FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1963
1963 - 0396.PDF
F-104G STARFIGHTER . . . results were outside these limits by a fair margin, but it has now been handsomely exceeded in a number of groups of flights. Each time the system has been operated by another unit or agency, the performance has had to be built up anew. This accounts for initial reports of poor performance in Europe, where the results at manufacturer level are now well within specification. The Luftwaffe at Norvenich is gaining experience in operating the system. Additional improvements now possible include a warm-up rate of 20°F/min and the use of air-bearing gyros offering random drift rates measured in thousandths, rather than hundredths, of a degree per hour. Air bearings can suffer severe wear during the starting and stopping phases, but Litton have already successfully run such a gyro through 85,000 stops and starts. During manufacturer's development flying at Palmdale, some 1,167 flights were made up to October 1961, and the c.e.p. of the LN-3 and PHI combined was a mile or so outside specification. The PHI contribution is impossible to assess, but the manufacturer's specification states that accuracy is ±1.4 n.m. plus 0.35 per cent of the total distance travelled. From October 1961 to January 1962 a further 123 flights at Palmdale were assessed, following incorpora tion of the -9 modifications, and the c.e.p. came almost up to speci fication. Initial test flights made at Edwards AFB with unmodified LN-3s showed somewhat poorer results. But further trials with modified systems, including a series of flights at low level with violent manoeuvring, intended to investigate the use of inertial systems in the F-4C Phantom 2, gave accuracies well within speci fication. Early F-104Gs assembled at Manching by Messerschmitt from US or Canadian parts showed a similar improvement in perform ance. During 17 flights from November 1961 to January 1962, the 50 per cent c.e.p. was relatively poor, but during a second series of 41 flights in April, achieved accuracies came well within specifica tion. In August Fairey and Sabca at Gosselies were reporting performance almost as good. Using the criterion of a "maintenance effort" representing about 5hr work, Edwards assessed the LN-3 as requiring 0.3 efforts per flight in the difficult "avionics marriage" period, when the equip ment was being matched to the aircraft. After about 100 flights the figure flattened out to 0.02 efforts (0.1 hr work) per flight, equivalent to 30 flights per effort or nine hours' flying per hour of maintenance. At Edwards AFB, during Category 2 testing, and at Palmdale during the "avionics marriage" period, mean time between failures of pre-9 systems was considerably below the 200hr specified, but the target has been exceeded since then. Litton give an MTBF for verified LN-3 failures during Category 2 flight testing, equivalent to an "airborne navigation effectiveness"—i.e., successful sorties against sorties flown—approaching 95 per cent. Whatever the apparent quality of this system, it provides a Components of the Honeywell MH-97G autopilot: I, 8G158 autopilot and damper computer; 2, control stick and grip; 3, roll-rate gyro pack age; 4, two-a/is rate gyro package; 5, pitch trim indicator; 6, pilot's control panel; 7, normal accelerometer; 8, automatic pitch control computer; 9, APC gyro package FLIGHT International, 14 March 1963 navigation capability unrivalled by that of any other tactical single- seater in Europe—with the possible exception of the F-105D. Pilots have reported very favourably upon it, pointing out that their knowledge of position after an hour's tactical flying is sufficiently good for them to be able to correct ground radar operators' position estimates. They claim it to be good enough for them to get within airborne-radar range of a ground target, thus allowing them to navigate a complete attack mission without any reliance on ground- based aids, or—even more important at high speeds—without the need for visual reference to the ground. The PHI allows the pilot to select by push-button immediate bearing and distance to any one of twelve different locations, including his home base. It is only with a system of this kind that the tactical use of the F-104G makes sense. MH-97G Autopilot Originally designed by Minneapolis-Honeywell for the F-104C, the MH-97G forms a comprehensive flight control system for the F-104G, containing a number of rate gyros of its own, but otherwise relying on the Litton LN-3, PHI and air-data computer for, res pectively, attitude and heading information, navigation reference and height and Mach data. Honewell defensively claim that, because the pilot's most direct contact with the aircraft is through the auto pilot, he attributes to this system many faults which actually origin ate in the more remote systems supporting it. In fact, the company is able to show some remarkably good MTBF figures for con siderable portions of the MH-97G. It provides three main sectors of flight control: autostabilizer three-axis damping; automatic pitch control (APC), which governs stall protection by stick shaking and then snatching; and flight-path control. Autostabilizer and APC normally remain engaged at all times, but the pilot can switch on the automatic flight-control system (AFCS) and the aircraft will be held in the attitude at the moment of engagement. Through a force-senser in the control column he can then set the aircraft in any attitude, under autopilot control; this is termed control-stick steering (CSS). If the bank angle at the moment of stick release is less than 7° the wings will be levelled and the existing heading held. Roll and pitch channels can then be separately slaved to the various holds by switches on the control panel. For example, a standard turn can be demanded, but pitch will remain under CSS. Height or Mach hold can be engaged, leaving heading under CSS. By setting the inertial navigator and selecting PHI setting, the aircraft can be flown under heading-coupling to a PHI-defined, inertially located destination; and either height, Mach or CSS pitch holds can be used at the same time. Navigation information can also be derived from Tacan and PHI. Numerous interlocks avoid superimposition of conflicting modes and allow for failures. The autopilot disengages if the APC or any stabilizer axis becomes disengaged, or if the monitors detect a fault. It will also revert to full CSS if more than 71b stick force is applied to a coupled axis. Bank angle in the standard turn mode is adjusted between 15° and 48° according to airspeed. The AFCS reverts to CSS if pitch or bank exceed 66° during coupled flight. The autopilot will not engage if the gear or landing flaps are ex tended, but it can be functionally checked on the ground, with the engine running, by the pilot and crew chief performing and checking certain actions, such as deflecting the angle-of-attack vanes and undercarriage-door microswitches. An auto-trim system ensures that no out-of-trim forces occur on autopilot disengagement. Angle of attack is one of the basic outputs of the air-data system, and the F-104G is notorious for requiring artificial stall protection. When the stalling angle is approached the APC stick-shaker operates. If the pilot continues to demand a nose-up attitude, a thruster applies a force of about 401b to push the stick forward and dis engages any autopilot modes then in force. The APC is auto matically engaged at all times, except when gear and flaps are down, and will operate on the ground if the angle-of-attack sensing vanes are appropriately displaced. The MH-97G will maintain pitch and roll attitude to within one degree, height to within 50ft or 0.2 per cent of indicated pressure altitude, and Mach number to within 0.02. All these features are necessary for effective use of the F-104G, especially during high-speed operation close to the ground or where maximum range is required. In the last-named case, for example, it is con sidered necessary to use the Mach-hold during the climb. CSS allows the pilot to devote minimum attention to actual attitude
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
