8 Industry 4.0 in Welding
273
in the welded sample, followed by thermo-mechanically affected zone (TMAZ)
and HAZ [78]. Thus, the mechanical and metallurgical properties of the welded
sample are dependent upon the temperature [78]. The classification of the welded
samples into defective and defect-free has been carried out by studying the temperature signature in time and time–frequency domains [79]. Significant variation in
the signature was found which helped in classifying the defective ones from defectfree welded samples. FSW technique has also been controlled in real time by using
the temperature signature [80]. A suitable controller was designed by collecting the
temperature variations across a range of tool rotational speeds. This study has been
further extended by considering force signature along with the temperature and tool
rotational speed for better stability [81].
8.3.2.5 Force and Torque Signals
The force signal has been found to be a crucial parameter for monitoring of the
spot welding and FSW techniques [50–52]. The force in spot welding technique is
exerted by the electrodes, and in FSW, it is because of the tool that plunges into the
workpieces. In spot welding technique, the force signal has been studied by using
descriptive statistics which provided useful information about the expulsion of the
material [50]. The force signal in FSW is responsible for the forging action, and, as
and when this quantity becomes insufficient, welding defects may occur. Since FSW
involves a rotating tool, the torque also carries information about the quality of the
weld. These signals were utilized to detect the occurrence of welding defects. For
instance, the gap between the mating edges of the two materials in FSW technique
gives rise to “gap defect". The tool constantly remains in contact with the workpieces;
thus, the force signature has been a suitable indicator. The study found a drop in the
signal amplitude whenever the tool crossed the region with a gap. Further, the signal
was also studied in the frequency domain by using FFT, and a significant difference
was found in between the welds, with and without gaps [82]. The gap defect has
also been detected by analysing the “power spectral density” (PSD) of the force
signal computed by using Discrete Fourier Transform (DFT) [83]. A few studies
have shown the capability of force signal for weld classification. This includes estimating the fractal dimension of the force signal by applying the Higuchi’s algorithm
[84]. The results obtained showed a higher value of fractal dimension for the defective weld, as compared to the defect-free weld. Secondly, analysis of the force signal
for classifying the welds has been carried out by using “Wavelet Packet Decomposition” (WPD) and “Hilbert Huang Transform” (HHT) [85]. Features such as the
“instantaneous frequency” and “phase angle” were extracted. While the instantaneous frequency was found to be higher in case of the defective weld with a
negative-scale of phase angle, it was comparatively low for the defect-free weld
with a positive-scale. Similarly, the torque signal has also been found to be effective
in detection of the welding defects. Studies have used WPD to analyse the torque
signal for weld classification by extracting features such as dispersion, excess and
asymmetry [86]. Further, both force and torque signals have been studied by using
273
in the welded sample, followed by thermo-mechanically affected zone (TMAZ)
and HAZ [78]. Thus, the mechanical and metallurgical properties of the welded
sample are dependent upon the temperature [78]. The classification of the welded
samples into defective and defect-free has been carried out by studying the temperature signature in time and time–frequency domains [79]. Significant variation in
the signature was found which helped in classifying the defective ones from defectfree welded samples. FSW technique has also been controlled in real time by using
the temperature signature [80]. A suitable controller was designed by collecting the
temperature variations across a range of tool rotational speeds. This study has been
further extended by considering force signature along with the temperature and tool
rotational speed for better stability [81].
8.3.2.5 Force and Torque Signals
The force signal has been found to be a crucial parameter for monitoring of the
spot welding and FSW techniques [50–52]. The force in spot welding technique is
exerted by the electrodes, and in FSW, it is because of the tool that plunges into the
workpieces. In spot welding technique, the force signal has been studied by using
descriptive statistics which provided useful information about the expulsion of the
material [50]. The force signal in FSW is responsible for the forging action, and, as
and when this quantity becomes insufficient, welding defects may occur. Since FSW
involves a rotating tool, the torque also carries information about the quality of the
weld. These signals were utilized to detect the occurrence of welding defects. For
instance, the gap between the mating edges of the two materials in FSW technique
gives rise to “gap defect". The tool constantly remains in contact with the workpieces;
thus, the force signature has been a suitable indicator. The study found a drop in the
signal amplitude whenever the tool crossed the region with a gap. Further, the signal
was also studied in the frequency domain by using FFT, and a significant difference
was found in between the welds, with and without gaps [82]. The gap defect has
also been detected by analysing the “power spectral density” (PSD) of the force
signal computed by using Discrete Fourier Transform (DFT) [83]. A few studies
have shown the capability of force signal for weld classification. This includes estimating the fractal dimension of the force signal by applying the Higuchi’s algorithm
[84]. The results obtained showed a higher value of fractal dimension for the defective weld, as compared to the defect-free weld. Secondly, analysis of the force signal
for classifying the welds has been carried out by using “Wavelet Packet Decomposition” (WPD) and “Hilbert Huang Transform” (HHT) [85]. Features such as the
“instantaneous frequency” and “phase angle” were extracted. While the instantaneous frequency was found to be higher in case of the defective weld with a
negative-scale of phase angle, it was comparatively low for the defect-free weld
with a positive-scale. Similarly, the torque signal has also been found to be effective
in detection of the welding defects. Studies have used WPD to analyse the torque
signal for weld classification by extracting features such as dispersion, excess and
asymmetry [86]. Further, both force and torque signals have been studied by using