Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Appendix C Advanced Composite Tape-Laying Bead The Advanced Composite Tape Laying Head automates the highly labor-intensive job of laying up laminated composite parts. The tasks the tape-laying machine is designed to accomplish are part of the overall composites production process, which consists of tool set-up, material orientation, material cutting, lay-up, cure, post-cure, and machining. The advanced tape-laying machine represents one set of concepts developed to achieve an automated approach to the process. The machine--of which the head is the most critical component-- comprises a bed on which the part is laid up, a gantry to carry the head, and computer control mechanisms for the drive and the head. The bead and its control determine the orientation of the fiber, the location of the ply and its termination, and the compaction of the entire laminate structure. The head consists of a tape roll supply system, a tape cutting system, a tape transport, tape laydown rollers, and compactors. All are controlled by a mini-computer. The following conditions, technical and market, underlie this case of technology transfer. Technical Conditions 1. General Dynamics holds several of the key patents for advanced composite tape laying. 2. Advanced composite materials are composed of either graphite or boron fibers in a resin base. They are unidirectional and must be laminated or woven to achieve the tremendous structural properties needed for airplane construction. When laid up and cured they have a strength-to-weight advantage of roughly 30 percent over aluminum. The unidirectional characteristic and the state before curing pose difficult handling problems in manufacture. For instance, the material comes on backing paper which protects its adhesive surface and allows it to be rolled, but the backing paper frequently gets out of alignment with the material itself. Furthermore' its perishability means it has to be dated and used in order of purabase. 3. Advanced composite materials have been changing rapidly since the early 1960s In format, cost, and composition. Accordingly, the processes to manufacture composite components have been highly unstable. Nevertheless there has been a great deal of pressure to stabilize production processes because composite materials offer immediate performance benefits in aircraft. The two driving forces 36
behind automation of composite manufacturing are labor cost and ease of handling, as larger and larger components are laid up. It is estimated that on the average lay-up costs account for 17 percent, and handling costs 47 percent, of total composite manufacturing cost. 4. Many types of composite material formats are available. The two principal categories are tape and broadgoods. Broadgoods can be unidirectional or woven. Tape comes in one-incb, tbree-inch, and six-inab widths, broadgoods in multiples of one inch. Adhesive systems differ from supplier to supplier and even from lot to lot. McDonnell's composite area, for example, deals with five different suppliers selling four different materials in 5-10 different formats with 8-10 different adhesive systems. 5. The cost of composite materials has decreased significantly since they were first introduced, but not as rapidly as was first predicted. In 1968 boron composites were S500/Pound. In 1972 graphite had supplemented boron at $100/pound, and now graphite is S40/pound and boron is $200/pound. Broadgoods are sold at premium prices, currently about S70/pound. 6. Originally composites could be purchased only in the form of tape. Broadgoods became available in the early 1970s, which gave rise to competing design and manufacturing philosophies. Some organizations maintained their preferences for tape, and other chose broadgoods instead. Today there are three schools of composite manufacture--tape, represented by General Dynamics; broadgoods, represented by Northrop; and hybrid, represented by Grumman. The tape school claims it is the low-cost approach, emphasizing the low scrap and easy handling properties of its format. The broadgoods school maintains that its format is more flexible to design requirements, and the hybrid school sees its approach as the most versatile. 7. The chief concepts of the advanced composite tape laying technology are the computer-controlled handling, laying up, cutting, and compaction of composite tapes. The embodiment of these concepts can be performed with equal effectiveness using a variety of different combined techniques. Thus the real value lies in the concept of automating these steps. Market Conditions 1. Until very recently advanced composites accounted for only a negligible part of every military airplane. The evolution of composites in McDonnell f ighters illustrates the rate of growth in composite use. The McDonnell F-15 Eagle has a boron and graphite empennage accounting for two percent of the mater ials in the plane . The F-18A Hornet now in prototype contains 10 percent graphite composites of which 800 pounds are produced by McDonnell and 400 by Northrop. The VTOL Harrier, still in early preproduction phase, 37
