A Technological and Hometown Appendix to Postblogging Technology, February 1950: The First Broken Arrow

I don't find any pictures of Port Hardy Airport's legendary derelict B-24.

It shouldn't really come as a surprise that the USAF lost three nuclear weapons in Canada in the early 1950s. If there's anything this blogging has brought home to me, it's that the aviation world remained just astonishingly lackadaisical in 1950. I didn't think anything was going to beat Bermuda Sky Queen or the BSAA saga, but then I noticed that no-one was noticing the four major KLM accidents in the two-and-a-half years beginning 26 January 1947. KLM is not an airline that occurs to me when the phrase "safety problem" comes up, perhaps because no-one brings it up. 

It's worth reflecting that the air search component of the Broken Arrow incident I will be discussing today was supplemented by aircraft called off the search for Douglas C-54 Skymaster 42-72469, which went down somewhere over the Yukon or British Columbia two weeks earlier with 8 crew, 34 members of SAC trooping to Montana, and someone's wife and baby (Joyce and Victor Espe, wife and infant son of Master Sergeant Robert Espe of 57th Air Installation Squadron.) The biggest air accident in Canada to that point and still the largest unexplained disappearance of American service personnel, you wouldn't even know about it from the aviation press. the much-urged policy of not mentioning passenger aircraft accidents in hopes that they would go away appears to have been implemented at some point in late 1949, and there you go. 

The same lackadaisicalness is seen in persistent reports that B-36B 44-92075 went down on Vancouver Island. The distinction between "Vancouver" and "Vancouver Island" seems to throw people for a loop and justify all sorts of geographical confusion, but the actual location of the wreckage, 6300ft up the side of Mount Kologet in the Nass Basin in the far northwest of British Columbia, is of very large significance to the story, and so is the island from which the survivors were recovered, Princess Royal Island in the central coast region, or, as we now say in a slightly twee way, "The Great Bear Rainforest." (Having given up on logging most of the region, we've settled for trying to attract the rare tourist who needs something more than 120" of rain a year to justify some adventure tourism.) 

On the other hand, the search headquarters was set up at Port Hardy, at least some of the survivors made it there before being hurriedly extracted to America, and for the first day of flying, the air search component was confined to the north end of Vancouver Island, because it was the only place where the weather allowed flying. Since Port Hardy is just a hop, skip and a bum-slide over from Port Alice, more than justifying slapping a "Port Alice" tag on this post. I had no idea that my hometown was so closely (that is, not closely at all) associated with a Broken Arrow incident, and, for that matter, the first one. 

(The first verse of the Skookumchuk Song. Best the Internet can do. Summer Wages, following River Road, is the same genre.)

The Convair B-36 is, to be precise, as Wikipedia is, "the largest mass-produced piston-engined aircraft ever built." the qualifiers suggest we're being pushed down a list to the point where the numbers aren't impressive, but that is emphatically not the case. The 162ft long, 230ft wide, 57ft high aircraft had an empty weight of 166,165lbs, and had a maximum auw of 410,000lbs. I won't cite performance statistics, which are a bit slippery and which in any case refer to later variants with auxiliary turbine engines and the armament stripped out, but 44-92075's flight profile, consisting of a run down the Pacific coast from Eielson Air Force Base near Fairbanks, Alaska, with a turn inland to carry out a practice bombing of San Francisco, followed by a return to Forth Worth, Texas, was chosen because it matched the aircraft's wartime assignment of delivering at least one atomic bomb against St. Petersburg. (44-92075 could theoretically carry up to 4 atomic bombs, but took off 51,000lbs overweight a recommended 278,000lbs with a single weapon on board, so colour me skeptical on that one.)

Near a ghost town then, and certainly one now, I'm a little surprised
that there's anything left of Butedale at all. It has its own
Wikipedia page, and yet another stranded hydroelectricasset.
There's a lot of them on the BC coast.  

Excuse me. A communications target in St. Petersburg.

The B-36B still carried its original armament of 8 twin 20mm gun turrets, six remote-controlled, retractable, two fixed in nose and tail, which meant that it needed a crew of fifteen. Throw in the weaponeer responsible for arming the bomb or bombs in flight, and the officer supervising the weaponeer, and you have the complement of seventeen that bailed out near Butedale on the night of Valentine's Day, 1950.  The flight was in trouble from the beginning, in fact, from well before its beginning. 

