Riding on the laser beam: 35 years developing laser applications for manufacturing processes
Editor’s Note: In the early 1980s, innovative work was being done by groups of process developers. Starting in 1986, with the founding of Industrial Laser Review, the predecessor to Industrial Laser Solutions, we offered a forum for these pioneers to promote their work to the manufacturing community. Early on, we met René Vierstraete at a European tradeshow, initiating a long relationship that culminated with his 2019 retirement. René’s career in industrial laser materials processing is an information look-back to the early introduction of high-power fiber lasers and 3D laser cutting and welding. The following is his review of that period.
Early in the 1980s after completing my Mechanical Engineering studies and with an uncle already in the machine-tool business, I launched a new firm—Vierstraete SA (Oignies, France). The plan was to design and manufacture special numerically controlled (NC) multispindle drilling machine tools for processing long beams and plates. After a promising start with two special systems a year, 1982 experienced a lack of system orders that, due to poor industrial investments in France, caused us to just run as a machining job shop.
In 1984, I decided to explore a new field—industrial laser machines. In its startup phase, I joined Laser Systemes (Beauchamp, France), a joint venture with Coherent (Palo Alto, CA) and Renault-Automation (the just renamed Machine-Tools and Robotic division of Renault Car in France). In fact, my first two payments were on Laser Optronic France, the French subsidiary of Laser Optronic (Munich, Germany), a manufacturer of laser marking and trimming systems (using Q-switched Nd:YAG), which had just been acquired by Coherent. I knew about NC systems but nothing about lasers, but fortunately I was rapidly educated by Coherent.
In August 1984, I was in training on laser system design and laser applications at Laser Inc. (Sturbridge, MA), the U.S. East Coast industrial Coherent base. There, I was welcomed by Al Batista (Managing Director) and Bill Shiner (Sales Director), two pioneers in solid-state laser processing. I was able to "play’’ with small Nd:glass pulsed lasers and long M46 CO2 lasers (450 W at 10 µm, slow flow) and designed machines with mirrors and telescopes.
At Laser Systemes, we opened an applications laboratory with NC 3-axis machines powered by the new Coherent General EFA 51 CO2 laser (1250 W at 10 µm, fast axial flow). With this, we investigated the 2D cutting market with great interest in the flying optics concept, promoted by two European pioneers—Bystronic (Bützberg, Switzerland) and Limoges-Précision (Limoges, France). In 1985, Ernst Zumstein (the CEO of Bystronic) showed me, in action at SARMO (Burgdorf, Swiss), one of the first Bystronic 2D systems (ByLas) powered by a 500 W at 10 µm Electrox laser (UK), the first fast axial flow laser. In 1983, Xavier Rouchaud (CEO of Limoges-Precision) had launched its second-generation flying optics cutter; in 1985, he integrated the Coherent EFA 51 laser (1250 W at 10 µm); and in 1986, when the market shifted from static to flying optics, he provided to Trumpf (Ditzingen, Germany) the 2D mechanics for the 10 first L3000 Trumpf flying-optics laser cutting systems.
In 1985, Renault was not interested in 2D cutting and pushed us to 3D. With the DTAA (Renault R&D robotic department in Boulogne-Billancourt, France) we designed an extremely accurate, five-mirror CO2-articulated robot (axes: 1x linear, 4x rotation) designed to work on one face of a car: 2 × 1 × 0.5 m and controlled by an R&D Geometric Command supported by an Intel Calculator. Unfortunately, this system could not be used in the field due to the lack of a fast industrial robotic controller.
The 3D cutting domestic demand came from prototype car body manufacturers that expect huge volume. For that, with the team of the SEIV Assembly department of Renault Automation (Evry, France), we designed two kinds of modular gantry machines (C.38-L for 2D & C.148-L for 3D; FIGURE) with a full flying-optics laser concept and a 3D probing facility (Renishaw tactile sensor). Such systems worked at Creusot-Loire-Industry (Saint-Chamond, France): In a 1987 Flexible Factory project for cutting armored steel with a 2 kW at 10 µm CILAS (Courtaboeuf, France) laser followed in 1991 by a welding system with seam tracker and wobbling optics facilities powered by a 14 kW at 10 µm United Technology Industrial Laser (East Hartford, CT), I had an opportunity to consult with Connie Banas, a high-power laser UTIL expert. In 1989 at Renault (Sandouville, France), on the R25 Renault body line, a C.1486L 5-axis laser gantry with an automatic laser-tools changer cut option holes with a 400 W at 10 µm Ferranti (Dundee, Scotland). For Aerospatiale-AIRBUS (Nantes, France), we built a huge (11 × 3 × 1.5 m) laser mask cutting machine for the chemical machining processing of plane panels with a 100 W at 10 µm Directed Energy (U.S.) laser. At ANSALDO (Milano, Italy), for power plant turbine manufacturing, we cut inclined wing shapes up to 30° from normal on 9-mm gauge stainless steel using a 2.5 kW at 10 µm Coherent EFA53 laser.
