Circling approaches. I agree with you and I'm amazed at the very low minima for circling specified in the approach charts in that era. However, with gear down and a lot of flap and so low speed it is manageable in FS. Just give yourself sufficient distance on finals, say 2 miles/600 feet. For practice try Hong Kong or the Idlewild Canarsie approach to 13L/R
Thank you to you both on circling approaches, which I had not heard of before. In terms of this particular landing on R18 at Shannon I approached from the north and this meant passing R22 and its associated NDB. When the NDB needle pointed to 220 this would have given me a position and the runway itself might have been visible. I suspect that a real navigator would then have used ded reckoning to calculate the time for the turn to R18, which might also have been visible. However, my simulated flight was at night and FS9 at night is not the real world at night. My installation of FS9 has some rather bright and apparently random lights in the vicinity of Shannon which have nothing to do with the airport and on one occasion I had difficulty picking out the runway concerned even though I was coming in on a navaid. If you look at the screenshot with the Jahn GCA gauge visible you can see the runway in the distance, but it can only be distinguished from other lights because the gauge shows it for what it is. Without that it would seem to be almost anybody's guess. Approach strobes make a considerable difference to FS9 landings at night, but I doubt if Shannon had anything like that at this date.
Post by connieguy on Sept 22, 2019 12:10:26 GMT -5
I think it would have been pretty difficult at night, Tom, but on the other hand it is quite true that I have never tried doing such a thing. One would have needed to be looking to the left (in this case) pretty frequently as well as forward and perhaps something like Track IR would make that easier in the virtual cockpit than using the joystick hat switch.
Last Edit: Sept 22, 2019 12:13:57 GMT -5 by connieguy
In addition to all of the above information, I think there is one thing here that FS9 does not replicate, but would be essential in many real life situations. If rwy18 is an option with regard to the weather, the much longer (and wider) rwy22 with instrument approach is almost in every case a better option, and real life ATC would assign or approve that for landing. I can think of only two reasons when it would not be: too much crosswind on 22, or it being obstructged by maintenance or a stranded aircraft. In the first case, rwy13 might in turn be a better option than 18. But FS9 does not know these things, it faithfully sticks to its simple rules and assigns 18.
You might even be able to land on 22 in this situation without being reprimanded, after obtaining landing clearance (for 18) by passing through final 18, close enough for Tower to think you are complying.
Post by connieguy on Sept 23, 2019 10:15:21 GMT -5
I don't usually take FS9 ATC too seriously and am quite capable of landing on a runway other than that assigned, and even of landing on a runway already occupied by another aircraft after ignoring an instruction to go around. I believe that programs like Radar Contact allow you to choose the runway, but this doesn't matter to me so much that I want to install it. On this flight the wind was not strong and I agree with Erik that R22 certainly would have been a better option. However, all that aside there is a question to which I would like very much to know the answer, and I suspect that somebody who actually flew passenger aircraft in that era is more likely than anybody else to know that answer. Did they if necessary land at night on runways which had no navaids? If they did then it may be one of those things which was feasible and safe in real life but which cannot easily be replicated in FS9.
Post by mrcapitalism on Sept 24, 2019 10:52:04 GMT -5
I by no means meant to tell you how to conduct your simulator flying.. I just wanted to point out an option that you might not have been aware of/realized, but looked appropriate for your description and screenshot of the approach into Shannon.
Since I've really stirred things up with the suggestion, I thought I would quote the Propliner Tutorial.
INSTRUMENT RUNWAY versus LANDING RUNWAY.
Now take a closer look at the 2D overhead view of the procedure.
The Instrument Runway at KIZG is always RWY 32 even when the landing runway is RWY 14. Get the difference between the instrument runway and the landing runway clear in your mind before going further. Of course the instrument approach runway is sometimes the landing runway, but that is by no means automatic. Since this is an NDB arrival to an entire airfield the only way to identify the instrument runway is to study the 2D overhead diagram. The instrument approach procedure points at RWY 32 so RWY 32 is the instrument runway.
The approach course is inline with RWY 32 and it would be easy to suppose that this approach is one in which we descend straight ahead to land on RWY 32 just like a modern radar vectored approach to an ILS.