Nearly nine years after the specification was first circulated to the industry and two years after the first flight, the B-36 still wasn't ready for prime time. 44-92075 was found with its gun turrets still in their factory sealing, the operators having sensibly decided not to borrow trouble by trying the highly ambitious mountings out on one of the only 36 aircraft deemed serviceable in an operational role. This is an aspect of the ongoing battle with teething problems that would continue for another two years. I'm not going to hold this against the aircraft and its builder given the extraordinary nature of the aircraft, but it does rather hang a lampshade over SAC's decision to send it up with an atom bomb, one of the first batch of 8 that SAC had managed to persuade the AEC to release to them after considerable bureaucratic wrangling. (Minus, I should say, the mixed plutonium/Oralloy core that makes the actual boom.) All eight were flying out of Eielson on Valentine's Day night, and since seven of them made it home, the loss of 44-92075 seems . . . more excusable? I mean, in wartime, seven out of eight would be tolerable attrition for a one-off mission, and SAC was trying to practice under wartime conditions, right?

44-92075 reported at Eielson on 3 February with a malfunctioning navigation radar and reported the flaps sticking in flight, a problem also reported by the rest of the squadron. Ground crews worked to fixed nagging problems with fuel leaks in the -40 Alaska cold, but reassembling the radar around presumably not-yet-on-fire replacement parts was left as a project for the crew once airborne. The flight profile called for a run at 12,000ft down the coast, a climb to 14,000ft over Washington and Montana, then a turn towards San Francisco and a final climb to 40,000ft to fox hypothetical Red Air Force jet interceptors assumed to have poor altitude performance on the basis of their fashionably sweptback wings. (Since air interception without AI radar was unlikely at best, RAF carping about poor high altitude performance vis-a-vis British designs sounds like sour grapes at best, to me. USN claims about being able to intercept with their jet fighters are a little unreal.) 

Haida Gwaii has very unusual geography for BC:It's flat. 

The upshot of all this is that 44-92075's climb performance hadn't really been tested before running into clouds south of Sitka. The preflight briefing had mentioned icing conditions (surprise!!!) off the coast, and the squadron was instructed to climb above the weather if it encoutered it. This proved entirely too much for 44-92075, which suffered a cascade of problems, beginning with the gigantic, 3-bladed Curtis Electric propellers going out of synchronisation, presumably due to icing. The enormous structural battering that resulted could hardly have helped the overall icing situation.  Eighty miles northeast of Sandspit, BC, at around 11PM, 44-92075 asked 44-92083, flying nearby, to relay a distress signal. Captain Barry reported difficulties holding an altitude of 12,000ft.  At 11:25, an engine was reported out, and at 11:30, Captain Barry reported that the aircraft was beginning to lose altitude. This was probably after Captain Barry ordered (a renewed?) climb to 15,000ft, leading to one or two engines failing. Crew recollected that first the No. 2 and then the No. 5 engines caught fire and had to be feathered, while Captain Barry reported one engine feathered and two losing power. 083 reported a final distress message indicating three engines on fire and feathered, and the other three on emergency power with no gain in torque, although the rate of descent was gradually declining, to only 100ft/minute, 135mph, sometime shortly before 11:41, when Captain Barry indicated that he would either ditch or have the crew bail out over land shortly afterwards. 

This left the situation of the aircraft somewhat ambiguous, which  could have had deadly consequences for the crew, but did not. This was because Captain Barry now had to deal with his cargo. One of the key goals of the mission was practicing the core insertion process that was intended as SAC's standard operating procedure. All the talk about cores not being inserted in Broken Arrow incidents slightly obscures the fact that it was all theoretical at this point. Core insertion was a relatively late desiderata in the Atom Bomb Mark 4's development, and the Air Force, SAC and Sandia Labs were working it out as they went. An insertion/extraction toolkit had just been developed, and the weaponeers were still practicing with it, the former method of manual handling having been found to cause rapid rusting in the plutonium of the core, with a drastic reduction in the weapon's anti-Communist capability. The kit, and the "birdcage" were found empty in the wreckage, indicating that Captain Schreier had already completed his task of inserting the practice core. 

This part is only of any relevance to the conspiracy theorists who remain convinced that a live atomic bomb was aboard 44-92075, or, in the most delicious version, two of them. But it remains of interest to the literally ones or twos of British Columbians who would have been inconvenienced in any way by a 44 kiloton explosion at 3300ft over the Inside Passage opposite Princess Royal Island at midnight on Valentine's Day. As  it was, it was merely the explosion of 5000lbs(!!!) of high-quality explosive (no Amatol or guncotton for Sandia!) that was noticed by exactly no-one. The No. 4 had been in service since March, after a rocky road to deployment paved by the limits of the B-29 bomb bay and under-fuselage clearance, a phrase only rescued from its sheer absurdity by the realisation that making an atom bomb go boom in those days actually required 5000lbs of HE and that the resulting weapon was a roly-poly ball of metal and exotic organic chemicals, only barely saved from centrifugal self destruction by the best aerodynamic train that current computers could design. It turns out that dropping anything less solid than a hunk of iron from 40,000ft is asking a lot of the science of external ballistics. 

