Automation development in the aerospace industry has quickened its pace, with the aviation and defense industries attempting to further automate manufacturing processes to meet growing OEM order backlogs and critical aerospace-defense program deadlines.
As aero/defense builders seek to speed up manufacturing processes, many companies continue to borrow ideas from the more highly automated automotive industry including the deployment of moving assembly lines and much greater use of robotic automation. In commercial aviation, aging jet fleets are being phased out in favor of more energy-efficient new aircraft just as the airline industry experiences an explosion in travel demand, necessitating a corresponding boost in aircraft production.
“The mantra for aerospace and defense had been, ‘We’re low volume, we can’t automate,’ and so they just did things manually because they were low volume, and they were also very demanding on their precision,” notes Joe Campbell, vice president, US products, ABB Robotics North America (Auburn Hills, MI).
“From a product standpoint, there is a consistent trend toward increased precision, accuracy, and rigidity. Every product’s a little bit better,” he adds. “What we’re finding to mirror that on the customer side is, their volumes are going up, whether you’re in the defense industry or you’re in the commercial aviation industry. They’ve never had backlog like they have now—Boeing has publicly reported that they are sold out for years. If you want to buy an airplane today, it’s going to be years before you can get it.”
With an expansion in air travel and commercial airlines looking to replace older, less-energy efficient planes, increasing automation in today’s aircraft plants makes much more sense, Campbell notes. “All of a sudden, their volume is not so low. He who has manufacturing capacity will get the order,” he says. “If you can take lead time off your product, you’re probably going to win some business.”
Nowhere is automation more crucial than for the high-stakes, high-budget Joint Strike Fighter (JSF) program to build the next-generation F-35 Lightning II jet fighter plane destined for the US military and its allies. Among automation efforts the JSF has deployed to date is the Integrated Assembly Line (IAL) installed by one of the program’s contractors, Northrop Grumman Corp. (NGC; El Segundo, CA), at the aerospace/defense builder’s Antelope Valley Manufacturing Center located in Palmdale, CA, where the F-35 center fuselage is being built.
The IAL, a highly automated manufacturing system to produce the center fuselage of the F-35, was installed at Northrop Grumman by Kuka Systems Corp. (Sterling Heights, MI, and Augsburg, Germany), earning Kuka an award as Northrop’s top tooling supplier for 2010-11 with its performance on the benchmarks of cost, schedule, quality and customer satisfaction. The Kuka Systems contract to design, build and install the IAL is worth over $100 million and covers more than 500 tools required for 78 tool positions occupying 200,000 ft2 floor space at Northrop Grumman’s Palmdale manufacturing center. The system includes automated assembly tool systems, transportation systems and manufacturing systems.
In citing Kuka for the supplier award, Northrop Grumman noted the automation integrator’s unique skill-set and approach to designing, layout, fabricating, validation, project management and the interfacing of a system of tools. The aero/defense builder also noted that Kuka Systems’ inspiration for the IAL are the fully integrated and highly automated and optimized assembly lines it has developed for major automakers around the world. The IAL required extensive development by Kuka Systems engineers and other technical staff working in close collaboration with their Northrop Grumman counterparts to design an aircraft manufacturing line that is heavily automated and fully integrated and optimized.
“Northrop Grumman is an innovation pathfinder whose support for this project is helping validate the potential of automated assembly concepts to change how military and civil aircraft are manufactured,” says Robert Reno, vice president, Aerospace Division, Kuka Systems. The IAL is being phased into service, and when F-35 production reaches a peak starting in 2014, it will be capable of manufacturing one center fuselage per day.
Automation Adoption Picks Up the Pace
With several industry collaborations in the last few years, adoption of automation efforts by the aerospace/defense industry appear to be gaining critical mass. The Aerospace Automation Consortium (AAC) spearheaded by Northrop Grumman helped start moves toward adopting automotive-style moving lines at aero/defense manufacturing (see “Modular Automation by the Aerospace Industry” in the March 2006 issue of Manufacturing Engineering). Similarly, a group of automation, metrology and software developers have banded together to further automation efforts in the industry with the Computational Manufacturing Alliance, or Compufacturing (see www.compufacturing.org), which last March demonstrated multiple technical advances at the AeroDef 2011 Manufacturing Exposition and Conference held in Anaheim, CA.
