This article was downloaded by: [de Resende, Andre Alves] On: 28 January 2011 Access details: Access Details: [subscription number 932877550] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK
Welding International
Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t778164493
A qualitative model to explain the polarity influence on the fusion rate in the MIG/MAG process Daniel Souzaa; André Alves de Resendea; Américo Scottia a Faculty of Mechanical Engineering, Centre for Research & Development into Welding Processes, Federal University of Uberlândia, Uberlândia, Minas Gerais, Brazil Online publication date: 09 November 2010
To cite this Article Souza, Daniel , de Resende, André Alves and Scotti, Américo(2010) 'A qualitative model to explain the
polarity influence on the fusion rate in the MIG/MAG process', Welding International, 24: 12, 934 — 941 To link to this Article: DOI: 10.1080/09507110903569032 URL: http://dx.doi.org/10.1080/09507110903569032
PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Welding International Vol. 24, No. 12, December 2010, 934–941 Selected from Soldagem & Inspec¸a˜o 2009 14(3) 192– 198
A qualitative model to explain the polarity influence on the fusion rate in the MIG/MAG process Daniel Souza1, Andre´ Alves de Resende2 and Ame´rico Scotti3 Faculty of Mechanical Engineering, Centre for Research & Development into Welding Processes, Federal University of Uberlaˆndia, Block 5H, Avenvu Joa˜o Naves de A´vila, 2121, 38400 902 Uberlaˆndia, Minas Gerais, Brazil
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
(Received 17 March 2009; final version received 23 May 2009) One of the main advantages of the MIG/MAG process is its high productivity. In most of the applications, positive polarity is used, due to its greater arc stability, generation of less splatter and formation of weld beads with suitable geometry. However, in some applications, there is a need for greater production capacity than that offered by conventional MIG/MAG welding. In the literature, it is stated that negative polarity provides a higher fusion rate than positive, despite leading to a high level of splatter and unsuitable formation of the weld bead. Unfortunately, there is not much information available on the effects of the process variables in this polarity, much less justification for such. Therefore, this work is an attempt to try to understand the reason why there is a higher deposit rate in negative polarity, as well as the related effect on the geometry of the weld beads. To do this, comparative MIG/MAG welds were produced in both positive and negative polarities, using two compositions of shielding gases at two current values. The transfer mode and the behaviour of the arc were analysed by synchronized profiling. The geometric profile of the weld bead was evaluated by means of metallographic procedures. From the results, which disagree in part with the current literature, it was seen that both the transfer mode as well as the morphology and the appearance of the weld bead are dependent on the composition of the shielding gas. To explain the phenomena inherent in the greater fusion rate of wire in DC2 , the suggestion is that the fact that arc scales the sides of the wire in this polarity may be the governing factor. Keywords: welding; MIG/MAG; metal transfer; negative polarity
1. Introduction The MIG/MAG welding process, also known as gas metal arc welding (GMAW), is defined by the ‘Welding Handbook’ of the American Welding Society1 as an arc welding process that uses as arc between a constantly fed wire electrode and a weld bead. It is used with an unpressurized external shielding gas. Despite the apparent simplicity of the definition, there are various factors that affect the operational features of this process, such as, among others, the chemical compositions of the electrode wires and of the shielding gases. These factors, in turn, directly affect the mechanical and metallurgical properties of the weld metal, the methods of metal transfer, etc. Welding with the conventional MIG/MAG process normally takes place using DCþ (with the electrode connected to the positive pole and the part to the negative pole, also known as inverse polarity). In accordance with the current literature, this configuration gives great penetrative depth, good arc quality and transfer quality and a relatively low number of splashes, as well as the option to weld with different metal transfer methods (short circuit, globular and the various forms of spray transfer). Welding with the electrode connected to the negative pole (DC2 ), also known as direct polarity, is said to have the following characteristics: low penetration and a high rate of wire fusion for a given current in comparison with DCþ welding, which leads to a lower quality of heat transferred to the part and a relatively high number of splashes. ISSN 0950-7116 print/ISSN 1754-2138 online q 2010 Taylor & Francis DOI: 10.1080/09507110903569032 http://www.informaworld.com
Welding using DC2 , according to Talkington2, is generally limited to globular transfer and is little used in practice because the resulting arc is unstable and the splashes produced are undesirable. In addition, according to this researcher, the greatest problem inherent in DC2 welding is the repulsive cathodic force that acts on the welded extremity of the electrode, given that the electrons are emitted from the electrode and transferred to the surface of the work (the piece). Lancaster3 describes these reaction forces as the main cause for drops being repelled asymmetrically during MIG/MAG welding in DC2 . He also describes the unstable positioning of the cathodic point that forms, on the extremity of the drop, an intense concentration of light emitted from a localized part of the drop that stands out in the middle of the arc. The cathodic point can be seen dancing around a large drop at the moment that this is being asymmetrically repelled, as shown in Figure 1. Scotti4, in a review of metal transfer, states that, as well as the effect of the polarity, repelled drops occur due to the shielding gas, even in DCþ . Initially citing Jo¨nsson et al.5, the effect of He as the shielding gas in this phenomenon stands out; deformed drops are formed due to a flow of gas (plasma jet) in the part-drop direction. This mass flow, in turn, is caused by electromagnetic forces (Lorentz force gradient) generated close to the surface of the cathode, as a function of a marked convergence of the current close to the weld bead (a very small cathodic point). However, this author also states that even with other gases, drops can be repelled in DCþ . Ushio et al.6, in a study into MIG/MAG with a 1.2 mm steel electrode
Welding International
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
Figure 1. Asymmetrically repelled drop and cathodic point. Adapted from Lancaster3.
