Affordable Machining of Unitized Aero-Structures
Proper selection of machining strategies can speed production and reduce shop-floor costs
By Jay Snider II, PhD
and Richard Guiler
Aurora Flight Sciences
By Glenn Sheffler
National Center for Defense
Manufacturing and Machining
The construction of complex aircraft structures such as the support skeleton within a wing is time-consuming and expensive. The main concerns when designing these types of structures are weight, strength, and cost. Traditionally, they are assemblies of sheetmetal parts, and machined, and/or composite parts, which are bonded and/or riveted together. For example, a modern aircraft wing structure consists of one or multiple spars that run the length of the wing, with ribs running perpendicular to the spar. Flanges are usually designed into these structures for the attachment of skins. Various other composite or metallic fastening points are also bonded or fastened into the wing. Finally, a composite or metallic skin is used to cover and closeout the other structures, while providing the aerodynamic surface.
There has been a trend toward using five-axis-machined aluminum or titanium components along with formed sheetmetal and composite parts in an effort to reduce the number of parts, increase the overall strength, and decrease the final weight of aircraft structures. Traditional five-axis machined components are expensive to produce, and the assembly of these structures remains a demanding task. Replacing an assembly of this type with a single machined piece can have the advantages of reducing the total number of build hours, decreasing weight, and increasing strength.
Large companies have made great strides in the production of big, complex, integral aircraft structures, but this has been done using large, expensive five-axis gantry mills. There have been a number of advances in machining techniques and equipment that can allow an entire structure like a wing skeleton to be machined as a single piece with a minimal number of setups, using much smaller and less-expensive machines. There are three main factors that need to be improved to fabricate complex machined aircraft structures more economically:
- Increase the cutting rate,
- Reduce the number of setups, and
- Allow smaller machines to more efficiently fabricate parts.
Aurora Flight Sciences of West Virginia (AWV) and the National Center for Defense Manufacturing and Machining (NCDMM) have worked in cooperation to develop and integrate five advanced machining technologies to achieve these goals. This research project has been funded by the Air Force Research Laboratory (AFRL) at Wright Patterson Air Force Base (Dayton, OH) through the Small Business Innovation and Research (SBIR) program in support of the F-35 Joint Strike Fighter program.
The goal of the first phase of this project was to integrate and demonstrate five advanced machining technologies in the fabrication of an aluminum proof-of-concept piece. Machining technologies investigated during this project include: plunge rough cutting, frequency designed and relieved cutting tools, heat-shrink tool couplings, variable and right-angle head attachments, and sonic/dynamic tuning of machine/tool combinations. The proof-of-concept piece was designed to include features that are typically found in a complex modern aircraft structure, such as pocketing, windows, “C” and “I”-type structures and thin webs and skins.
Plunge roughing allows a relatively small milling machine to rapidly remove metal by feeding only in the Z axis and never in X and Y. The tool makes a series of overlapping, drilling-like plunges to remove cylindrical plugs of material. For the same programmed feed rate as conventional milling, the increased rigidity of a Z-axis move lets the tool remove a larger cross section of material.
One technique that increases efficiency in machining all materials is dynamic tuning. Typically, different materials require very different machining techniques. High spindle speeds combined with heavy cuts are possible with aluminum, however, appropriate spindle speeds need to be combined with relatively light cuts for titanium. There are points where the frequency of a tool’s cutting edge hitting a part becomes in phase with the resonant frequency of the machine. When these two frequencies are in phase, the vibration between them goes away, and the loads on the machine and the part become almost constant. This situation allows smooth cuts at much higher material removal rates.
There are a number of points where the tool frequency is in phase with the machine’s resonance frequency, and machinists call this phenomenon the “sweet spot.” These sweet spots can be calculated quickly by dynamically tuning the machine, tool, and coupling combination. Once frequency data exist for the various tool/machine combinations, the sweet spot with the highest RPM and greatest possible cutting depth can be chosen for the highest material removal rate.
