Behaviour of spreading molten metal drops deposited by fusion

Experimental Thermal and Fluid Science 48 (2013) 29–36

Contents lists available at SciVerse ScienceDirect

Experimental Thermal and Fluid Science journal homepage: www.elsevier.com/locate/etfs

Behaviour of spreading molten metal drops deposited by fusion J. Chapuis ⇑, E. Romero, F. Soulié, C. Bordreuil, G. Fras Laboratoire de Mécanique et Génie Civil (LMGC), Université Montpellier 2, CNRS, cc048, Place Eugne Bataillon, 34095 Montpellier Cedex, France

a r t i c l e

i n f o

Article history: Received 29 June 2012 Received in revised form 8 January 2013 Accepted 3 February 2013 Available online 26 February 2013 Keywords: Capillarity Wetting Heat transfer Metal transfer Liquid metal droplet Welding

a b s t r a c t Liquid droplet deposition on solid surfaces has an important role in the industrial and research activities. The behaviour of such deposit is influenced by volume and interfacial phenomena and involves a large number of mechanisms such as gravity effect, mass transfer, capillary forces, and wetting. For the case of metal deposition the analysis of the problem is more complex because of the importance of thermal effects, involving steep gradients and phase changes. A unique experimental approach is presented in order to study the evolution of the spreading of a large drop of liquid metal called ‘‘macro-drop’’. The objective of this work is to supply qualitative and quantitative information during the deposit of liquid metal in relation with process parameters. The overall shape of the macro-drop, especially its spreading and contact angles are studied in detail. The gradual spreading of the macro-drop is mainly governed by mass and heat transfers. The initial rapid spreading is due to kinetic energy of depositing droplets and direct arc heating on the solid target. All experimental results are analysed in the light of process parameters to identify the physical mechanisms involved and appreciate their effects on the behaviour of such a macro-drop. Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction Deposition of small liquid droplets on solid surfaces is an active research topic in a large panel of industrial applications such as ink printing, thermal spray coating, micro-fabrication, soldering or welding [1–4]. The impact of droplet on solid surface has been extensively studied and the evolution of spreading base and height of impacted drop have been described. Whatever the material used (ink, wax, solder, etc.), one of the most important shape parameter is the contact angle, which affects the processes of flattening or wetting on the solid surfaces. The contact angle could be considered as an equilibrium contact angle in the case of isothermal problems or as solidification or apparent dynamic contact angles in the case of non-isothermal problems. For example, Schiaffino and Sonin studied apparent dynamic contact angles and the shape of drops obtained by continuous droplets deposition of wax [5]. When inertia effects are negligible and the evolution is mainly dominated by capillary and viscous forces, they show that the apparent dynamic contact angle seems to obey Hoffman’s law [6]. The spreading and wetting are then closely linked to the behaviour of contact line and contact angles. The presence of solid–liquid–gas interfaces, and related interfacial phenomena, play also an extremely important role in the process called ‘‘to high temperature’’ such as welding [7,8], even in normal conditions of

operating. In Gas Arc Metal Welding (GMAW) process particularly, these interfaces take an important place in the metal transfer both in the arc and in the weld pool (Fig. 1). The shape of the droplets is directly linked to the competition between volume forces, such as gravity or electromagnetic forces, and interfacial phenomena such as surface tension or drag forces. The heat and mass transfers and the above mentioned interfacial phenomena have a great influence on the evolution and behaviour of the weld pool. The shape of the weld pool thus determines the final quality of the welding operation. The main goal of the present work is to better understand the physical mechanisms involved in the behaviour of liquid metal deposition in relation with pulsed Gas Arc Metal Welding parameters (P-GMAW). Their relative importance can be appreciate through the study of the spreading and wetting of a stationary weld pool (called macro-drop in this paper) obtained by liquid metal droplets deposition on a solid target. These problems are mainly studied by numerical approach and compared with experimental results [3,9,10]. A specific experimental study will be presented in this work; it focuses on the effects of the modification of the process parameters on the behaviour of the macro-drop. The experimental approach is described (Sections 2 and 3) and the results are discussed in relation with the welding parameters (Section 4). 2. Deposition of liquid metal droplets on a solid target