contains 25 percent composite materials. Composites experts foresee that by 1990 there will be military planes they call ~blackbirds,. constructed of 55 to 60 percent composite materials. At the same time the size of components is increasing rapidly. The new Harrier design, for example, calls for 28-foot wing skins. 2. The first use of composites has been for airplane skins, but some companies are beginning to experiment with composite structural parts, as in the McDonnell Barrier. Designers disagree as to whether airplanes will incorporate significant structural use of composites in this century. The cost sensitivity in recent military acquisitions has led to a countervailing trend in which weight has been sacrif iced for cost. 3. Three firms have staked out leadership positions in various aspects of the composites area. Two of these, General Dynamics and Grumman, have been leaders in composites manufacturing--General Dynamics for the fuselage and Grumman for the wing. McDonnell also claims leadership, but its leadership has to do with the intricacies of compos ite des ign and with the types of advanced compos ite structures its designers are incorporating into aircraft. 4. Adoption of composite equipment for limited production use is not quite as dependent on major program commitments as the adoption of other forms of automation has been in conventional assembly areas. Most airframe companies recognize a need to gain experience in automating the s new technical area in advance of volume production. Visions of future composites factories differ according to the production philosophies of different companies. The chief philosophical split seems to center on the balance between design and manufacturing e McDonnell and Grumman, for instance, bold the philosophy that the highest priorities should be design enabling and airplane performance, while General Dynamics tends to emphasize manufacturing considerations. 5. The machine tool companies play a pivotal role in composite automation. Tbey have been responsible for producing come of the key parts of equipment, in some cases transferring the technology from other industries such as adhesive tape producers and the garment industry. Since the materials suppliers made broadgoods available, a whole new set of equipment makers, previously focused on the garment industry, have entered the market. The result has been that the previous suppliers perceived smaller markets and became less willing to commit themselves to equipment design without charging custom prices. The Air Force Materials Laboratory began to sponsor development of composite production technology in the mid-1960s when it funded General Dynamics to develop and improve its f irst composite tape-laying machine. In the late 1960s the Air Force Scottsdale Conference devoted a good deal of attention to composite automation. Both General Dynamics and Grumman announced that they expected h igb volume use of 38
automated composite production in the next decade. General Dynamics projected a need for 15 tape laying machines for its F-16 program, and Grumman anti cipated that 5 would be needed for its F-14 program. After funding a second General Dynamics contract in the early 1970s, the AFML acknowledged the proliferation of composite manufacturing technologies by funding a group of four different integrated composite production system programs in 1976-79. These projects had two objectives: to find out where the costs were in composite manufacturing and to encourage complete automation of the entire labor-intensive process. The case described here is a program to improve still further the General Dynamics advanced compos ite tape laying head, one of the first projects funded under the AEML manufacturing technology group's composites production integration program. General Dynamics General Dynamics (see description in Appendix B) was a pioneer in automating composites production. It developed its first mechanized device for lay-up of advanced fiber composite tape in 1965, using corporate funding, and took out basic patents covering the technique. At that time composite materials were available only in tape form. General Dynamics was the first user of three-inch tape instead of the one-inch format previously available. Having proved the feasibility of its tape-laying concepts, General Dynamics requested support from the AFML for a development program to build and demonstrate a full-scale tape-laying machine. The machine was desi gned and built by Conrac Corporation, a subcontractor to General Dynamics. The Conrac machine contained computer control, a head that moved along three axes, a six-inch guillotine ~hear, and a sprocketed tape guide system that used the tape backing paper to orient the tapes. The machine would lay a strip of tape no less than 9.75 inches long. It was a prototype, but one that procured sturdy enough for production use. In 1972 General Dynamics, by this time well into the prototype phase of the F-16, secured another Air Force contract to develop a new tape laying head capable of positioning the tape without a sprocketed guidance system. The improved head went into production on F-16 prototypes in May 1974. In the early 1970s materials suppliers began making broadgoods available. Many companies saw in broadgoods the possibility for greater design flexibility--albeit at greater cost--but General Dynamics maintained a steadfast commitment to tape on several grounds. . Broadgoods cost more per pound. 39
