The crash of a 41st Transport Squadron C-121C in 1956

[connieguy]

Browsing through the individual aircraft histories in Peter Marson's book on the Lockheed Constellation tends to bring home fairly forcibly the considerably greater risks of being involved in an aviation accident in the 1950s compared with such risks as there are today. These ranged from minor collisions which resulted in minor damage to total loss of the aircraft, and one of the latter was the demise of Construction Number 4184, a C-121C of the 41st Transport Squadron, which crashed at Dhahran, Saudi Arabia, at about midnight on 30/31 December 1956. Marson, who has had access to USAF records, says that it was on a regular MATS flight from McGuire AFB to Tripoli (Libya) and Dhahran, that the event took place in thick fog when the airport's GCA system was not working, and that 15 out of the 41 on board were killed. This entry in Marson's book first attracted my attention some time ago because the entire schedule seemed suitable for replication. This was partly why a little while ago I encouraged Al Von Pingel to create period scenery for Dhahran, and since then he has also released his scenery for McGuire, to add to that for Charleston which was the base of the 41st and was the original departure point for the aircraft and crew.
       I also recently thought that I would like a little more detail on the entire sequence of events, and so began looking around for what else I could find. This has not been fruitless, but it has also been confusing. There is to begin with, the information supplied by the Bureau of Aircraft Accidents here:

www.baaa-acro.com/crash/crash-lockheed-c-121c-super-constellation-dhahran-12-killed

This gives the schedule as Charleston - Hamilton - Horta (Azores) - Lisbon - Tripoli and Dhahran. It says that the crew numbered 7 and that all the 12 fatalities were suffered by the 31 passengers, while 26 others were injured. These numbers do not coincide with Marson's and nor does the route, though Hamilton looks a fairly clear error for McGuire. This is the only source known to me which says there was a stop in Lisbon. It also says that the aircraft was carrying high-ranking officers from the U.S., Pakistan and Iranian Army and that GCA was inoperative in a northwest wind of 5 knots in a visibility just before the crash of 3 kilometres.

There is also the material given by the Aviation Safety Network here:

aviation-safety.net/database/record.php?id=19561230-0

This concurs with the Bureau of Aircraft Accidents in saying that there were 12 fatalities out of 38 on board. It also says that the flight began at Charleston, landed at McGuire for an overnight stop and then called at Lajes (AFB) in the Azores and Tripoli before flying on to Dhahran. It adds that two more Super Constellations also flew on to Dhahran from Tripoli and landed at Bahrain after the accident to their fellow. It also says that 'reports' indicate that GCA was inoperative.

To muddy the waters even further there is additional material published in a Wikipedia article which draws on reports in the Charleston Evening Post on Monday 31st December 1956 and Tuesday 1st January 1957. It is available here:

en.wikipedia.org/wiki/List_of_accidents_and_incidents_involving_military_aircraft_(1955%E2%80%931959)

According to this account, the C-121 was carrying UN troops into the Suez Canal zone and was also intended to transport Hungarian refugees back to Charleston. Why the UN troops were being taken to Dhahran is not clear. 12 out of 38 on board were killed and a USAF source at Wiesbaden in Germany said that a manifest showed that there were 38 people aboard, 11 crew and 27 passengers. Moreover - and this is where a source published in Charleston seems likely to be accurate - four of the crew were killed and seven were listed as among the survivors. The dead comprised the aircraft commander and first pilot and two flight engineers, the names of all four being given. The other seven crew members survived and included three flight attendants, one of them female. It also mentions Lieutenant Colonel Ali A. Raft, a transportation observer of MATS operations, from Iran. There might be an echo here of the high ranking officers mentioned by the Bureau of Aircraft Accidents, and if such officers did comprise the passengers that might explain why three flight attendants were carried. This source too says that two further aircraft also flew from Tripoli (having possibly started their schedules from somewhere else?).
Finally, the Charleston Evening Post lists some of the injured as 'critical'. If some of them subsequently died that might explain why Marson's figure for the dead is 15 rather than 12.
          Any thoughts on all this would be welcome, and the recreation of the flights, starting in Charleston, will follow.

