Proceedings of the 2012 9th International Pipeline Conference IPC2012 September 24-28, 2012, Calgary, Alberta, Canada

IPC2012-90214
DEVELOPMENT OF LAMINAR PLASMA SHIELDED HF-ERW PROCESS 1

ADVANCED WELDING PROCESS OF HF-ERW 3 -

Hideki Hamatani, 2Fuminori Watanabe, 2Nobuo Mizuhashi, 2Sunao Takeuchi, 2Yoshiaki Hirota, 2 Shigeharu Matsubayashi, 2Kazumoto Tsukakoshi, 2Yasushi Hasegawa, 3Takuya Asano, 3 Takashi Motoyoshi, 4Takao Miura, 5Kimiharu Tanaka, 6Kazuto Yamamoto, 2Tetsuro Nose 1 Nagoya R&D Lab. Nippon Steel Corporation, Tokai City, Aichi, 476-8686, Japan, Contact Author 2 Technical Development Bureau, Nippon Steel Corporation, Futtsu City, Chiba, 293-8511, Japan 3 Nagoya Works, Nippon Steel Corporation, Tokai City, Aichi, 476-8686, Japan 4 Kimitsu Works, Nippon Steel Corporation, Kimitsu City, Chiba, 299-1141, Japan 5 Oita Works, Nippon Steel Corporation, Hikari City, Yamaguchi, 743-8510, Japan 6 Head Office, Nippon Steel Corporation, Tokyo, 100-8071, Japan Oleg P. Solonenko, 7Andrey V. Smirnov Institute of Theoretical and Applied Mechanics, a Siberian Branch of Russian Academy of Sciences, 4/1 Institsuaya av, Novosibirsk, 630090, Russia.
7

KEYWORDS Line pipe, HF-ERW pipe, weld defect, shield, plasma ABSTRACT High frequency - electric resistance welded (HF-ERW) pipe has been successfully used for many years for a number of applications. The benefits of HF-ERW pipe are considerable, including a higher dimensional tolerance and lower prices than seamless pipe and UO pipe. The conventional weld seam produced by HF-ERW, however, often has a relatively low toughness. We have developed an automatic heat input control technique based on ERW phenomena that relies on optical and electrical monitoring methods and has been shown to result in a significant improvement in the toughness. Shielding of the weld area must also be considered as a key factor in the formation of a sound weld. It has been shown that an inert cold gas (e.g., at room temperature) shielding technique is effective for maintaining a stable low oxygen state in the weld area that inhibits the formation of penetrator, a pancake oxide inclusions. Compared to the cold gas shielding technique, high temperature gas shielding, due to its higher kinetic viscosity coefficient, should make it easier to sustain a higher laminar flow, thus leading to a rather low air entrainment in the shielding gas. In addition, plasma is a much higher temperature state (~6000 K), and the dissociated gases can react with the entrained oxygen; plasma jets should, therefore, enhance the overall shielding effects. Moreover, oxides on the strip edges can be expected to melt and/or be reduced by the high temperature plasma jets. Nippon Steel has developed a plasma torch that can generate a long and wide laminar argon nitrogen (hydrogen) jet. This paper describes the results obtained from our investigation of the effects of a plasma jet shield on the weld area of high

strength line pipe with a yield strength grade of X65. Preliminary attempts in applying this novel shielding technique has been found, as expected, to demonstrate extremely low numbers of weld defects and a good low temperature toughness of the HF-ERW seam. NOMENCLATURE : gaseous dynamic viscosity 0: gaseous reference dynamic viscosity T: gaseous temperature T0: gaseous reference temperature n-value: empirical viscosity value S: Sutherland constant Re: Reynolds number Q: quantum collision cross section : gaseous density V: gaseous mean velocity D: characteristic linear dimension : gaseous kinematic viscosity P: gaseous pressure R: gas constant ts: the time to completely melt oxide particles s: particle density r: particle radius diameter C: particle heat capacity q:particle latent heat Tin: particle initial temperature Tmp: particle melting point : coefficient of heat transfer M: particle molecular weight Aw: width of shielded area Al: length of shielded area in the parallel direction 1 Copyright 2012 by ASME