⇑ Corresponding author. E-mail addresses: [email protected] (J. Chapuis), [email protected] (F. Soulié). 0894-1777/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.expthermflusci.2013.02.005

The purpose is to study the shape and the spreading of the macro-drop according to the heat and mass transfers supplied by

30

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

Nomenclature Ca Bo cp h L R T V u

capillary number Bond number specific heat height of the macro-drop enthalpy of melting base radius of the macro-drop temperature volume of the macro-drop speed of the contact line

the deposition of droplets from feed solid wire with the stationary P-GMAW process. This operation produces a direct heating at the surface of the macro-drop (liquid–gas interface) and heat transfer by convection in the liquid macro-drop and conduction in the solid. The macro-drop shape evolves with the heat and mass supplied by the GMAW process. The formation and spreading of the macrodrop result from the energy (heat) and mass conservations; the arc supplies direct heating to the liquid–gas surface of the macro-drop and to the solid target whereas liquid droplets input heat and mainly mass in the macro-drop. The global shape and behaviour are also influenced by the balanced effects of several physical mechanisms such as gravity, capillarity, inertia, thermic effects, and viscosity. The macro-drop is assumed to be axisymmetric with low penetration (the liquid metal forms a drop of liquid at the surface of the solid target). Fig. 2 presents the main geometric parameters necessary to describe the evolutions of the shape of the macro-drop and of the contact line (line at the solid–liquid–gas interface). The geometry of the macro-drop is defined by the height h, the base radius R defining the base length and the contact angle h at the solid– liquid–gas interface. These measured parameters are used to determine other parameters describing the behaviour of the macro-drop

We

Weber number

Greek symbols h contact angle / wire diameter q density l viscosity c surface tension k thermal conductivity

such as its volume V or the speed of spreading u corresponding to the contact line speed. The physical characteristics required to describe and to analyse the phenomena involved in the behaviour of the macro-drop are given in Table 1. 3. Experimental method This section presents the unique experimental setup dedicated for multiphysics studies, the experimental matrix for the study of macro-drop behaviour and the developed numerical libraries to analyse experimental data by a systematic way. 3.1. Experimental setup Experiments are realised on a platform dedicated to the study of arc welding processes [11]. This platform allows the synchronised data acquisition of different kinds of signal (process, thermal, mechanic) and high speed video during arc welding. Arc welding is a really harsh environment because of the perturbation due to electromagnetic noise and radi-

Fig. 1. Physics description of Gas Metal Arc Welding (GMAW) during a static operation.

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

    

31

welding time: 4 s, wire feed speed: 6 m/min, frequency of droplets: 113 Hz, initial temperature of solid target: 293 K gas composition: 8% CO2 and 92% Argon.

In order to study the behaviour of the macro-drop, we vary several parameters as those cited for the reference test. The choice of these different parameters is given in Table 2; ‘R’ corresponds to the values for reference test, ‘S’ for variation of welding durations, ‘WFS’ for variation of wire feed speed, ‘F’ for variation of frequency of droplets deposit, ‘T’ for variation of initial temperature of solid target and ‘G’ for variation of chemical gas composition (mixture of Argon and CO2). Fig. 2. Geometrical description of the deposition and spreading of macro-drop during GMAW static operation.

Table 1 Numerical values of physical characteristics [3].

c q l cp Tf L k

Properties (units)

Values

Surface tension (N m1) Density (kg m3) Viscosity (kg m1 s1) Specific heat (J kg1 K1) Fusion temperature (K1) Enthalpy of melting (J kg1) Thermal conductivity (W m1 K1)

