Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Laser-vapour interaction in high-power cw Nd:YAG laser welding (October 2003)

   
J Greses, P A Hilton, TWI, Granta Park, Great Abington, Cambridge, UK; C Y Barlow, Engineering Department, University of Cambridge, Cambridge, UK; W M Steen, Material Science Department, University of Cambridge, Cambridge, UK and Laser Group, University of Liverpool, Liverpool, L69 3GH, UK

Paper 1607 presented at ICALEO 2003 Conference, 13-16 October 2003, Jacksonville, Florida, USA

Abstract

During high-power cw Nd:YAG laser welding a vapour plume is formed containing vaporised material ejected from the keyhole. Spectroscopic studies of the vapour emission have demonstrated that the vapour can be considered as thermally excited gas with a stable temperature (less than 3000K), not as partially ionised plasma. In this paper, a review of temperatures in the vapour plume is presented. The difficulties in the analysis of the plume spectroscopic results are reviewed and explained. It is shown that particles present in the vapour interact with the laser beam, attenuating it. The attenuation can be calculated with Mie scattering theory, however, vaporisation and particle formation also both play a major role in this process. The laser beam is also defocused due to the scattering part of the attenuation mechanism, changing the energy density in the laser beam. Methods for mitigating the effects of the laser beam-vapour interaction, using control gases, are presented together with their advantages and disadvantages. This 'plume control' has two complementary roles: firstly, the gas must divert the vapour plume from out of the laser beam path, preventing the attenuation. Secondly, the gas has to stabilise the front wall of the keyhole, to prevent porosity formation.

1. Introduction

This paper deals with the laser-vapour interaction in high power cw Nd:YAG laser welding. In a previous paper [1] , also included in these proceedings, a review of the laser-vapour interaction in high power cw CO 2 laser welding was presented.

2. Review of the vapour temperature when Nd:YAG laser welding

Compared to the quantity of experimental results available on electron temperature and density in CO 2 laser welding [1] , only a few researchers have investigated the plume vapour temperature and electron density when pulsed or cw Nd:YAG laser welding.

Peebles and Williamson [2] detected atomic Fe lines - iron is a major contaminant in aluminium alloys - when pulsed laser welding of Al 1100. For an incident laser irradiance of 1.6MW/cm 2, the plume electron temperature was 3400±300K at a height of 0.5mm above the weld pool. The absolute upper limit of the plume temperature under any of their conditions was 4200K, with an electron density of 5x10 15cm -3. This corresponds to an ionisation fraction of only 0.3%. They calculated that, even in the case of a fully ionised Al plasma (at more than 10000K), the total inverse Bremsstrahlung absorption of the laser beam by the Al plasma would not exceed 2% for the Nd:YAG laser wavelength.

Matsunawa [3] investigated the spectroscopic emission from a Al-Mg alloy plume induced by pulsed Nd:YAG laser in air and in argon atmospheres. At 1mm above the surface, an electron temperature of 3280±150K was calculated using atomic chromium (Cr I) lines in the Boltzmann-plot method. Analysing the atomic and ionic magnesium (Mg I and II) lines, an electron density of 1.85x10 13cm -3 was calculated.

Lacroix [4] studied the plumes generated when pulsed Nd:YAG laser welding stainless steel and pure iron. In their experiments, electron temperatures based on the atomic Fe I lines varied in the range of 4500-7100K, and electron densities in the range 3x10 16cm -3 - 6.5x10 16cm -3, depending on height, pulse duration, average power and shielding gas flow rate.

Mueller [5] analysed the plume spectra when cw Nd:YAG laser welding galvanised steel. At 1mm above the keyhole aperture, a few iron line peaks were visible, but difficult to identify. Fitting the spectra emission to a blackbody radiation curve, the calculated temperature for the plume was about 2700K, higher than the weld pool temperature. However, atomic emission lines of aluminium, magnesium and manganese could be observed when cw Nd:YAG laser welding of 5000 and 6000 aluminium series. In addition, Mueller [5] pointed out that the lower melting point of these materials could account for the formation of a keyhole at much lower surface temperatures and even lower plume temperatures. However, no indication was given about the spectra being calibrated against the background and that particular spectrometer. In a similar way, Hurup [7] for the case of 2kW cw Nd:YAG laser welding of stainless steel, was able to fit the measured radiation to a blackbody radiation shape with a blackbody temperature of 1300K.

Greses [8] also applied the blackbody emission method for a calibrated plume spectrum. He stated that a plume consisting of a thermally excited gas (at around 2000K) could be expected when 3.5kW cw Nd:YAG laser welding mild steel, with the plume shape characteristics changing depending on the gas used as a control mechanism for the vapour. In addition, it was found that the plume temperature was stable and independent of the control gas parameters.

