Thank you, Manfred. I certainly very much look forward to giving it another try. In the meantime I cannot use a gauge which switches on the autopilot and immediately alters the aircraft's trim when I already have it trimmed for the approach. Please let us know when the new version of this remarkable gauge is released.
Shannon to Gander. Part Two. The Flight.
I am nervous about the side winds I know that I shall encounter on this flight, especially if the skies cloud over and it is impossible to use the sextant. In the past I have done flights between Iceland and the United Kingdom which have strayed seriously off course in these circumstances. However, as I have noted before these Constellations did carry Loran. Flight (18th July, 1946, p. 57) had the following to say about navigation in connection with the return BOAC proving flight from New York to London, with a photograph of the Loran unit, 'Over the American continent, navigation is strictly controlled by the use of Radio Range (and identification at each station). Across the ocean the radio compass and W/T fixes are used up to medium ranges, and Loran and astro-nav. are also used. Loran suffers from a dead area for about two hours flying about mid-ocean. On nearing the Irish coast, Radio Range again comes into use.'
I would be interested to know what exactly is meant by 'W/T fixes are used up to medium ranges'. Lines of position plotted from radio beacons, perhaps, or perhaps not? Lew Price, who navigated for MATS in the early 1960s, reports that Loran was subject to tube failure although in-flight repairs sometimes worked. He also says that 'The LORAN antennae was a long wire on top of the fuselage and would often take on ice. This was not usually considered a reason to change altitude; so, the navigator would be without this nav aid when icing occurred there'; and that LORAN 'could be accurate to within 50 feet when close enough to its broadcasting stations. It was better than nothing when farther from those stations.' This does not sound very encouraging and I am not entirely sure that he may not have been using a more advanced form of Loran than that available in 1947. Nevertheless, I decide that if things get really sticky I will simulate Loran by pressing Shift Z and reading my true position in FS9.
I am using a dynamic FSGRW real weather file starting at 22:48 on 2.11.2018. I have a weather report on this file, and as I said in Part One I also downloaded reports for 04:00 and 06:00 the following morning which I consulted at the appropriate UTC times in the flight. I have superimposed the later two reports on the first here, and it will be seen that the winds changed very little, at least according to FSGRW. All these readings are for FL 180.
The tanks are full with 28,140 lbs of fuel and with a half passenger load gross weight at start up is 83,664 lbs. Take-off, not surprisingly given that it is the runway we landed on a short time ago, is on Runway 18. I use the 1957 Traffic and an Air France Super Constellation precedes us there. In 1947 Shannon in the late evening must have been a hive of activity, with Constellations and other aircraft landing from various airports in Europe and filling their tanks ready for the crossing of the North Atlantic, while their passengers disembarked and ate in the terminal buildings. I taxi along the runway and then have some difficulty in doing the 180 degree turn necessary for take off, as the sort of differential engine power which presumably played a significant role in this sort of operation is not available with my throttle set up. However, eventually I am where I need to be and notice as my virtual crew go through the before take-off checks that I can see stars through the windscreen. Long may they last.
