Trends of Research Activities on Welding Process in Japan

This paper gives a recent progress of welding process research in Japan, especially, an instance in the capability of the visualization and prediction of welding arcs, though the dynamic observations of spectral image by a high-speed digital video-camera with a monochromator and also the calculations by numerical models. The dynamic observations lead to visualizations of metal vapor behavior in arc plasma and then, for example, reveal the temperature minima at arc axis in GMA welding. Finally, plasma phenomena are discussed in welding arcs through numerical calculations, using the basic conservation equations of mass, energy, momentum, current and electron density of plasma physics. There is close interaction between the electrode, the arc plasma, the weld pool, and also the metal vapor, which constitute the welding process, and must be considered as a unified system. Introduction The progress of numerical analysis in recent years greatly contributes to quantitative understanding of phenomena in the arc welding. Especially, in Gas Tungsten Arc (GTA) welding, the numerical analysis model which enables to predict heat and mass transfer, weld shape and so on accurately was reported [1]. Conversely, in Gas Metal Arc (GMA) welding, a numerical analysis model which represents physical phenomena accurately has not yet been developed. It is because metal transfer phenomenon consisting of melting at the tip of the welding wire, droplet formation and droplet detachment is very complex and the state of the arc changes greatly due to transport of metal vapor evaporated from the wire tip to the arc through the plasma jet. However, the latest research activity on the numerical analysis model in GMA welding has achieved to predict the total system consisting of melting at the tip of the wire, droplet formation and droplet detachment with time-dependent calculation of arc plasma including metal vapor [2]. In GTA welding, the metal vapor is evaporated only from the weld pool surface because of the non-consumable electrode. The metal vapor is swept away along the weld pool surface due to the plasma jet from the upper stream, and then inclusion of the metal vapor into the arc is little. Consequently, influence of the metal vapor on the arc properties is relatively small. Therefore, results of numerical analysis using the numerical model without consideration of influence of the metal vapor agree well with actual phenomena. On the other hand, in GMA welding, it is easily guessed that the inclusion of the metal vapor into the arc affects radiation loss, temperature distribution, electron density distribution, current density distribution, heat flux distribution and so on dramatically and consequently changes the thermal efficiency to be important as welding heat source. Rouffet et al showed that plasma temperature became approximately 8,000K near the arc axis through an experimental observation with the Boltzmann plot method [3]. Furthermore, they succeeded in quantitative measurement of the metal vapor concentration and found that molar fraction of the metal vapor reached 0.7 near the arc axis. However, information about the metal vapor concentration obtained from the experiment was only radial distribution at a height of the arc column. The overall phenomena of GMA welding are not yet understood accurately though the visualization of experiments. A recent progress of research activities in Japan on the visualization and prediction of welding arcs gives the overall phenomena consisting of time-dependent temperature and metal vapor concentration in arc plasma of GMA welding, through the dynamic observations of spectral image by a high-speed digital video-camera with a monochromator. And, it shows the temperature minima at arc axis in GMA welding. Technique of visualization Figure 1 shows an example of experimental setup for the visualization of welding arcs consisting of three monochromators and three high-speed video-cameras [4]. The monochromator is the Czerny-Turner type and has a diffraction grating. Time-dependent and monochromatic images of arc plasma through three monochromators with three high-speed video-cameras are simultaneously captured at the rate of 2,000 fps. The images are Ar I spectrum (696 nm) and two Fe I spectra (537 nm and 538 nm) with the wavelength resolution of 0.4 nm in FWHM. After the Abel conversion of the captured images, plasma temperature for the arc axis region is obtained by the relative intensity ratio method from the two iron spectra. Also, plasma temperature for the arc surrounding is obtained by the Fowler-Milne