As a general result, plume temperatures when pulsed Nd:YAG laser welding seem to be higher (around 4000K) than the plume temperature when cw Nd:YAG laser welding (well below 3000K). The electron density of the plume has only been calculated when pulsed Nd:YAG laser welding, and with much lower results than when CO 2 laser welding. The methods for analysing both temperatures are different and that could account for some of the differences. However, atomic iron lines could not be identified by Mueller [5] nor Greses [8] when cw Nd:YAG laser welding mild steel, making the Boltzmann-plot method unusable under those circumstances.

In addition, the method of fitting and comparing the plume spectrum to the theoretical blackbody spectrum at any given temperature also presents some difficulties and problems. Blackbodies are theoretical objects with emissivity equivalent to one, meaning they both are perfect absorbers and emitters of radiation. [9] The plume vapour could be considered as a grey radiator or 'greybody', implying a spectral emissivity lower than one, but also constant over a particular spectral range. Since spectroscopic methods are also used to analyse the plume spectrum, the same problems and difficulties [1] explained in the review of electron plasma temperatures in high-power cw CO 2 laser welding apply to the use of Nd:YAG lasers. Nevertheless, apart from the assumption of considering the plume vapour to be a perfect blackbody, the blackbody method is much simpler and less prone to misinterpretation than the Boltzmann-plot method used to analyse the plasma when CO 2 laser welding.

In general, spectroscopic methods are expensive to perform and require considerable precision to apply correctly. Furthermore, interpretation of the data is problematic and based on questionable assumptions (such as local thermodynamic equilibrium in plasma conditions). A high-temperature Langmuir probe system10 could partially solve some of the problems encountered with spectroscopic methods to accurately determine the plume temperature.

3. Vapour and particle formation when Nd:YAG laser welding

The absorption of a high intensity beam by a metal can transform the solid structure of the target area into a melted one. The melt surface temperature can exceed the boiling temperature because the laser interaction, under typical conditions is accompanied only by surface evaporation, and volumetric evaporation, i.e. boiling, does not occur. [11] A description of the particle formation, clustering and size distribution, for Nd:YAG laser welding, was given by Greses. [12] The measured diameter of the evaporated particles present in plasma vapour when cw CO 2 laser welding at 3.5kW was around 4nm. Tu [13] used a different approach to measure and analyse the particles present in the plasma when 20kW CO 2 laser welding, but also found a mean particle diameter of 4nm. Greses [12] argued that the plasma effect over the Knudsen layer, which controls the vaporisation process, reduces and homogenises the size of the CO 2-induced particles, although the mechanism for this is not clear. The sizes of the evaporated particles when welding with 3.5kW of Nd:YAG laser power were also measured in the previous work. [12] The measured average particle diameter was 40nm, with the particle size distributed over a wide range of diameters up to a maximum size of 160nm, although 98% of all particles had a diameter smaller than 100nm.

Researchers have used different methods for collecting the particles generated when laser welding. Therefore, it should be taken into account that the measured size of the particles could be affected by a number of factors, including the distance from the weld pool/keyhole (where the particles are generated), the type of the collector, the material of the collector (i.e. plastic, glass, etc), the laser energy intensity, the mode of laser operation (i.e. pulsed or cw) and other laser or welding parameters. Invariably, the preferred method for analysing particle size is by measuring the size from high-resolution TEM (transmission electron microscopy) photographs. Therefore, since each experiment is unique in its own way, direct comparison between results from various researches is difficult and may explain the observed size variations.

4. Extinction coefficient

From the temperature and electron density results presented in section 2, plasma absorption by inverse Bremsstrahlung is not possible when cw Nd:YAG laser welding due to the Nd:YAG wavelength being 10 times shorter than that of CO 2 laser light. [1] Changes in the refractive index of the vapour, which are dependent on a gradient of electron density, are not as steep as in CO 2 laser welding, if present at all. Nevertheless, similar effects to those caused by plasma formation when CO 2 laser welding, i.e. loss of penetration and wine-glass cross-section shapes at low speeds, have been noticed when cw Nd:YAG laser welding at high power. [14] Therefore, despite lower temperatures and uncertainty about the electron density in the plume, it is apparent that the plume still plays a major role in attenuating and/or defocusing the laser beam.

