Lasers, a Bright Spot for Automotive Welding
Technology emigrates from Europe
By James R. Koelsch
If you're a Tier One supplier in the auto industry, what do you do when one of your customers in Germany can't fit side-impact beams inside the doors of a new model? If you're Magna Cosma, the forming group of 14,000-employee Magna International Inc. (Aurora, ON), you invest in laser welding.
The manufacturing engineering team at the Cosma plant in Heiligenstadt, Germany specified laser welding because the automaker's designers had trimmed away so much material that other methods simply would not work. Because the pieces had too little mass to absorb much heat, MIG welding caused excessive distortion, and would not fit in the allotted space, according to Franz Divjak, engineering manager at Heiligenstadt. And the flanges were too small for spot welding.
Working with Trumpf GmbH (Stuttgart, Germany), the engineering team designed and installed a welding cell around a YAG (yttrium aluminum garnet) laser. As the operator loads two mating 1.3-mm and 1.0-mm stampings of high-strength steel into a fixture, a vision system verifies that the parts are in the fixture and oriented properly. When the table indexes, it presents the part to the laser-welding unit. After welding, the part goes to a third station, where a robot transfers the welded assembly to a measuring system that checks a number of features. Then it goes into a marking station where an engraving tool scribes a serial number into it for tracking the part and the process parameters used to make it.
Not only does the laser weld the 574-mm seam in about 12 sec, but it also does so without distorting the high-strength-steel assembly outside the specified tolerances. Moreover, only one person is needed to make the parts. "If we had done it with conventional welding, we would have needed two or three cells to do the same number of parts," says Divjak. The reduction in outlays for equipment and labor more than paid for the large capital expenditure required by lasers.
Because of this success, the Heiligenstadt plant now has three laser-welding cells putting together side-impact beams, each for different cars. Moreover, the company's North American operations, like many of its competitors in this market, are looking to import a technology that has been very successful in European automotive factories.
"We're trying to promote laser welding here," says Harish Mistry, engineering manager of joining processes in the Research & Development Group at the US headquarters in Troy, MI. To do so, he is commissioning a new remote laser-welding cell over the summer for R&D, and any prototyping work for programs where lasers might be the welding technique of choice.
For example, the manufacturing engineers at Heiligenstadt had to ensure that the edges being welded together were within 0.2 mm of each other, which is much closer than is customary in MIG and spot welding. "So we launched a two-pronged program of developing strong and accurate fixtures, and installing a restrike station in the die to produce parts to tighter tolerances in the stamping operation that precedes the welding cell," says Divjak.
Because of their ability to weld from only one side of the material, and to produce relatively clean welds without distorting the material, lasers are finding more work welding steel joints on Class A surfaces and other sheetmetal assemblies with surfaces. "More hollow sections made by hydroforming are being incorporated into the body-in-white systems now," notes Peter Busuttil, manager, process development, lasers, Comau Pico (Southfield, MI), a builder of automation for the automakers and their suppliers. Lasers have a big advantage in these applications because the inside surfaces are usually inaccessible to spot-welding guns.
The system Comau built for welding motorcycle gas tanks illustrates the suitability of lasers for Class A surfaces headed for the paint shop. "Laser welding still leaves a little scar, but it gives you the least amount of distortion on the panel," says Busuttil. "So very little subsequent work needs to be done in order to produce a paintable surface." The motorcycle gas tank, for example, needed only a little brushing and polishing.
Another application proving the ability of lasers on bodies in white is brazing of roofs to body sides, which has been one of the body shop's biggest headaches. "With this joint, the laser brazes it, and you're done," says Busuttil. "After a little bit of polishing and inspection, you send it to the paint shop." Although a fine line is still visible, there is no ditch to seal and cover with epoxy or plastic trim, as there is with resistance spot welding and laser welding. Consequently, laser-brazing these joints saves anywhere from $6 to $8 per vehicle.
For this reason, the technique has been catching on quickly, even though it's only about five years old and is still evolving. Having already taken root in Europe, laser welding has begun to migrate to North America. Comau Pico reports that it has its first order on this side of the Atlantic.
Lasers also have made significant inroads in welding aluminum alloys, which historically has been a particularly tough nut to crack. Automakers working with aluminum sheets typically joined them by MIG welding or riveting. In applications that lend themselves to lasers, laser welding often costs half as much as resistance spot welding, according to Busuttil. "For rivets, the ratios are even better," he says. "Laser welding can cost 20 to 30% of the cost of riveting."
A good illustration of the cost-effectiveness of laser welding over riveting occurred at Freightliner, a division of DaimlerChrysler (Portland, OR). One of its plants joins extrusions and roll-formed sections to make floor pans for the cabs of its Argosy trucks. "One laser cell replaced a host of stations that did nothing but riveting," says Busuttil. Laser welding replaced 93 pop and self-piercing rivets on each side of the engine compartment, reducing cost by a little more than $16 per part, or 70%.