Like many others this arrival and approach procedure does not terminate at a runway. It terminates at an entire airfield. It is the means by which airliners locate the entire airfield before joining one of the visual circuits to land. If this were an instrument approach to a runway the plate would show the runway landing length (Rwy ldg) and touch down zone elevation (TDZE) of RWY 32, not just the general airfield elevation in the top left boxes.
Some airfields have no published approach procedures. Some airfields have published approach procedures, but they do not point at any particular runway, in which case no runway is defined as an instrument runway; in which case none of the runways may have any lights. If an airfield has one or more runways with approach lights, the one with the most visible lights will normally be chosen to be the instrument runway and usually, but not always, the published approach procedure will point us at its lights.
When flying propliners in we will normally be flying from and to airfields that have published arrival procedures. They will usually have at least one instrument runway and they may have many, but never assume that the instrument runway is the landing runway.
The approach procedure is there to point us at the brightest approach lights on the airfield, from a safe position, at a safe starting altitude and most importantly a known orientation; known to us and to everyone else. It is the only way to stay safe until either we can see all relevant obstructions and the airfield, or we know we are 'inside' the obstructions and on the extended centreline of the instrument runway. We will not 'blunder' into the active side of the visual circuit. We will not be confused about which runway is which or where left hand downwind for RWY 14 begins. The more runways an airfield has the more important it is to arrive with known orientation so we know exactly which runway the nose is pointing at when we first tally the airfield. Then we can 'count round' the runways and determine which one is the landing runway and figure out where downwind left hand begins.
I noticed that Shannon has an ILS for runway 22. An option you could use involving the existing navigation aids, and only perhaps monitoring from a GCA, would be to conduct the ILS 22 Approach, and then upon contact with the airport (while still above Circuit Height MDA), turning right to join a left base for runway 18. If traffic was a little busier, maybe making the same right turn, never descending, and joining an upwind for runway 18, or turning left to reverse for a downwind entry. This is just another tool at your disposal for future flights. Enjoy!
EDIT: I'd also like to clarify that the procedure quoted above is the "NDB or GPS-B" for KIZG. While this approach is aligned with Runway 32, it only publishes Circling minimums and has no runway designation in the approach title (no "straight-in" minimums given) so legally this approach is not designed to be flown straight in. Straight in approaches will very frequently (always?) publish circling minimums.. which allow pilots the option of safely maneuvering to land on another runway...i.e the 'landing runway'. (other than straight-in).
Many thanks for this reference to a very relevant section of the propliner tutorial, which fairly clearly I had not read. FS Aviator is doubtless most knowledgeable and classic flight simulation owes him a great deal, but I have always found the manner in which he addresses his audience not entirely to my taste and consequently have tended to limit my exposure to it. However, I have recently been doing a little more research in the Flight archives and this has turned up one or two things of interest in terms of issues raised in this thread. On cruising heights, for example, and although it was not a transatlantic flight, it seems worth noting that a Constellation which lost an airscrew on a flight between New York and Miami in February 1948 was cruising at 22,000 feet (Flight, 19th Feb, 1948, p. 200). On navaids at Shannon, Flight reported on 21st August, 1947 (p. 199) that G.C.A. was going to be installed and operated at Shannon by American Overseas Airlines 'on a scheme similar to' that already being operated at Gander, and on 19th February, 1948 (p. 199) that International Aeradio Ltd were by then operating G.C.A at Shannon. As it turns out, then, my use of the G.C.A gauge at Shannon in November 1947 was probably if fortuitously legitimate. Runway 22 (then called 23) had ILS by April 1948, when a Pan Am Constellation (CN 2058) crashed when trying to land on it; striking the ground more than 2,000 feet short of the runway, there was only one survivor. A TWA Constellation had also crashed at Shannon in December 1946 as a result of a faulty altimeter when trying to land at night. Visibilty was one mile and there was 10/10 cloud clover at 400 feet and 4/10 at 250 feet. ' While turning to the left for final approach to runway 14 [13 on my chart], the aircraft passed behind a low hill which blocked the airport lights from the pilot's vision'. Assuming that there was no navaid on 14 this shows that aircraft did do circling approaches at night in demanding meteorological conditions, though the requirement that the airport should be visible to the pilot was clearly being observed when the faulty altimeter came into play (information from the link below).