Mie scattering applies to the scattering of light by a group of spheres. All the spheres should be of the same diameter and composition and they should also be randomly distributed and separated from each other by distances that are large compared to the wavelength of the light. [15] The rigorous demonstration of Mie scattering theory for spheres (i.e. particles) of arbitrary size is outside the scope of this paper, but can be found in [16] . Van de Hulst [16] also gives some simple formulae for general use in a medium containing a number of particles per unit volume ( N). The extinction coefficient ( α EXT ) in that medium, independently of the state of polarisation of the incident light, is caused by scattering and absorption (also in [17] ) and can be defined as

α EXT = Q EXT Πr 2N    
[1]

 

where r is the average radius of the particles, N the number of particles per unit volume, and Q EXT , is given by

spjgoct2003e2.gif
[2]

where m is the complex refractive index ( m = n + ik) and λ the laser wavelength.

The first part of equation (2), is recognised as the absorption term ( Q ABS ), and within this only the first fraction is generally significant. The second part is the scattering term ( Q SCA ), where again only the first part of the expanded series is relevant and has been included in equation (4). This formula is valid if the second part of the scattering term [16] is less than 0.2. In this case, Q EXT = Q ABS + Q SCA , with

spjgoct2003e3.gif
[3]
spjgoct2003e4.gif
[4]

Equation (4), Q SCA , is the well-known Rayleigh scattering formula, which is a special case of Mie scattering theory. [15] The term Q SCA was also derived by Matsunawa and Ohnawa [18] , but using the complex refractive index rather than the dielectric constant. Although in equation (4), the scattering term is inversely proportional to the 4 th power of the wavelength, this does not necessarily mean that the scattered loss is also proportional in the same way, since the refractive index of most metals, including iron, strongly depends on the wavelength. [16]

The extinction, i.e. absorption plus scattering, coefficient, approaches a definite limit if the size of the particle increases. As indicated before, equations (1-4) will be valid if the second part of the scattering term in equation (2) is less than 0.2, which translates [16] into the term 2 Πr 2/ λ being less than 0.3. This is true when the particle diameter is smaller than 1/10 th of the wavelength of the light. [9] For the Nd:YAG laser wavelength (1.064µm), the maximum particle radius for general validity of the equations (1-4) is therefore 50nm. Since the particles in the plume (see section 3) have been measured with an approximate radius of less than 25nm (less than 50nm in diameter), the given equations appear applicable. It should also be stressed that the evaporated mass, i.e. number of particles per unit volume ( N), plays the main role in the attenuation.

For pure Fe radiated with 1µm, the complex refractive index [19] ( m) equates to 3.81+4.44i and considering an average particle radius of 20nm, then from equations (3) and (4), Q SCA = 5.492x10 -4 and Q ABS = 4.497x10 -2. The scattering ( Q SCA ) can be considered practicably negligible compared to the absorption, for very small particle radius.

A different problem faced when using this refractive index is the issue of the composition of the particles. For the attenuation results in the paper by Greses [12] , the particles were presented as being pure iron and the appropriate refractive index has been used. However, an analysis of the particles showed a combination of particles made of pure iron and its oxides, as also indicated by Tu [13] for CO 2 results. Only a very small quantity of oxygen in the plume is required for the iron particles to oxidise. Matsunawa [20,21] analysed the gas composition inside the pores, showing that entrained trapped oxygen reacted with the base metal forming oxides on the wall of the pore. However, most of the trapped gas inside the pore came from the gas used as plume control or shielding gas, indicating that due to the instabilities of the keyhole it is possible for the gas, even the atmospheric gas which contains oxygen, to enter the keyhole. Although it can be assumed that oxidation takes place both during welding and after welding is completed, it may be a good approximation to use the complex refractive index of pure iron for the attenuation calculations.

In general, the extinction coefficient is very difficult to determine with accuracy, not only due to the difficulty of calculating the refractive index of the evaporated particles that form the plume, but due to the uncertainly in the values of particle density and radii.

5. Plume effects

5.1 Attenuation

The laser intensity I(z) transmitted through the plume-laser interaction zone along a path length z can be described by the Beer's law as

spjgoct2003e5.gif
[5]

where I 0 is the incident intensity of the laser beam and α EXT the extinction coefficient from Mie scattering theory, equation (1).

Tsubota [22] directly measured the attenuation of a probe laser (different laser types were tried: CO 2, Nd:YAG and He-Ne) crossing a plume perpendicular to the main Nd:YAG laser beam. For the CO 2 probe wavelength, the beam attenuation was nearly constant, at 5%, along the length of the plume. For the Nd:YAG probe, the beam attenuation varied from over 50%, at a height of 3.5 mm from the weld pool, to around 20% at 15 mm. For the He-Ne probe laser, the attenuation was even higher than observed for the Nd:YAG wavelength. [18] Both Tsubota [22] and Matsunawa [18] pointed to Rayleigh scattering as the main mechanism for beam attenuation.

