FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1962
1962 - 0293.PDF
fLIGHT International, 22 February 1962 295 concep* itself about which Smiths have so far revealed relatively little—and left their rivals wondering just how they have made it work effectively. A great deal of new information is to be released at the end of this month during the joint RAeS-Institution of Electrical Engineers symposium on electrics in aircraft. Here, Smiths will finally reveal some of the details of multiplex working. All that can be said at the moment is that the technique largely relies for stability and failure-sensing on datum- and gain-equalizers applied to each channel in a given control axis. Considerable com puter studies of the stability of a multiplex channel were made; and these proved to be substantially correct in the first triplex system flown in the Smiths Dakota. Once the channels are actually run ning in synchronism, any difference in output will indicate a fault. Each group of channels therefore has a torque switch which is live in the closed position and will open to cut out one channel if full torque is applied. The switch is in fact opened mechanically pro viding maximum safety of operation. Power sensing is also applied to identify a "dead" servo and isolate this from the system. The three channels are symmetrically arranged, there being no master channel. Pitch, roll and yaw axes are also separated so that two faults do not necessarily fail the entire system. Two faults might reduce two axes to simplex level and not completely disengage one channel, but reliability forecasts have been based on the supposition that they would. The separation of the three axes considerably simplifies the matching of the SEP.5 autopilot to another type of aircraft. Smiths have now also completed detailed study of the divided control situation, when one axis is under automatic control and others under manual. All the possible combinations have been tried in actual poor visibility conditions, that is, one channel manual and two automatic, or two manual and one automatic. If azimuth control is automatic, and pitch manual, the kick-off drift (or decrabbing) manoeuvre is extremely difficult to time. It becomes relatively easy for the pilot to allow the aircraft to drift right off the runway in cross-wind components as low as eight or nine knots. If, on the other hand, pitch is automatic and azimuth manual, the timing of the whole manoeuvre is dictated by the pitch programme, which remains constant. The pilot is able to use the most effective visual aids for azimuth guidance, and he can easily learn the timing of the flare manoeuvre and judge when to decrab the aircraft. Other combinations are easily seen to defeat themselves. Smiths are therefore confident that the autoflare stage of the Trident system will work effectively. Another flight programme has covered manual take-over in the presence of the ground. Under these psychologically more pressing conditions, passive faults have been shown to be relatively easy to handle. If, for example, such a fault occurs close to the ground, but above the beginning of the flare-out, at 50 or 60ft, the pilot has about five seconds in which to take action. A reaction time of two seconds is not unreasonable and the margin is therefore adequate. If the fault occurs after initiation of the flare-out, the rate of descent has already begun to decrease and very little corrective action is required to produce an adequately gentle touchdown. The SEP.5 is designed so that at any instant during the actual flare-out manoeuvre, the air craft is in pitch trim. The pilot would therefore find the aircraft trimmed at whatever point he had to take over manual control. In addition to the test and production work for the V-bombers, the Belfast and the Trident, Smiths have contributed to automatic landing programmes abroad. A DC-7 belonging to the Federal Aviation Agency has been fitted with couplers and computers so that it can demonstrate the single-channel Autoland system in the USA. Automatic landings have been accomplished with this aircraft at RAE Bedford and it is shortly to return to the FAA test establish ment at Atlantic City for a full evaluation programme. The German WGL-Fachausschuss Blindlandung at Hanover is to be supplied with Autoland equipment for a Noratlas and, later, for a Transall. Demonstrations have been given to potential foreign customers. PVD, which forms an integral part of the Trident system, but can also be used by itself, has made considerable progress abroad. Boeing have been evaluating PVD in the 707-80 prototype; and BOAC have completed an installation in one of their 707s. KLM has flown PVD in a DC-8 for some time and is now preparing to add a flare-out computer to take the system a stage further towards low-visibility operation. The ramifications of the many developments which Smiths pro duced specifically for automatic landing are many and have resulted in a new range of components applicable to a great variety of systems. On the present schedules, BEA and de Havilland can be Assembling the mechanical unit of a Trident master compass. Note shaft indications brought to the front of the case. The electronic unit bolts to the rear to form a § ATR case expected to make the first operational trials of a civil duplex system for autoflare. This will represent the first major step in practical application of what has remained for so long a tantalizing prospect for the airlines. But the first routine trials of the ultimate triplex system are now certain to be made in the military Belfast, which is to be equipped from the outset with this level of equipment. There are at the moment no firm plans for Tridents to be up-graded to triplex level. The RAF has shouldered the burden of first trying this equipment, and civil customers are likely to await their verdict before committing their own aircraft. In the meantime, RAF Bomber Command will soon begin to amass squadron experience with single-channel Autoland and will doubtless add usefully to the fund of practical knowledge. ELLIOTT BROTHERS (LONDON) A CONSIDERABLE amount of information has already been published on the Elliott Brothers (London) equipment for the VC10 (Flight for October 9, 1960). In designing the overall system, the company have incorporated many of the components and techniques employed in the Bendix PB-20 autopilot, the 100 series Flight Director System, and the Polar Path compass. It was considered that it would be expedient to approach full automatic landing in two phases. In order to gain experience of the problems of automatic control during the landing phase—and to gain pilot confidence in the manoeuvre—automatic control to touch down would at first be applied solely in pitch below the normal break-off height, the pilot retaining control in azimuth. When the experience gained in perfecting autoflare is consolidated, and the azimuth ground guidance is resolved, it is envisaged that full automatic landing will be incorporated on the VC10. In order to meet the Air Registration Board safety requirement for automatic landing—the overall system must be such that the risk of a fatal accident during the automatic landing manoeuvre shall be less than once in ten million landings—Elliott Brothers adopted the principle of a monitored duplicate control system. It has been shown that by adopting this system the mean time between failures required of one autopilot reaches practical levels. Extensive con sideration has been given to the isolation between the two systems in order to minimize the risk of a common fault affecting both channels at the same time. The design of the basic autopilot has now reached completion and rig tests are well under way at the Vickers-Armstrongs factory at Weybridge, where a full-size aircraft systems rig, including complete control runs and power controls, has been set up. Tests are at pres ent being carried out in conjunction with the Elliott aerodynamic simulation facilities and representative circuitry and hardware of the automatic control system. A great deal of statistical performance data and operating experience will have been accumulated on the rig by the time autoflare flight trials begin. Continued overleaf
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