Ep: Plate voltage of vacuum tube in power supply for welding Dp: Plasma jet diameter Ds: Diameter shielding gas arranged outside of plasma jet : irradiation angle between the torch centerline and the plate L: distance from the exit of the torch to the plate surface V0: geometric weld position Vp: plasma irradiated point on the plate IP: irradiation position as defined by the distance from V0 to Vp INTRODUCTION HF-ERW uses a very high intensity current as a joule heating source for welding with upsetting molten metal. Thus, the welding speed of ERW can be relatively higher than other welding processes, such as arc welding and laser welding. In addition, HF-ERW uses a hot or cold rolled plate as the base material, therefore the thickness of the ERW pipe can be much more homogeneous than a seamless pipe. The weld seam produced using HF-ERW often has a relatively low toughness compared to other pipes. The low toughness may preclude the use of ERW pipes in high integrity onshore and offshore applications requiring high weld reliability. The toughness of an ERW weld seam depends substantially on features such as the microstructure and the weld defects, namely the cold defect, an oxide inclusion in the form of a continuous film that results from a lack of heat input for the welding speed. The Penetrator, which results when excessive heat input occurs for the welding speed, is another fatal imperfection of the weld seam. The excess molten metal on the butting strip edges is exposed to the atmosphere during welding, and oxide inclusions sometimes remain without being squeezed out, if sufficient material upset is not achieved. In particular, when using pipe containing large amounts of Mn, Si, Cr, or other elements that readily form oxides, penetrators are easily formed. These weld defects, therefore, must be avoided by providing sufficient heat input for the welding speed. To overcome these problems, we have developed an automatic heat input control technique based on ERW phenomena that uses optical and electrical monitoring systems [1-5]. Even though there is only a narrow range of operating conditions that are adequate for ERW, the toughness of the weld seams in our HF-ERW pipe is greater than that of other ERW pipe. Conventional HF-ERW is, in addition, carried out under an oxidizing environment due to the presence of not only oxygen, but also coolant water for the forming rolls, and thus often includes penetrators in the weld seam. In order to exclude oxygen and water vapor from the weld point, gas shielding techniques at the weld point are widely used for ERW [6,7]. Incomplete shielding of the weld area is, however, often a primary reason for the narrow ERW operating conditions. In order to expand the operating conditions, an improved shielding technique for the weld area that inhibits penetrator formation by maintaining a stable low oxygen state at the weld area must, therefore, be considered as another key factor for the formation of a sound weld.

A turbulent shielding gas has disadvantage such as ambient air into the shield gas. The higher laminar flow must therefore be preferable as a shielding source. Compared to cold gas shielding techniques, a high temperature gas shielding approach should easily sustain a higher laminar flow due to the higher kinetic viscosity coefficient. A plasma jet, which has a much higher temperature state, allows for a higher laminar flow [8] and has the further advantage that the dissociated gas can react with the entrained oxygen. As a result, a plasma jet should serve as a more effective shielding source. Moreover, the oxides on the butting end faces can be expected to melt and/or be reduced by the high temperature plasma jet. In general, non-transfer-type electric-arc plasma torches are currently most widely used. These torches are comprised of a cathode and anode, and are configured so that an electric voltage is applied between the cathode and the anode to generate the plasma. The conventional plasma torch has problems, however. A turbulent plasma jet flows out from the torch. Because the turbulent plasma jet actively mixes with the surrounding low-temperature atmosphere, the length of the zone over which there can be a high temperature and a low entrainment of ambient air cant exceed five to seven times the length of the inner diameter of the torch in the axial direction. This length is insufficient for effective shielding of the process. Moreover, the temperature and the velocity field of the plasma jet become non-axisymmetric. As a solution to this problem, the magnitude of the swirling force of the working gas is usually increased, which causes the arc spot to rotate at the surface of the anode. When the flow velocity of the working gas is small, i.e., when the Reynolds number is small, however, due to the pressure of the gas, a swirling effect cant be obtained. Therefore, increasing the magnitude of the swirling force of the working gas cant be applied effectively. Another solution is to install a solenoid covering the anode, thus applying an electromagnetic swirling. When this solution is applied, however, the structure of the plasma torch becomes intrinsically complicated, while the problems described above are not adequately solved. Furthermore, since a swirling plasma jet is usually turbulent, the length of the plasma jet is relatively short. In addition, due to the turbulent plasma flow, the noise level becomes extremely large, and be as high as 120-130 dB. Based on the technical requirements for the shielding process, it is necessary that the plasma jet be low-velocity, lownoise, relatively long, and have a large diameter. We have modified a cascade-type plasma torch [9-11] comprised of a cathode, an anode, and a cascade (an electrically insulated inter-electrode insert) between the cathode and the anode, wherein the cathode, the anode, and the cascade are insulated from one another. As a result, the distance between the cathode point on the cathode and the anode point on the anode becomes longer. Consequently, the electric voltage becomes higher, and a quasi laminar plasma jet can be more readily created. The objective of the present study is to develop a plasma shielded HF-ERW process in order to decrease the formation of penetrators by using the newly developed plasma torch that can