A different approach for understanding effect of the plume on the laser beam, was studied by Russo. [23] For a single Nd:YAG laser pulse (i.e. no keyhole formation), a Huygens-Kirchhoff diffraction model was used with several plume shapes, at atmospheric pressures. The plume temperature decreased from the boiling point of the metal at the top of the weld pool, to ambient values several cm above the surface. The laser power at the centre of the plume was significantly reduced and the laser power intensity redistributed over a larger radius. Unfortunately, the effect of plume diffraction when cw keyhole Nd:YAG welding was not included in this research, but this topic should be further investigated for a better understanding of the effect of the plume on the laser beam.

Greses [12] measured the attenuation of a 9W Nd:YAG probe laser beam, horizontally incident across the plume generated by a high-power Nd:YAG laser, at various positions with respect to the beam-metal interaction point. Up to 40% attenuation, that is absorption plus scattering, was measured at positions corresponding to zones of high concentration of vapour plume. The attenuation decreased along the plume, away from the top of the keyhole as the volume ratio (particle density) decreases. It was found that an increase in attenuation of the incident beam significantly reduced weld penetration and fusion area [24] , although the use of a gas side jet to 'blow' the plume away from the laser interaction path was proven as an effective method of reducing the attenuation.

A knowledge of the evaporated mass contained in the plume is critical to estimate the attenuation of the laser beam by the plume. Greses [25] found that the evaporation rate when welding approximately doubled, from ~0.012g/s to ~0.024g/s, when changing from a Nd:YAG laser of 2kW to 3.5kW. The focal spot was kept constant at 0.6mm. However, since the vaporisation rate is dependent on the energy density, the plume formation and associated attenuation will also affect the vaporisation rate. As explained in section 4, the plume formation changes the energy density via attenuation of the laser beam.

5.2 Scattering

Lacroix [26] modelled the scattering effect in the plume using a radiative transfer model in a semitransparent medium. The scattering produced a significant defocusing effect on the laser beam, which increased with high particle volume ratios. For larger particles (around 1µm) the plume temperature decreased (due to less absorption as reflected in equations (3) and (4)).

Another aspect that has to be considered when using Mie scattering theory is the direction of the scattered radiation. Following the same approach that derived equations (3) and (4) from Mie scattering theory (equation (2)), then the intensity of the scattered light for perpendicular ( I Perp ) and parallel ( I Parall ) polarisation is defined as [16]

spjgoct2003e6.gif
[6]
spjgoct2003e7.gif
[7]

where r is the particle radius, λ the laser wavelength, m the complex refractive index of the particles and θ the scattering angle.

Equations (6) and (7) represent Rayleigh or elastic scattering. For a particle radius of 20nm, m=3.81+4.44i and λ =1.064 (Nd:YAG), the intensities are plotted in polar co-ordinates in Figure 1. This was generated using MiePlot v2.0 developed by Laven. [27]

Fig. 1. Intensity of the scattered light for perpendicular and parallel polarisation based on equations (6) and (7) for Rayleigh scattering using MiePlot v2.0 from Laven [27]
Fig. 1. Intensity of the scattered light for perpendicular and parallel polarisation based on equations (6) and (7) for Rayleigh scattering using MiePlot v2.0 from Laven [27]

Lacroix [27] calculated that for the Nd:YAG laser wavelength, a particle diameter of 50nm and m=1.21+1.3x10 -3i, the incoming beam was scattered isotropically in the forward and backward direction, with minor deviations on the side direction of the incoming beam. For bigger particles of 1µm diameter, keeping the same wavelength and complex refractive index, there is a departure from the plane of symmetry, with more light being scattered in the forward direction. This phenomenon is often called the Mie effect. [15] Some researchers have tried to control the welding process by monitoring the back-reflected light through the optical delivery system of the laser. [28] However, it is not clear if the collected light is coming from the Mie effect, emitted by the plume vapour or directly reflected from the metallic surface or the weld pool. Hansen and Duley [29] also investigated the angular distribution of the scattered light, leading to similar results to those obtained by Lacroix. [26]

6. Plume control

When Nd:YAG laser welding [12] , the interaction (i.e. attenuation) of the laser beam by the plume above the surface of the workpiece reduces welding efficiency. However, as opposed to the many different ways of controlling the plasma formation when CO 2 laser welding [1] , only one method of control, by suppressing or restricting the plume vapour formation above the keyhole using a gas side jet, has been applied successfully to date when cw high-power Nd:YAG laser welding. Olivier and Gerritsen [30] and Greses [12,24] showed that when Nd:YAG laser welding, the ionisation potential of the gas was not as important as in CO 2 laser welding but that the position of the side jet was critical for high penetration and quality in the weld.