Take-off is straightforward. 'Gear up'. 'Gear up, captain', and then 'Gear up and locked. Landing lights off and retracted'. Stuart, my flight engineer, tells me that he is setting climb power' and I adjust the controls to produce 2200 RPM and 134 BMEP as we begin the turn to starboard. Climb to FL100 takes 13:50 minutes and to FL180, with blower shift at FL145, 29:15 minutes. Initial cruise at a gw of 82,000 lbs and OAT of -10C is at BHP 1169, RPM 2193, BMEP 126, FF 546, MIX 16, f/a 0.068, KTAS 239. I do not monitor the engine settings constantly, as a real flight engineer would have done, but periodically I open the Connie Team engine information gauge to check that all is well. At 03:04 UTC this leads to a significant adjustment at 75,000 lbs gw and an OAT of -16C; that is, to BHP 1200, BMEP 130, RPM 2193, FF 600, MIX 19, f/a 0.083, KTAS 252. This was not full throttle, because this would have taken BHP over 1200. When I fixed the settings originally a fuel to air ratio of 0.068 produced the highest BMEP, unless I made a mistake, which I don't think I did; in other words, if I set it richer or leaner BMEP fell. Yet later in the flight at the same altitude, but not the same OAT, a richer mixture produced a noticeable increase in power and a higher KTAS, the highest I have achieved in this aircraft at this height. There was a brief bout of propeller icing and wing icing shortly after this change. At a later stage I descended to FL80 to take advantage of more favourable winds. Cruise settings then (in Low Blower, of course) at a gw of 68,000 lbs and OAT of -10C were BHP 1198, RPM 2206, FF 528, MIX 25, f/a 0.064, KTAS 232. The fact that dropping 10,000 feet only cost me 20 knots in true airspeed came as something of a surprise, but I shall not forget it.
A most welcome feature of this flight was that apart from a brief period of blackout shortly after I descended to FL80 the stars were visible, if sometimes through patchy cloud, all the time. On the starboard beam The Plough and its attendant Pole Star shone all night. This was particularly valuable because a sextant shot of the Pole Star alone will allow one to fix latitude, which is the crucial thing on an East-West flight. Astern was a full moon and this allowed me to identify the stars around it without difficulty in the program Stellarium which was running on my separate laptop.
The night skies on the morning of 3rd November, 1947. The Plough and Pole Star, followed by the moon astern. These screenshots have been brightened in photo editing.
After take off I track out on the reciprocal of one of the Shannon NDBs, and it is immediately obvious that the southwesterlies are strong because in order to keep the needle at the correct point I need a heading of 262 magnetic, almost 20 degrees from the plan heading of 280. However, the net result of my course corrections is that they are too big and I end up south of the track for much of the flight, although not seriously so. The following images are of the flight plan and then the results of the sextant shots plotted on the Google Earth plan, with UTC times when I remembered to write them down (taking sextant shots is always exciting).
The weather station NDBs confirmed some of this. I picked up Ocean Station Juliette (FS9 range 112.5 nm) at 00:56 and the needle pointed slightly to the right rather than to the left, as it would have if I had been exactly on course. It was abeam at 01:35 and the signal ended at 02:10. Ocean Station Juliette registers at 00:56
Ocean Station Charlie (same range) was picked up at 04:02 and lost at 04:54. After that the sextant showed me to be pretty close to waypoints ATL 006 and ATL 007 but I drifted south again thereafter at the lower altitude of FL80, although this had almost no effect on my approach to Gander, where I had to do a downwind leg to land on Runway 04. I had put an NDB with a range of 150nm at Gander to simulate the radar system installed there in 1947 (see Part One) but did not pick it up until I was 120nm away according to my AILA gauge. If this is correct, it looks as though NDBs with a range of 150 nm do not work in FS9. Conditions at Gander were as clear as they had been throughout the flight and there was no need of G.C.A. for a landing or even the ILS possessed by this runway, although I did use it after employing the AILA gauge to line up on it. It was also the runway which at this date was exceptionally wide (Flight, 29th August, 1946), perhaps a wartime answer to the problem of landings in poor visibility. Approach was consequently straightforward and my landing so gentle that I didn't feel the wheels touch. This was at 05:50 Local Time after a flight time of 9 hours and 8 minutes and with 7,892 lbs of fuel still on board. We are early, but there is plenty of time for the passengers to stretch their legs and make use of the new restaurant facilities (Flight, same page).
Almost there - the landing. The runway end strobes helped too.
G-AHEN over the North Atlantic, 3rd November, 1947. Again, this screenshot has been digitally enhanced. Real flight engineers set the mixture partly according to the colour of the exhaust flames.
So, we have done the hard work and the final leg to La Guardia is next.