Get Ready for a More Efficient Future
New machining concepts and a drive for determinate assembly are changing aerospace manufacturing
By Randy Von Moll
Aerospace Product Manager
It's no secret that aerospace OEMs continue to study the automotive industry intensively in their search for a more efficient manufacturing strategy. Two major lessons have emerged from this effort:
- Machine utilization must be maximized to achieve efficient manufacturing
- Parts must fit properly at assembly without modification.
While neither of those concepts will be major revelations to anyone with an automotive or general-manufacturing background, they are, nevertheless, the drivers for sweeping changes in the way aerospace components are designed and manufactured, and for how aircraft are built. The transformation is already well underway, and the pace of change is sure to accelerate as global competition drives aerospace OEM's and upper-tier suppliers to adopt ever more efficient manufacturing processes and technologies.
For many years, the core of aerospace manufacturing was a suite of highly specialized special-purpose machine tools like spar mills, skin mills, and profilers that had little or no flexibility in application range. With the possible exception of stringer mills and large, specialized systems, aerospace-specific machine tools are rapidly becoming historical artifacts.
Those specialized machines are being replaced with machining centers that have features commonly found in the general-manufacturing environment, albeit on the gargantuan scale typical of the aerospace industry. The essential features of current-generation aerospace machines are:
- An A/C head,
- Zero spindle downtime for part loading,
- Effective chip management,
- Automatic toolchanging,
- Advanced accuracy levels.
With the exceptions of the A/C head and improved accuracy, all of these features are aimed at maximizing machine utilization—keeping the spindle cutting as close to 100% of the time as possible. That's why machining centers equipped with pallet changers and/or dedicated safe zones for part loading are preferred over older configurations.
Another consequence of the pursuit of maximum uptime is a growing preference for horizontal-spindle architectures, as this configuration facilitates effective chip management with an assist from gravity. Chip management has always been a challenge in the aerospace industry, and will continue to be as parts become more monolithic and lighter than ever. There are a number of chip-management solutions available, ranging from simple manual evacuation to complex automated systems. While a horizontal machine presents fewer challenges, numerous factors can be evaluated when determining the best way to manage chips. For parts more than 4-m long, a strong argument can be made to use vertical spindle five-axis machines due to much lower capital investment required. Another major advantage of the vertical configuration is the opportunity to employ two-spindle machines like our company's HyperMach, which has twin independent five-axis spindle heads to double the machined-parts throughput.
Automatic toolchanging obviously improves machine availability, but even more important is the fact that it supports efficient processing of different parts. It's not at all uncommon for a modern aerospace machine to be loaded with a complete package of 100 or more tools, only a few of which are needed for any particular part. In addition to providing greater flexibility for a wider range of parts, automatic toolchanging is one of many features that positively impacts the human factors of manufacturing. Such features not only protect workers from unfavorable work conditions (i.e. loading and unloading heavy tools bending over uncomfortably while standing in chips and coolant), but also greatly reduce process variability, another key element that leads to greater efficiency and lower cost.
The A/C head and improved accuracy capabilities of modern aerospace machines support the concept of Determinate Assembly, which is simply another way of saying that all of the parts fit together properly at assembly without modification. While this may seem an obvious requirement, the fact is that most aerospace components were, and still are, shimmed or modified at assembly before the full-sized fastener/assembly holes are drilled.
Traditional A/B heads used on many machines offer a limited rotary axis range, which is sufficient for parts that will be custom fit at assembly. While A/B heads have significant advantages in manufacturing certain types of parts, they cannot support some Determinate Assembly operations. To achieve Determinate Assembly requires a true five-sided machining capability to finish edges and drill assembly and fastener holes. An A/C head makes this possible in a single setup without using a right-angle attachment on the spindle.
Obviously, parts produced for Determinate Assembly have to be dimensionally accurate, and this is another area where modern aerospace machines differ from their predecessors. As recently as the 1980s, tolerances for major aircraft components were measured in hundredths of an inch, with (0.030" being a reasonably common value. Today, (0.005" or in many cases, better, is typical for aircraft like the Boeing 787, the Airbus A380, and the new F-35 Joint Strike Fighter variants. The trend is clear and shows no signs of abating.
One major consequence of this enhanced accuracy requirement is the fact that rebuilt and/or remanufactured machine tools will be increasingly unable to achieve the desired result. For a multitude of reasons, the thousands of machines designed and built during the 1970s and 80's simply cannot deliver the kind of accuracy needed to meet the requirements of the 21st century. When rebuilt by a top-of-theline service, those machines may still have a place producing parts for legacy aircraft and those not designed to 21st century tolerances, but they are not the answer for manufacturers looking for long-term solutions.
The concept of Minimum Quantity Lubrication (MQL) should also be mentioned in relation to today's aerospace machines, as it's an existing technology that is certain to find more widespread use in the future. Developed initially as a response to environmental concerns about conventional coolant/lubricants, MQL is slowly evolving into a replacement for traditional high-pressure/high-volume systems. Much more work is required, however, to make it a viable option, but certainly the wishes of users—and environmental factors—will urge builders to continue research into this new processing method.
So the question that comes to mind is: Where do we go from here? The truth is that there probably aren't any radical mechanical breakthroughs on the near to mid-term horizon that will alter the essential nature of the machine tools used to build tomorrow's aircraft components. What will change radically, however, are the systems that support those machines as they, and the whole manufacturing process, become smarter and evermore efficient.
The "smart" aspect will be the result of improved on-machine data collection, monitoring, analysis, and adaptability. In a real sense, tomorrow's machines will be able to "learn" and change their behavior in response to changing conditions. One of the major benefits of this change will be to minimize the impact of human error on system performance, while significantly reducing process variability from all causes.