Reilly [31] demonstrated, in a theoretical approach to the interaction between a side jet and a supersonic vapour emerging from the keyhole, the importance of the dynamic pressure of the side jet on the plume. However, the vapour jet is not supersonic during laser welding. [32] Nevertheless, the dynamic pressure of the side jet is very important and is influenced by many parameters, such as type of gas, nozzle diameter, side jet position, etc.

Douay [33] presented a study of a free jet, modelling the behaviour of the gas as it exits the tip of the side jet nozzle. When this model was tested against an idealised plasma emerging from a keyhole when CO 2 laser welding, it was found that a helium side jet flow with a Mach number of only ~0.4, far from supersonic, was necessary to divert the idealised plasma by 90degrees. The idealised plasma vapour was based on experimental data that estimated a flow rate of 0.04g/s and a velocity of 150m/s. Following on Douay's work [33] , Hamadou [34] also simulated the interaction between the plasma vapour and the gas side jet, by 3D modelling of how the thermo-convective effects from the heated surface and the flow induced by the ejected plasma could perturb the control gas.

Kamimuki [35] studied the prevention of welding defects by the use of a gas side jet when cw Nd:YAG laser welding. Gas delivery angles which are too small (10°) and too large (40°) result in humping formation. With an increase of the gas side jet pressure, the bead width became narrower and the penetration depth became deeper, although too much pressure also caused humping bead formation. It was found by Kamimuki [35] that when the position of the gas side jet is optimum, the front wall of the keyhole changes the direction of the vaporised flow. The resultant backward vapour flow pushes the molten metal to the rear, and the width of the keyhole is increased. A deep hollow is thus formed in the molten pool just behind the keyhole. The molten metal near the surface flows rapidly backwards along the backward vapour flow. By the formation of this hollow, the distance between the molten surface and the tip of the keyhole shortened, allowing bubbles generated at the tip of the keyhole to easily come up to the pool surface. However, by the weak molten metal flow along the bottom of the molten pool, some of the bubbles move along the bottom part of pool and are trapped at the solidifying front, remaining as porosity. This backward liquid flow in the weld pool, which is generated by the side jet, enlarges and stabilises the keyhole opening. As a result, the Nd:YAG laser beam can reach directly to the bottom of the keyhole and thus the penetration depth increases. Reduced porosity formation when CO 2 laser welding, results more from increased plasma stability than the widening of the keyhole. [1]

The enlargement of the keyhole and weld pool by dual [36] , and triple focus beam welding [37] , has been emerging lately as a method to control the keyhole and/or the plume formation. Dual and triple focus beam welding was originally intended as a porosity control mechanism, although it was found that this could also benefit plume control. Filler wire addition could play a similar role to that of multi-focus beams, enlarging the keyhole and weld pool. [38,39]

The gases used in plume control do not differ from those whose characteristics were described in a previous paper [1] , although the ionisation property is not as important. [1] However, there is more use of heavier gases such as to 'blow away' the plume, when Nd:YAG laser welding.

7. Discussion and conclusions

It is reasonable to state that the temperature of the plume vapour when Nd:YAG laser welding mild steels lies between 2000 and 3000K. In contrast to the plasma temperature when CO 2 laser welding, the calculation of plume temperature is not sensitive to the ratio between atomic/ionic lines since these lines cannot be detected. Furthermore, it is believed that the blackbody method to calculate the temperature in the plume is less problematic than the Boltzmann-plot method used to calculate plasma temperature.

As in CO 2 laser welding [1] , the use of different control gas conditions when Nd:YAG welding does not affect the plume temperature, but does affect the size and extent of the plume vapour. The size of the plume is extremely important as this determines the particle density within it, and therefore the attenuation coefficient due to Mie scattering. As discussed in section 4, the laser beam will be absorbed by the plume and part scattered. Due to the size of the particles forming the plume when Nd:YAG laser welding (less than 50nm in radius), the attenuation of the Nd:YAG laser beam is mainly due to absorption, with only a small part of the beam being scattered. However, enough absorption and scattering is produced by the plume when high-power Nd:YAG laser welding, to defocus the laser beam and change its energy density. A defocusing of the laser beam will reduce the maximum penetration that can be achieved at a certain energy density and processing speed. Defocusing is also responsible for changes in the weld shape, mainly the wine-glass cross section, and porosity formation, due to the plume instabilities causing keyhole collapse. It was found that optimisation of the penetration can be achieved using a side jet of argon gas delivered at a high flow rate and using a small diameter, precisely positioned nozzle. [24] However, the tolerance in the gas side jet position when Nd:YAG laser welding was very narrow (less than 1mm with respect to the impingement position of the laser beam with the workpiece). This is due to the delicate balance between the two roles of the control gas. Firstly, the gas side jet diverts the plume ejected from the keyhole from the interaction path with the laser beam. Secondly, the gas side jet further opens the keyhole opening stabilising the keyhole front wall.

