It may appear difficult to think back to a time when everything was not continuously connected to the internet or some sort of information or management system. The VC10 is definitively from this earlier age and in the late '60s trials were carried out that looked at two new developments: AUTONAV and data link. Both are now common on airliner flight decks (although the terms have changed since 1967) and owe some thanks to those early days of VC10 testing. This article provides some information about the trial fit on G-ASGH and the testing that was carried out.
What is AUTONAV?
When the VC10 entered service, navigation was a task for a dedicated crew member: the navigator, who would update the position of the aircraft based on position fixes from radio beacons, astro fixes using the periscopic sextant and/or other means at his disposal. The idea behind AUTONAV was that an automated system could take over this task, continuously updating the position of the aircraft and relieving the navigator of this full-time task. The initial systems used various radio beacons such as the LORAN stations and their hyperbolic position lines but the big step forward came when inertial navigation was perfected, combined with DME/DME position checks, which eventually evolved into the integrated navigation and flight management systems that have become a standard fit on modern flight decks. The VC10 trials were a midway point in development that aimed to evaluate a number of different techniques.
G-ASGH trial fit
Super VC10 G-ASGH was the airframe used for these trials and was modified with a specially designed junction box mounted in the ceiling of the equipment bay. Chris Mitchell still has the scars on his head from the days when he forgot that it was there. This junction box, referred to as the 'Stott box' after its designer John Stott, provided an easy way to interface with various systems in the VC10 without having to modify all the different wiring looms. In 1967 'GH was fitted with a Decca Omnitrack computer that was able to automatically use Decca signals, LORAN information, Doppler data and VOR/DME signals to compute the present position of the aircraft. The combined output of this computer could provide a lat/long position but also bearing, distance to go and other relevant information.
Another innovation, although not new, was the presentation of this position information on a moving map display on the overhead panel. This was fitted at the bottom of the overhead panel, providing a quick reference for the pilots without having to refer to a paper chart and hand-drawn position lines. A moving map display was not new, but the innovation was that the pilots could use several inputs and select which one suited their needs, enabling them to see their position at all times. Previous moving map displays, such as used on the Hawker Siddeley Trident fleet, were only useable within range of Decca stations, needed a complicated set up or used Doppler radar to show an estimated position based on drift and ground track. This new version on G-ASGH was designed to also include information from an Elliot E.5 Inertial platform but this had not been produced yet.
This was part of the Decca Company's drive to get their 'Dectra' system, short for 'Decca Track', accepted as a viable long-range navigation system over the North Atlantic. Additional transmitter sites in East Newfoundland and Scotland were equipped with high-power transmitters and larger antennas and would enable airliners to stay within a defined track over the North Atlantic sector of their flight. Development of Inertial Navigation Systems (INS) would provide the same functionality (and more) without the need for these ground stations and because of this, Dectra never really took off.
The overview shows the various systems that were interconnected in this trial fit, enabling various options for the pilots:
In the early days of the VC10's career the only communication between aircraft and ground was by VHF or HF radio, transmitting and receiving voice messages. In Birdlip, Gloucestershire, there was an HF communication station that was used to test the first version of a datalink setup. This would enable other information, such as position information from the aircraft, to be automatically transmitted and received without a human in the loop. Several BOAC aircraft had been fitted with Decca Mk.5 test boxes that would respond to a request from the ground station and transmit a fixed format test message in return. The test setup in G-ASGH took this one step further as this was the first instance of real aircraft position information being sent from an aircraft to a ground station. The previous messages did not contain any useful information. The Decca Omnitrack Mk.2 computer was connected to the datalink test box and the next time G-ASGH passed over the Birdlip station, position information would be sent out in response to the ground station interrogation message. The transmission used the VHF band and the message contained latitude, longitude, altitude and airframe identity at a data rate of 600 bits/second. In comparison, modern 5G data speeds typically start around 150 Mbits/second with a possible maximum of 1 Gbits/seconds. For the duration of the Decca Omnitrack testing on 'GH this datalink test ran simultaneously.