1.67 7200 0.06 753 1798 2.77  105 26

ation from the arc. Specific developments were realised for this platform to allow measurements to be done in this environment. Stationary spot welds are made using the Pulsed Gas Metal Arc Welding process (Oerlikon CitoWave 500) to observe non-isothermal spreading of weld pools. The target is a steel disk of 10 mm in thickness and 150 mm in diameter. The contact tip to workpiece distance is equal to 20 mm and ER70S steel welding wire (/ = 1 mm) is used for welding experiments. Welding current and arc voltage are recorded at 30 kHz sampling rate. During the experiments the temperature is measured on the surface of the solid target at a distance of 20 and 30 mm from the centre of the metal drop by means of a Type K thermocouple of 0.5 mm in diameter. This temperature measurement allows us to check the axisymmetry configuration of the deposit. Moreover, the low penetration of the weld (1–2 mm at the centre) is controlled post-mortem by macrography. A high-speed camera (Phantom V5.0) with a back-lit shadow graphic method records weld pool images at a rate of 4000 frames per second so that liquid metal drop profile radius and apparent liquid–solid contact angle histories can be measured (see Fig. 3). A 650 ± 10 nm band pass filter is used to attenuate arc radiation for clear images of weld pool growth and weld metal transfer. Some examples of these pictures are given in Fig. 4 that illustrates the macro-drop formation and spreading. Three main periods can be identified. The first one correspond to the arc initiation (Fig. 4(1)) with short-circuiting and erratic globular metal transfer (metal spatters) corresponding to the first droplets deposit (Fig. 4(2) and (3)). During the second period, the pulsed metal transfer is well established and a regular spreading of the macrodrop until the arc extinction can be observed (Fig. 4(4)–(6)). The last period concerns the solidification of the macro-drop after the arc extinction. 3.2. Experimentation The test campaign is based on a reference test (marked ‘R’), for which the parameters are the following:

3.3. Data processing: numerical libraries In complement to the experimental platform of synchronised data acquisition, two numerical open source libraries were developed. They offer the possibilities of easy management of the large flow of multi-physical experimental data (up to 2 GB a test) in order to compare and analyse the experimental results. The library BAME1 is used for the analyses of the whole data [12], whereas the library erCv2 is dedicated to image processing [13,14]. Measurements of weld pool dimensions are made using the erCv (cf. Fig. 5) and BAME libraries. The erCv library is designed to determine the contour lines of specific domains like droplets or weld pools in images, with huge flows of processed images (up to 30,000 frames a test). Several functions have been developed (using the BAME and erCv libraries) in order to extract the values of several geometrical characteristics; this extraction is done during the ‘‘cold’’ periods of welding operation (periods with lower energy) in order to decrease the disturbances due to arc radiations. It is then possible to measure the evolution of the height h, the base radius R with time on the base of the pictures supplied by the video recorded. The apparent contact angles h are determined with linear regression or least square method on the filtered profile of the macro-drop. This method allows the treatment of more than 1000 frames per test to be done. It is then possible to study the evolution of geometrical parameters with welding time but also computed information such as the welding energy or the volume of the macro-drop. 4. Experimental results This section presents first the overall behaviour of the macrodrop based in results for ‘R’eference tests. Then it is showed and discussed results illustrating the effect of several chosen parameters of process controlling the deposition and spreading of liquid metal on the target. 4.1. Global behaviour: ‘R’eference tests Some characteristics of evolution can be observed for all the experiments. These main characteristics are presented and discussed for tests in reference configuration (‘R’ tests). 4.1.1. Contact line The evolution of the shape of the macro-drop can be divided in three main periods. The first one corresponds to the creation of the macro-drop, the second one to a regular spreading and the last one to the solidification. At the very beginning of experiments, the di1 2

https://subver.lmgc.univmontp2.fr/BAME. https://subver.lmgc.univmontp2.fr/erCv.

32

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

Fig. 3. Experimental set-up for back-lit shadow graphic recording.

(1) - 0.1 s

(2) - 0.4 s

(4) - 1.2 s

(5) - 2 s

2 mm

(3) - 0.8 s

(6) - 3 s

Fig. 4. Examples of frames that can be extracted from the high speed video recording during the P-GMAW operation.