8. Comparison of results of Nd:YAG and CO 2 laser wavelengths

As explained earlier [1] , the aim of this work was a better understanding of the physical processes involved in the interaction between the laser beam (of different wavelengths) and the vapour ejected from the keyhole, and its effects. In summary, the differences between temperature and composition in the vapour outside the keyhole for CO 2 and Nd:YAG laser welding of mild steel have been established. These differences in temperature and composition are responsible for the effects that the plasma/plume vapour produces in the weld, being mainly the weld shape and its penetration. When CO 2 laser welding, defocusing of the laser beam due to a gradient of electron temperature, density and refractive index in the plasma changes the beam energy distribution and focal position. For Nd:YAG laser welding, defocusing due to scattering is also largely responsible for the weld shape, although the absorption part of the attenuation is responsible for changes in penetration. These conclusions should help in understanding the steps necessary to provide higher quality and repeatable keyhole welds under high-power laser conditions.

As a final conclusion, a summary of the overall project, highlighting the plasma/plume formation characteristics and its effects in CO 2 and Nd:YAG high-power laser welding, are presented in Table 1 and 2.

Table 1 Summary of plasma/plume effects in CO 2 and Nd:YAG high-power laser welding.

Laser sourceCO 2CommentsNd:YAGComments
Characteristics
Wavelength 10.6µm   1.06µm  
Vapour temperature 6500K to 12000K Partially ionised plasma formed due to inverse Bremsstrahlung absorption in the laser beam 2000K to 3000K Thermally excited gas due to the low radiative transfer from attenuation
Measured particle diameter in vapour outside the keyhole ~4nm The plasma affects the Knudsen layer (formed between the liquid and vapour phase of the keyhole), minimising and homogenising the particle size. ~40nm The particle size is distributed over a large range due to lower temperatures in the vapour, possible caused by localised high temperature areas in the keyhole or vapour.

Table 2 Summary of plasma/plume effects in CO 2 and Nd:YAG high-power laser welding (continuation Table 1).

Laser sourceCO 2CommentsNd:YAGComments
Characteristics
Vaporisation Rate Not measured in this work The effect of the plasma on the Knudsen layer also reduces the vaporisation rate. ~0.024g/s Depends on the energy density of the laser beam.
Loss mechanism Plasma absorption causing beam defocusing Although the plasma absorption is small when using an optimised control gas, the absorbed energy is enough to create a gradient of electron temperate and density, resulting in a gradient of refractive index in the plasma. Attenuation (absorption and scattering) Mie particle scattering theory is applicable to calculate the energy loss in the plume. The major part of the energy is absorbed, with a small part being scattered. Attenuation mainly depends on the vaporised mass in the plume and the refractive index of the particles.
Recommended process control gas Helium A high ionisation potential is important. Argon The molecular weight of the gas is important. CO 2 and N 2 can also be used depending on the material characteristics.
Energy transfer efficiency with optimised control gas at low speeds (< 1 m/min) 100% with a helium control gas Plasma formation may enhance the energy transfer between the beam and the workpiece. The incident energy density is modified due to beam defocusing. 85-90% with an argon control gas Plume formation attenuates the energy transfer between the beam and the workpiece. The incident energy density is modified due to beam defocusing caused by Mie scattering.
Energy transfer eff. at high speeds (> 1 m/min) 100% or close The laser-plasma interaction does not effect to a great extent the transfer efficiency. 100% or close The laser-plume interaction does not effect to a great extent the transfer efficiency.
Penetration By doing a comparison between CO 2 and Nd:YAG laser welding at 3.5kW using optimised control gas conditions, the penetration was higher for the CO 2 laser light than for the Nd:YAG laser light. The higher penetration could be due to the higher beam quality of the CO 2 laser.
Weld shape Wine-glass at low speeds (<1m/min) Wine-glass at low speeds (<1m/min)
The weld shape is caused by the defocusing of the laser beam as it crosses the plasma. The plasma has its higher electron temperature and density inside the keyhole, close to the top part, re-radiating part of the absorbed energy into the keyhole, maybe stabilising the front keyhole walls. The weld shape is caused by the defocusing of the laser beam via the Mie scattering mechanism as it crosses the plume.
The geometrical characteristics of a gas side jet help to stabilise the front keyhole wall to prevent porosity formation, as well as diverting the plume away from the path of the laser beam.