Using heat-shrink couplings for cutting tools can greatly increase the rigidity of the machining system, which reduces vibration and increases accuracy. Sonic/dynamic tuning of a machining system can only be effective if frequencies of the system in operation are repeatable. One major source of variation is derived from the attachment of the cutting tool to the machine coupling. To minimize variation and allow for balanced operation, a standard collet or hydraulic toolholder must be manufactured to tight tolerance. A shrink-fit toolholder uses thermal expansion and contraction to securely clamp the tool. This tool/toolholder combination provides great clamping power. Typical runout using a heat-shrink coupling is 4 µm as compared to 5 and 10 µm with a collet or hydraulic coupling, respectively.
A major consumer of time, and cause for increased complexity in machining aircraft structures, is the number of setups needed. Employing a five-axis machine center can reduce the number of setups, but to minimize the setups further, pocketing in the X and Y directions will be necessary. Aircraft structures like the skeleton of the F-35A make extensive use of “C” channel or “I”-shaped structures for spars, ribs, stiffeners, and stringers. These are usually assembled from multiple components either bonded or riveted together. Such structures can now be created as part of a one-piece machined unit. Recently, there have been significant advances in both gear-driven and fluid-driven right-angle heads, which can allow the rapid removal (pocketing) of material from “C” channel or “I”-shaped structures.
The proof-of-concept part was designed and fabricated from 6061-T651 aluminum plate on a five-axis mill at AWV. The design has a length of 24" (610 mm), chord of 20.73" (526 mm) and thickness of 4.0" (51 mm). Its ribs include pocketing and windows, which allow passage of hardware and reduce weight in an actual aircraft structure. An I-beam structure was used for the spar, and pocketing was incorporated to help reduce the overall part weight. Running the length of the training edge is a C-channel structure of the type typically used to create attachment points for rudders, flaps, or ailerons in an actual aircraft wing.
To fabricate the proof-of-concept workpiece, an NC program was developed. During the programming process, it was determined that three machining setups would be required to fabricate the design. The first setup, Operation 1, consisted of roughing out the inside of the aluminum billet to remove most of the stock and help relieve any stress in the material. Operation 2 required the piece to be turned over on the mill bed for machining of clamping points and the outside airfoil profile. In Operation 2, the billet was clamped in the locations installed, and the airfoil profile was programmed to be machined to its final dimensions. Then the C-channel slot was made using a fluid-driven right-angle head made by Eltool Corp. (Cincinnati, OH). The final machining operation required the use of a vacuum fixture to hold the contoured piece in place for finish profiling of the inside of the airfoil and installation of the spar and rib pockets and windows.
Operation 3 was programmed to be completed using ball end mills and the right-angle head. Because of the limitations of the programming software and the process used to post this program to the machine, the right-angle head could only be used in the ±Y directions. This required that the proof of concept piece be rotated 90º to install the rib pockets and windows in the final step of Operation 3.
Programming the workpiece determined that plunge rough cutting would not result in the highest material removal rates for this aluminum proof of concept design, because of the feed rates and depths of cut possible in aluminum with traditional approaches and tooling, such as large-diameter face mills. Material removal rate was maximized by making a series of passes over the piece with a 2.00" (51-mm) face mill. To study the plunge-roughing technique with titanium, representative titanium aircraft fittings were fabricated at AWV with the help of NCDMM. The results of these tests indicated that the machining time required to fabricate this representative piece in titanium could be reduced as much as 21% when using plunge roughing.
Representative titanium concept fitting is shown being machined on a three-axis mill using plunge rough cutting.
All of the tools and their respective couplings used in the fabrication of the proof-of-concept piece were dynamically and sonically tuned in conjunction with AWV’s VF-11 five-axis mill (Haas Automation Inc.; Oxnard, CA) by BlueSwarf Manufacturing Laboratories (Clearfield, PA). Dynamic tuning is performed by attaching an accelerometer to the end of the tool while it’s mounted in the mill, and then exciting the tool with a hammer containing a piezoelectric transducer. This method of tuning is referred to as a “tap test.” When the machine/tool combination is tapped with the piezoelectric hammer, the frequency response and compliance of the system are recorded. Using a FEA module in conjunction with the number of flutes on the tool, the milling machine’s torque curve, desired tool stepover, and material of fabrication, the optimum spindle speed, depth of cut, and feed rate is predicted. By determining the optimum spindle speed and depth of cut for a machine/tool combination, the system comes in resonance, and any chatter in the system is virtually eliminated. Dynamically tuning a machine/tool combination yields a stability map that shows the stable depth of cut as a function of spindle speed at a specific tool step-over or width of cut.