Manufacturers who already are well along in the acquisition and use of current-generation aerospace machines are going to have a significant advantage over those playing catchup as these new technologies come to market over the next few years. The essential features of tomorrow's aerospace machines are:
- Condition-based management.
- Integrated predictive maintenance.
- Interactive documentation.
- Auto N/C.
- Lights-out manufacturing capabilities
Interestingly enough, these features all have the potential to impact both aspects of the drive for increased manufacturing efficiency: machine availability and machine precision.
Condition-based machine management is the natural next step in the evolution of machine monitoring and control systems. Many machines are already equipped with extensive sensor networks to monitor temperatures, pressures, accelerations, and other operating parameters, but most do little with the data beyond alerting operators and managers to problems. That is about to change.
If a machine can monitor itself and issue alerts when parameters go out of tolerance, it's only a short step to a system which will automatically apply corrective actions as these parameters drift toward unacceptable values. It's an unalterable fact that machine performance will change over time due to wear and other factors, but those changes do not have to automatically translate into out-of-tolerance parts or unscheduled machine downtime.
That's the first step. All of that sensor data can also be used to identify the machine's sweet spots, and build a library of information on the dynamic performance of a cutting tool assembly in a specific spindle/machine tool combination. That is very powerful information that can be used to take the black art out of programming the machine for optimum productivity, while eliminating both wasteful trial-and-error experiments and on-machine testing.
If you already know what works, the programmer can simply keep the whole process in the machine's sweet spots. Better yet, the condition-based management system can be programmed to keep it operating in those sweet spots even as the machine's performance changes over time, by adjusting machining parameters to maintain optimum performance.
As that happens, a condition-based management system will monitor the changes and apply corrective actions, while it also notifies the appropriate individuals and functions in the plant of the situation. So, instead of having to deal with an unscheduled breakdown, or a batch of bad parts, the manufacturing engineer or manager can schedule maintenance downtime in advance, pre-order the repair parts, and see to it that the necessary personnel are on hand. That's integrated predictive maintenance.
Many of the tools needed to make this happen already exist. Through our Maintenance Technologies group, for example, our company has a product called FREEDOM E-LOG which provides the essential sensor monitoring, data organization, and communication functions. Machines equipped with FREEDOM E-LOG will call designated personnel via cell phone to deliver notifications of out-of-range events that include information on why the machine is down. The program is also Internet-enabled, so designated personnel can monitor plant activities down to the level of individual part programs in real time—from anywhere in the world.
For customers who choose to share FREEDOM ELOG data with us, it's already possible for factory technicians to perform predictive analyses on the information and recommend proactive solutions. Those technicians also can monitor machine trends and speed the process of diagnosing machine problems.
One day, all of those capabilities will be resident on the machine's own condition-based machine management and integrated predictive maintenance systems. These capabilities will build on the work of the AMT (The Association for Manufacturing Technology; McLean, VA) machine availability standard for data characterization, and the ongoing efforts of the Smart Machine Initiative and other groups around the world pursuing this goal.
The same technology that makes a machine smart enough to take care of routine, predictable events, makes it much easier to handle the inevitable unplanned events that occur in production. Today's controls have moved well past the reporting of a simple alphanumeric error code that had to be looked up in a manual before action could be taken. Now, operators get specific data, including sensor readings, and suggestions of possible causes and fixes.
In the future, the control will provide an instant diagnosis and detailed recommendations for remedial action supported by digital photographs and engineering drawings. This capability will be resident in the control itself.
One impediment to widespread use of this technology in the aerospace industry is development cost. For a high-volume machine, such as a HMC, the cost can be spread among many units. But for a low-volume aerospace machine, the cost is often prohibitive using today's development technology. This situation will almost certainly change in the future, however, as aerospace machines continue to evolve toward a more machining-center-like model.
Even today, Cincinnati Machine and other builders have Interactive Tech Support systems in operation that deliver voice, data, and video of a machine in the field to factory technicians over ordinary telephone lines. This capability can often save the hours or days typically required for a factory technician to arrive at a customer site to effect repairs.
The so-called art-to-part software that today only works three-axis parts has no direct counterpart in the five-axis world of aerospace manufacturing. Several attempts have been made to bridge the gap, but none have been notably successful to date. That must change, and it certainly will. The productivity implications of being able to move directly from CATIA, Unigraphics, or Pro-E design models to code that Siemens or Fanuc controls can use to cut parts are so compelling that the problem will certainly be solved sooner rather than later.
This is another area in which the machine performance and tool libraries mentioned above will make a major contribution to machine productivity. Not only will art-to-part software reduce the time required to move a part into production, it will also take full advantage of the other machine management technologies to make sure the part is produced in the most efficient process of which the machine is capable.
Lights-out manufacturing and snap-together aircraft represent the Holy Grail for manufacturing in general, and aerospace manufacturing in particular. They are the natural culmination of all the various efforts to maximize machine utilization through automated processes and intelligent control systems. But, they're only half of the story.
The other half is that the parts produced in the dark factory will be precise enough to fit without modification at the assembly location. That capability, when it is achieved, means that aircraft and automobiles will be assembled using the same production model, most likely on a moving assembly line with parts produced just-intime. This capability is already being employed to a limited extent on a few Boeing and Airbus commercial aircraft assembly lines.
Twenty years ago the aerospace and automotive industries were at opposite poles in terms of manufacturing strategies, systems, and processes. Today, they are rapidly converging in all three areas. Twenty years from now, it may well be the world's automakers who are studying the aerospace industry intensively in their search for a more efficient manufacturing strategy. Stranger things have happened.
This article was first published in the March 2007 edition of Manufacturing Engineering magazine.