Copyright 2012 by ASME

generate a long and wide laminar argon nitrogen (hydrogen) plasma jet. In this paper, we report the results of our investigation of the effects of a plasma jet shield on the weld seams of SUS304 and high strength line pipe with a yield strength grade of X65. PLASMA SHIELD EFFECT CONCEPT With an inert cold gas (e.g., at room temperature) shielding technique, because ambient air becomes entrained in the inert gas, it is difficult to maintain a stable low oxygen state in the weld zone. On the contrary, as is well known, a common approximation of the viscosity of a dilute gas is the power law

Most importantly, if the plasma jet consists of a polyatomic molecular gas system, the disaggregated or ionized atomic gas should readily react with oxygen gas, and thus the oxygen content in the plasma jet should be, simply, reduced as, X 2 2X (7)

T = 0 . T 0
3

(1)

From kinetic theory, using S, the viscosity can also be expressed as

T 2 T + S = 0 T T + S 0 . 0
Table 1 power low viscosity parameter for gases 0, Ns/m2 T0, K n Ar 273 2.13E-05 0.72 N2 273 1.66E-05 0.67 H2 273 8.41E-06 0.68

(2)

The empirical values of n, T0, 0, and an effective temperature for argon, nitrogen, and hydrogen are listed in Table 1 [12].
S 144 107 97

In the plasma state, the expression of the viscosity of a pure gas can be defined by [13] mkT . 5 (3) = 16 Q In equations (1-3), the dynamic viscosity increases with increasing temperature. The Reynolds number, which represents the turbulent level, is determined by

2 X + O2 2 XO (8) As another advantage of the high temperature gas, it is expected that high melting temperature oxides on the surface of the strip edges should be melted. The time required to completely melt an oxide particle is expressed by T Tin rC s rq (9) ts = s + log M T Tmp M (T Tmp ) The formula suggests that SiO2, one of the most significant oxides for forming penetorator and/or inclusion of flying steel scale and spatter from the environment into the weld seam, particles with a diameter of 50 um will be completely melted in 5 msec, assuming an argon plasma with a temperature of 4000 K. Fig. 1 shows photographs of the plasma jets from various plasma torches with similar, but not exactly identical, discharge conditions. These photographs are useful for identifying the high temperature region (over 3000 K) in the plasma jet and indicate that the developed laminar plasma torch with a nozzle diameter of 18 mm had a larger high temperature zone than that of the turbulent plasma torches. Furthermore, the tail of the laminar jet seems to be unstable because of the air entrainment resulting from the turbulence. For the laminar plasma torch, however, it was found that the core flame extending 300 mm from the exit of the torch is stable.
(a)

(b)

Re =

VD .

(4) (c) . (5)

Fig. 1 Comparison of plasma jet photos. (a) Argon plasma jet by commercial Praxair 2086A spraying torch with an inner nozzle diameter of 8mm (b) Argon plasma jet by in-house turbulent plasma torch with an inner nozzle diameter of 16 mm (c) Argon-Nitrogen plasma jet newly developed cascade plasma torches with an inner nozzle diameter of 18mm.

The density of a gas is roughly given by

P RT

Simply selecting the viscosity definition as equation (1), equations (1), (4), and (5) yield

Re =

PVDT0n . R 0T n +1

(6)

This equation infers that a high temperature gas shielding should easily sustain a relatively low Re number and a higher laminar flow, thus leading to a rather low air entrainment into the shielding gas.