For use in the proof-of-concept design, a 1.00" (25 mm) insert end mill was tuned under various conditions. First, a hydraulic 1" standard duty Powergrip Mill Chuck (manufactured by Kennametal Inc.) with two different tool reaches was tested with reaches of 3.75" (95 mm) and 3.50" (89 mm). The tuning performed in these tests was determined with a constant tool stepover of 0.75" (19 mm), and was limited to spindle speeds no greater than 7500 rpm (the maximum spindle speed possible with the current gearing of the five-axis mill used in this research). This test determined that by reducing the specific tool’s length of reach from 3.75 to 3.50", the depth of cut can be approximately doubled from 0.025 to 0.050" (0.64–1.27 mm). Chatter frequencies collected for both configurations across the range of spindle speeds were greater for the 3.50" length of reach.
A series of sonic/dynamic cutting tests were performed using the 1.00" insert end mill mounted in a side-lock holder in 6061-T651 aluminum to determine the maximum depths of cut possible with this configuration. These tests were performed by recording the sound produced by the tool while making a test cut, and running a Fast Fourier Transform (FFT) of the resultant data. The tool was run at a spindle speed of 5000 rpm with a feed rate of 100 ipm (2540 mm/min) and depth of cut of 0.100" (2.5 mm). The results of this test indicated that, at these operating parameters, chatter was present.
Chatter was also evident to the operator due to the machine sound and the surface finish on the test piece. A threshold magnitude of 65 was used to determine the presence of chatter in the machine/tool combination based on experience. This test was performed again using the same parameters as listed above, except that spindle speed was increased to 6500 rpm. Test results indicated that little or no chatter was present at these operating parameters. Comparing the collected data showed that the chatter evident at a spindle speed of 5000 rpm was virtually eliminated at 6400 rpm.
A heat-shrink tooling system with a CV-40 taper spindle was selected for use on this project. Shrinker system heat-shrink couplings from Kennametal Inc. (Latrobe, PA) come factory-balanced at a minimum of 2.5g, which makes them well-suited for higher spindle speeds and feed rates. The selected tools have relieved cutting surfaces that are designed to negate interference with the surface of a part, reducing possible undesired harmonics in the system. Gripping torque of these heat-shrink toolholders is significantly higher than that produced by standard toolholders, especially as the tool diameter increases. In addition to providing higher gripping torque, heat-shrink couplings reduce tool run-out by up to 60%.
The proof-of-concept design was completed in approximately 33 hr of machining time with 6.5 hr of setup time required. Operation 3, finishing of the inside of the airfoil and installation of the rib and spar pockets, accounted for about 2/3 of the machining time. A total of 1.26 ft³ (0.036 m³) was removed from the piece, with the largest amount removed during the roughing of the airfoil carried out in Operation 1. The tolerances achieved in the final piece were ±0.003" (0.08 mm) at any location. Additionally, the airfoil skin thickness, designed at 0.050", had a final tolerance of ±0.002" (0.05 mm).
The proof-of-concept piece is representative in design and structure to the outboard section of an aircraft tail produced by AWV. It was fabricated in about 60% of the time required for the tail component. The tails contain aluminum ribs with a forward, aft, and main box spar contained within two composite skins. This assembly is bonded and riveted together using typical aerospace fabrication techniques. A representative structural section of this tail structure requires approximately 70 hr of machining, lay-up, finishing, trimming, sanding, curing, bonding, and fastening touch-labor hours. Such a representative section is approximately the same size as the proof of concept piece, which required only 39.5 hr of recurring time to fabricate to final dimensions.
This article was first published in the March 2006 edition of Manufacturing Engineering magazine.