Table 2 Values of experimental parameters used in tests. Test code

Welding time (s)

Wire feed speed (m/min)

Frequency (Hz)

Target temp. (K)

Gas (vol.% CO2)

R S WFS F T G

4 4, 6, 8 4 4 4 4

6 6 4, 6, 8, 10 6 6 6

113 113 113 40, 70, 113, 200 113 113

293 293 293 293 293, 573, 873 293

8 8 8 8 8 0, 8, 18, 30

rect arc heating creates slight changes in the solid target with the creation of a small fusion zone corresponding to a low penetrated weld pool. The droplets (1 mm diameter) do not then directly impact the solid target but a liquid layer of metal in a first time. Fig. 6 shows the evolution of the base of the macro-drop during spot welding. The base radius of the macro-drop increases like a square root of time in agreement with the inertial spreading of liquid drops [15]. The beginning of experiment corresponds to transient period: arc initiation and creation of the macro-drop (initiation of the droplets deposit) with a rapid increase of the base radius. This initial rapid macro-drop spreading is due to kinetic energy of depositing droplets on flat surface and direct arc heating of the solid target. This direct arc heating directly contributes to facilitate the rapid spreading of the macro-drop. At the beginning of welding operation, during the process of impact, the first droplets that create the liquid macro-drop are forced to spread by inertia force, and this spreading is opposed by surface tension and viscous forces. A huge increase in the base radius between 0.5 s and 0.65 s is ob-

served. This evolution is a combined effect of the rapid increase of the base of the macro-drop at the beginning of welding and of the numerical method based on threshold elevation to detect the macro-drop on frames. The initial macro-drop spreading speed is quite huge in the beginning of spreading. The average speed of displacement of the contact line is about u = 4.06 mm/s (during the first 1.4 s, cf. Fig. 6). After this rapid evolution, the spreading of the macro-drop is more gradual, and the heat flow is not directly ensured by the arc heating but by conduction from the melting zone to the solid target. The regular growth of the macro-drop is supplied by the mass transfer due to wire feed. The average spreading speed of the macro-drop is three times lower than in the first step (u = 1.23 mm/s, for t > 1.4 s, cf. Fig. 6). The capillary number C a ¼ lcu is lower than 2  105 and then underlines the minor effect of viscosity in regard of surface tension. 2 The Bond number Bo ¼ R Dcqg increases from low values at the creation of the macro-drop to 2.5 at the end of experiments. These values underline a quite well balanced contribution of gravity and capillary effects on the macro-drop, with a progressive higher

33

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

Fig. 5. Contour detection with erCv library for several steps of macro-drop spreading.

influence of the gravity. This evolution is in adequation with the increasing mass of the macro-drop due to the droplets deposit and the observed evolution of the macro-drop shape from a spherical cap to a paraboloid shape. At the end of welding operation, arc heating and droplets inputs stop and the macro-drop spreading does not continue by default of energy and mass supplies. Only the protective gas remains active during few seconds in order to avoid corrosion phenomenon. We can observe that, at the arc extinction, the weld pool spreading is stopped with arresting of the contact line. The inertial effect can then be supposed negligible in regards of surface tension effects. 2 Moreover, the Weber number W e ¼ quc R in the range 103–102

for all experiments shows the importance of surface tension effects on inertial effects in the governing of macro-drop shape and spreading.

4.1.2. Contact angle The evolution of contact angles during the spot welding is similar for both left and right sides (Fig. 7). The evolution of contact angle is influenced by the physico-chemical properties of liquid and solid, by the cleanliness and the roughness of the solid target [16,17]. But it is also affected by fluid dynamics and deformation of the macro-drop; the contact angles can then vary quickly in dy-

8

Contact angle (deg)

Base radius of the macrodrop (mm)

80 9

7 6 5 4

60

40

20 Left contact angle Right contact angle

3 0

2 0.5

1

1.5

2

2.5

3

3.5

4

Time (s) Fig. 6. Evolution of the base radius of the macro-drop with time. This evolution is characteristic of the ones obtained for all the tests, only data for ‘R’ test are plotted.

