Experimenting with beam oscillation for metal hardening
Handika Sandra Dewi explores the effects of circular, square and triangular oscillation strategies when treating microalloyed steel
Laser surface treatment can be used to increase the resistance of metals to environmental attack when sufficient resistance cannot be attained by alloying addition. For example, hardening the surface of a metal using a laser is one strategy for improving its fatigue properties. The laser increases local surface temperature and induces martensitic transformation, which increases hardness and induces residual stress characteristics that hamper fatigue crack propagation through the material.
Laser hardening is a precise and energy-efficient method that offers flexibility when hardening complex structures, such as crankshafts for internal combustion engines. However, the commonly used Gaussian laser beam profile produces non-homogenous depth in the hardened area.
The depth homogeneity of laser-treated zones is one factor that can define the quality and efficacy of altered mechanical properties in the material. Since the depth homogeneity depends on the laser beam profile and shape, customised laser beam shapes can tackle this problem.
For example, doughnut and top-hat beam profiles can be of particular interest for laser hardening. A doughnut beam profile, which has higher energy intensity at the edge compared to the centre of the beam, has been reported to produce homogeneous treated areas. A top-hat beam profile, which has homogenous energy intensity across the beam shape and therefore delivers equal energy input to the specimen, also creates a homogenous treated area.
Another developing technique that could be of interest to laser hardening is beam oscillation, in which special optics use mirrors to oscillate the laser beam at a specific frequency and pattern, while moving across a material’s surface.
The technique influences the distribution of input laser energy, and has been shown to increase efficiency in laser welding by increasing weld depth and influencing both temperature gradient and solidification rate.
With efforts to optimise laser hardening still underway, in addition to the continual advancement of oscillating optics, combining the two fields – using beam oscillation to influence laser hardening – becomes an interesting topic to study.
My colleagues and I have therefore experimented with laser surface treatments on microalloyed steel disks using a variety of oscillation strategies. This was so that we could observe their effect on the quasi-static laser beam profiles and the geometry of the treated areas. Microalloyed steels are commonly used in industry as a base material for crankshafts.
In this work we used oscillating optics recently developed by Permanova Laser System. The optics have two main components: the optical unit and the control system. The optical unit consists of a collimation and focusing lens, water cooling, cover slides, temperature and safety sensors, and two additional oscillating mirrors. The mirrors perform very fast rotations around a central axis, resulting in a movement of the focus spot in two directions relative to the laser head.
The control system features software where adjustments can be made to laser power, oscillation speed and oscillation pattern. The software provides predefined settings that can be modified as desired. The laser power and scan speed can also be set to change during the oscillation to achieve variable energy density over the scanning area.
These optics have a 1:1.76 optical ratio, a scanning area of +/- 10mm, can oscillate at up to 500Hz for 1mm amplitude and can support laser power up to 8kW for wavelengths from 1,030 to 1,070nm.
Figure 1a illustrates our experimental setup. The oscillation optics were mounted to a moving robot and integrated with Trumpf’s 1,030nm TruDisk laser. The robot moved along the X-axis, producing single, straight, lasertreated tracks. Circular, square, and triangular oscillation strategies (illustrated in figures 1b, c and d) were performed using constant laser parameters and a 1mm beam spot.
Figure 1: (a) illustration of the experimental setup for (b) circular, (c) square and (d) triangular oscillation strategies including top view (e), (f), (g)
Top views of the tracks (see figures 1e, f and g) show the characteristic shapes achieved with each oscillation strategy. While the beginning and end sections of the tracks were not fully treated, the middle section of each track was fully treated for each oscillation strategy. Small regions of melted material were found at the edge of the square track, while three vivid lines were found on the edge and centre of the triangular track. Respective cross-sectional images (figures 1h, i and j) show that the circular oscillation strategy produces homogenous depth, while the melted regions and vivid lines of the square and triangular strategies result in a locally deeper treated area.
Figure 1 continued: Cross-sectional (h), (i), (j) images of the resulting tracks and quasistatic laser beam profiles at different interval times (k) – (s)
Quasi-static laser beam profiles of the oscillation strategies were visualised by reconstructing the laser beam pathway at 0.5s, 1s and 2s (shown in figures 1k to s). A vector graphic editor was used to stack semi-transparent drawings of the oscillation strategies along the robot’s path. The amount of assembled drawings represents the number of oscillations at a specific interval time. As a result, figures 1q, r and s have similar features to the top view of the track in figures 1e, f and g.