The distance between the anode and the cathode, the nozzle design, and the gas flow rate are fundamental factors affecting the formation of a laminar plasma jet [14], which is not surprising since Kolmogorovs law indicates that turbulent

Copyright 2012 by ASME

energy decreases along the plasma axis due to the dissipation of the energy downstream of the maximum turbulent position. An estimation of the temperature, gas content, and velocity profiles in a plasma jet, displaying a significant shielding effect was carried out using Fluent, a commercial fluid dynamic finite element method (FEM) program. An asymmetric 2D model for the plasma jet geometry, a model for the turbulence, and a combustion model for the gas content were employed. Here, in the case of laminar plasma flow, even though selecting minimum and values in the Fluent, due to the over estimation of the turbulence, the calculated high temperature region, shielded area, was shorter than the actual one. Thus, the calculated air entrainment was larger than the actual one. Fig. 2 shows the oxygen content distribution inferred from the calculations for turbulent and laminar plasma jets with a hydrogen content of 10%. The oxygen content is plotted vs. the lateral displacement from the centerline. For this calculation, to express the laminar jet, a turbulence of and a fixed area model were used. This figure illustrates that the choice of the plasma torch significantly influences not only the temperature field but also the oxygen content field. Although the partial oxygen pressure at the centerline in the turbulent plasma is nearly 0.1 atm, half the value in the atmosphere, in the laminar jet, the oxygen content is 4 to 5 orders of magnitude lower. In addition, it is easily understand that, in the case of the turbulent plasma, the irradiation distance must be set closer, and the irradiation position must be more severe than in the case of the laminar plasma.

The ratio of the partial pressures of H2 gas to H2O vapor, PH2/PH2O present at the local thermal equilibrium (LTE) was estimated in order to validate the shielding effect. Because air entrainment reaches an LTE condition, in the present work, the PH2/PH2O values are assumed to be confirmed only for a specific level of air inclusion into the plasma jet at given hydrogen content and gas temperature. The H2 content and the temperature were varied in the ranges of 4 to 35% and 0 to 6000 K, respectively. The calculated PH2/PH2O values as a function of the gas temperature are shown in Fig. 3. In all cases, the PH2/PH2O value increases as the temperature increases, but the H2 content becomes saturated above 5000 K. The saturated PH2/PH2O value is approximately 102 in the case with an air entrainment of 1%. Cr is one of the most significant elements in line pipe steel and readily forms penetrator. The critical PH2/PH2O value for Cr reduction/oxidation in an Ellingham diagram is nearly 102. If it is assumed that the oxidation occurs in the chemical equilibrium regime, which is an overestimate because that the oxidation time is not long enough to reach chemical equilibrium during welding, then, no oxidation of Cr should occur in the laminar plasma jet during welding.

(a) Air entrainment = 1%

Fig. 2 Comparison of the estimated radial partial oxygen pressure profiles for the turbulent and laminar plasmas. Hydrogen content: 10%, nozzle diameter: 16 mm, axial position: 150 mm.

The partial oxygen pressure, mentioned above, in the turbulent plasma is nearly half the value in the atmosphere, then, the air inclusion is estimated to be 50%. Contrary, in the laminar jet, even though, the calculated oxygen content is 4 to 5 orders of magnitude lower, the air inclusion therefore must be lower than 1%. For the next estimation, the air entrainment was set at 1% and 50%, corresponding approximately to the laminar and turbulent cases, respectively.

(b) Air entrainment = 50%

Fig. 3 Estimated PH2/PH2O values as a function of the temperature.

Copyright 2012 by ASME

PLASMA SHIELD EFFECT ON THE INCIDENCE OF WELD DEFECTS IN THE LABORATORY Fig. 4 shows the experimental setup highlighting the plasma shielding HF-ERW process and the components used for the experiments. The HF-ERW source consisted of a pair of mechanically moving steel plates that, virtually, created a strip edge from a flat strip that was deformed gradually to a cylindrical form through a series of rolls; a pair of contact shoes that supplied the electrical power to the steel plates via directly introduced electric resistance, and a squeeze roll that pressed the heated and molten two plate edges. The electrical power was supplied by a high frequency power supply with a maximum oscillator tube plate voltage of 14.5 kV, and applied to the SS304 plate with a thickness of 6mm via the contact shoes. The HF-ERW power supply was operated in a voltage constant power mode; thus, first an applied oscillator tube plate voltage, Ep, value was selected, then the current value was determined by the welding current path or the impedance. In present works, only Ep value was varied ranged of 70 to 90% for the maximum value, 14.5kV, therefore, Ep was substituted for the HF-ERW input power