0.5

1

1.5

2

2.5

3

3.5

4

Time (s) Fig. 7. Evolution of the left and right contact angles of the macro-drop with time. Data are plotted for ‘R’ test.

34

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

namic configuration and are very difficult to precisely measure. The observed disturbances and minor asymmetry are mainly due to the oscillations of the weld pool because of the dynamic configurations and of the non-symmetric arc and input droplets. We can also observe two periods of evolution. The first one, corresponding to the macro-drop creation, is characterised by quite huge variations of contact angles due to arc initiation and droplets impacts on flat surface with small amount of liquid metal. The second one, with lower variation in magnitude, corresponds to the growth of the macro-drop with constantly increasing volume and regular spreading. 4.1.3. Solidification The arc extinction stops the welding solicitations on weld pool such as heat transfer and arc pressure on the liquid surface but also

the mass transfer supplied by the droplets impingement in the macro-drop. Just after the arc extinction, the macro-drop remains totally liquid before the solidification process begins. This characteristic time, corresponding to ‘overfusion’, equal to 0.25 s in this case and translates the necessary time for the cooling down to the temperature of phase change. Some oscillations of the liquid surface can be observed; they indicate return to mechanical equilibrium after the welding solicitation stops. This oscillatory behaviour is attenuated with thermal equilibrium and the beginning of the solidification (the effect of viscosity increases). The solidification begins at the bottom part of the macro-drop, at the level of the contact line, and progressively reaches the top of the macrodrop (Fig. 8). The necessary time for global solidification is 1.6 s. The observed speed of solidification front is quasi-constant and equals to 2.2 mm/s. The post-solidification shape of the macro-

Fig. 8. Different steps of solidification process: from the end of welding to the complete solidification of the macro-drop.

Fig. 9. Contour detection of the macro-drop with erCv for several chemical compositions of the welding gas (mixture Argon and CO2, ‘G’ tests). Pictures are extracted from several experiments, at same time (2.5 s) during a low current period.

35

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

4.2. Wetting: ‘G’as tests The gas composition can significantly influence several physical characteristics and the global behaviour of the macro-drop. It modifies the gas–liquid interface behaviour through the surface tension and the efficiency of the gas contributing to arc and plasma creation. When the shielding gas is only composed of argon, it is called ‘inert’. It changes to an ‘active’ gas by the introduction of CO2 in the composition. The CO2 notably modifies the ionisation potential of gas and changes the distribution of energy supplied by the arc. The wetting of the macro-drop can also be influenced by the chemical composition of materials involved in the formation of the macro-drop. This phenomenon is clearly shown in Fig. 9 that illustrates the improvement of wetting with the amount of CO2 in the gas mixture. The experiments are realised in the same power conditions. This wetting improvement is clearly underlined in Fig. 10; at a same level of the droplets deposit operation, the values of contact angle decrease with the increasing amount of CO2 in gas mixture. The amount of CO2 directly modifies the surface tension magnitude at the liquid–gas interface, and thus directly modifies the wetting of the macro-drop. This wetting effect also theoretically influences the spreading of the macro-drop; the base of the macro-drop is modified with the amount of CO2 as shown in Fig. 11, but it is less significant than the effect of contact angle. We can notice a particular behaviour of the evolution of the base radius in the case of pure argon gas; some steps appear during the evolution of the macro-drop. They correspond to contact line arresting; the spreading of the macro-drop stops. During these periods, the contact angle considerably increases until its limit value for which the contact line starts moving again. 4.3. Mass and Heat transfers: ‘WFS’ tests The increasing wire feed speed has a clear influence on the evolution and shape of the macro-drop as shown in Fig. 12. The growth

12 10

Base radius (mm)

drop does not correspond to a spherical cap, a flattening can be observed on the top part. This shape is mainly due to the axisymmetric shrinkage because of solidification and can also be increased by degassing phenomenon at the end of welding. The same type of characteristics was observed for all types of experiments. Some differences can appear in relation with the variation of experimental parameters and will be presented. The ‘F’ and ‘S’ tests did not show specific behaviour and are not discussed.