At Birdlip, the messages were printed out and compared to check for errors. This showed that the data quality was good for as long as the aircraft was above what is known as the radio horizon, or the maximum range at which VHF communication is possible. With a VC10 at 30,000 feet this will be just over 200 nautical miles. Tests over HF radio were continued using the fixed format messages (after the Decca navigation equipment was removed from G-ASGH and replaced with another test setup from Ferranti) and after some initial troubleshooting acceptable results were obtained at ranges up to 800 nautical miles. The main problem was receiving the data messages from the aircraft as transmitting power was limited in this direction.
Chris Mitchell recalls installing the test setup on G-ASGH as well as some of the early results:
The unit I installed was connected to the Intercom system so that it could be connected to either the VHF or HF transceivers. It was meant to respond to interrogations from a base station at Birdlip which could transmit on VHF and HF SSB and I believe DSB.
I flew on several trials which were done on commercial services to the USA and it occasionally seemed to work on VHF but I never got any luck on HF and the trials were abandoned. In my time as the Development Engineer responsible for VC10 Comms and Nav we did not install Decca Omnitrac or Dectra and long range navigation was carried out using the Doppler Radar and Loran A. Although later we eventually installed a dual INS (Delco Carousel) but if these were linked to DataLink it was after I had moved on to be the Assistant Engineering Project Manager for B707s. Had such trials taken place I would probably have got to know about them.
The downside of the concept was that it occupied one of the main communication channels and this inhibited the trials as the Pilots would insist on talking to ATC etc. The limited bandwidth and hence data rates was also a limiting factor. I can't complain however as I did get a few 'jollies' to the states, but most of my time in the USA on these trips was spent re-positioning up and down the east coast to pick up Golf Hotel on its next trip which invariably was to Philadelphia or Washington when the outbound trip was to New York. But the expenses came in handy as I was recently married ie. broke.
Capt. Brian Calvert, sadly no longer with us, was the Flight manager Technical and a very good one too and we worked together on this and on several other projects. AIRINC held a conference in London on Digital Data which we both attended, but Brian tended to switch off when it got technical rather than operational, and insisted on playing Battleships with me. During this conference we were graced by a flying visit from an American guru on the subject who lectured us all on Quadraphase Phase Shift Modulation where several digital and audio channels could be used on one 50Khz channel. When he got to the Fourier Analysis and second order Bessel Functions of the third type even I was glazing over. The Chairmen, a Pan Am skipper called Bob Bohannen, an old mate of Brians, saw what he was up to and when the Professor finished he immediately called on Brian for his views on the operational significance. Brian just said "It's all done by bloody magic if you ask me. Lets have lunch".
The datalink testing was a sideline alongside the Autonav testing and from the pilots' point of view, it was the navigation part that was of interest. Datalink testing would continue on the 747 fleet as the INS setup on that type had finished flight testing. One of the 747s had been modified with a larger capacity computer for its primary INS system, which could be combined with the datalink equipment from G-ASGH. The plan was to have a complete data-communication ground station ready for evaluation by mid-1971 and expand the system to two-way communciation at a later stage, hoping to include maintenance data and ATC purposes as well.
Further Autonav testing and implementation
The second autonav option that was evaluated was a system by Ferranti, based on the Argus 400 digital computer. The Argus family of computers developed by Ferranti were process control computers, used in many applications such as production lines and nuclear powerplant control systems. They were also used for the Bloodhound missile system and the basic computer formed the core of this Autonav system. The computer used various navigation sensors, processed the outputs and presented position and guidance to the pilots. The schematics show an INS system as one of the sensors but, although installed on G-ASGH, it was not connected to the Ferranti system during the trial. The Ferranti system enabled the pilots to input up to fourteen waypoints using lat/long coordinates and it would use the various inputs from radio navigation, Doppler and an automatic air plot to track the aircraft's progress and figure out guidance to any programmed waypoint. Without INS input, the precision of the system was only as good as the input sensors but instead of the traditional method of flying from radio beacon to radio beacon, any position could now be used as a waypoint and the system would sequence from one to the other automatically. All functions that current pilots would recognise, but with a somewhat more cumbersome interface. The datalink was apparently still part of the test, as page 40 explains that position reports could be transmitted over the no.2 VHF transmitter using 134.8 MHz or using the no.2 HF transmitter using 4676 KHz or 8837 KHz when further out over the Atlantic Ocean.