0.05%, which is considered to be the oxide inclusion ratio in the base metal. During welding, the welding area was shielded by high temperature gas using several heat sources, such as a commercial TAFA JP5000 HVOF (High velocity oxygen fuel) spraying torch, a commercial Praxair 2086A spraying torch with an inner nozzle diameter of 8 mm, an in-house turbulent plasma torch with an inner nozzle diameter of 16 mm, and newly developed cascade plasma torches with an inner nozzle diameters of 18 and 25 mm. Kerosene and oxygen were used as the combustion sources for the HVOF, and the atmosphere in the flame was varied from oxidizing to reducing via the mixing rate. Contrary, for the plasma torches, a mixture of argon, nitrogen, and hydrogen was selected as the plasma operating gas. Varying the oxygen/kerosene ratio produced different oxidation atmospheres under the shielding conditions. For shielding purposes, the benefit of a lower oxygen/kerosene ratio was the lower oxygen content. At a stoichiometric composition, around 275, the exceeded CO and/or hydrogen gas serves as an effective shield source.

Fig. 4 Experimental setup highlighting the plasma shielding HFERW process.

The high temperature gas torch bodies were located between the contact shoes, and the high temperature shielding gas was irradiated from the exit of the torch onto the heated and molten plate edges. The shielding width and length of the high temperature gas are roughly written by (10) A w Dp and

Fig. 5 Method for evaluating the weld defect area ratio on a weld seam, and typical observed penetrators on the fracture surface.

A l D p / sin ( )

10 30 ,

(11)

respectively. In this study, the welding speed was selected to be 30 m/min, and the rate of flow of the injected water on the welding area was 0.8 l/min through two tubes with an inner diameter of 4 mm set forward of the weld point and the contact shoes. In order to understand the effect of the high temperature gas shield on the incidence of weld defects as compared to that observed in conventional HF-ERW, Charpy tests were carried out on the weld seam at a testing temperature of 433 K, as shown in Fig. 5. At this testing temperature, most of the fracture surfaces of the weld seam become ductile, and the weld defect portions become brittle. Consequently, based on the weld defect area ratio, which is defined as (Penetrator area + cold defect area) / weld seam area, the weld seam quality was evaluated. The acceptable weld defect area ratio is less than

Fig. 6 shows the effects of the combustion gas shield on the weld defect ratio using the HVOF system; material: SS304

Fig. 6 shows the effects of the combustion gas shielding on the weld defect ratio using the HVOF system. As the oxygen/kerosene ratio increased, the oxidation rate accelerated,

Copyright 2012 by ASME

while the use of a lower oxygen mixing ratio in the shield combustion gas resulted in an overall lower number of weld defects. However, for a lower mixing ratio, exceeded kerosene caused being unstable of combustion and increasing of generating CO gas, effective for shield but toxic for environment. Thus, in the case of using combustion gas shielding, HVOF, even under the reduction conditions, a sound weld could not be obtained. On other hands, in the case of using plasma jet as a shielding gas, it was expected to form lower oxygen content shielding source due to that we dont have to use oxygen as the operating gas. The effects of an argon-hydrogen commercial turbulent plasma jet with an inner nozzle diameter of 8 mm and a side shielding gas on the weld defect area ratio is shown in Fig. 7. The plasma conditions for the results shown here were an argon gas flow rate of 17 l/min, a hydrogen gas flow rate of 10 l/min, a direct current of 350 A, and a distance of 90 mm from the exit of the plasma torch to the steel plate surface. It can be seen in the figure that, in the case of the conventional HF-ERW with no plasma shield, the minimum weld defect area ratio, nearly 0.3%, is achieved with an intermediate Ep value of 80%. A lower Ep condition results in a generally lower strip edge weld temperature, which results in the cold defects. For larger Ep values, excessive molten metal was observed to be the cause of more penetrators. When a plasma shield is used with the best irradiation position, the weld defect ratio was reduced to less than 0.01%. This value is considered to be lower than the oxide inclusion ratio in the base metal. For other irradiation positions, however, this enhancement was canceled. A possible explanation of this outcome is that the effective inner nozzle diameter is a primary factor in shielding the weld area during welding. In general, it is observed that a larger the inner nozzle diameter generates a larger diameter of the plasma jet under the optimum plasma conditions. Thus, the effect of the plasma diameter on the tolerance of the irradiation position was investigated.