In parallel, in 1987 I started to work on special systems for steel processing for Usinor (Florange, France) producing Textured Roll-Mill (LASERTEX) with a 2.5 kW at 10 µm EFA 53 laser. I had an opportunity to train on the EFA 53 laser at Coherent and was introduced to Gaussian beam propagation and the M2 beam quality concept. The LASERTEX process required ultrastable laser power at a very short time period. It was not the case of many lasers based on Roots pumps, so I thought about a gas flow laser based on a triangular or square laser cavity surrounding a gas-turbine equipped with magnetic bearings exited by microwave oven generators—in fact, a kind of early-Trumpf TruFlow secon-generation design, although my boss said we were there to make systems, not lasers. In 1989 at Lectra-Systeme (Cestas, France), I had seen the prototype of an affordable laser steel cutting station (laser flat cutter with integrated CAD solutions) as Lectra did for clothes, unfortunately they had to focus on their historical Market.
At the end of the 1980s, the 10 µm laser system was less rugged than robotics systems due to the robot’s maturity, so Renault stopped industrial inline CO2 laser processing (cutting). For Renault R&D, we did multiaxis-compliant welding tools used to continuously laser weld the car body door-ring contour. Some R21 models were processed (one to two a day per body) in a difficult process that showed no significant advantages for rigidity, but a significant improvement in fatigue, so Renault stopped its 3D laser welding development and focused on easier processes like 2D laser welding of a tailor welded blank (TWB). With the Renault Body R&D team, I performed my first TWB in 1989 on a Renault-Automation system. Our first industrial TWB laser system started in 1992 at Renault (Douai, France). It was TWB pioneer time with UTILASE (Detroit, MI), VIL (Version Industrial Laser; Illinois) and right after with Noble, Prototec & Laser Welding International (Detroit area, MI), Automated Welding System (AWS; Toronto, ON, Canada), and SOUDRONIC (Neftenbach, Switzerland).
Although CO2 lasers were not compatible with standard robots, we investigated early Nd:YAG fiber-optic beam delivery for robotic cutting. In 1991, we experimented with a Laser Cheval (Besançon, France) 600 W continuous-wave at 1 µm with two ways to beam-switch, two fibers, and two robots for car BIW cutting at Citroen (Aulnay, France). In that development phase, I burned a lot of fibers (spot size and fiber diameter are both in the 1 mm range) and I dreamt of a small 15 W TEM00 Nd:YAG spot that could safely enter into a Nd:glass-doped fiber used as an amplifier (a kind of MOPA fiber-laser design), but many "laser guys" said it was not realistic.
In July 1995, my first article on Laser Welding Process Quality Control was published in Industrial Laser Review.
In Europe, except for Renault, TWBs were mainly produced by steelmakers, so in 1998, I moved to the French steel maker Usinor (now ArcelorMittal) at the Welding R&D Center (Dunkerque, France), followed by the Automotive Applications R&D Center (Montataire, France) to develop laser processes for TWB. Doped fiber and fiber laser were not a dream, as in 1999 I discovered a milliwatt fiber laser from JDSU (Santa Clara, CA). Then fiber lasers moved fast, especially with IPG Photonics (Burbach, Germany). In 2003, I started IPG Photonics investigations and in 2006, we ran 24/7 a 6 kW IPG Photonics fiber laser on an industrial TWB system at AMTB (Liege, Belgium). In 2008, most of our systems were still based on high-power CO2 lasers from Trumpf (Ditzingen, Germany) or Rofin (Hamburg, Germany), so I optimized the 10 µm butt welding process on thin sheets using beam shaping with a ring-shape spot instead of single or twin spot to significantly increase the welding speed. In 2012, we started to produce TWBs with fiber-coupled direct diode lasers and in Europe on our industrial TWB systems, Laserline diode generators (Mülheim-Kärlich, Germany) replaced the old CO2 lasers.
In 2004, we had a new welding technical challenge. For aluminum silicon-coated hot-stamped steel, we developed a laser ablation cleaning process. For this ablation, in 2006 we investigated the use of one of the first industrial high-power nanosecond lasers—a 450 W at 1 µm DQx45 Rofin (Hamburg, Germany) with peak power up to 1.9 MW. Aluminum silicon ablation for a BIW upgrade was the subject of my second Industrial Laser Solutions article in the March/April 2010 issue, and a third article in the May/June 2014 issue introduced a 1.6 kW nanosecond laser from Powerlase (Crawley, England) to the field. With the new steel generation and new ablation process, I had the opportunity to deploy new up-to-date laser systems in Europe, America, and China until my retirement in July 2019.
I’m now retired, but still dreaming on "new beams to ride," like 1 µm laser beam shaping.
RENÉ VIERSTRAETE ([email protected]) has been retired from ArcelorMittal since July 2019.