, which is similar to the rotation speed of 1,500 rpm in Figure 6(a). However, it can be seen that strong AE signals are detected across the entire frequency range (frequency 50 to 200 kHz) for a short period of time around 22.5 s during the pressing end process (time region D (enclosed by dotted line)). On the other hand, looking at the graph for a rotational speed of 3,000 [rpm] in Figure 8 (see the broken line in circle 2), we can see that, just like at a rotational speed of 1,500 [rpm] (circle 1), the AE signal strength (voltage RMS [V]) rises gradually from the start of indentation through the travel process, but at around time 22.5 [s] during the indentation end process (time region D (circled by dotted line)), the AE signal strength rises sharply, then falls within a short period of time, and then, once the indentation process ends (time region D has passed), the AE signal is no longer detectable.

[0039] That is, when the rotation speed was 3,000 rpm, strong AE signals were detected across the entire frequency range for only a short period of time (around 22.5 s) during the final push process. This is thought to be because, as the push force of the FSW tool 20 against the workpiece 22 decreases during the final push process, burrs that have formed around the joint rub against the sides of the FSW tool 20, and this rubbing generates strong AE signals with high frequencies up to 200 kHz. Therefore, it is clear that the occurrence of undesirable burrs at the end of push can be detected by detecting the frequency and intensity of the AE signal.

[0040] Next, at a rotational speed of 6,000 [rpm] as shown in FIG. 7, the rotational speed becomes even faster than 3,000 [rpm], causing the workpiece 22 to overheat, and conversely, as shown in the photograph on the left, from the weld mark 132a during pressing to the weld mark 132b during running, the plastic flow becomes too large and the center part drops, causing a void 132d, and at the weld mark 132c at the end of pressing, it can be seen that the material of the workpiece 22 located around the FSW tool 20 has flowed out as a burr 132e that is larger than when the rotational speed is 3,000 [rpm].

[0041] Additionally, in the right figure, from the travel phase after the pressing phase to the pressing end phase, strong low-frequency AE signals of approximately 50 to 80 kHz are detected, similar to rotational speeds of 1,500 rpm and 3,000 rpm. Specifically, the pressing phase is performed from 2.5 to 21.5 seconds, and the travel phase is performed until approximately 23 seconds, after which the FSW tool 20 is removed (withdrawn) from the workpiece 22. At approximately 22.5 seconds during the pressing end phase (time region D (enclosed by dotted lines)), strong AE signals are detected across the entire frequency range (frequencies of 50 to 200 kHz), similar to the case of a rotational speed of 3,000 rpm. Unlike the case of a rotational speed of 3,000 rpm, high-frequency AE signals of up to approximately 200 kHz are still detected after pressing end (after time region D). On the other hand, looking at the graph for a rotational speed of 6,000 [rpm] in Figure 8 (see the broken line in circle 3), we can see that, just like at a rotational speed of 3,000 [rpm] (circle 2), the AE signal strength (voltage RMS [V]) rises gradually from the start of indentation through the travel process, but at around time 22.5 [s] during the end of indentation (time domain D (dotted line)), the AE signal strength rises sharply, and even after that, a high-intensity AE signal is detected even after the end of the indentation process (time domain D has passed).

[0042] This is presumably because not only is the force of the FSW tool 20 against the workpiece 22 decreasing during the final plunge process, but also during the process of separating the FSW tool 20 from the workpiece 22, the burrs standing tall around the joint continue to rub against the sides of the FSW tool 20, generating strong high-frequency AE signals of up to about 200 kHz, even during the separation process. Therefore, by detecting the frequency and intensity of the AE signals, it is possible to detect in real time from the AE signal that undesirable large burrs are occurring from the end of plunge from the separation process when the rotational speed reaches an excessively high speed (excessive plunge amount) of about 6,000 rpm.

[0043] As described above, from the measurement results using the AE signals shown in Figs. 6 to 8, it was found that it is possible to detect in real time the favorable plastic flow of the workpiece 22 during joining from the push-in joining to the end of the push-in while traveling in the X direction, which cannot be detected by sensing only the weld mark 132 (or cross-sectional photograph), temperature, or acceleration, and to externally control in real time the rotational speed of the FSW tool 20 that can maintain the appropriate plastic flow.

[0044] The above describes an embodiment of the friction stir welding monitoring system of the present invention as an example, but the present invention is not limited to this configuration, and those skilled in the art will easily understand that other embodiments exist from the description and concept of the claims. [Explanation of symbols]

[0045] 10 Friction stir welding equipment (FSW equipment) 12 spindle 14 Tool holder (rule holder with measuring device) 14a Chuck part 14b Storage section 14c Grip 20 Rotary tools (FSW tools) 20a probe 20b Shoulder 20c channel (thermocouple insertion and fixing hole) 22 Workpiece (base material to be joined) 24 tables 26 Foundation 28 Dynamometer 30 AE sensors 32, 132, 132a, 132b, 132c joint marks 132d void 132e Bali

Claims

1. A friction stir welding monitoring system that uses an AE sensor located near the workpieces to detect AE signals generated from the rotating tool during welding, including at the start and end of welding of the workpieces, and quantitatively evaluates the friction stir welding status in real time using a CNC control means that receives the AE signals and controls the operation of the friction stir welding device, and an external device that communicates bidirectionally using a high-speed, low-latency, multi-connection compatible wireless communication method.

2. When a predetermined specific AE signal is detected from among the AE signals received and monitored by the external device, it is determined that a specific friction stir welding situation has occurred, 2. The friction stir welding monitoring system according to claim 1, wherein the external device has an external control means for notifying the CNC control means and correcting an NC program of the friction stir welding device that controls the operation of the friction stir welding device and rotary tool appropriately to be executed.

3. 3. The friction stir welding monitoring system according to claim 1, wherein the external device detects the received AE signals by excluding AE signals in a range other than a specific frequency range that is caused by the occurrence of a specific friction stir welding situation.

Images