6 4

CO 2 : 0% CO 2 : 8%

2 0 0.5

20

CO 2 : 0% CO 2 : 8% CO 2: 18% CO 2 : 30%

2

2.5

3

3.5

4

of the macro-drop is clearly linked to the mass transfer and kinetic energy due to droplets deposits but not only. When the wire feed speed increases, the global free shape of the macro-drop changes from a spherical cap to a paraboloïd cap. The mass transfer and surface tension effects cannot explain alone this change, it is also due to the increasing energy associated to the wire feed speed increase. The increased welding power supplies more energy to the macrodrop system that facilitates the spreading of the macro-drop. 4.4. Heat transfer: ‘T’ tests This effect of heat and energy can also be observed on the results of T-type experiments. Fig. 13 shows the evolution of the average contact angle with the increasing volume of macro-drop for several initial temperatures of the solid target. We can observe a clear difference in the values of contact angle between no-preheating configuration (initial temperature equals to 293 K) and preheating configurations (593 K and 893 K). The preheating facilitates the wetting of the macro-drop. Nevertheless, it seems there are no difference inbetween the two configurations of preheating. The different experimental results show the importance of capillary and surface tension effects in the behaviour of the macrodrop, mainly in the definition of its global shape. Nevertheless, as

Height of the macrodrop (mm)

40

1.5

Fig. 11. Evolution of the base radius of the macro-drop for several chemical compositions of the welding gas (mixture Argon and CO2, ‘G’ tests). This graph describes the obtained wetting for the several chemical compositions.

5

60

1

Time (s)

80

Contact angle (deg)

8

Wire speed: 4 m/min Wire speed: 6 m/min Wire speed: 8 m/min Wire speed: 10 m/min

4

3

2

1

0 0.5

1

1.5

2

2.5

3

3.5

4

Time (s) Fig. 10. Evolution of the contact angle for several chemical compositions of the welding gas (mixture Argon and CO2, ‘G’ tests). This graph describes the obtained wetting for the several chemical compositions.

0

2

4

6

8

10

12

14

16

Base radius of the macrodrop (mm) Fig. 12. Evolution of the height of the macro-drop in relation with its base radius for several wire feed speeds (‘WS’ tests). The measured data give indication on the global shape of the macro-drop and its spreading.

36

J. Chapuis et al. / Experimental Thermal and Fluid Science 48 (2013) 29–36

Left contact angle (deg)

80

60

40

20 T: 293 K T: 573 K T: 873 K

0 0.2

0.4

0.6

0.8

1

Dimensionless volume Fig. 13. Evolution of the contact angles in relation with the normalised volume of the macro-drop for the different initial values of the temperature of the solid target (‘T’ tests).

target that leads, with kinetic energy, to an initial rapid spreading of macro-drops, (2) a more gradual spreading of the system piloted by the conduction heat flow from the melting zone to the solid target, and (3) a time of ‘overfusion’ just after the extinction of the arc and the beginning of the solidification of stationary weld pool. Also, it is observed the important effects in function of the arc nature and initial heat content in the metal target onto the wetting of the macro-drop for a constant volume and an equivalent power delivered by the process. Then among the numerous of physical phenomena involved in the behaviour of this system as shown by the obtained data, two main phenomena have been identified: surface tension effects and thermal effects. The qualitative and quantitative analysis of experimental data in such a ‘basic’ configuration show the potential of this approach. Further developments are in progress in order to extend to more complex problems such as undercutting or humping that can happen in high speed welding operations.

Acknowledgements This work was partially supported by the ANR project 2007JC3838, France. It is a pleasure to thank Denis Cervellin for helpful discussions.

Contact angle (deg)

70

References 60

50

40

1x10-6

1.5x10-6

2x10-6

2.5x10-6

3x10-6

3.5x10-6

Capillary number Fig. 14. Contact angle vs capillary number Ca. The graph synthesises the data obtained for the different types of experiments. The values of contact angle and spreading speed required for the calculation of Ca are the average final values obtained before the welding operation stops.