Tape was, by General Dynamics's estimate, the lowest cost approach by 10 percent. Since tape was laid up and cut to near final shape there was less scrap than with broadgoods. The extra handling problems caused by the material's limited shelf life and the varied mix of materials required a sophisticated storage, inventory control,-and retrieval system. The tape could be inspected as it was laid down; broadgoods required more sophisticated inspection ahead of time or higher scrap parts later. Flexibility and intelligence could be incorporated in the machine, not in skilled workers. Still another important consideration was General Dynamics's organizational investment and proprietary position in tape equipment. The composite production philosophy that emerged as General Dynamics defined its choices relative to those of other composite users emphasized three factors. First, it was of highest importance that each composite part be fabricated in the most cost-effect~ve manner. Second, the investment should be made in sophisticated equipment rather than in materials handling and control systems. Third, General Dynamics would instruct its designers to design composite components that were manufactured using the tape-based process, even if that required them to limit their designs as to weight or performance to some degree. The embodiment of this philosophy was a system using ply-on-ply, near net shape laminates with automatic process control. As the F-16 production program began in 1976, the Air Force funded General Dynamics to develop new manufacturing concepts for tape laying to overcome the drawbacks of its previous tape layers and provide for a f ully automated composite production system. The chief objectives of the contract were to eliminate hand laying, cut any form or angle (which the guillotine shear would not do), and lay pieces shorter than the 9.75 inch minimum. While the program was underway the Conrac improved head was used in production for lay up of vertical stabilizers. This program continued under changing leadership from 1976 until after the final report in March 1980. The head remains in the laboratory. Through it is said to work, General Dynamics has so far had difficulty getting a machine tool builder to build it at an acceptable price. Ingersoll-Rand has meanwhile built a non-prototype version of the Con rac which is being debugged for production. A new version of this equipment, capable of laying one-inch and six-inch tapes, is out for bid, but bids so far are more than twice the previous machine cost. To adopt the latest improved head supported by the AFML into production would require debugging it on the Conrac which, through nominally a prototype, has been the most reliable production equipment 40
General Dynamics teas had. Even provisional adoption of the new head, therefore, will probably await the successful and reliable operation of the Ingersoll-Rand. Grumman Grumman thee description in Appendix B), like General Dynamics, bad pursued mechanization followed by automation of composite production since the early days of composite materals availability. In the late 1960s it borrowed the Conrac machine for evaluation. It rejected the Conrac as too inaccurate in lay-up and cutting and purchased instead a Metro tape head. The Grumman lay-up approach differed from that of General Dynamics. While General Dynamics laid up its plies one over the other on a large template, Grumnan inspected each ply separately and then f it each ply into a stepped frame. Grumman 's plies had to have precision edges in order to fit whereas General Dynamics' could be near final shape and then trimmed after curing. The Metro machine had been developed by the Metro Company for use at 3M. Grumman's Advanced Development Group modified the Metro machine for centerline track ing to eliminate the sprocket boles and four-inch paper. In 1969 it built its own mechanical Flintstone machine for the F-14 program. In anticipation of work on the unwieldy B-1 bomber structure, the horizontal stabilizer, it modified the Flintstone to handle larger F-14 mylars . From the beginning, then, two factors drove Grumman to pursue composite production technology. One was accurate lay-up leading to consistent, high quality production; the other was handling of large parts. In the early 1970s Grumman saw broadgoods as an opportunity for increased design flexibility. Beginning to analyze the entire composite production process as a system, the Advanced Production Process Group began on its own to develop concepts for an integrated laminating center (ILC) in 1974-76. The reports of GD's improved tape head were available in 1974. Grumman compared the General Dynamics head with two other ~class. machines designed to do similar tasks, the LTV head and the Boeing Vertol head. This time versatility was the decisive criterion because the LTV head could deal with shorter minimum lengths of tape. Grumman once again rejected the General Dynamics alternative in favor of the LTV head to incorporate into its planned laminating center. In 1977 Grumman received AFML funding for evaluation of and demonstration of ILC concepts as applied to the B-1 stabilizer. The ILC combined tape laying capability with broadgoods production capability in the same facility. It eventually installed a laser cutter to cut the broadgoods. It was rated to produce eight stabilizers per month. While the previous case (Appendix B) shows that there were poor communications between Grumman and General Dynamics during the period in question, the decisive factor in Grumman's non-adoption of the 41