wire, observed regular transfers only in globular and spray modes in their respective operational scales, when a 0 –10% CO2 –Ar mixture was used. Transfer predominantly changed to repelled (flexed) in the globular and goticular modes when the CO2 index exceeded 10%. Rhee and Kannatey-Asibu Jr.7 reported repelled drops with 100% CO2 shielding, an effect that is associated with electromagnetic forces and with fumes. Finally, Scotti4 also cited the trend of repelled globular transfer being expected to decrease with an increase in current, but at that time there was no proof of this hypothesis. Despite the undesirable characteristics of welding in DC2 , this method offers some useful characteristics for GMAW welding. The main benefit of DC2 will be the reverse balance of the heat produced in the arc. According to Talkington2, the characteristic of GMAW welding in DCþ is that approximately 30% of the heat generated in the arc is transferred to the electrode and the remaining (approximately 70%) to the base metal. This energy balance is reversed for DC2 welding, in which approximately 30% of the energy is transferred to the base metal and 70% to the electrode. Some researchers, such as Lancaster3, have studied the various factors that affect the fusion speed and have shown that greater speeds of fusion (greater volume deposited per unit of weld length) and lower penetrations are obtained in DC2 when
Figure 2.
935
compared with DCþ . The bead assumes an extremely convex shape that is not suitable for welding. Figure 2 summarizes the relative differences expected (based on the descriptions by both the researchers) between the profiles of the weld beads resulting from MIG/MAG welding in DC2 and DCþ . In light of the above, it seems to be the consensus that welding in DC2 produces a higher fusion rate, but with a repelled globular transfer and an extremely convex weld bead with very low penetration. However, exploratory research in the Centre for Research & Development into Welding Processes (LAPROSOLDA) at the Federal University of Uberlaˆndia have provided evidence that DC2 does not always lead to low penetration, or that a transfer is always in repulsive globular mode. In addition, other experiments show that the addition wires ensure that the fusion rate is the same in both polarities. As a result, the explanations referred to above in the literature about the reason for the higher fusion rate based on heat distribution in the two polarities, or that DC2 causes repelled transfer are, at the least, not generic. There is, therefore, a gap in the bibliography about the real role of polarity in metal transfer and on the geometry of the weld bead in the MIG/MAG process. As a result, this work is intended to add to the knowledge base in these areas, with a view to providing a better understanding of the phenomena that occur during MIG/MAG welding in DC2 .
2.
Experimental procedures
A total of eight experiments were carried out with the aim of verifying the influence of the composition of the shielding gas, the current level and the polarity (DCþ and DC2) on the metal transfer modes in MIG/MAG welding. These experiments were carried out using a water-cooled commercial torch and a multi-process electronic source with a secondary key. The source was programmed to work in constant direct current MIG/MAG mode, in order to guarantee the same current value in all the comparative experiments. In this way, the value of the current and of the feed speed were regulated, with the voltage being the result of the charge (characterized for each arc). To try to
Characteristics of the weld bead for welding in DC2 and the inverse (adapted from Talkington2).
