Contents
Chapter 2 Literature Review
2.3 Monitoring of FSW process
Chapter 2
(Mehta et al., 2017; Fujii et al., 2006; Boz and Kurt, 2004). Higher surface area and better mechanical contact with the plasticized materials around the tool pin during the welding process makes this profile one of the most suitable option for FSW process. The modification on the straight cylindrical pin is obtained through producing threads on the pin. Various threaded profiles has been investigated and reported in literature and many have confirmed successful joining with square thread profile (Imam et al., 2013;
Elangovan and Balasubramanian, 2008; Sundaram and Murugan, 2010).Apart from the profile of the pin the other parameter associated with the tool that highly influence the FSW process is the geometry that includes shoulder diameter and shoulder profiles. The process of FSW process was investigated over various ranges and design of shoulder diameter. The common conclusions made in each contribution is the influence of shoulder diameter is maximum on FSW process (Scialpi et al., 2007; Leal et al., 2011;
Rajakumar et al., 2011; Liu et al., 2008; Mehta et al., 2011). Shoulder diameter is mainly responsible for frictional heat generation between the rotating tool and the rigidly clamped workpiece material. This frictional heat softens the material around the tool and the stirring action of the tool help in plasticization of the materials. Thus contribution of shoulder diameter is observed to be more in heat generation during the process after tool rotational speed and welding speed. However, effect of pin diameter is found to be insignificant as reported by various researchers (Tozaki et al., 2007; Sharma et al., 2012;
Mehta et al., 2017; DebRoy et al., 2012). The other parameter associated with FSW tool is the tool pin length. However, the available articles effect of tool pin length reported to be negligible compared to other parameters (Elangovan and Balasubramanian, 2008;
Mishra and Ma, 2005; Gibson et al., 2014; Garg et al., 2017, Mehta and Badheka, 2016).
The survey on effect of tool geometry on FSW process provides the insight that straight cylindrical and square tool pin profile results in better mechanical properties compared to other tool pin profiles. Apart from the tool pin profile, shoulder diameter of the tool also plays significant role in FSW process. However, higher shoulder diameter might results in deterioration of joint quality. On the other hand tool pin diameter does not seem to have significant influence on FSW process.
Literature Review
of the process can be evaluated continuously or periodically with the acquired information. Monitoring of a process involves two crucial steps: one is acquisition of valuable information through observation and the other one is the estimation or prediction of outcome from the acquired information. To have better control over the outcome of the process, the monitoring should be very effective to capture every possible change during the process. Process signals can be an effective option for detecting changes occurred during the process over time. Moreover, acquired signals give a more direct representation of the time dependent behaviour of the physical process.
Monitoring a process over time can effectively reduce process faults, defects in the final product and can help in controlling the output quality in the desired level. FSW being an automated process monitoring the process over time would be useful in controlling the output as desired. Efforts made for development of methods for monitoring of FSW process are outlined in the following paragraphs.
Monitoring of FSW process with acoustic emission signals was reported in the work presented by Soundararajan et al. (2006). Fast Fourier transform (FFT), short time Fourier transform (STFT) and wavelet transform had been applied for processing of acoustic signals. The research work claimed that among the three signal processing methods adopted wavelet transform results in more information than FFT and STFT.
Processing of acoustic emission signal with wavelet transform for monitoring of FSW process was also presented by Chen et al. (2003). The conclusions made leads to the impression that various defects formed during the FSW process have different signal band energy characteristics and the same can be an effective indicator for identification of defects in the welded samples. In the research work presented by Subramaniam et al.
(2013) a signal frequency analysis based approach for monitoring of FSW process was presented with acoustic emission signal and Fourier transform. The chances of noise contamination with acoustic emission signal are high and the presented research work hardly describes the methods adopted for elimination of noise and extraction of suitable band in the desired frequency range.
Features extracted from force signals acquired during FSW process were presented as the indication for observing in-process gap between the mating surfaces by Yang et al. (2008).The presented results showed that with the gap present in the mating plates during the process the computed power spectral density would yield higher values.
In an another attempt by Fleming et al (2008) force signals features based methodology
Chapter 2
was proposed for identification of gap in welded samples. Possibilities of defect monitoring in FSW process using vertical force signal information was attempted by Kumari et al (2016). Continuous wavelet transform was implemented for analysis of the vertical force signals and it was reported that the developed methodology was able to detect defects in FSW process. However, the authors studied only surface level defects and did not extend the idea to identify internal defects those are more challenging in detection. Vertical and transverse force signals during FSW process were acquired and Fourier transform was implemented in order to develop defect monitoring method by Boldsaikhan et al. (2011). Artificial neural network based models were developed for classification of defective welds from defect free welds with signal information. Input electrical signatures of the driving motors in FSW system were captured and processed for monitoring of torque and transverse force during FSW process by Mehta et al.
(2013).Monitoring of weld quality in terms of UTS of the joints was presented by Jene et al. (2008) with force signals features. Frequency spectrum of the force signals were reported to have significant information regarding process behaviour. It was reported that high frequency amplitudes in frequency spectra were observed for welds without any notable imperfections or defect.
Presence of defects in friction stir welded samples are attempted to monitor by Longhurst et al. (2016). The frequency information of current signal acquired from the main spindle motor of a FSW system is presented as the key feature for monitoring defect occurrence in the welded samples. The authors claimed that inclusion of void in the welded sampled would result in increase in frequency band of the signal. However, the authors did not emphasize on monitoring weld quality with the extracted signal features. Applicability of infrared thermography in online monitoring of FSW process has been presented by Serio et al. (2016). A new indicator termed as maximum heating slope value has been proposed to correlate tool rotational speed and welding speed during the process for monitoring the behaviour of the joining process. Imam et al.
(2013) developed a methodology based on temperature distribution of friction stir welded joints for monitoring of the process. The study discloses that weld nugget zone temperature below 350 °C can result in formation of tunnel defect in joining 6063-T4 aluminum alloy. Image processing based monitoring of first mode metal transfer in FSW process is attempted by Sinha et al. (2008). Images of welds are captured to obtain image features in terms of grey level distribution, texture, pattern and contours. The authors
Literature Review
reported that the proposed methodology is effective and can be extended towards prediction of process parameters for unknown materials.
The survey of available literature provided the information that few works has been carried out for monitoring of FSW process. However, in other manufacturing processes monitoring has already gained significant attention. For instance, in the comprehensive review of turning operation over a decade by Sick (2002) around hundreds of literature on monitoring of the process is presented. In comparison to this the effort made for developing schemes for monitoring of FSW process is quite less. The demand of FSW process is increasing across various automated industrial sectors. In industry low cost, effective and reliable monitoring methods are useful for achieving the desired quality of the product. Monitoring schemes can help in decision making regarding the process outcome for the betterment of the process. The very less attempts made for monitoring of FSW process the current research work is motivated to develop effective methodologies for monitoring weld quality.