method from the argon spectrum. Then both plasma temperatures are superimposed in the same frame, which leads to take a visualization of the overall plasma temperatures in GMA welding. After taking the overall plasma temperatures, metal vapor concentration distribution is estimated from the radiation intensity distribution of Fe I spectrum (538 nm) under the assumption of Local Thermodynamic Equilibrium (LTE) in the arc plasma. The CTWD (contact tip to work distance) was 25 mm. The pure iron wire with diameter of 1.2 mm and mild steel (SS400) plate were employed as the anode and cathode, respectively. The welding current and arc voltage were 270 A and 37 V. The shielding gas was pure argon and its gas flow rate was 20 L/min. Visualization of welding arcs Figure 2 shows dynamic variations of plasma temperature, iron vapor concentration, and monochromatic images of Ar I (696 nm) and Fe I (537 nm) in GMA welding with metal transfer [4]. The plasma temperature near the arc axis is about 7,000 K, but the plasma temperature apart from the arc axis is about 12,000 K. The temperature minimum at the arc axis shown by Rouffet [3] is confirmed and the dynamic variation of the whole plasma temperature distribution in GMA welding has been revealed. It is found that the arc has a double structure consisting of a high temperature region apart from the arc axis and a low temperature region close to the arc axis in the pure argon shielding gas, i.e., in the MIG welding, while the plasma temperature varied under the influence of the metal transfer. The iron vapor concentration around the wire tip is higher than that near the base metal surface. This result shows that the plasma jet carries the iron vapor ejected from the wire tip to the base metal surface. The iron vapor concentration reaches to 100 mol% around the wire tip. On the Fig.1 An example of experimental setup for visualization of welding arcs [4]. Fig.2 An example of dynamic visualizations of plasma temperature, iron vapor concentration, Ar I image and Fe I image in MIG welding [4]. other hand, the iron vapor concentration reduces below the droplet. The stream of plasma jet is prevented by the droplet, and then the iron vapor ejected from the wire tip can’t be carried well behind the droplet. Discussion of arc plasma phenomena through numerical calculations A numerical calculation is very useful for discussion on physics of arc plasma phenomena which have been observed through the experimental visualization. An example of numerical simulation model has been shown under the assumption of LTE in arc plasma with steady state [5]. The shielding gas is assumed as pure argon in this model. Figure 3 shows calculation results of iron vapor distribution. In case of non metal vapor, the metal vapor does not exist in the arc at all. In case of metal vapor only evaporated from the base metal, large part of the metal vapor is swept away along the weld pool surface to outside of the arc because the metal vapor source is located at a downstream side of the plasma jet. As a result, the metal vapor hardly migrates into the arc center. On the other hand, in case of metal vapor only evaporated from the wire, the metal vapor is transported into the arc center by the plasma jet because the metal vapor source is located at upstream side of the plasma jet. In case of the metal vapor evaporated from the wire and base metal corresponding to the actual situation, it is seen that mainly the metal vapor from the wire tip is also transported into the arc center by the plasma jet. Figure 4 shows calculation results of temperature distribution in the same conditions with Fig. 3. It is found that the arc has the double structure consisting of the high temperature region apart from the arc axis and the low temperature region close to the arc axis if the metal vapor exists at the arc Fig.3 Calculations of iron vapor concentration [5] Fig.4 Calculations of arc plasma temperature [5] Fig.5 Net radiative emission coefficients for mixtures of iron vapor in argon plasma [5]. Fig.6 Prediction of arc plasma temperature and iron vapor concentration on the assumption that the arc plasma has the net radiative emission coefficient for pure argon [5]. center, which agrees with the experimental visualization. Figure 5 shows comparison of the net radiative emission coefficients for mixtures of iron vapor in argon plasma [5]. It can be seen that the presence of just 1 mol% of iron vapor greatly increases the radiative emission at all temperatures. Figure 6 shows a prediction of plasma temperature and iron concentration on the assumption that arc plasma has the net radiative emission coefficient