Prestwick had G.C.A by October 1948 (and quite likely considerably earlier) when a KLM Constellation which had used it opted to do a visual circuit as well (G.C.A apparently not having delivered a perfect result) and crashed into power cables with the loss of all on board (Flight, 28th October, 1948, pp. 504-5). There is also information on how aircraft received weather reports, but I will save that for the account of my flight from Shannon to Gander, which ought to happen in the next few days.
As the flight from Shannon to Gander in early November is likely to prove the most challenging I have ever done my account of it will be divided into two parts - firstly the planning and secondly the actual flight. Here is Part One.
This flight was scheduled to leave Shannon at 00:15 (UTC+1) on 3rd November, 1947 and arrive at Gander at 07:15 (UTC-2:30), a flight time of 10:30 for a distance of 1721 nm. I shall carry a half passenger load of 3,424 lbs including crew and full tanks at start up. These aircraft did carry cargo but I am not doing so on this occasion. It will be done with a dynamic FSGRW weather file for 2nd November 2018 starting at 22:48. Weather, and especially the winds, will be crucial. It was possible for the weather to be so bad that flights were cancelled. This happened to the BOAC proving flight scheduled for 21st June, 1946, which was cancelled because of bad weather at Shannon. 'It would be quite within the Constellation's capacity to over-fly Rineanna, but this was not in any case desirable on a proving flight, and there is always a possibility in the event of unusually strong headwinds of running uncomfortably close to the fuel safety mark on East to West crossings.' (Flight, July 11th 1946, p. 31). These turned out to be prophetic words, not least because they assumed competent flight planning. On 13th October 1947 the flying boat Bermuda Sky Queen left Foynes for Gander but when still 600 miles from its destination sent out an SOS saying that because of a severe headwind it had only 2.75 hours of fuel left and was therefore returning to land near a weather ship. Subsequently it landed in a rough sea three miles west of the weather ship Bibb [which I assume was Ocean Station Charlie] and then taxied for two miles to be nearer the ship. Many hours later, after rough seas, gale force winds and what must have been an extremely frightening experience, all the passengers and crew were safely aboard the Bibb (Flight, 23rd October 1947, pp. 479-80, and there is film material on You Tube). Although this might seem to suggest that the weather had not been forecast correctly the report by the Civil Aeronautics Board rejected this as the prime explanation of what had occurred and concluded that errors by the crew were the most important factors responsible:
That the importance of accurate forecasts was fully appreciated is shown by the fact that the year before, on 14th March 1946, Flight (p.268) had reported that a North Atlantic Route Service Conference in Dublin intended that 'A grid system of meteorological stations, incorporating stationary ships, merchant shipping, data from aircraft on the routes and from special meteorological flights and land stations, is to be adopted and standard forms for flight forecasts established.'
It is not possible to replicate in FS9 the detailed meteorological planning before flights that all this would have made feasible, but I go some way towards it in taking from FSGRW not only the weather reports for the various waypoints for my time of take-off but also those for four hours later and then six hours later. I do not look at the later ones, but I save them as pdfs in my weather report folder so that I can consult them in the cockpit four hours and six hours into the flight. This should give me some idea of weather changes since take off.
As in the past I have used Plan G to create waypoints at intervals on the flight plan, on this occasion with a frequency rather greater than previously so that I shall get more detailed weather reports from FSGRW. The plan below also shows that in navigational terms I ought to get considerable assistance from the NDBs at the weather ships Ocean Station Juliette in the first third of the flight and at Ocean Station Charlie later on.