Automation developers have had some recent successes, observes Chris Blanchette, manager, aerospace integration channels, Fanuc Robotics America (Rochester Hills, MI), including the mobile robots installed by Fanuc integrator Comau Aerospace (Southfield, MI). Comau’s mobile robot systems are installed at the Bell Boeing rotorcraft factory and the company also developed an automated robotic drilling workcell for the F-35’s inlet duct hole process at Northrop Grumman (see “Automating Aerospace Processes” in the March 2011 ME issue.)
“They have one installation of mobile robots in the field and it’s working well,” Blanchette says, “and now they’re starting to get some repeat orders. That’s one of the challenges. A lot of the OEMs are holding back; educating them on automation’s capabilities is still a problem. The organizations are so large and the infrastructures are very deep.”
A new group of automation suppliers, the Great Plains Robotics Alliance (GPRA) in the Wichita, KS, area, has a similar directive to the AAC in trying to pull together aerospace-related parties, Blanchette notes. “They’re working with robotic integrators and OEMs, doing robotic evaluations, running off machines and testing systems,” Blanchette says. “For the OEMs, the challenge is identifying technologies that they know they will work. They don’t have a long track record and they’re still a little bit apprehensive to accept the data—the parts in aerospace are very expensive, and there are often very long delivery times.”
The large backlog in orders also makes it more challenging for aerospace/defense builders that might otherwise explore newer automated processes, he adds. Faced with deadlines to meet, the builders hold off on implementing newer technology in order to be able to meet production requirements.
Robots Gaining Traction
While technical advances in robotic automation offer improvements in precision and rigidity, the primary robotic applications remain in drilling, riveting and fastening operations for both commercial and aerospace/defense, Blanchette adds, with typical tolerances reaching 0.010– 0.030″ (0.25–0.76 mm) for commercial aviation and 0.005″ (0.13 mm) for defense. Painting and coating applications also make up a large chunk of automation applications, he says, and there is very large market for aircraft engine components that requires tight tolerances on complex geometric surfaces. “To be able to inspect the microfinish of the surface of the blades is also very challenging,” Blanchette notes.
AV&R Vision & Robotics (Montreal, Quebec, Canada), a Fanuc Robotics integrator, has developed automated visual inspection systems for critical rotating parts in aircraft, as well as offering automated robotic systems for finishing, profiling, deburring, polishing and grinding. The company works with all the major aircraft engine builders including GE, Pratt & Whitney, Rolls Royce.
“We’ve been working on robotic and vision systems for 15 years,” says Eric Beauregard, AV&R Vision & Robotics’ president and CEO, noting about 95% of the company’s business is aerospace. “We have developed two core competencies: automated visual surface inspection and expertise in finishing parts, mostly aircraft engine parts, which at one point have to be finished and many of those finishing processes are done manually because of the variation of the parts.”
One of AV&R’s customers uses its automated polishing systems to finely finish the aluminum rings on the turbine engine housing’s outer edge. “We have developed an adaptive polishing process for airfoils that removes surface defects either from the forging process or human manipulation,” says Beauregard. “If you look at the engine, around the blades there is a shiny surface that allows air to get into the engine, which is called the lip skin. After being forged, it needed to be polished manually, which is tough, and aluminum dust is explosive and very dangerous to humans.”
With adaptive finishing, aircraft engine builders can add inspection to the process, he notes, which is an important ingredient for mission-critical aircraft engine applications. “You want to take a part and adapt your process to it to reach a high precision level, since each part has some variation,” Beauregard notes. “If you want to do a perfect process on a part, you need to not do the job blindly. Each part is a little bit different and our automated systems can adapt to it.”