9. Acknowledgements

TWI equipment and the dedication of the staff of the Laser Department made this work possible. Part of this research was carried under EPSRC grant No. 99313089.

10. References

  1. Greses, J., P.A. Hilton, C.Y. Barlow and W.M. Steen. 2003, Laser-Vapour Interaction in High-Power cw CO 2 Laser Welding. Proceedings of ICALEO´03. Paper P546.
  2. Peebles, H.C. and R.L. Williamson. 1987, The Role of the Metal Vapor Plume in Pulsed Nd:YAG Laser Welding on Aluminum 1100. Proceedings of LAMP'87, pp. 19-24 and SAND86-2219C.
  3. Matsunawa, A., J.D. Kim, T. Takemoto and S. Katayama. 1995, Spectroscopic Studies on Laser Induced Plume of Aluminum Alloys. Proceedings of ICALEO'95, pp. 719-728.
  4. Lacroix, D., G. Jeandel and C. Boudot. 1997, Spectroscopic Characterization of Laser-Induced Plasma Created during Welding with a Pulsed Nd:YAG. J. Appl. Phys. 81:6599.
  5. Mueller, R.E., Gu, H., Ferguson, N., Ogmen, M. and W.W. Duley. 1999, Real Time Optical Spectral Monitoring of Laser Welding Plumes. Proceedings of ICALEO'99.
  6. Schauer D.A., W.H. Giedt and S.M. Shintaku. 1978, Electron Beam Welding Cavity Temperature Distributions in Pure Metals and Alloys. Weld. J., pp. 127-s, May.
  7. Hurup, K., A. Glansholm, M. Hietanen, P. von Nandelstadh and K. Schröder. 1995, Proceedings of the 3 rd EUREKA Industrial Laser Safety Forum, pp. 217-322.
  8. Greses, J., P.A. Hilton, C.Y. Barlow and W.M. Steen. 2001, Spectroscopic Studies of Plume/Plasma in Different Gas Environments. Proceedings of ICALEO 2001, Jacksonville, Florida (USA), paper 808.
  9. Meyer-Arendt, J.R. 1984, Introduction to Classical and Modern Optics. Prentice-Hall Inc., New Jersey.
  10. Fanara, C. and I.M. Richardson. 2001, A Langmuir Multi-probe System for the Characterization of Atmospheric Pressure Arc Plasmas. J. Phys. D: Appl. Phys., 34:2715.
  11. Semak, V.V., R.J. Steele, P.W. Fuerschbach and B.K. Damkroger. 2000, Role of Beam Absorption in Plasma during Laser Welding. J. Phys. D: Appl. Phys, 33:1179.
  12. Greses, J., P.A. Hilton, C.Y. Barlow and W.M. Steen. 2002, Plume Attenuation under High Power Nd:YAG Laser Welding. Proceedings of ICALEO 2002, Scottsdale, Arizona (USA), paper 808.
  13. Tu, J., I. Mimayamoto and T. Inoue. 2002, Characterizing Keyhole Plasma Light Emission and Plasma Plume Scattering for Monitoring 20 kW class CO 2 Laser Welding Process. J. Laser Appl. 14:146.
  14. Olivier, C.A., P.A. Hilton and D.K. Russell. 1999, Materials Processing with a 10 kW Nd:YAG Laser Facility. Proceedings of ICALEO'99, p. D233.
  15. Born, M. and E. Wolf. 1999, Principles of Optics. 7 th (expanded) edn. University Press, Cambridge, England.
  16. van de Hulst, H.C. 1957, Light Scattering by Small Particles. Dover Publ., New York.
  17. Nonhoff, C.J. 1988, Material Processing with Nd-Lasers. 1 st edn. Electrochemical Publications Ltd., Ayr, Scotland.
  18. Matsunawa, A. and T. Ohnawa. 1991, Beam-Plume Interaction in Laser Materials Processing. Trans. JWRI, Vol. 21, No. 1.
  19. American Institute of Physics Handbook, 1972, ed. by E.G. Gray, MacGraw-Hill.
  20. Matsunawa, A. 2001, Problems and Solutions in Deep Penetration Laser Welding. Sci. Technol. Weld. Joining, 6:351.
  21. Matsunawa, A. 2003, Beam-plasma Interaction and Weld Pool Dynamics in Laser Welding. Presentation to the Laser Department at TWI, 20-1-2003.
  22. Tsubota, S., T. Ishide, M. Nayama, Y. Shimokusu and S. Fukumoto. 2000, Development of 10 kW Class YAG Laser Welding Technology. Proceedings of ICALEO'00, pp. C219-229.
