High-velocity machining often defies conventional wisdom about how metal should be cut and is revolutionizing how aircraft parts are designed and produced.
A London taxi driver must pass a stringent test to get a license. The preparation for this test is known as “acquiring The Knowledge.” The aspiring licensee must be able to recite from memory the most effective route between any two points in the city. Since London’s streets are essentially ancient animal trails, widened a bit and paved, it is a formidable challenge. And it will often take the driver, several years to master “The Knowledge.”
Companies aspiring to do high-velocity machining, or HVM, face a similar challenge. Some of the rules governing this emerging technology are unconventional. Some even fly in the face of accepted machining principles and practices. Coupled with these technical challenges, HVM shops must also pass the ultimate test — one of bottom-line economics. Once an HVM operation is running successfully, however, the results indicate that part costs, including labor and documentation, will all be significantly lowered. And as the technology matures, it is squaring up to revolutionize some long-standing manufacturing practices in its path.
International competition demands continued productivity improvements. HVM is proving to be a potent force in increasing productivity. Some say it could prove to be almost as important as the inception of numerical machine control a few decades ago. It is driving a quantum decrease in cost and complexity in the aircraft industry that will spread to other manufacturing areas if and when current work in the application of HVM to alloy steels and titanium proves successful.
One of the machining job shops successfully climbing the HVM learning curve is Remmele Engineering Inc., New Brighton, Minn. Working closely with key component builders, Remmele is getting 6,000 to 6,500 hr/yr use from its vertical milling machines built by Forest-Liné, France. The company also anticipates being able to run its new 93,000-sq-ft Plant 60 “lights out” (unattended, round-the-clock machining).
Eight years of development work were needed to reach this level. And according to Red Heitkamp, Remmele’s director of advanced engineering manufacturing, it wasn’t an easy journey, nor was it inexpensive. With HVM machines “you can’t just move ’em in and turn ’em on” as is done with commodity machines. A close working relationship with the machine builder and machine shop is required. Some machine builders have adapted better than others in providing this hand-in-glove relationship.
An HVM milling machine pushes technology to the limits in many areas. The broad range of specialized capabilities required is difficult for a single company to master. As a result, HVM pioneers tend to be small to medium-sized machine builders that possess one or more of the key skills required. They then integrate their own technology with the other critical subsystems from specialized suppliers to meet the needs of each HVM component. As a result, the commodity machine builders are noticeably absent from the current field of HVM suppliers.
The major drivers influencing the technology are economic. Currently, HVM methods are not broadly used in manufacturing consumer products. Remmele’s major HVM focus, like others in the industry, is “hogging-out” high-value aircraft components. For example, McDonnell-Douglas (now Boeing) is producing parts for the F-15 and F-18 aircraft as long as 18 ft. Likewise, British Aerospace is also expecting to benefit from HVM when machining parts, which include complete wing spars. Working with Ingersoll Milling Machine Company, Rockford, Ill., they’ve developed a machine dubbed the “High Velocity Profiler,” which runs at 20,000 rpm, considered to be a speed which is near HVM capability. The machine measures in at 107 3 48 ft, is three stories high, and uses hydrostatic bearings to run two spindles on opposite sides of a part simultaneously.
The window for high-velocity machining
Competing forces within the industry often blur the definition of HVM. However, researchers at the University of Dramstadt in Germany have defined a standard for HVM which is widely accepted by the industry. For aluminum, the preferred HVM working envelope uses spindle speeds of 25,000 to 50,000 rpm and surface speeds of 5,000 to 18,000 sfpm depending on tool diameter. The high rotational speeds must be carefully matched with higher than traditional feedrates, 600 to 900 in./min.
At these spindle speeds and feedrates some unusual things happen at the “cutting edge” where the tool meets the workpiece. Instead of cutting the workpiece by inducing a shear failure at the working edge, the cutting tool drives the workpiece material to the plastic state allowing the “cut” chip to be more wiped away by the tool than truly cut. Given the right combination of spindle speed and feedrate, the tool forces are greatly reduced. The reduced force is the key to routinely producing parts with walls and webs of 0.020-in. thickness and in special cases, even thinner. Concurrently, the high speed at the working edge doesn’t give the heat generated time to transfer to either the tool or the workpiece, and most of the heat generated leaves with the chip.