Laser welding can supplant MIG welding and riveting in these and other applications because of advancements in the mid-to-late 1990s that boosted the power rating of solid-state lasers to 4 kW. "Even at 4 kW, we're still somewhat starved for power," says Busuttil. "We can't do some of the thicker gages, but we can do the thin gauges that are used for automotive panels." He predicts even wider use of solid-state lasers for welding once their power output jumps to the 6 - 10-kW range.
Even though CO2 lasers have been able to generate beams with power exceeding 4 kW for more than a decade, their longer wavelengths are unsuitable for welding aluminum. "Once you create a molten puddle, the reflectivity of aluminum is three times that of steel," explains Busuttil. Aluminum absorbs the shorter-wavelength beams of solid-state lasers much better. The biggest disadvantage to using CO2 lasers, however, is that they cannot deliver the beam through fiber optics for robotic manipulation.
Even so, aluminum alloys are a diverse group of materials. "Sometimes, there might need to be a little bit of development for the particular flavor of aluminum that a customer is using," says Busuttil. Even then, laser welding will not replace all MIG welding and riveting.
The most important trend in laser welding over the last five years or so has been a steady improvement in beam quality. The improvement has been the most dramatic in solid-state lasers. A technological breakthrough has allowed manufacturers of these lasers to replace the 6" (152-mm) long rods of YAG used to generate the laser beam with a much thinner disk made of a slightly different crystal. "The disk has a fraction of the volume of the rod, allowing us to cool it much more efficiently and, therefore, achieve much higher beam quality than was possible previously," says Tim Morris, technical sales manager at Trumpf Inc.'s Laser Technology Center (Plymouth, MI).
For CO2 lasers, on the other hand, the improvements in beam quality have not been quite as dramatic. "It's been a number of small, continuous improvements in the resonator, such as more consistent temperatures and better alignment of optics," says Morris. Perhaps another reason that the improvements do not seem as great is that CO2 lasers started with better quality beams and higher power ratings and, therefore, have less room for improvement.
Nevertheless, the beams in both types of lasers are steadier and tend to diverge less than in the past, giving modern lasers the ability to focus their energy into smaller spots, use longer focal lengths, or do some combination of both. Smaller spot sizes offer two advantages, the abilities to generate higher energy densities and to fit in smaller places. The higher energy densities tend to benefit cutting by making the process faster, and the smaller spots tend to benefit welding by allowing lasers to join very fine components.
Although some automotive suppliers weld small components and can profit from smaller spots, they and the automakers reap much greater returns from long focal lengths in sheetmetal fabrication. The reason is that the work can be farther from the welding head than was possible with older laser technology. Because the greater standoffs keep the optics farther away from the welding spatter and smoke, users spend less time and money cleaning and replacing them. A bigger advantage of longer focal lengths, however, is that laser welders can use the same scanner technology used by laser markers to direct the beam and perform what the industry often calls remote welding. Rather than using a robot or a gantry to manipulate the welding head, a welding unit using this technology directs the beam from the head with a system of mirrors. "You can think of remote laser welding as a very high-power laser marker," says Morris. "Instead of single-digit watts, we're using thousands of watts over longer distances."The mirrors use these distances as leverage to move the beams at very high speeds, because the actual length of travel multiplies as a function of the distance between the weld and the mirror. Standoffs in remote welding are such that even small amounts of mirror rotation move the beam a significant distance. Moreover, the servomotors moving the mirrors can do so quite rapidly because the mirrors are relatively light, and consequently have low inertia. Coupled with the leverage that distance provides, the laser can not only weld more rapidly, but also jump between welds almost instantaneously.
Because of the ability to jump from weld to weld, the applications that see the greatest gains are sheetmetal assemblies containing several welds. "The longer the distances, the more time you're going to save," says Morris. Because automotive plants process a variety of components with lots of spot welds, he believes that remote welding has great potential there.
"Because processing speeds are so fast, you have to use creative material handling to avoid starving the system," says Morris. In most cases, that means loading parts in and out of a welding fixture robotically. A more sophisticated arrangement is for a pair of robots to present work sequentially to the scanner. In this scheme, each robot would have a welding fixture at the end of its arm. While one holds work in front of the scanner for welding, the other would put down the welded part it's holding, grab another assembly for processing, and get ready to present it to the scanner.
Comau Pico deploys a different concept in its Agilaser remote-welding cell. Rather than mounting the scanning head overhead in a stationary position or on a simple one-axis linear slide, its engineers put it on a five-axis gantry. "Not only can the remote welding head move to wherever you want it to go, but it also can move while it's welding," says Busuttil. "Coordinated motion is part and parcel of the control architecture."