[chris_c]

Here's a video of the crash of Trans Canada Airlines L-1094E CF-TGG during an attempted landing at Toronto, Malton on December 17th 1954. All aboard survived.

TCA's 1954 "Christmas Miracle"

The presenter is (apparently) a RyanAir training captain and he has a number of air accident videos in case you are unfamiliar with his YouTube channel "Mentour Pilot". Mostly modern stuff there though. His knowledge of Canadian geography is off a wee bit though, since the destination was not actually Brampton then a sleepy small town, which just happened to be the location of CYYZ, then Malton Airport serving Toronto.

Chris

[Tom/CalClassic]

Hi Ken,

I have run into this kind of thing very often doing research for various projects. The information is often contradictory, and confuses the issue. Basically you just have to decide which sources you deem more reliable and use those, ignoring information that conflicts with those sources. It's often the best you can do...

Nice video, I've seen a few of his before but not this one.

[Al]

Looking forward to this one Ken since it involves several of our sceneries. Do you have a decent LPLA installed for the trip? We never got around to doing that one. I had to install a more modern version that I began back dating as I have so much ai traffic passing through. Good luck on the bad weather and no GCA at your final destination.

[connieguy]

Tom,
I have often found that establishing the truth about things can be extremely challenging, and it can be very interesting when you see media reports of things you know about independently, which inclines you not to believe anything too readily. One might have thought this would not extend to air accident reports, but clearly not!

Al,
There is a scenery for Lajes by John Stinstrom which is rather good but of course not period, and I guess that is the one you mean. There is also a decent looking freeware landclass for the whole of the Azores on Avsim.  I shall use real world weather for the appropriate dates in 2022 and already know that it will blow me across the Atlantic. The fact that they were attempting to land at midnight may not have helped, but I suspect that is because they wanted to use astro navigation when crossing a desert bereft of radio aids.

[connieguy]

According to the public information officer at Charleston the aircraft left there with a crew of 12 on 27th December 1956; by the time of the crash it had a crew of 11. This comprised the aircraft commander, Major C.W. Ellis, a first pilot and a co-pilot, two flight engineers, three navigators and three flight attendants. The flight from McGuire to Dhahran was a regular MATS service via Lajes and Tripoli, the former calling point of Port Lyautey in Morocco having been dropped by late 1956.
For our flight I am using dynamic real world weather files for the equivalent dates in 2022, and they are not without interest. Ventusky shows that on 27th December a weather system east of Newfoundland was generating westerly winds of 90 mph at FL180 which extended almost to the Azores. Moreover, these winds held for another two days. The flight from Charleston to McGuire on the 27th will benefit from strong south-westerlies at altitude, taking off about 14:00 Local Time (UTC -5).
Just before start up at Charleston. Scenery by Al Von Pingel and myself, available from CalClassic. This is USAF 54-156 (Lockheed Construction Number 4175), delivered to the 41st Transport Squadron in November 1955, a month before 54-165, lost at Dharhran. It survives today as the Breitling Super Constellation.





Awaiting take off at the threshold of KCHS Runway 15



The flight to McGuire is about 500nm. I track out on the Charleston Range Station and will track in on that at McGuire. The main runway there did not have ILS in 1956 but one of the arms of the Range Station to the south was aligned exactly upon it. I cruise in low blower at FL150 pressure altitude at BMEP 160 and 1700 BHP. This is a part of the world with a large number of radio aids and the offsets necessary to keep the rmi needles in the right place are an indication of the strength of the wind behind us and to port. We pick up McGuire 100nm away and begin the descent immediately.
I have Plan G open on my second monitor and follow the flight from its map and what I can see of the landscape from the cockpit on a day when visibility is good. Some of the places have names familiar from what I know of the Civil War - Richmond, Fredericksburg, Wilmington. Here we are over Delaware Bay with Philadelphia beyond us to starboard, and beyond that McGuire AFB.



