Helmut Huegel

Aspects of keyhole/melt interaction in high-speed laser welding

The increasing availability of highest power lasers broadens the economic use of deep penetration laser welding by increasing maximum welding depth and by increasing welding speed. The problems encountered in those two areas of interest differ significantly. Whereas the maximum welding depth is strongly affected by the absorption mechanism1,2, the maximum welding speed is limited by fluid dynamic phenomena . Theoretical modelling allows to seperately investigate the different processes.

Modeling and investigation of melt ejection dynamics for laser drilling with short pulses

The presence of melt during the laser drilling process always signifies a balance between an efficient material removal in molten form and a reduction of quality due to recast on the hole walls and near the crater entrance. Earlier investigations have demonstrated that by reducing the laser pulse duration the amount of produced melt can be decreased and hence, the precision increased. Nevertheless, they also demonstrate that melt can never be avoided completely. Therefore, to achieve an optimum balance between efficiency and quality by a preferably complete expulsion of melt the physical fundamentals of its generation and ejection have to be understood. By applying several different analytical and numerical models ranging from simple estimations to multi-dimensional simulations, the authors will outline the peculiarities of the melt formation and dynamics during the drilling with short and ultra-short laser pulses. Since these calculations demonstrate the importance of the consideration of melt acceleration and geometric aspects, special interest will be taken in these matters. While the evaporation stops soon after the laser pulse, the melt ejection may continue until the complete solidification of the material. For a better understanding and verification, the results of the models will be compared to experimental data.

Fast and compact adaptive mirror

Adaptive mirrors are used in laser material processing machines in order to control the propagation of the raw beam through the beam-guiding system as well as the geometry of the focused beam. As modern machine concepts for laser material processing tend to operate at increasing working speed, fast and lightweight adaptive mirrors with excellent optical properties and a large range of radius variation are needed. In this paper, a new approach for a fast, compact and lightweight adaptive mirror is discussed. In the approach the deformation of a circular plate under pressure is used. The pressure is generated by a small piezoelectric actuator which presses against a small water layer behind the mirror plate, thereby deforming it. The deformation of the mirror surface is almost parabolic across a large area of the overall surface, which results in a high optical quality even at large deformations. Based on this concept, an adaptive mirror was designed that can be operated at a bandwidth up to 400 Hz at maximum deformation (which results in a range of radius-variation from -20 m to +20 m) and about 2 kHz at smaller deformations. Furthermore, extensive investigations have been performed using FEM analysis of the adaptive mirrors in order to understand the influence of mirror thickness, diameter and operating pressure on the deformation and the resulting radii of curvature. The result is an analytical method which allows to calculate the mirror dimensions in order to achieve a desired characteristic of deformation for given beam diameters.

Influence of wire addition direction in C02 laser welding of aluminum

Laser beam welding of aluminum alloys is expected to offer both the technical and economical advantages. In most cases welding wire addition is necessary from the viewpoints to suppress hot cracks, improve mechanical properties of the weld as well as to reduce demands on the edge preparation, fit-up tolerance, and beam alignment. In practice, the filler wire could be added to the weld pool either in the leading or trailing direction. In this experimental work reported, the influence of the wire addition direction on the weld efficiency and process stability was investigated by using a 5kW CO2 laser to weld aluminum alloy 6009 with a plate thickness of 3mm. Beam-on-plate welds were made either in the autogeneous mode or with filler wire AISi12. Illuminated by a double frequency Nd:YAG laser, the weld pool dynamics and the wire melting process were observed applying a high speed camera. The experimental results demonstrate that welding with filler wire in the trailing direction is more efficient and stable than in the leading direction. High speed camera photographs show that the filler wire is mainly melted through weld pool heating and plasma heating in the former case and through direct laser irradiation and plasma heating in the later case. The weld pool is of vibration when welding in the autogeneous mode or with filler wire in the leading direction. However, the weld pool is much calmer when the wire is added in the trailing direction to the weld pool.

Shadowgraphic and interferometric investigations on Nd:YAG laser-induced vapor/plasma plumes for different processing wavelengths

Shadowgraphic and interferometric analysis of vapor/plasma plumes generated by pulses of a ns-Nd:YAG laser display a distinctive morphology for each of three different processing wavelengths (1064 nm, 532 nm, and 355 nm). Analytical models of shock wave expansion can be applied to determine the total energy stored within the plumes. Electron density distribution measurements reveal the regions of absorption and, using a simple model of inverse bremsstrahlung and photoionization, allow us to estimate absorption coefficients.