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
936
D. Souza et al.
find a similar working condition whether DCþ or DC 2 was used, the feed speed in each case was increased until it started short circuits. The work point was then defined with a feed speed of an instant less than when the shorts would start. It is worth pointing out that for small weld bead lengths and in the laboratory conditions imposed, no significant variation was seen in the average length of the arc during welding. The use of arc length control was, therefore, discarded, since it would have an impact on the phenomenon being studied. The electrode wire used in the study was of AWS ER70S-6 class with a diameter of 1.2 mm. Mixtures of 92%Ar þ 8%CO2 and 98%Ar þ 2%O2 were used as shielding gases. The flow used in all the tests was 13 l/min. For comparison purposes, all the tests were carried out using simple deposit onto a plate. The test pieces were made from plates of ABNT 1020 steel, 200 £ 25.4 £ 9.5 mm in size. Table 1 shows the parameters regulated for each test. Tests 1 –4 were carried out to establish the influence of the shielding gas, while Tests 5 –6 verified the influence of the current on the transfer mode, in comparison with Tests 1 –2. Tests 7– 8, compared, respectively, with the Tests 1 –4, were carried out to establish the effect of the polarity, using the same feed speed/weld speed ration, with the aim of obtaining the same amount of material deposited per unit of length and comparing the profiles of the weld beads in both DCþ and DC2 modes. To visualize the metal transfer and the related phenomena, the profiling technique was used with high132#speed digital filming at 2000 fps and a laser8 length of
Table 1.
Regulated parameters for the tests.
Test
Shielding gas
Polarity
1 2 3 4 5 6 7 8
Ar þ 2%O2 Ar þ 2%O2 Ar þ 18%CO2 Ar þ 18%CO2 Ar þ 2%O2 Ar þ 2%O2 Ar þ 2%O2 Ar þ 18%CO2
CC2 CCþ CC2 CCþ CC2 CCþ CCþ CCþ
Figure 3.
Profiling principles applied to welding10.
632.2 mm. In welding, the term ‘profiling’ can be used to refer to the formation of a shadow projected by the various elements (torch, electrode, drops, weld bead and plate) on a photographic film or directly onto the lens of a camera, a technique also known as backlighting (Figure 3). The experimental array was set out in accordance with Figure 4. To enable the arc also to be seen, the drops in formation and transfer had to be affected, due to the intensity of the optical filters used. To correlate the variations in voltage and current with the formation and detachment of the drops, the electric signals were synchronized with the film frames, using the technique described in another article9. The effect on the geometry was studied by means of conventional macro graph techniques, on cross sections taken from the welded test plates. 3.
Results and discussion
Table 2 sets out the values for voltage and current monitored during the tests, while the transfer mode and the behaviour of the voltage and current signals in Tests 1– 6 are shown in Figures 5 –10. It was initially observed that, for the same current, the use of DC2 led to a higher fusion rate, quantified by a regulated feed speed to give stable welds for the given currents (Table 1). In these welding conditions, globular transfer was obtained (lower current and goticular (higher current) in DC2 , which contradicts the contents of the literature principally for gases rich in CO2. By looking at the graphs drawn up for DCþ welding, for the two gas mixtures and current values used in the tests, a smaller arc length can be defined when compared
Feed speed (m/min)
Welding speed (cm/min)
CTWD (mm)
Gas flow (l/min)
Reference current (A)
11.7 7.2 9.8 7.0 7.2 3.2 7.2 7.0
34.8 34.8 34.8 34.8 34.8 34.8 21.5 25.0
20 20 20 20 20 20 20 20
13 13 13 13 13 13 13 13
250 250 250 250 150 150 250 250
Welding International
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
Figure 4.
Details of optical laser system used10.
to welding in DC2 (the arc length was not measured and was only recorded by visual assessment, due to the difficulty in defining what the arc length would be, given that the two polarities and different gases have distinct coupling characteristics and light emission). The tendency for the arc to scale the wire, thereby causing an increase in the arc length, was confirmed. This effect is more pronounced when a small percentage of oxygen is used as the active gas. Taking into account, as well as these results, the concepts of field emission (or cold cathode emission) and the known phenomenon of cathode cleaning in TIG AC aluminium welding (the sputtering zone), this behaviour was considered in the present article as being mainly responsible for higher rate of consumption of addition material when compared to welding in DCþ (rather than just considering that during an anode connection more heat is generated that in a cathode connection). As a result, the model proposed by the authors to explain the higher fusion rate in DC2 would give greater heat efficiency leading to the arc reaching a greater part of the electrode point in a search for oxides for field emission. This behaviour of the arc when welding in DC2 explains some of the particularities of this process, such as the greater arc instability that consequently causes
Table 2.