The basic weather situation is that as I take off there is quite a deep depression to the NW of the British Isles. 24 hours later (the Metcheck historic pressure charts only deal with 00:00 UTC) this has not moved far but it has filled slightly, although another deep depression has formed just north of Newfoundland. These conditions seem to be reflected in the data with which I am supplied by FSGRW, as it is evident from the weather report that I am going to be facing some winds of over 80 knots, which raises immediate issues about the best cruising height. In examining them I am unable to use my manual E6B flight computer to work out groundspeeds and course adjustments to allow for drift because it does not seem possible for this to calculate the effects of winds stronger than 60 knots. Although there is an E-6B app for Android, there is also an online electronic version here:
To facilitate the calculations I get the flight planner of FSGRW to identify the winds at individual waypoints at two heights - FL8 and FL180 (the latter second):
What is rather unusual in my experience is that in the first part of the flight the winds are very strong at both flight levels, rather than significantly lighter lower down, and therefore it is far from obvious that I will be better off at FL8. Hence, I use a piece of graph paper to tabulate the results from the flight computer in two columns for the two heights based on an assumed TAS of 230 knots, although at Maximum Cruise, which is what I intend using in these conditions, the TAS at FL180 ought to be at least 10 knots higher than that at FL8 and probably more. It is clear that between waypoints ATL004 and ATL 005 the wind is going to change from SW to NW. Prior to that there is little to choose in terms of groundspeed between the two altitudes, although at FL180 I shall need to offset the course rather more because of the effect of the wind, and at ATL667 by as much as 20 degrees. At ATL005 I should be travelling 30 knots quicker at FL180, at ATL006 20 knots quicker and at ATL007 at about the same speed. However, I shall then need to descend to FL8 because at ATL008 the winds are far stronger at altitude and I should be doing 200 knots at FL8 compared with 145 knots at FL180; similarly with ATL009. Working all this out takes a little time but it does give a clear result. I am a little concerned about the extent of the corrections which some of these winds will demand in terms of course heading, but as I noted above the two ocean stations should come in extremely useful here, as well as any star shots that I am able to take with the sextant. In the latter regard, the skies may be clearer at FL180 than 10,000 feet lower.
It remains to say something about Gander. 'Flight' for September 25th 1947 reported that:
JAMES H. SMITH, Atlantic Vice-President for Pan American Airways, has announced the appointment of Mr. P. R. Dickinson as liaison officer for the new all-weather radar landing system being installed at Gander Airport, Newfoundland. Mr. Dickinson will co-ordinate the operation of the unit which Pan American have been operating for all major American and foreign transatlantic airlines since last December, and operate in addition, a new unit which is being obtained from the American Army Air Forces. The new device, a search radar unit with a range of about 150 miles, plus the present GCA system, should eliminate the necessity for ''stacking'' incoming aircraft ofter the airport in bad weather. The search radar unit will permit air traffic control to locate aircraft approximately 150 miles away and space their arrival time to allow uninterrupted landings. The unit is expected to be in operation by about November 1st. The Newfoundland Government is aiding in the installation cost and the Air Transport Association is sending engineers to Gander to train the GCA operators in the use of the search device. Airlines using the present GCA system which are expected to share in the use of the search unit include Air France, American Overseas Airlines, B.O.A.C., K.L.M., Sabena, Scandinavian Airlines System and T.W.A.
In order to simulate the installation of this system just prior to the date of our flight I have created an NDB wih a range of 150 miles at Gander (it will be interesting to see whether an FS9 NDB will work at this range) and clearly the use of the Jahn G.C.A. gauge will be allowable. I shall not be using the greater range of the Gander VOR, upon whose existence at this date I have no information.
Last Edit: Oct 1, 2019 10:36:19 GMT -5 by connieguy
Speaking of GCA I am in the process of tweaking the PAR code a little so as to generate a more rhythmical flow of alternating course/glideslope commands, following a request by a RL calibrator pilot. Yes, they are still doing PAR's. PAR now sounds and feels better, but will also check if the course algorithm can be improved.
Many thanks indeed, Manfred. Would there be any possibility of doing a version of the gauge which does not switch the autopilot on automatically, or of offering an option as to whether it is switched on automatically? Best wishes, Ken
Hi Ken, I assume you are referring to the Approach Pattern Control (APC) module, which directly posts heading and course data to the AP, whether on or off. If you turn the AP off then you have to handle all of it on your own, which is tough. I assume you still want to use the AP, but without interference from APC, manually entering the data you are verbally given. In the PAR module we have a 'green' mode (via middle-click on "PAR") that allows something like this, with AP on but waiting for the pilot to enter any data. Would it help if we had a similar mode for APC? Should be possible...