The Remmele machine has a 40,000-rpm spindle operating in a 60 3 10 3 3.28-ft gallery. The tool can be positioned to within 0.010 in. throughout the gallery’s entire 1,968 ft3 volume. However, for parts that don’t require full translation along the full 60-ft length, the tool can be positioned within 0.001 in. This accuracy does not come without a price. The building itself is almost a part of the machine. Remmele’s Plant 60 is held to working temperature with an accuracy of ±2°F. Multiple heating and cooling units are positioned to minimize temperature gradients, and the plant is windowless to prevent solar radiation from falling directly on the machine ways.
During the roughing phase, aluminum is removed at a rate of 200 to 250 in.3/min. As the part approaches completion the removal rate slows so that overall, from aluminum plank to finished part, material is removed at an average of 30 in.3/min. Monolithic aircraft parts commonly constitute only 3 to 5% of the mass of the original plank. Remmele routinely machines part web thicknesses to 0.020 in. However, they are machining web thicknesses as thin as 0.007 in. for space applications and have experimentally achieved webs down to 0.005 in. This is approaching the thickness of heavy-duty aluminum cooking foils.
Because most of the heat generated in the process is transferred to the chip, initial concerns about possible alteration of properties in the machined T-7 aluminum proved unfounded. The workpiece sees almost no heating from the HVM process. Hot chips falling on the workpiece, however, will cause heating and distortion of the part, so chip removal is an important part of the operation. Chips are carried from the machine galleys to a hydraulic compactor, reducing the scrap to hockey-puck-sized disks. The high-density “pucks” are considered ideal feed stock for reprocessing by aluminum manufacturers.
A Little History
A 1931 German patent describes experiments by Dr. Carl J. Salomon from 1924 to 1931. Using helical milling cutters, Salomon reached surface speeds of 16,500 m/min (54,200 sfpm) in aluminum. Salomon’s work also included copper and bronze, but was not further pursued until 1958 when Lockheed Aircraft Corp. conducted systematic studies. It was not until 1970, and a Navy sponsored program, that Lockheed Missiles & Space Co. pushed HVM forward technically.
Following suit, the U.S. Air Force awarded a series of contracts to General Electric Co. in the late 1970s and early 80s. GE’s work created a data base for HVM machining of aluminum alloys, titanium alloys, steels, and nickel-based superalloys. In the 80s the literature base exploded with GE test results as well as those from educational institutions such as the University of Darmstadt, MIT, and Purdue, to name a few.
During this period aircraft manufacturers, primarily Airbus Industries in Europe and Boeing and McDonnell-Douglas in the United States, brought HVM methods into day-to-day operations. In 1984 Forest-Liné showed the first commercial HVM machine at the International Machine Tool Show in Chicago. The two-axis machine operated at 40,000 rpm and had feedrates in excess of 5,000 sfpm. However, it was not taken seriously by the industry at large.
Curtailing the paper chase
Not all the benefits are rooted in making chips; there are bureaucratic benefits as well. Tracking parts makes manufacturing aircraft components and assemblies a lot more costly by the need to be able to trace the origin of every part virtually back to the mill or mine. In the event of an accident resulting from a part failure, a paper trail must exist, linking it to all other parts manufactured from the same lot of material. “Coupons” (small metal samples) are cut from every batch of material and must be stored for decades so that the exact composition of base metals can be reviewed should it ever be called into question.
Record keeping on individual components is part of a complex accident analysis and prevention protocol that has been in place for more than 50 years. For example, just the options associated with a Boeing 737 landing gear leg currently requires 465 pages of documentation. Part consolidation and the elimination of assembly operations, through HVM, play a significant role in reducing the amount of paperwork needed to track a part through production and final assembly. It is reported that HVM reduces the amount of paperwork by factors as high as 50 to 100.