On the other hand, the Mark 4 [automatically downloading pdf] had been improved over previous iterations with a radar and mechanical fuze, an elevated core (a fancy way of saying that it was in a wire basket so that the imploding charge could get a running start) and some detailed improvements in the layout of the detonators that prefigured wholesale improvements in the conventional charge and firing harness that would pave the way to the newer, lighter atom and thermonuclear weapons of to-day. All of this leaves one rather happy that Captain Barry was able to get the bomb away, while putting in perspective the USAF's alarm at hearing three years later that the aircraft had been discovered on a BC mountainside, and not safe from following, living gales under the North Pacific waves.

Even before the crash, the cynic and skeptic had occasion to doubt the crew's version of events. B-36 crews were rather implausibly expected to wear exposure suits, flight suits, Mae Wests and parachute harnesses attached to personal dinghies. Some of the confusion over just who jumped and when is linked to the crew's disingenuous response to any question about whether they were actually wearing this implausible rig. (They weren't.) More importantly, at least from the perspective of Convair engineers, is whether the "anti-icing" system had been running when the plane first hit icing conditions. By all accounts the system, which pumped massive electrical current through the wing plates and prop blades, was neither particularly airworthy nor conducive to crew comfort. It also strikes me that it probably couldn't be turned on when required, and that this part was obscured in Captain Barry and the crew's report. Since Captain Barry was long dead in yet another B-36 accident (this one down to the negligence of a P-51 squadron of the Kansas National Guard) by the time the wreck was discovered, we will never know for sure. In any case, the core icing issue was in the carburetors of the big Wright R-3350s, which carried the dual handicap of being typical Wright  Aeronautical pieces of shit (hurrah for the brain-rotting effects of patent trolling!) and of being installed in an unconventional pusher configuration that meant that incoming air had to be carried through an S curve bathed in the engine's re-radiant heat before entering the chokes. 

Again, this could have been a tragedy, but wasn't. Icing conditions on this coast are often due to Arctic outflows, so the crew was parachuting into an unexpected offshore breeze, and this is almost certainly why the first five men to go were never found. The conspiracy theorists are persuaded that Weaponeer Schreier remained aboard the plane to nurse his live core, and took time out to take control, turn the aircraft back towards Alaska, and follow the Nass river valley through the Coastal ranges to a controlled crash landing in a cirque halfway up Kologet, but I find this unlikely. The remaining twelve men were sprinkled over Princess Royal Island. Ten landed close enough to the coast for eight to be spotted and picked up by the unsung heroes of the Canadian Fishing Company's troller, Cape Perry. Alerted to the crash by the hastily-formed search command while it was still in the process of setting itself up at Port Hardy Airport, Cape Perry's crew had its eyes peeled for smoke as they made their way down the Inside Passage. Considering the sheer fiasco overtaking the air search component that included helicopters flying up the I-95 at rooftop height, landing on Qualicum Golf Course, and crashing in the Fraser Canyon trying to follow the railroad tracks, it's just as well that there were some professionals on the job. 

Cape Perry? Vancouver city archives are having
denial-of-service problems today. 

Adding some further competence to the story, Tribal-class destroyer HMCS Cayuga made a fast run up from Esquimalt in time to rescue the last two on the beaches, and form rescue parties under local boys to rescue one man inland, known to be immobilised by a broken ankle, and a second man who had landed in the same condition, alone. One can only imagine the chagrin in the upper echelons of the USAF when the USN also came through, putting a PBY in the water quickly enough to extract the crew of 44-92075 before they'd been able to give more than a single interview to the Canadian media and whisk them away to hermetic solitaire in the United States. As the USAF wasn't yet prepared to tell even the AEC that it had misplaced a bomb, never mind the public, a cover-up was high on the agenda, and it is unfortunate that it could not have been extended to the USN as well. 

In later days, weather relented enough for the air search to be pressed home, helicopters included. Various teams of local outdoorsmen and fishers of the Gitga'at Nation at Hartley Bay, no strangers to marine rescue missions in the Inside Passage, combed the island and surrounding waters for stray survivors and human remains, eventually finding what was probably a skeletal leg of weaponeer Ted Schreier. By some minor miracle, no-one was lost on the ground in the treacherous conditions of the winter rain forest or the air search.

And that was it until three years later, when an RCAF aircraft looking for the missing personal de Havilland Dove of "Texas millionaire oilman Ellis Hall"spotted 44-92075 in its actual last resting spot. I did mention that aviation back in the day was a bit lackadaisical, right? Also, how quaint does "Texas millionaire oilman" sound these days? It's also worth noting, not to rag on Wikipedia, that Hall's wife, two daughters, and "17 year-old Patrick Hibben" were lost with Hall, and that apparently some wreckage was found only 35km east of originating airport Ketchikan in the fall of 1953.  