The plasma diameter significantly affected not only the tolerance, but also the adequate Ep, plate voltage, range. Fig. 8 displays the effects of the plasma jet shielding on the weld defect ratio using an in-house turbulent plasma jet with a nozzle diameter of 16 mm. The plasma conditions for the results shown here were an argon gas flow rate of 40 l/min, a hydrogen gas flow rate of 5 l/min, and a direct current of 380 A. These conditions are very common for our in-house plasma torch studies. As expected, an increase in the plasma diameter drastically increased the irradiation position tolerance, and the misalignment of the irradiation position, within 20 mm, was not as significant on the incidence of the weld defects for the plasma shielded HF-ERW process. Moreover, the margin for the suitable ERW input power seemed to be 2% in the oscillator tube plate voltage.

Fig. 8 Effects of an argon-hydrogen turbulent plasma jet with a plasma diameter of 16 mm on the weld defect area ratio; material: SS304

Fig. 7 Effect of an argon-hydrogen commercial turbulent plasma jet with an inner nozzle diameter of 8mm on the weld defect area; material: SS304. Plasma Best = best IP case, see Fig.4, Plasma Best-10 = 10mm far from the best IP case.

In this example, the distance from the exit of the plasma torch to the sheet plate surface was 90 mm. If the distance is greater than 90 mm, the benefit of the plasma shield is lost due to air entrainment. For shielding purposes, however, the distance must be greater than 90 mm, because of the mechanical interference from the top roll and squeeze roll. In addition, extremely loud and disturbing noise levels (120-130 dB) have been observed for turbulent plasma systems in industrial ERW mills. A laminar plasma shielding effect on the weld defect area ratio is shown in Fig. 9. The plasma conditions were a total argon gas flow rate of 8 l/min, a nitrogen gas flow rate of 20 l/min, a hydrogen gas flow rate of 0 or 3 l/min, a current of 200 A, and an L of 150 mm in the hydrogen gas case. In both gas systems, the plasma shield enhanced the weld quality, and the weld defect area was lower than 0.05%, which satisfied our goal. The margin of ERW input power, also seemed to be relatively larger than in the turbulent cases. In addition, the range of noise value was between 75 dB and 90 dB. Our newly developed laminar plasma torch demonstrated that the core length of an argon - nitrogen plasma or an argon

Copyright 2012 by ASME

nitrogen hydrogen plasma is more than 300 mm and 150 mm, respectively. It has also been determined that the shielding enhancement of the argon nitrogen hydrogen plasma system is often superior to that of the argon nitrogen plasma system, although the hydrogen gas is not always necessary.

plasma jet can melt the burr and/or steel scale during welding, the number of related defects will be reduced. Fig. 10 shows the effect of a turbulent plasma jet on the number of defects caused by burrs and bonded scales on the strip edges before welding. As shown in this figure, those defects might be decreased by using plasma irradiation during welding.
PLASMA SHIELD EFFECT ON WELD DEFECTS IN THE NAGOYA MILL A preliminary attempt at applying this novel shielding technique to pipe line production was carried out at the Nagoya 16 inch mill. A high strength line pipe with a yield strength of x65 grade was employed in this study. After welding, to reduce the residual stress and improve other mechanical properties, heat treatment in a continuous process was carried out. The set point level (SPL) that expresses the welding state monitored by measuring 1/f [4,5] is substituted for an HFERW heat input, and ranges from an SPLmax-10% to an SPLmax+15%. Here, the welding state of SPLmax means a lower power limit for the second welding type phenomenon, and thus a narrow gap forms between the weld point and the convergence point, leading to the highest short circuit [4,5]. In order to evaluate the plasma shield effect, the weld defect area ratio was measured, a Charpy test was carried out over a testing temperature range of -80C to 0C, and macrosection qualifications on the weld seams were performed. The plasma conditions used here were a total argon gas flow rate of 8 l/min, a nitrogen gas flow rate of 20 l/min, a current of 200 A, and an L of 180 mm.