for pure argon even if iron vapor exist in the plasma. If the arc plasma has no intense radiative emission coefficient like the pure argon, the temperature minima disappear. It is clearly concluded that temperature minima are caused by the energy loss due to the intense radiative emission of the iron vapor. This cooling effect was previously predicted by Schnick [6]. The future trends of visualization The rotational symmetry around the arc axis was assumed in the above whole examples. However, actual welding arcs appear at a three dimensional (3D) space. Actually, it is very difficult that we can meet a beautiful situation in which the arc has a perfect axial symmetry. It means that we naturally need 3D visualizations for welding arcs. The latest research activity on the visualization has achieved to show 3D-argon distribution and 3D-metal vapor distribution in arc plasma of GMA welding, through the simultaneous and multidirectional measurement system by 12 CCD cameras which can capture such axially asymmetric arc phenomena [7]. Figure 7 shows a schematic illustration of experimental setup for the visualization of 3D-arcs using 12 CCD cameras with narrowband interference filters for Ar I spectrum (696 nm) and Fe I spectrum (537 nm). Figure 8 is an example of 3D visualizations. It has been shown that there is a clear deviation of Ar I image from Fe I image in the globular transfer mode because of the 3D-phenomenon of metal transfer, whereas there is the axially symmetric double structural distributions consist of Ar I image and Fe I image in the spray transfer mode ideally close to the axial symmetric phenomenon [7]. On the other hand, the latest research activity on the numerical analysis model in GMA welding has also achieved to predict the 3D-phenomena consisting of droplet formation and detachment at the tip of the wire with time-dependent calculation of arc plasma including metal vapor [8]. Conclusions In Japan, as one of the “Structural Materials for Innovation” in the SIP (Cross-ministerial Strategic Fig.7 Schematic illustration of experimental setup for visualization of 3D welding arcs [7]. Fig.8 An example of 3D visualizations of Ar I image and Fe I image in MIG welding [7]. Innovation Promotion Program) by the Japanese Government, the “Development of Simulation Techniques for Performance Assurance of Weld Joints (PL: Prof. Akio Hirose, Osaka University)” has started since 2014, and the diverse experts have been impelling joint researches as an ad-hoc team. It is the 5-year national project aiming at the completely predictable techniques of welding, which gives the assurance of the performance and quality of weld joints by the numerical simulation before actual welding in manufacturing. The recognition of “the understanding of phenomena can innovate the future welding” is increasing and leading to a big wave of all-Japan, which means the strong collaboration between government, industry and academia. This is one of the future trends on welding research in Japan. In any period, the faculty to see through the essence of something is necessary. Continuous efforts based on the patient observations and profound considerations are included there. In addition, the hearts filled with expected dreams for the future will be required. The visualization is attractive because it gives the opportunities to every expert to see through the essence of things from each viewpoint and to draw each dream. The visualization makes it easier to combine the knowledge and skills which various companies and research institutes have. Thevisualization has great potentials for more diversity. We should explore the new fields of welding technology withadvanced visualization techniques and lead to the development of the future innovative welding technologies. References[1] M. Tanaka, K. Yamamoto, S. Tashiro, K. Nakata, E. Yamamoto, K. Yamazaki, K. Suzuki, A.B. Murphy andJ.J. Lowke: J. Phys. D: Appl. Phys., 43 (2010), 434009.[2] Y. Ogino and Y. Hirata: Quarterly J. Japan Welding Soc., 341 (2016), pp.35-41.[3] M.E. Rouffet, M. Wendt, G. Goett, R. Kozakov, H. Schoepp, K.D. Weltmann and D. Uhrlandt: J. Phys. D:Appl. Phys., 43 (2010), 434003.[4] Y. Tsujimura and M. Tanaka: Quarterly J. Japan Welding Soc., 304 (2012), pp.288-297.[5] Y. Tsujimura and M. Tanaka: Quarterly J. Japan Welding Soc., 301 (2012), pp.68-76.[6] M. Schnick, U. Fussel, M. Hertel, A. Spille-Kohoff and A.B. Murphy, J. Phys. D: Appl. Phys., 43 (2010),022001.[7] K. Kataoka, K. Nomura, K. Miura and Y. Hirata: Quarterly J. Japan Welding Soc., 333 (2015), pp.233-241.[8] Y. Ogino and Y. Hirata: Quarterly J. Japan Welding Soc., 331 (2015), pp.1-12.