  23. Russo, A.J., Akau, R.L. and J.L Jellison. 1991, Thermocapillary Flow in Pulsed Laser Beam Weld Pools. Weld. J. 69:23s.
  24. Greses, J., C.Y. Barlow, P.A. Hilton and W.M. Steen. 2002, The Effect of Different Gas Environments on CO 2 and Nd:YAG Laser Welding Process Efficiencies. Proceedings of International Congress on Laser Advanced Materials Processing (LAMP2002), Osaka, Japan, May 2002. To be published by SPIE.
  25. Greses, J. 2003, Plasma/Plume Effects in CO 2 and Nd:YAG Laser Welding. Ph.D. Thesis, University of Cambridge.
  26. Lacroix, D., G. Jeandel and C. Boudot. 1998, Solution of the Radiative Transfer Equation in an Absorbing and Scattering Nd:YAG Laser-induced Plume. J. Appl. Phys. 84:2443.
  27. Laven, P. 2003, MiePlot v2.0, downloaded from www.philiplaven.com [Accessed January 2003].
  28. Steen, W.M. and V.M. Weerasinghe. 1986, High-Power Beam Diagnostics: Part 2. Proceedings of the 6 th International Symposium on Gas Flow and Chemical Lasers. Israel, Sept. 1986. Springer-Verlag, pp. 323-330.
  29. Hansen, F. and W.W. Duley. 1994, Attenuation of Laser Radiation by Particles during Laser Materials Processing. J. Laser Appl., 6:137.
  30. Olivier, C.A. and C.H.J. Gerritsen. 2000, Study into the Use of Process Gases in High Power Nd:YAG Laser Welding of C-Mn and Stainless Steels. TWI Report 746.01/00/1070-2.
  31. Reilly, J.P. 1990, Debris Plume Phenomenology for Laser/Material Interaction in High Speed Flowfields. Proceedings of SPIE 1397:661.
  32. Bungartz, K. 1992, Bestimmung der Stroemungsgeschwindigkeit des Metalldampfes im Bereich des Keyholes beim Schweissen mit CO 2-Lasern. Diploma Thesis, RWTH Aachen.
  33. Douay, D., F. Daniere, R. Fabbro, and L. Sabatier. 1996, Study of High Power Laser Welding: Effects of Plasma Blowing. Proceedings of SPIE 2789:202 and Proceedings of ICALEO'96, pp. D54-63.
  34. Hamadou, M. R. Fabbro, G. Caillibotte, K. Chouf and F. Briand. 2002, Study of Assist Gas Flow Behaviour during Laser Welding (Poster). Proceedings of ICALEO'2002, paper P538.
  35. Kamimuki, K., T. Inoue, K. Yasuda, M. Muro, T. Nakabayashi and A. Matsunawa. 2002, Prevention of Welding Defect by Side Gas Flow and its Monitoring Method in Continuous Wave Nd:YAG Laser Welding. J. Laser Appl., 14:136.
  36. Gref, W., B. Hohenberger, F. Dausinger and H. Hügel. 2002, Energy Coupling and Process Efficiency in Double-focus Welding with Nd:YAG. Proceedings of ICALEO'2002, paper 807.
  37. Fabbro, R., F. Coste and F. Janin. 2002, Deep Penetration Laser Welding with Nd:YAG Lasers Combination up to 11 kW Laser Power. Proceedings of LAMP'2002, paper 2368.
  38. Salminen, A., V.P. Kujanpaa, and T.J.L. Moisio. 1996, Interaction between Laser Beam and Filler Metal. Welding Journal, January, pp. 9s-13s.
  39. Salminen, A. 2002, The Effects of Wire Feed Parameters on Filler Wire Laser Beam Interaction. Proceedings of ICALEO'2002, pp.1105.

Meet the authors:

José Greses joined the Laser and Sheet Processes Group of TWI (UK) in May 2003, after completion of his PhD on laser welding at the Engineering Department of the University of Cambridge (UK). Paul Hilton is Technology Manager at the Laser and Sheet Processes Group at TWI (UK), Claire Barlow is Senior Lecturer at the Engineering Department of the University of Cambridge (UK) and Bill Steen is Emeritus Professor at the University of Liverpool and Distinguished Research Fellow at the Univ. of Cambridge (UK).

For more information please email:


contactus@twi.co.uk