FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1961
1961 - 0842.PDF
854 FLIGHT, 22 ./««*> 1961 Missiles and Space flight MISSI RELIABILITY BY A CORRESPONDENT THE elements that make for success in the development ofa major defence system, such as Thor, Atlas and Titan,were outlined recently by a group of men best qualified in the United States to discuss the subject. They spoke before theannual meeting of the Aviation/Space Writers Association in New York last month. Leading the discussion was Lt-Gen B. A.Schriever, newly appointed chief of the USAF Systems Command, who told the writers that, while "flight tests are glamorous andcertainly conspicuous," they are analogous only to that part of an iceberg that we see above the surface. For every flight test thereare many more development steps, plans, actions, designs, confirma- tions and design changes which in total make or break a new missileprogramme. Gen Schriever noted that the Germans had found it necessary totest a great many V-2s before they got a workable design. Today's missiles, which were much more complex, might require 10,000flight tests if they were to be proved in the same manner. Yet, missiles such as the Titan and Atlas were proven for operationaluse after only a few dozen test flights. "To achieve the performance that would be gained from the other 9,900 test flights, a series ofsteps must be taken during the development of a new missile to measure its anticipated performance indirectly and thus avoid thetrial-and-error, 10,000-flight method," Gen Schriever said. This is part of the programme of concurrent development, betterknown as concurrency, which Gen Schriever has fostered in the development of missiles and without which, in his words, "apriority weapon system cannot be developed today."' He further denned concurrency as carrying on simultaneously the research anddevelopment effort for a new missile, its operational planning, logistic support and the training of operational personnel. "Concurrency," he told his audience, "puts a premium on beingtechnically correct," and he cited the Thor programme as a good example of what can be done. "Thor was on the stand for firstlaunching within a year after inception of the programme, and in operational sites ready for firing by military personnel in threeyears," he said, and then added: "Today the reliability and readi- ness of these missiles exceed that of manned bombers." J. R. Dempsey, president of the Astronautics Division of GeneralDynamics Corp, who directed the Atlas missile programme for that company, explained how the principle of concurrency was employedin the development of Atlas. "Our team had to make decisions on a missile that was still many months away from its first test flight—it was a good deal like betting a poker hand in the dark until the fifth card." "But the concept has paid off," said Mr Dempsey, "and one ofthe reasons is that we hedged our bets with an intensive reliability programme that has been a basic part of our work on Atlas fromthe very beginning." He indicated that the reliability programme had been accom-plished in two principal phases: First, by defining in the early design stages those areas which seemed most likely to give technicaltrouble. Intensive theoretical and applied analyses were made of those areas so that information would be available to designers asthey worked out details of the actual hardware. Second, the re- liability programme was set up so as to evaluate each bit of technicalexperience as Atlas was developed and proven on the test stand, and this evaluation was then pumped upstream to manufacturingand design phases so that it could be worked into all new produc- tion. As a result, declared Mr Dempsey, about a third of the 25,000employees of GD-Astronautics were engaged at one time or another in some form of testing or quality control. The permanent reliability organization at GD-Astronauticsconsists of 116 people, divided into two groups. The first is Human Engineering, where psychological studies of man-machinerelationships are made to anticipate and avoid the design of machines or procedures that run contrary to human behaviour.The second is a Reliability Diagnostic Laboratory where critical weakness tests are made. New components not yet used on a missilesystem are tested under appropriate environmental conditions and the information is fed to design engineers. And, when somethingalready installed in the missile system does not perform as it should, the components may be funnelled back through this laboratory. Failure-analysis reports are handled promptly, with feedbackto all interested parties for immediate corrective action. There is End-product of missile reliability, a perfect launch from Cape Canaveral (Titan G8 with Avco RVX-3 nosecone, September 29,-1960) even a follow-up system. When action is not taken within theprescribed time, a report goes to top management, which sees to it that appropriate action is taken. Mr Dempsey cited several examples of the detailed, precisereliability standards that were being used in connection with Atlas. "Even cleanliness has its reliability standard," he noted, "for therequirements of missiles in this respect in some ways surpass those of the human system. Fluid systems on Atlas are chemically cleanedand inspected by black light, which makes impurities glow for easier detection. Hydraulic and fuel systems are not up to standardif impurity sizes exceed 175 microns in the fuel system and 30 microns in the hydraulic system. We are working with mere specks,and maintaining cleanliness standards which surpass those of most hospital operating rooms to assure reliability," he said. Mr Dempsey noted that 81 Atlas missiles had been fired for theUS Air Force as of April 10 and, of these, 53 were totally successful, 20 were partially successful, and only eight were totally unsuccessful.He said the results to date "demonstrate the painstaking effort that goes into each missile long before it reaches the test base." Test launches yield a vast amount of information that is fedback into the reliability programme. Each test launch provides about two and a half miles of data recorded on magnetic and papertape. Before an Atlas test flight is made, 145 measurements are transmitted from the missile to recorders by land lines. Afterlaunch, 185 measurements are taken, 32 continuously throughout the flight and the rest intermittently. The intermittent readingsprovide over 600,000 data points. Recorded at Cape Canaveral, the data then is flown to San Diego and within 24 hours the designengineers meet to review each significant element of the missile and its performance during flight. Therefore, in little more than aday after a test flight, GD-Astronautics personnel can be working out improvements for subsequent missiles. Dr Harold W. Ritchey, vice-president of the Thiokol ChemicalCorporation, a leading producer of solid propellants and rocket engines, discussed some aspects of reliability as related to missilepropulsion systems. He began by saying that reliability was only one of four problems a manufacturer faced in producing systems forthe armed forces. In addition, cost, performance and schedule must be given equal attention. If costs were too high, funds mightbe exhausted before a system was ready; if performance did not meet requirements the system might not gain acceptance: ifschedules were not met or surpassed the system might be obsoles- cent before it was ready. "We literally are on a tightrope, trying tobalance the four elements as we go along," declared Dr Ritchey. "However, the task of balancing these four factors can be greatlysimplified by extensive test programmes on the components, on each sub-system as it is developed, and finally on the propulsionsystem itself. Then we no longer need to walk a tightrope, but instead can make our way along a bridge that has some width Almost any degree of reliability was attainable at cost, dechredDr Ritchey, who illustrated his point with some reliability-.'ost curves. However, he added, the dollars spent to achieve a certaindegree of reliability for any given system would be less if, h lr>e early stages of the programme, an extensive test programm' 1°
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