Inspecting and repairing turbine blades, which are very expensive and are consumed in high volume, is a huge market for maintenance, repair and overhaul (MRO), he adds. The airfoils found in the cold compressor section of engines are made of titanium alloy and are very light, and the hot section uses a ceramic-coated alloy that protects it from heat and prevents it from melting.
“The most critical and important part is the leading edge’s profiling of an airfoil. The hot section’s castings are good enough, they don’t have to re-profile it, but when you look at the cold section’s part, those are forged, with no profiler at all,” he adds. “What we do is we measure the shape of the blade, from the first part to the last part, within ±0.002″ tolerance,” Beauregard says. “Other than the universities testing it in labs, we’re the only company automating the final quality inspection in shops, therefore creating standards.”
The quality inspection of turbine blades today is done manually, he notes. “Our goal is to automate the visual inspection and the finishing,” Beauregard says. “Our ultimate goal is to inspect the part and correct it. The only thing we’re missing is the 3-D inspection. We take a part and manipulate it in our software behind the camera to find the surface defects, such as pits, dents and scratches.”
The company last year signed a technology transfer with the National Research Council of Canada to work on integrating a noncontact 3-D measurement system that will offer greater precision and speed for applications. This technology previously was used for the inspection of the heat-absorbing tiles on the NASA Space Shuttle. “What we were looking for is speed and precision in 3-D data acquisition, which you cannot find in industrial environments today,” Beauregard says.
Automated Welding for Shipbuilding
Shipbuilding applications aren’t considered high volume for heavy automation, but with the multipass high-deposition process from robotic welding integrator Navus Automation Inc. (Knoxville, TN), an ABB Robotics integrator, the overall welding process time has been cut about 60% for welding large steel sheets for the US Navy’s next-generation DDG-1000 Zumwalt-class destroyer.
“The system we build is about 120′ [36.6-m] long with six ABB robots moving along a track,” notes Don Bernier, vice president, Northeast Region, Navus Automation, of the robotic workcells used to speed high-quality welds for the destroyer under construction at a shipbuilder located in the northeastern US. “We were looking for some manufacturing process improvements in welding the parts, which weigh in excess of 30 tons.”
For this application, Navus employed six ABB IRB 4600 robots on two 30-m tracks using a Fronius GP tandem system, Bernier says. The system does interpass cleaning and grinding and removes silicate that comes to the top in order to meet the required finish quality, Bernier says. “We maintain an interpass temperature of 200–300ºC in the weld,” he notes. “The part needs to be pre-heated. In two cells, there are a total of six robots, two welding per cell and two shared cleaning robots.
“The workpiece is a steel plate that’s 2″ [51-mm] thick and has to be flipped four times total to do welding. Cleaning has to be done before the next pass is done,” he adds, “and we’ve qualified this to do a UT [ultrasonic testing] class weld, checking for what they call deviations or inclusions in the weld itself. It’s allowed only so many inclusions, which are a little pocket of slag or silicate because of the welding wire. That’s the reason for the cleaning.”
Digital manufacturing enabled Navus to program the parts designed for the customer’s 3-D CAD modeling software. “They send it through the line in a nesting process, save it as a STEP file, and the robotics engineers retrieve that data and they do offline programming to do the part,” Bernier explains. “A huge amount of time and energy went into the programming,” adds Mark Oxlade, ABB welding manager. “This kind of job can’t be done without offline programming, which was done in RobotStudio.”
With this system, the builder was able to cut the manufacturing process from a minimum of 28 weeks to build one part down to eight weeks, according to Bernier. “They were building three or four at a time, and I’d say that they do a ship set now in roughly 60–70% less time,” he adds. The shipbuilder is currently working on creating a library of programs that will help with future ship designs.
“Right now, you’re in the infancy stages of using these types of systems,” Bernier notes. “Some of these shipbuilding programs are so old that the designs were done either in 2-D or on paper. With 3-D, it’s much easier.” ME
This article was first published in the March 2012 edition of Manufacturing Engineering magazine.