General Dynamics tape laying machine seems to have been the existence of alternatives that more closely matched Grumman's preferred manufacturing philosophy, which emphasized consistency, accuracy, and versatility for design performance over manufacturing cost effectiveness and volume production. McDonnell McDonnell Douglas (see description in Appendix B) pursued a leadership position of a different kind from Grumman and General Dynamics in the area of composite materials. Known as an engineering-dominated design house, McDonnell favored the production technology tbat was the most flexible from the design point of view. By the late 1960s it was clear that McDonnell designers would find many uses for composites in numerous con f figurations. McDonnell began tracking tape-laying machines in 1969 when Boeing Vertol demonstrated its version of a tape layer using 2 1/2 inch fiberglass tape. After the Scottsdale Conference, where General Dynamics and Grumman projected big needs for tape-laying equipment on their next programs, McDonnell monitored the Air Force's sponsored developments of the Conrac used by General Dynamics. Its composite engineers made trips to Vertol and to LTV to see their versions. But all had unacceptable gaps and overlaps by McDonnell' s standards. In May 1972 the McDonnell manufacturing process engineers became aware of a new approach to composite production using broadgoods cut by a laser cutter. After evaluating this approach in comparison with the General Dynamics and LTV tape layers, McDonnell ordered a laser cutter in 1974. McDonnell's chief reasons for opting for the laser were its need for accuracy in cutting, laydown accuracy, and the ability to cut irregular shapes. The firm also wanted ~ throughput rate that none of the tape layers available offered, since it needed to cut single layers and often laid up 55 layer plies. It intended to keep single layer cutting to give maximum scope to designers. The raw material cost in broadgoods form, which it also calculated at a 10 percent total cost differential from tape, was a factor; but it seemed likely that McDonnell could eventually begin to make its own broadgoods in the 1 2-foot long bites that seemed optimum for the type of nesting McDonnell typically did. McDonnell 's immediate reason for non-adoption of the General Dynamics or other tape layers was its dissimilar manufacturing philosophy. In ~ ime, when high volume composites production becomes commonplace, McDonnell could well invest in some form of tape layer to fabricate small components, but for the present the main capital investment will continue to be in the broadgoods area. 42
Northrop Northrop, McDonnell's partner in the F-18 program, came to composite production in the mid-1970s, later than the other companies. As a result, it was in a position to choose between tape and broadgoods. It has a low volume of composite components to make, and they are fairly small in size. Although it is known as a low-cost producer, Northrop also places a very high priority on flexibility and it iodses broadgoods to be cons istent with that philosophy. not pursued the idea of ~ tape layer. ~ ~ Consequently it has bike Grumman it has received ANAL support for an integrated composites production facility, but its approach to the concept has been much dif ferent. In 1976 Northrop gathered a group of people f rom several levels of management to visualize what a composites factory might look like in 1990. From that point its engineering department assembled a series of building block s to achieve this objective gradually. . _, _ _ Instead of the laser cutter which will cut only single plies, it teas adopted the Gerber cutter which will handle multiple plies. other equipment that has been used successfully outside the industry but has hitherto been unfamiliar in air frame manufacturing. In general the characteristic approach Northrop has taken to manufacturing innovation is to adopt feasible technology from any industry and adapt it for its purposes. It has added to the t Northrop' s reason for non-adoption of the General Dynamics equipment is the clearest example of a mismatch in manufacturing philosophies, to the point that Northrop did not consider the tape laying concept at all. 43
Composite Tape Laying Chronology 1965 General Dynamics begins mechanization of composite tape laying 1966-67 First AEML contract results in Conrac machine 1969 19?0 The AFML holds the Scottsdale conference and publicize s expectations for future automation in composites. Grumman builds Flintstone machine for F-14 program to handle larger F-14 mylars in anticipation of B-1 hor izontal stabi li zer s . Early Composites become available in broadgoods form. 1970s 1972-73 General Dynamics has AFML contract to improve Conrac head. 1974 McDonnell commits to broadgoods and laser. 1975-76 Grumman develops Integrated Laminating Center (ILC) using LTV tape head concepts. 1976 The AFML sponsors General Dynamics in another tape laying head improvement, this time under heading of integrated composite production. 1 977 Grumman gets contract f ram the AFML for application of ILC ideas to B-1 stabilizer and Northrop receives support for its IFAC (Integrated Fabrication for Advanced Composites). 44