Fig. 9 Enhancement of laminar plasma shielding for the HFERW of SS304. Inner nozzle diameter is 18mm.

While it was necessary to optimize the experimental conditions to avoid overheating and spattering of the melting metal, the present method is applicable for achieving a lower weld defect ratio and a wider ERW input power range than in other shielding techniques for HF-ERW.

Fig. 10 Additional effects of plasma irradiation on defects caused by burrs and scales on the strip edge; material: SS400. Defect area = Penetrator area + cold defect area + scale area.

As mentioned above, cold welding and penetrators are the fatal defects that result basically from a lack of ERW input power control. In welding, there is also a problem with anomalous electric shorts during welding, since burrs at the corners of the strip edges come in contact with the other side of the strip edge. Furthermore, inclusion of flying steel scale and spatter from the environment into the weld seam often causes other types of weld defects. If a high temperature

Fig. 11 Effect of the laminar plasma shield on the defect area ratio for x65 at the Nagoya mill.

As expected, the weld defect area ratio was strongly dependent on the plasma shield and the heat input power. Fig. 11 shows the effect of the plasma jet shield on the weld area ratio For an HF-ERW heat input, It can be seen that, for a conventional HF-ERW, the weld defect area ratio is minimized at an ERW heat input of SPLmax+5, a welding phenomenon of the second type. A weld defect area ratio of less than 0.05%

Copyright 2012 by ASME

seemed to be a sufficiently low value. The adequate ERW heat input power tolerance was about 7%, however. On the other hand, in the case of the plasma shield, no defect could be detected with a heat input range of SPLmax+5 to SPLmax+10, and thus the adequate ERW heat input power tolerance was more than 15%.

CONCULUSION To suppress formation of weld defects and expand the HFERW operating conditions, we have developed a laminar plasma-shielded HF-ERW process. It was found, as expected, that preliminary attempts of this novel shielding technique resulted in extremely low numbers of weld defects and good low temperature toughness of the HF-ERW seam for x65 grade line pipe. The laminar plasma jet shield is a promising method for the protection of pipes from weld defects. These results have guided our selection of the materials and configurations of the HF-ERW pipes needed to extend this process to high integrity onshore, offshore, and more challenging structural applications.

Fig. 12 Enhancement of the laminar plasma shield in the Charpy test for x65 at the Nagoya mill.

It is expected that the defect area ratio and cross sectional microstructure will directly affect the toughness of the welds and as such, the absorbed energy value determined from Charpy impact testing A comparison of absorbed energy of the weld seam as a function of the Charpy test temperature at a fixed ERW input power (SPLmax+15%) is displayed in Fig. 12. In this study, the obtained absorbed energy of the sample with a thickness of 7.5mm is converted to the full size value on the assumption that the absorbed energy is proportional to the thickness. The vTrs of the plasma shielded seam exhibited a value approximately 20C higher than that of the conventional seam, because the number of defects was lower.

REFERENCES 1. H. Haga, The Welding Journal, Research Supplement, July (1980) 208. 2. H. Haga, The Welding Journal, Research Supplement, June (1981) 104. 3. H. Haga, Nippon Steel Technical Report, No.26 (1985) 27. 4. T. Fukami Proc. Int. Pipeline Conf. 2012, (2012) to be published. 5. N. Hasegawa Proc. Int. Pipeline Conf. 2012, (2012) to be published. 6. R. K. Nichols, Practical Welding Today, March/April (1988) 29. 7. Y. Fukuda, Sumitomo Metal Industries Ltd., September (1992). 8. X. Chen, Pure Appl Chem., Vol.78 No. 6 (2006) 1253. 9. V. I. Kuz'min. Proc. of the 8th National Thermal Spray Conference, Houston (1995) 83. 10. O. P. Solonenko, HTPP9, abst, Petersburg (2006) 10. 11. K.Osaki Vacuum 59 (2000) 47. 12. Hilsentrath, Table of Thermodynamic and Transfer Properties, Nat. Bur. Std. Circular (1960) 564, Pergamon. 13. M. I. Boulos, Thermal Plasma, (1994) 269 Plenum Press. 14. H. Hamatani, Proc. National Meeting of Japan Welding Society, Autumn (2011) 376.

Copyright 2012 by ASME