Values monitored during the Tests. Voltage
Test 1 2 3 4 5 6 7 8
937
Current
Average
RMS
Average
RMS
32.4 28.5 43.4 34.7 24.0 23.5 32.3 43.0
32.7 28.7 43.6 34.9 24.4 24.3 32.6 43.2
249 250 255 251 155 149 250 251
249 250 255 251 158 149 250 251
greater variations in the current and voltage signals (see Figures 5– 10) and the increase in the amount of splatter observed when welding in this condition. The instability of the arc and the splatter has already been seen by Talkington2, who also stated that transfer in this polarity is normally limited to globular mode. This affirmation was only put in doubt by Test 6 (Ar þ 2%O2), in which the welding took place at a low level of current (150 A). Test 1 (Ar þ 2% O2) and Test 4 (Ar þ 18% CO2), carried out with the highest level of current (250 A), show that such an affirmation is not always valid, given that transfer in goticular mode was seen. Figure 11 shows the appearance of the weld beads and their respective cross sections revealed by macro graphic attack, while Table 3 sets out the values of the geometric parameters measured. Initially, it was analysed that the effect of the shielding gas (see the characteristics resulting from Tests 1 £ 4 and 2 £ 3, all at 250 A) confirm that shielding with 18%CO2 leads to increase in the width of the weld bead, providing weld beads that are less convex, but that also have a worse appearance. It is important to note that shielding with 18%CO2 also reduced the fusion rate of the electrode wire (or feed speed) in comparison with welds using Ar þ 2%O2 as shielding at the same current level (this aspect and the reduction in convexity is most evident in DC2), which must tend to encourage a reduction in width. The use of shielding with Ar þ 18%CO2 also led to increases in the welded area (i.e. greater efficiency in fusion of the base material) in the two polarities. Therefore, in regard to penetration and the profile of the weld bead, the effect of the shielding gas was shown to depend on polarity. In DCþ (Tests 2 £ 3), the influence of the composition of the shielding gas seems to be minimized by other effects, given that there were practically no differences in the depth of penetration and in the profile of the weld bead. This became clear in DC2 (Tests 1 £ 4), when the penetration fell by practically 100% when the shielding gas was changed from Ar þ 2%O2 to Ar þ 18%CO2.
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
938
D. Souza et al.
Figure 5. Behaviour over time of the voltage and current signals with a synchronized illustration of arc formation during the growth and detachment of drops, for Test 1 (Ar þ 2%O2; 250 A; DC2).
Now making comparisons to see the effect of the polarity (Test 1 £ 2 and 4 £ 3, all at 250 A), the use of DC2 with a shielding gas of Ar þ 2%O2 (Test 1 £ 2) led to a totally convex weld bead (in contrast to that when DCþ was used), but with a penetration level that reached almost 60% of the penetration obtained in DCþ. On the other hand, with shielding of Ar þ 18%CO2 (Test 4 £ 3), the use of DC2 now produced a weld bead with an acceptable convex level, but with poor appearance and low penetration (around 34% of that obtained with DCþ), eliminating the cup-type penetration format. The welded areas are much higher in DCþ . As a result, the question of penetration is also linked to the type of gas and not just to the polarity, as referred to in the literature (this reduces the probability of the heat generated in the anode connection being higher than that in the cathode connection, regardless of polarity, as illustrated).
In regard to the splatter, a greater amount of splatter can be seen to be formed in negative polarity when the shielding gas mixture contains CO2, confirming the consensus between the authors who have written about the subject. However, in relation to the mixture containing O2, this large quantity of splashes was not seen. This phenomenon can be explained by the fact that for the mixture containing CO2 (Figure 8) the arc ‘danced’ around the drop, thereby altering its direction, which did not occur with the use of the mixture with O2 (Figure 5), where it can be seen that the drops have a well-defined direction. By comparing Tests 1 £ 6 and 2 £ 5, one can try to verify the influence of the current, given that the gas was the same (Ar þ 2%O2). Naturally, the reduction in current led to a reduction in the volume of material deposited for each unit of weld bead length and the heat imposed, as well
Figure 6. Behaviour over time of the voltage and current signals with a synchronized illustration of arc formation during the growth and detachment of drops, for Test 2 (Ar þ 2%O2; 250 A; DCþ).
Welding International
939
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
Figure 7. Behaviour over time of the voltage and current signals with a synchronized illustration of arc formation during the growth and detachment of drops, for Test 3 (Ar þ 18%CO2; 250 A; DCþ).
Figure 8. Behaviour over time of the voltage and current signals with a synchronized illustration of arc formation during the growth and detachment of drops, for Test 4 (Ar þ 18%CO2; 250 A; DC 2).
Figure 9. Behaviour over time of the voltage and current signals with a synchronized illustration of arc formation during the growth and detachment of drops, for Test 5 (Arþ 2%O2; 150 A; DCþ).