ATIS tells me that it will be virtually a straight-in landing on Runway 06. Winds are calm in rain, but there is a surprise in store in the final moments. My ice gauge lights up to tell me that wing ice is forming and although switching on the de-icing will often clear this immediately it does not on this occasion. However, the aircraft is handling normally - carburettor icing would have been a different matter - and we carry on with the landing.



Touchdown. Flight Time 2:01. Pretty quick.



Parked near the close of the short winter's day. McGuire scenery by Al Von Pingel and Dan French, available at Cal Classic.

[Tom/CalClassic]

Nice trip. Who’s textures are you using? I didn’t see that info, sorry if I missed it.

[connieguy]

Tom,
The runway and taxiway textures are Airport Environment Upgrade 7, released in 2005 and available on Flyaway Simulation. They are the best I have come across. The rest are Flight 1's Ground Environment Professional,
Ken

[Al]

Excellent recount of the first leg with great pictures. This will be a great story to follow with some excellent scenery along the way, if I do say so myself. There will also be a new one so stay tuned.

[Tom/CalClassic]

And the plane's livery textures? Thanks.

[connieguy]

One of five C-121 paints available here:
www.flightsim.com/files/file/205650-fs2004-usaf-mats-lockheed-c-121/
However, the rather beautiful bare metal reflective shine needs the env.map file by Tom Knudsen recommended on here some time ago by Dave McQueen. It will do wonders for other aircraft too.

[Al]

Tom, all these liveries painted by Ken (along with some others) come as an ai setup as well in the Charleston scenery in the HiDef version of the traffic. I had created a separate aircraft folder with separate entries for both passenger and cargo versions. This aircraft folder is used by several of our later sceneries.

[Tom/CalClassic]

Thanks.

[connieguy]

I decide that the flight from McGuire to Lajes will leave at 9:00 Local Time (14:00 UTC). Parked at McGuire before start up. Scenery by Al Von Pingel and Dan French, available at Cal Classic.





Ready to go at Runway 24 at exactly 9:00.



Climbing out and heading east over the coastline. At the moment we are flying on a reciprocal from the McGuire Range.



But once the signal from the Range is lost navigation will become a lot less easy.



Here is the flight plan and the FSGRW dynamic weather file will give me data for all the waypoints upon it. I also have data not only for 14:00 UTC but for 19:00 and 22:00 too; in the light of what happened I should probably also have had it for 17:00. I look at this later data when the appropriate time arrives. I have loaded 29,673 lbs of fuel, which previous experience has suggested is correct for the distance of approximately 2,190 nm. Climb to a pressure altitude of FL180 takes 30 minutes and after that I decide to use power settings for long range cruise because it seems likely that the crew in 1956 did that. As the fuel burns off and gross weight drops power is slowly reduced in such a way that true air speed and trim remain more or less constant. The accuracy of the Luis Pallas flight model for this Manfred Jahn/Connie Team aircraft is little short of breathtaking. Here we are at an early stage at a gross weight of 120,885 lbs. The tables in the real world USAF manual for the C-121C are as follows with the readings on the model's engine information gauge in brackets. Density altitude FL180 (OAT temperature is -20C and pressure and density altitudes are almost the same).

BHP 1706 (1702/1696) RPM 2340 (2342) BMEP 172 (172) FuelFlow 705 (705/703).



Once established in cruise it is also necessary to pay attention to the fuel tanks. Tanks 1-4 were all full at start up, though 1 and 4 have approximately twice the capacity of 2 and 3. Tank 5, the centre tank, was 35% full, while the two wing tanks were empty. It is good practice to empty 5 as quickly as possible, so its fuel pump is now switched on and the appropriate lever depressed to open the tank. I then open crossfeeds 2 and 3 and close tanks 2 and 3, so that engines 2 and 3 are both being fed from 5. It is then necessary to keep a close eye on 5 so that before it is exhausted the sequence is reversed and tanks 2 and 3 become active again. There are choices as to what happens next, but I always adopt what I regard as the simplest method, making a note of the fuel in 2 and 3, opening all the crossfeeds and then closing tanks 2 and 3 again, so that all engines are being fed from 1 and 4. Some hours later (at 19:16 in this case, to be precise) their level will have fallen to that of 2 and 3 and I can reopen their tanks, close all the crossfeeds and each engine will be supplied from its own tank until we land.