Autostable injection locking of a 4x4 VCSEL array with on-chip master laser

Stable phase-locking of a VCSEL array of N equals 16 emitters with the master laser on the same chip is demonstrated. To accomplish the injection-locking a very small portion of the master radiation is seeded frontally into the 16 slave lasers. All VCSELs are driven by one voltage source with small individual series resistors to compensate for frequency differences. The beams are collimated by a microlens array and are commonly focused. A factor 14 increase of the peak power density of the superimposed beams compared to the unlocked operation is achieved (overall system coherence 85%) without any phase control. The coherent operation is stable for hours without any sophisticated voltage or temperature control. By slightly varying the current of each emitter within the locking range a maximum phase shift of (pi) can be achieved for each emitter. In this way residual phase differences between the individual beams can be compensated. The fraction of the master power necessary to lock one slave laser is below 10-3. Therefore, scaling to very large arrays is possible. Apart from simply increasing the peak power density of the chip, a promising perspective for data transmission applications is the GHz-modulation of the system coherence. The locking of the array can be switched by a very small (2%) modulation of the master current thereby switching the peak power density by a factor of nearly N.

Studies on condensation phenomena and refraction index distributions in excimer laser-induced plasma/vapour plumes

In this contribution, the condensation theory of I. M. Lifshitz and V. V. Slyozov is used to calculate the formation of metal clusters in the plasma/vapor regime above an aluminum target in a nitrogen atmosphere of 1 bar irradiated with 248 nm, 30 ns FWHM pulses. This theory takes into account formation of clusters by fluctuations in the supersaturated medium (plasma/vapor). It also accounts for both cluster growth due to atom-cluster-collisions (nucleation) and collisions between clusters (coalescence state). This condensation model is integrated in a gasdynamic code which calculates the flow field and the thermodynamic properties of the plasma/vapor. In addition, the density of electrons, clusters, and neutral particles are determined and, finally, with the appropriate polarizability the refraction index can be calculated. The theoretical data are discussed together with the results of interferometric measurements.

Energy transfer in the laser ablation of metals

We outline results obtained from Schlieren and dye laser resonance absorption imaging of the plume ejected from an aluminum target in a nitrogen atmosphere of 1 bar and 100 mbar by a KrF excimer laser (lambda equals 248 nm, FWHM equals 30 ns). The results show that for relatively low and high laser fluences (14 J/cm2 and 36 J/cm2), the plume closely follows the shock wave which is generated by the ablated material pushing against the surrounding gas. Calculations of the evolution of the ambient gas and ablated material show that the temperature and electron density vary greatly depending on the laser fluence and the external gas pressure. We report maximum plasma temperatures of 39888 K and electron densities of 4.2 multiplied by 1026 m-3 for a laser fluence of 36 J/cm2. These results indicate that inverse bremsstrahlung may play a very significant role in how the laser pulse energy is distributed in the plume for high laser fluences.

Coherent fiber coupling of laser diodes

Laser diodes with diffraction-limited beam quality offer high power densities of the order of 107 - 108 W/cm2, but are limited in output power to some watts. Scaling to higher powers has to be realized by superposition of a number of laser diodes. Coherent superposition allows us to further increase the power density in the far field. This is realized by injection locking of three slave laser diodes (Toshiba TOLD) 9140, 20 mW, 690 nm) by one master laser diode (TOLD 9140) and superpositioning of the three slaves lasers by a lens array. The feedback of the slaves into the master is suppressed by two Faraday isolators. For superpositioning the light of the slaves while maintaining the high beam quality, the light of each diode is coupled into an optical single-mode fiber. Phase shifts due to mechanical or thermal disturbances of the single-mode fibers for frequencies up to 1 kHz are compensated by a single-mode optical fiber piezoceramic phase modulator and an electronic control circuit. A phase stability with a maximum phase error smaller than 6 degrees is kept over an hour. The power-density distribution in the focal plane of the focusing lens shows a peak power 2.6 times that of the incoherent superposition and a modulation corresponding to the Fourier transform of the nearfield distribution of the lens array.

Diagnostic techniques and process monitoring of pulsed laser welding

Different diagnostic techniques were used to investigate melt pool instabilities and droplet generation during the welding process with pulsed lasers. In order to find the causes and the dependencies for these effects, an experimental setup was built up that allows the measurement of three different quantities. The luminosity of the welding plasma together with the instantaneous laser power were recorded by a personal computer. High speed video pictures were taken concurrently and were temporally correlated with the other signals to understand the basic phenomena. The results of experiments, which were done for a variety of several parameters, are discussed and serve as boundary conditions for a theoretical model, which helps us to understand the complex hydrodynamic mechanisms.