940
D. Souza et al.
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
Figure 10. Behaviour over time of the voltage and current signals with a synchronized illustration of arc formation during the growth and detachment of drops, for Test 6 (Ar þ 2%O2; 150 A; DC2).
as reducing the effect of the amount of movement of the drops had an impact on the weld (greater drop diameter, with lower speed and frequency), reducing the penetration, width and welded area. The profiles of the weld beads were similar, although the scaling effect was lower.
Table 3. Geometric parameters measured at the cross sections of the test plates. Test 1 2 3 4 5 6 7 8
Welded area (mm2)
Width (mm)
Reinforcement height (mm)
Penetration (mm)
7.2 16.8 9.8 19.2 0.8 4.1 26.7 30.3
8.0 9.4 12.4 11.7 6.3 6.5 13.6 17.1
5.3 3.0 3.0 3.0 4.5 2.2 3.8 3.1
2.0 3.3 1.1 3.2 0.4 1.2 3.6 2.5
Note: The welded area and penetration refer to measures taken in area below the surface of the test board.
For the tests carried out using the same feed speed/weld speed ratio (compare Tests 1 £ 7 with Tests 4 £ 8), in other words with the same amount of material deposited per unit of length, the same trends seen in Tests 1 £ 2 and 4 £ 3 were once again observed. As a result, the use of shielding gas with Ar þ 2%O2 leads to an improvement in terms of penetration, generation of splatter and finishing, but it does not correct the lack of a suitable weld bead profile.
4.
Conclusions
For the welding conditions and parameters used in this work, it can be concluded that:
Figure 11. beads.
Appearance and transverse section of the weld
(1) The metal transfer mode in DC2 is dependent on the type of shielding gas used, but it is possible to obtain transfer without repulsive drops (globular and goticular), to the contrary of what has traditionally been said in the literature. (2) The high level of splatter in DC2 also pointed to as a feature in the literature was only observed when a mixture of Ar þ 18%CO2 was used. (3) The feature mentioned in the literature of there being a higher fusion rate in DC2 was confirmed, but the
Welding International results indicate that this was mainly due to the arc rising up the sides of the point of the wire in a search for oxides (for field emission), increasing heat efficiency, than by a greater amount of heat generated in the cathodic connection. (4) The shielding gas influences the geometric parameters in a different way, depending on the polarity used: Ar þ 2%O2 provides a weld bead profile in DC2 that has low malleability (unsuitable format) but with a certain amount of penetration. Acknowledgements The authors would like to thank CNPq for the financial support in providing grants and FAPEMIG for the financial support for infrastructure provided under the TEC 604/2005 project.
Downloaded By: [de Resende, Andre Alves] At: 15:49 28 January 2011
Notes 1. 2. 3.
Email:
[email protected] Email:
[email protected] Email:
[email protected]
References 1. AWS. Welding handbook: welding process. Vol. II, USA: AWS; 8th edn. 1991, chapter 4, 955 p. ISBN 0-87171-354-3.
941
2. Talkington BS. Variable polarity gas metal arc welding [masters dissertation]. The Ohio State University, 1998. 113p. 3. Lancaster JF. The physics of welding. 2nd ed. New York: Oxford: Pergamom; 1986. 4. Scotti A. A review on special metal transfer modes in GMAW. Rev Bras Mech Sci – RBCM, ABCM. 1998 Sep;XX(3):465 –478. ISSN 0100-7386. 5. Jonsson PG, Eager TW, Szekely J. Heat and metal transfer in gas metal arc welding using argon and helium. Metall Mater Trans B. 1995 Apr;26B:383 – 395. 6. Ushio M, Ikeuchi K, Tanaka M, Seto T. Effects of shielding gas on metal transfer. Weld Int. 1995;9(6):36 – 40. 7. Rhee S, Kannatey-Asibu E Jr. Observation of metal transfer during gas metal arc welding. Conference on Welding and Joining Processes, ASME, Atlanta, Georgia; 1991 Dec. p. 203– 213. 8. Lin Q, Li X, Simpson SW. Simpson metal transfer measurements in gas metal arc welding. J Phys D: Appl Phys. 2000;347 – 353. 9. Ba´lsamo PSS, Vilarinho LO, Vilela M, Scotti A. Development of an experimental technique for studying metal transfer in welding: synchronised shadowgraphy. Int J Join Mater. 2000;12(1). The European Institute for Joining of Materials (JOM), Denmark, pages 1 to 12. ISSN 0905-6866. 10. Vilarinho LO. Development and evaluation of an alternative algorithm for synergy MIG welding of aluminium [masters dissertation]. Federal University of Uberlaˆndia, Minas Gerais; 2000. 111 p.