The most important feature through figures 1k to s is the contrast. The brightness corresponds to how often the laser beam passes through the area. The brighter the area, the more often the laser beam passes through. Therefore, brighter features on the image can be associated with higher local energy input due to multiple passes of the laser beam. It is now clear why the beginning and end sections of the tracks were not fully treated. The laser beam only passes through this area once, and thus the energy input was not sufficient to induce martensitic transformation in the material.
The microstructure of the melted regions at the edge of the square oscillation strategy resulted in increased softness compared to the non-melted regions. This means that the hardness is not homogeneous across the treated area.
For the triangular oscillation strategy, while a homogenous microstructure was achieved across the treated area, due to the vivid lines it demonstrated nonhomogeneous depth, which hampers the improvement of fatigue properties. Therefore, the circular oscillation strategy, which demonstrated both homogenous depth and hardness, was considered the best choice for performing laser hardening via beam oscillation.
Mechanical limitations cause additional energy input
Temporal aspects and mechanical limitations of the oscillating optics could explain the melted regions and vivid lines on the square and triangular tracks respectively. Imagine the oscillating laser beam passing through a standing point at the crosssectional position of the tracks during the process. The energy input that standing point experiences over time ranges from X0 to X1 on figures 1q, r and s. The centre of the triangular oscillation track received energy input from 0s, while the neighbouring edges did not. Accordingly, the temporal accumulation of energy input is higher at the centre of the track compared to the neighbouring edges.
However, the speed of the oscillating laser beam decelerates at the vertices of square and triangular oscillation strategies due to mechanical limitation, thus causing additional local energy input at these points. This additional energy input contributes to the melting of the edges of the square oscillation track and the production of the vivid lines on the triangular oscillation track.
In general, the beam oscillation strategies produced shallower treated areas compared to a common Gaussian beam or other beam profiles such as top-hat or doughnut, as well as those produced using diffractive optical elements. In terms of the energy needed over the volume of the treated track, beam oscillation consumes more energy than the above beam shaping techniques. However, we will conduct further experiments using a larger beam spot diameter and different oscillation strategies that may be able to improve the efficiency of beam oscillation. Despite its lower efficiency, beam oscillation has shown flexibility in tailoring energy input, and thus the geometry of the treated area without the need for changing any optics. Therefore, beam oscillation should indeed be considered for laser hardening and other surface treatment applications.
Oscillated laser beams influence both spatial and temporal energy input distribution during laser surface treatment through overlapping oscillation patterns. While triangular and square oscillation strategies produce quasi-static laser beam profiles with equal distribution and an abrupt drop in the energy intensity from the edge peak of the profile, a circular oscillation strategy produces one with higher energy input at the edges and gradual intensity decrement towards the centre. The latter results in homogeneous microstructure and depth over the treated area, making it the best oscillation strategy of the three for laser hardening.
Beam oscillation therefore shows application potential in surface treatment, although further improvement of the efficiency and resultant treated depth is required. Temporal aspect and mechanical limitation of the oscillating optics also need to be considered for choosing an oscillation strategy for surface treatment. Experimenting with a larger beam spot and other oscillation strategies to improve process efficiency is on our future research agenda.
Handika Sandra Dewi is a PhD student at Luleå University of Technology
Edmund Optics is a world-class manufacturer of high-precision optical components and an innovator in the optical manufacturing field. Each of Edmund Optics’ global factories is specialised in a particular area of optical manufacturing for both off-the-shelf and custom parts.
Conventional subtractive optical polishing methods leave behind subsurface damage on the optic as a natural result of the process, leading to scattering and an overall decrease in system efficiency. Many laser systems
and applications demanding an increased throughput and lower losses require optical components that minimize this scattering.
Edmund Optics has developed a process for optical surface polishing down to an RMS surface roughness of sub-Angstrom ≤1Å (10-10 m) for parts-per-million-level scattering, all supported by a comprehensive in-house metrology suite. This superpolishing manufacturing process produces ultra-low scattering optics ideal for any materials processing application where a short wavelength or high laser power is required, such as surface treatment, cutting, welding or marking.
Superpolished optics are equally ideal as substrates for ion-beam-sputtering (IBS) coatings.