At 14:55 UTC we are level with the Nantucket NDB to the north (FS9 range 112nm) and at 15:19 the signal is lost. After that I set the headings using my E6B computer and the wind data in the weather reports, although as time proceeds this may become increasingly inaccurate. There are strong tailwinds but initially they are almost directly behind me and therefore there is little in the way of drift. However, later on the winds strengthen further and they also veer to starboard in a way which is going to produce significant drift. At one stage the E6B suggests a heading correction of 15 degrees. There is no assistance to be had from the stars, because we shall fly into the night on this short winter's day, but it is not yet. Also in the distance is Ocean Station Dog, but that is not yet either. However, the real aircraft was carrying LORAN and three navigators - an indication of just how crucial navigation was. There is no LORAN available for FS9 but Plan G is running on my secondary monitor and connecting it with FS9 will give me my correct position instantly. When I eventually do this I am level with Fix 5 but 70nm north of track. Later investigation showed that the wind had both strengthened and veered further than I had allowed for, and as result I was blown further north than anticipated. I used LORAN again at 18:40 and by then I was just beyond Fix 7 and still 70nm north of track, the heading adjustments I had made having worked only to the extent of preventing further deviation. During all this I had been paying insufficient attention to Ocean Station Dog in the belief that it was unlikely to be within range. Tom's Ocean Stations have an inevitably restricted range because of the limited range of FS9 NDBs. However, I have the receiver adapted for FS9 by Jorge Rechani from one originally made for FSX. At 18:55 this picks up a strong signal from Dog which of course gives the same impression as the LORAN fixes and also suggests it could have been picked up considerably earlier. In simulation terms LORAN might not have have been necessary at all, though one can be certain that the real crew used it to its maximum potential - that is why it was there. At 19:10 we check Dog again and by this time it is almost on the port beam, the light is draining from the sky and night is near.



By 19:47 Mars is visible ahead (or it would be if the FS9 depiction of the stars contained the planets) while the North Star is on the port beam. A passenger who looked out of the window to the north in 1956 would have seen it as part of The Plough, itself part of Ursa Major. The Plough is recognisable here, just above the wingtip.



The resultant starshot produces a result of Latitude 41 35 Longitude 39 15. We are now much nearer to the correct track and at 19:57 this is confirmed by the third LORAN fix.



We can now relax as far as navigation goes and wait to pick up the NDB (FS9 range 112nm) at Flores in the Azores, which we do at 20:34. There is still some way to go to Lajes but the NDBs off both ends of the main runway make lining up on it easy (there was no ILS in 1956) and winds are calm. The landing approach at 22:08.

We are parked with engines cut by 22:18 (21:18 Local Time). Flight Time 8:11. Fuel remaining 8,835 lbs - by no means a large amount considering the strength of the following winds. Lajes scenery by John Stinstrom backdated by Al Von Pingel and myself, but there will be much more of it to be seen in the next post.

    Information on Lajes and its navaids in the mid 1950s from another valuable Flickr album by John Hewson.

[Tom/CalClassic]

One of the hardest tasks on a propliner, flying across an ocean to a tiny dot of land.

I didn’t realize that Jorge never released our LORAN gauge, I had finished it and it only needed user beta testing. But to be honest it’s a long process that yields the same thing as the other things including the GPS - your current lat and long as reported by FS. It is kind of fun the first few times, but eventually becomes a chore. And as I remember it was only useful over the Atlantic Ocean, since that’s where we had LORAN charts.and station locations We had plans to expand the coverage later by creating our own, but that never happened.