Powering up in San Fran

Share this on social media:

Issue: 

The benefits of AI and machine learning in laser processing were discussed by a panel of experts at the show

Increases in power for ultrafast lasers, green lasers and blue lasers were highlighted by experts at the Photonics West conference in San Francisco this year.

Discussing the development and use of ultrafast lasers of increasing average power was Dr Arnold Gillner, of the Fraunhofer Institute for Laser Technology (ILT).

Back in 2018 it was announced at The International Laser Technology Congress, AKL, that the ILT was looking to ramp up the average power of ultrafast lasers to a staggering 20kW by 2022. For comparison, ultrafast laser manufacturers, such as Light Conversion and Amplitude Laser, are currently able to offer commercial solutions at anywhere up to 80W and 300W respectively. The institute’s work has progressed well so far, with Gillner informing Photonics West attendees that an average power of 10kW has already been achieved in the lab by the ILT and its partners. He noted that pulses ranging in duration from 10ps down to 200fs will be achievable using the developed systems. 

The reason for such large power increases is so that the processing throughput of ultrafast lasers can be upscaled dramatically, enabling the technology to be applied on an industrial scale. This will be possible by using diffractive optical elements or beam splitting optics to divide the beams of high-power ultrafast lasers into multiple, lower-power beams, which will enable larger areas of material to be processed. This makes the technology far more suitable for applications such as texturing moulds for tool production in the automotive industry, or for creating functional surfaces – such as those exhibiting hydrophobicity – on a large scale, for example on wind turbines in order to prevent ice build-up.

Gillner noted that micro-drilling is another application set to benefit from such beam-splitting strategies. The ILT is currently involved in a research project in which filters are being made to remove micro plastic particles from wastewater in treatment plants. To produce the filters, ultrashort laser pulses are being used to drill holes only 10µm in diameter into thin metal foils, which are large enough to allow water to flow through while being small enough to filter out any micro plastic particles. The project partners are currently looking to upscale the process and make it more affordable for industrial use, and are therefore investigating the use of beam-splitting optics to create an array of more than 100 parallel beams in order to increase process throughput. 

Dr Arnold Gillner spoke on the benefits of increasing ultrafast laser power.

Drilling 100 holes simultaneously could introduce challenges, however, as this could lead to the melting and distortion of the filter foil. Therefore, in order to ensure that carefully aligned process parameters and suitable processing strategies are used, the project partners are combining a process simulation developed at Fraunhofer ILT with optimisation software from German firm OptiY. In addition, a measuring system developed in collaboration with Lunovu will be used to ensure that all the holes are all drilled correctly, and that water can flow through the filter at a normal rate.

Boosts to green and blue

Anthony Prugar, of Trumpf’s laser division, was at the conference discussing the many applications of laser processing in the growing sector of e-mobility. In addition to informing attendees that the stripping and welding of copper hairpins for stators in electric motors is one of the most prominent applications for lasers right now, he also mentioned that a new application currently seen by Trumpf is the laser coating and drying of copper foils used in battery cells. Lasers are already used to clean, drill, cut and weld such foils.

Regarding the processing challenges posed by copper’s high reflectivity, Prugar referenced Trumpf’s development of a green version of its TruDisk laser, the wavelength of which exhibits a 40 per cent absorption in copper – compared to the approximate 5 per cent absorption of the infrared wavelength used by fibre lasers. The latest 2kW version of this technology was on display at Trumpf’s booth during Photonic West, and Prugar informed attendees that an upcoming 3kW version has already been produced in the firm’s R&D lab. He noted Trumpf is looking for partners with which it can develop uses for this higher power, such as the welding of copper and other materials at higher thicknesses.

Prugar also acknowledged the suitability of blue laser technology for copper processing. However, he remarked that while blue wavelengths do exhibit higher absorption than green, there does tend to be a drop in beam quality and diode lifetime when using the technology. He added that while Trumpf is currently in the process of evaluating the two wavelengths, so far the firm is finding the trade-offs to be more beneficial for green laser technology.

This doesn’t mean that blue lasers are set to be outclassed by green in e-mobility, however – far from it. Nuburu and Laserline were also at the show discussing the advances they have made to blue laser technology for copper processing. Both have upscaled their technology to offer 1.5kW power: Nuburu with its AI-1500 and Laserline with the LDMblue 1500-60. 

Nuburu has recently introduced a 1.5kW blue diode laser with increased brightness, allowing it to be used for remote welding (Nuburu).

Speaking with Nuburu’s co-founder, Jean-Michel Pelaprat, I learnt that the 1.5kW laser is the first model of the firm’s new AI product line (AI meaning ‘deep blue’ in Japanese) whereas previous lasers have been part of its original AO line (meaning ‘blue’ in Japanese).

The new product line is based on a very different light engine design to that of the AO line, Pelaprat explained. It is a third of the size, more than twice as powerful, up to three times brighter (offering BPP < 11mm.mrad), and 30 per cent more efficient than its predecessor.

The high brightness offered by this new laser enables it to be used for remote welding, which, according to Pelaprat, is when the distance to the workpiece is 50cm or more. He noted that this is not possible with any other blue laser on the market. The high brightness also enables the use of scanning – for which the firm also provides technology – with a power density optimal for deep penetration and speed. ‘Scanning and remote welding are very critical delivery processes for the key markets: lithium ion battery welding, e-mobility, consumer electronics and general electronics packaging,’ Pelaprat said.

Speaking with Laserline’s head of product management, Stefan Aust, I learnt that the LDMblue 1500-60 was first shown by the firm in November at Blechexpo, in Stuttgart, where it was presented as part of a hybrid concept for copper keyhole welding that sees blue and infrared laser light combined. The new technique connects infrared and blue laser light via a special optic, enabling a large, blue spot to be used to form and stabilise a melt pool in the copper, and then a centred infrared beam – from a 4kW diode laser – to create and maintain the keyhole. This enables welding depths larger than those possible when LDMblue is used on its own. The LDMblue can also be offered with scanning technology, according to Aust, with the firm having presented the ‘LDMblue plus special scanner’ solution alongside its hybrid solution at Blechexpo last year.

Artificial intelligence in process control

A Q&A session followed the ‘Lasers in Manufacturing’ industry session. The speakers, in addition to Gillner, of Fraunhofer ILT, and Trumpf’s Prugar, consisted of Chi-Woo Kim, president of APS Holdings, which uses industrial laser technology in the manufacture of OLED displays.

I took the opportunity to ask the panel whether artificial intelligence and deep learning were poised to bring benefit to industrial laser processing, as it’s a topic I’ve been following ever since Dr Ben Mills, of the Optoelectronics Research Centre, shared his work with deep learning in Laser Systems Europe’s spring issue last year.

Gillner said this is definitely the case, and that the ILT has been exploring tools, including machine learning and AI, for two years for controlling laser processes. High-speed cameras are used by the institute to monitor parameters such as plasma formation, irradiance, temperature measurement and melt flow behaviour.

Using AI enables upper and lower limits of such parameters to be identified in the large amounts of captured data, which, when breached, can be used to automatically detect failures in welds or cuts. 

Kim added that AI is also very important in display manufacturing for controlling the multitude of process parameters involved. He said that it is very hard to achieve high-quality, optimal displays without being able to control all the parameters. AI is therefore playing a key role in ensuring the stringent quality demands of the display industry are being met. 

Prugar commented that, at least in the battery world, the holy grail now is to find an in-process method to qualify every single weld non-destructively. While he said that such a capability does not yet exist, when it eventually does, tools such as machine learning and AI will undoubtedly play an important role.

Navigation

Navigation

Navigation

Navigation

Navigation

Navigation