Fig. 14 shows, direct relationship between contact angle evolution and capillary number Ca like in Hoffman’s law [6] cannot be observed. Attinger et al. [4] did not find too quantitative agreement with Hoffman’s law in their work of solder droplets impacting on wafer. The spreading of the contact line seems to be not directly linked to capillary effects. Rather, the spreading of the macro-drop appears to be governed by mass, heat and energy transfers. 5. Conclusion A specific experimental approach was used in static P-GMAW process to measure geometry histories of a macro-drop of liquid metal during its deposit. The experimental results allowed us to study the spreading of this macro-drop in the very noisy environment of arc welding. The results show that the evolution of the shape of macro-drops passes through three phases: (1) a direct arc heating onto the solid

[1] F. Gao, A. Sonin, Precise deposition of molten microdrops: the physics of digital microfabrication, Proceedings of the Royal Society of London, Series A (Mathematical and Physical Sciences) 444 (1922) (1994) 533–554. [2] R. Li, N. Ashgriz, S. Chandra, J. Andrews, Solidification contact angles of molten droplets deposited on solid surfaces, Journal of Materials Sciences 42 (2007) 9511–9523. [3] Y.C. Lim, D.F. Farson, M.H. Cho, J.H. Cho, Stationary GMAW-P weld metal deposit spreading, Science and Technology of Welding and Joining 14 (17) (2009) 626–635. [4] D. Attinger, Z. Zhao, D. Poulikakos, An experimental study of molten microdroplet surface deposition and solidification: transient behavior and wetting angle dynamics, Journal of Heat Transfer 122 (2000) 544–556. [5] S. Schiaffino, A. Sonin, Formation and stability of liquid and molten beads on a solid surface, Journal of Fluid Mechanics 343 (1997) 95–110. [6] R. Hoffman, A study of the advancing interface, Journal of Colloid and Interface Science 50 (2) (1975) 228–241. [7] T. DebRoy, S. David, Physical processes in fusion welding, Reviews of Modern Physics 67 (1) (1995) 86–112. [8] K. Mills, E. Hondros, Z. Li, Interfacial phenomena in high temperature processes, Journal of Materials Science 40 (2005) 2403–2409. [9] H. Fan, R. Kovacevic, Droplet formation, detachment, and impingement on the molten pool in gas metal arc welding, Metallurgical and Materials Transactions B 30B (1999) 791–801. [10] M. Cho, Y. Lim, D. Farson, Simulation of weld pool dynamics in the stationary pulsed gas metal arc welding process and final weld shape, Welding Journal 85 (12) (2006) 271s–283s. [11] J. Chapuis, Une Approche Pour l’optimisation des Opérations de Soudage à l’arc, Ph.D. Thesis, Université de Montpellier 2, March 2011. [12] J. Chapuis, C. Bordreuil, F. Soulié, G. Fras, Structuration et archivage de données lors de campagne d’essais en soudage à l’arc, Soudage et Techniques Connexes (2011) 31–35. [13] E. Romero, J. Chapuis, C. Bordreuil, F. Soulié, G. Fras, Edge detection of weld pool, macrodrop and metal transfer drop in a GTAW and GMAW process, by a new weld image processing library: Ercv, in: Proc. of the Int. Conf. on Advances in Welding Science and Technology for Construction, Energy and Transportation, 2010, pp. 565–569. [14] E. Romero, J. Chapuis, C. Bordreuil, F. Soulié, G. Fras, Image processing and geometrical analysis for profile detection during p-gas metal arc welding, Proceedings of the Institution of Mechanical Engineers, Part B, Journal of Engineering Manufacture (2012). [15] A.-L. Biance, C. Clanet, D. Quéré, First steps in the spreading of a liquid droplet, Physical Review E 69 (016301) (2004) 1–4. [16] P.-G. de Gennes, Wetting: statics and dynamics, Reviews of Modern Physics 57 (3) (1985) 827–863. [17] D. Quéré, Rough ideas on wetting, Physica A 313 (2002) 32–46.