The Air Force, obviously looking for another way to look bad, tried to send a ground crew up to the crash site directly after the discovery was reported. After all, on the map it was a half-day's walk from the trailhead on the old Yukon Telegraph Trail, and, yes, it is up a mountain, but, still, it's only 13km!

Pictures suggest that the pioneers of the old Northwest lived in barely
reconditioned landfills, and this Telegraph Trail base camp could easily
stand in for some neighbourhoods of  Vancouver 70 years ago. It's the
clearance. There's just so bloody much ground cover. You can imagine
what it must have been like hauling stretchers up and down slope in a
Princess Royal Island winter. 

. . . A second, massively supported effort, in the summer of 1954, reached the site. Supporting helicopters were unable to land, as they lacked the lift to take off again from 6300ft, but they could drop demolition explosives to support the ground team's effort to destroy what could not be salvaged. 

On the evidence, it did not go well. B-36s are notorious for burning fiercely in crashes, but this did not happen. Parts of the fuselage were widely scattered, but a definitive 1997 follow up discovered the "improved Norden bombsight,"and turret fire control computers still intact, and 20mm shells scattered widely across the site. In the rear fuselage, they found the weaponeer's station, complete with the "birdcage" for the core (or practice core), and associated suitcase arrangement for holding the detonators that had to be removed and replaced during the core installation process, including three spare detonators. A considerable amount of unfired demolition explosives were collected by Canadian Forces personnel and lit off, following in the tradition of the 1956 Canadian Geographic Survey party who were the first to discover the site after the USAF had had its way, and which amused itself by setting off flares and signal grenades from the plane's survival stores. (Geologists are dumb.) 

I'm sorry, that kind of slipped out. It's the pure science physics>geology thing again. Although considering my academic transcript I shouldn't be calling myself a physics major. Anyway, It looks as though the search party settled for clearing out the radars, particularly the tail warning/fire control set, and bits of the engines, probably in hopes of settling the carburetor problems. If Commie infiltrator mountaineer artificers wanted to see what Sperry had come up with in the way of the automatic fire control systems, they were bloody welcome to them.   The conspiracy theorists are convinced, of course, that the team extracted a Mark 4, or possibly a Mark 4 and a Mark 5, and if that isn't Byzantine enough, did it before the recorded first sighting, perhaps because Ted Schreier was able to radio in a fix on the landing spot. Also, just to make sure, the USAF swore everyone in the Bulkeley Valley to secrecy, which is why the locals don't talk about this secret mission that totally for sure happened because spies are awesome! 

It's not clear to me how the conspiracy theorists got the idea that a Mark 5 could have been aboard, since it only entered service in 1952, but it was both smaller and much easier to arm and disarm in flight, so it was probably the first feasible weapon for a second load on a B-36B. 

It's still not clear how 44-92075 ended up where it did, but on the evidence, the circular course that Captain Barry plugged into the autopilot to take the plane out to sea, somehow turned into a straight course (perhaps the flaps froze?) as it rounded to a north-by-northeast heading and hit a patch of dry air that took off the icing load and allowed the plane to climb to 12,000ft at the last engine settings. Once over the coastal range and once again assailed by icing, descending gradually to its final encounter with Kologet. 

So that's the story of the first Broken Arrow, only ten months after the Revolt of the Admirals, eleven after the Mark 4 entered the nuclear arsenal, at a time when there were only B-36Bs in service, and the promised B-36D, with the jet engines that gave it a chance of eluding a MiG-15, was still months away. It turns out that the only secret the whole massive cover-up managed to keep was the wiring of the tail-warning radar, which actually makes sense given that the Russians could home on the frequency, if they knew it. It's enough to make you think that someone should work on some kind of omni-jamming technology and frequency-hopping radar. 

That being said, there are heroes  here. Without the Cape Perry it's quite likely that the entire crew would have perished of exposure before the search effort came near to completion. The search had no idea that they needed to be looking anywhere near Princess Royal, and the terrain of the central coast is as wild a wilderness as exists outside of Antarctica. the plane itself would have been found. 44-92075 was too close to the Yukon Telegraph Trail to remain undiscovered for very long, but as a matter of record, it took three years and the whole of the Korean War. That is a long time for the American security establishment to remain in the dark about the fate of a real, actual atom bomb at any time, never mind the McCarthy era. God only knows what this might have turned into without the sharp eyes of the lookouts on Cape Perry. 

Naturally, the names of the men of the Cape Perry are all but lost in the contemporary coverage. It is certainly to be hoped that they got a nice payout from the USAF in return for being shot by amnesia darts.