The scanning technology allows the CO2 laser to make 150 - 200 stitch welds a minute. "We've been particularly successful with this technology in subassembly applications, mainly doors," says Busuttil. He adds that CO2 lasers tend to suit flexible manufacturing cells particularly well because they typically need to be mounted above the work. Consequently, remote laser welding typically finds greatest application among suppliers.
The situation is different for solid-state lasers. Because their beams can travel through fiber-optic cables, they allow yet another arrangement not possible for their CO2 counterparts. Rather than hovering over the work in a welding cell, a scanner on solid-state systems can reside on a robot's end-effecter. Consequently, remote welding based on solid-state lasers can replace a spot-welding gun or even a conventional laser-welding head of a robot working on the line in a body shop. In fact, attaching a two-axis scanner to a six-axis robot adds degrees of freedom.
In the simplest application, the robot goes to a predetermined location and sits there while the scanner shoots the beam to the appropriate joints in its 250 X 250-mm range. A more sophisticated application would coordinate the robot's motion with the scanner's. The programming takes more effort, but the gains in productivity can make it well worth the trouble, because the scanner can free the robot to travel much faster than the welding speed would allow otherwise.
To understand the gain, consider a conventional welding robot that moves along a weld at 5 m/min, the required welding speed for the sake of this illustration. There's always dead time while the robot travels to the next weld, even if engineering is able to program the robot to accelerate between welds to traverse the distance as quickly as possible.
The robot, however, could do the same job in half the time if it traveled at 10 m/min, and relied on the scanner to weld in the opposite direction at 5 m/min to create the required net-welding speed of 5 m/min. If the next weld is within the scanner's range, the scanner could jump to it instantaneously, eliminating the dead time between welds. Moreover, the welds need not be exactly in the same line. The robot can travel in a straight line or along a smooth arc, and let the mirror do the jumping back and forth.
"If the next weld is outside the scanner's working envelope and the robot has to move it to the next weld, then we would still have some limitation on speed," admits Morris at Trumpf. "Typically, though, we're looking at small stitch welds spaced every 2 or 3" [51 - 76 mm], which are within the scanner's envelope."
Comau Pico plans to exploit robotic remote welding to the fullest and perhaps even to push it along. It initiated the first phase over the summer by introducing a version of its Agilaser cells built around a disk laser and a robot carrying a scanning head. Later this year, the schedule calls for unveiling a remote laser robot capable of welding in three dimensions. The builder believes that the robotic laser will be a bright spot for automotive welding.
Hybrid Laser Welding: Two Heads are Better than One
As the old adage says, two heads are better than one, even in welding. Users often can reap some important benefits from a two-headed joining process called hybrid laser welding. The process combines laser technology with conventional gas metal arc welding (GMAW), using the laser as the primary energy source and the GMAW torch as the secondary source.
Because the laser beam is the primary driver, the process still offers the faster speeds, small heat affected zones, and narrow, attractive welds associated with laser welding. "GMAW as a secondary energy source improves overall process energy efficiency, lowers fit-up cost with enhanced ability to bridge gaps, slows cooling rates, and improves the energy-coupling efficiency of aluminum," says Richard Green, product manager, Concoa America (Virginia Beach, VA), a manufacturer of gas-pressure and flow-control equipment.Although adding the secondary energy source makes the process more complex to set up, it can lower cost, because the additional power source reduces the size of the resonator needed to perform the weld. Since the power supply for GMAW is much cheaper, the cost of the hybrid system is less than a laser big enough to do the job itself would be. Moreover, the dimple created by the GMAW arc at the front edge of the weld pool reduces the overall depth that the laser beam must penetrate. The GMAW's wire can be located either before or after the beam, but Green notes that placing it after the beam allows faster welding speeds and lowers the amount of power necessary to melt it.Although helium is the best gas for minimizing the attenuation of the laser beam (by controlling the size of vapor particles coming from the weld), argon is heavier and cheaper. Consequently, Green recommends a 40 - 50% mixture of argon and helium for an optimum combination of plasma suppression, vapor particle evacuation, and cost. "The mixture also provides an inert atmosphere for a longer duration as the weld pool solidifies, enabling greater travel speeds," he says.
Small amounts of carbon dioxide, oxygen, or both can enhance the characteristics of the bead further. "Helium-argon mixtures tend to produce higher arc voltages, which subsequently yield wider bead profiles and greater arc instability," explains Green. Adding 3 - 10% carbon dioxide can stabilize the transfer and constrict the arc. In some cases, a concentration of 1 - 5% oxygen can stabilize the arc further, and offer better tie-in (wetting) at the edge of the weld.
Handling the details might take more thought up front, but it's often worth the effort. Two heads are indeed better than one.
This article was first published in the September 2005 edition of Manufacturing Engineering magazine.