In the end the Delco Carousel IV was selected for a number of reasons, one being that it was already in use on the 747 fleet. Chris Mitchell remembers this time:
So in my time in Development long range navigation was by Doppler Computer with position fixes from Loran A. We did a trial of Loran C, which was fun as the only North Atlantic chart was from the US Air Force. This had an area slightly larger than the entire floor area of the rather spacious flight deck. There must be a large navigator's office in a B-52.
Although Loran C gave excellent fixes there was no suitable equipment available at the time to turn Hyperbolic Intersections into lat and long (I don't know what had happened to Omnitrac) and would still have required the carriage of a Navigator, unlike the Inertial Navigator (AC Delco) that we eventually settled on. This was already fitted to the B747s and so there were lots of logistical advantages from using the same equipment.
There was obviously lots of internal politics going on at the time, of which I as a mere development engineer was blissfully unaware. But I was given the job of managing the installation of INS at the last minute after all the initial design work had been done. Being dropped into it at the last minute shows how far down the pecking order I was. It was a bit of a nightmare, especially when I insisted in changing the gearing between the computer and the autopilot. My instrument colleagues did not believe that a Radio man could understand servomechanisms, even thought my previous career was as a research engineer working entirely on control systems not Radio.
We did the first installation on a normal Major Check, with several wrong connections that I hadn't spotted when I was checking the installation drawings and found the rest! There was no special test flight and the aircraft went straight out of the hangar into service, where it was rescheduled at the last minute from JFK (convenient) to Australia via Beirut. Anyway the aircraft departed with a standard crew plus two techies, me and a colleague from the 'Instruments' department. Strict instructions were issued that, under no circumstances was the autopilot to be connected to the INS until my initial check had been completed and I was replaced in the jump seat by my Instruments colleague and we were on an overwater sector. Unfortunately, the first VOR waypoint in France was u/s (big surprise) and ATC wanted us to re-route, so we keyed in the VOR Lat and Long as an additional waypoint. The Captain agreed just to use the steering guidance but after ten minutes we thought 'what the hell' and coupled the autopilot. All the other waypoints (VORs) were added and we flew on INS guidance until we reached the Med. Where the autopilot was solemnly switched to heading mode and I was replaced on the flight deck for it to be 'officially' switched on by my counterpart. I retired to the First Class just in time for a glass of champagne followed by Real Turtle Soup, Lobster and fillet steak (white Burgundy and a passable Bordeaux I seem to recall). I jumped ship in a pre-disaster Beirut, a lovely city which disappeared in blood and dust shortly afterwards.
But I enjoyed a couple of days stopover before picking up the return flight. The INS worked very well on the VC10 with low deviations at the waypoints even with large course changes, I thick it was better than the B747 in that respect, but that was probably due a lot more to the aircraft's performance than my small contributions to the program.
For a longer version of Chris Mitchell's experiences, click on Radio Development on the VC10. Once INS was in use on the early 747s, some limitations of the system became apparent. The Delco Carousel system was not able to use "Turrets" with pre-programmed locations, you needed to type in the full latitude and longitude coordinates and there was room for only nine waypoints in the system. Because of this, once you were past waypoint number eight, you should really have started adding the coordinates for the next couple of points on the flight plan. Otherwise, after passing the ninth waypoint, the system would automatically cycle back to the first waypoint in the list and start navigating back to the start of your route. Another possible outcome would be for the system to send you to a point south-west of Lagos, Nigeria, as that is where latitude zero, longitude zero is located. Several experienced pilots on the widebody fleet would jokingly refer to that location as the '747 graveyard'.
The photos above show the final configuration on both the BOAC Super VC10s and the RAF VC10s, who would continue to use the 1970s Delco INS together with the Rockwell-Collins FMS-800 up to the final flights of the type in 2013. Some of the INS systems on the Super VC10s ended up being reused for the 1982 Black Buck raids, click here for more on that story.
Sources: CATC Electronics News, February 1971, BOAC Horizon, May/June 1967, VC10 AutoNav II system by Ferranti, BOAC, with thanks to Chris Mitchell, Colin McKeeman, John McCrickard.