On 8 June 1966, Al White and Maj. Carl S. Cross took AV-2 up on a flight to make 12 subsonic airspeed calibration runs and one supersonic test. Afterwards, the XB-70 was scheduled to rendezvous with a contingent of jet aircraft, all powered by General Electric engines, for a public relations photo session. The other aircraft included a Northrop YF-5A (59-4898) flown by GE test pilot John M. Fritz, a Northrop T-38A (59-1601) piloted by USAF Capt. Peter C. Hoag with Col. Joe Cotton in the rear seat, a McDonnell F-4B (Bu.No. 150993) flown by USN Commander Jerome P. Skyrud with E.J. Black in the back, and Lockheed F-104N (N813NA) flown by NASA chief research pilot Joseph A. Walker. A Gates Lear Jet flown by H. Clay Lacy, loaded with photographers, recorded the event.

It was Walker's 13th XB-70 chase flight and took place just two days before he was scheduled to make his first flight in the XB-70 itself. In this case, he did not perform chase duties during the test mission itself. After the number two XB-70 had completed its test activities, Walker (in NASA 813) joined up in formation near the bomber's right wingtip.

Rendezvouz began at 0827 hours with the aircraft arranged in a V-formation. The Lear Jet was positioned about 600 feet out and to the left of the formation. Walker's F-104N was off the starboard wing of the XB-70. The photo session lasted from 0845 to 0925. Suddenly, at 0926, the F-104 collided with the Valkyrie's right wingtip. The F-104 rolled inverted and passed over the XB-70 shearing off the bomber's tail fins. The F-104N burst into flames, disintegrated, and fell towards the desert 25,000 feet below. Walker was killed instantly. Several seconds later, the XB-70 tumbled out of control and began to break apart, impacting 12 miles north of Barstow. The F-104 fell two miles northeast of the XB-70. The cockpit section fell half a mile northwest of the fuselage. It was the blackest day in the history of Edwards Air Force Base. A motion picture camera had recorded all events until just before the collision, and then picked up again shortly afterwards. There were also numerous still photos of the accident sequence. The images showed that the F-104 pitched upward from its position in the formation, tearing through the XB-70's right wingtip. From there, the F-104 rolled inverted as it passed across the top of the XB-70, shearing off both of the ship's large vertical tails. By this point, the F-104 was in several large pieces and trailing a ball of fire from its ruptured fuel tanks. General Electric chief test pilot John Fritz, flying on Walker's right wing in the YF-5A, saw nothing amiss prior to the collision. After the impact, the XB-70 continued in straight and level flight for 16 seconds as if nothing had happened. Someone got on the radio and shouted, "Mid-air, mid-air, mid-air!" Colonel Joe Cotton, riding in the back seat of Captain Pete Hoag's T-38, called frantically to the XB-70 crew. "You got the verticals came off, left and right. We're staying with you. No sweat. Now you're looking good." But, the giant ship rolled ponderously to the right and entered an inverted spiral. As it shed parts, fuel poured from its broken right wing. Cotton and the others began shouting for the stricken crew to eject. "Bail out. Bail out, bail out!"

F-104N breaking up in midair, after collision with the XB-70 on the left (click to enlarge)

After some difficulty, Al White ejected in his escape capsule. Maj. Carl Cross, possibly incapacitated by extreme g-forces, was less fortunate. Additionally, a seat retraction mechanism failed, making automatic encapsulation impossible. He remained trapped in the stricken jet and died when it struck the ground.

The accident investigation board determined that once the horizontal tail of the F-104 came up under the wingtip of the XB-70, it was caught in the wingtip vortex from the bomber. The F-104 lost its trim and pitched up violently, rolling inverted across the top of the bomber. The accident board concluded that the swirling wake vortex only became a contributory factor in the accident after the F-104's tail was so close to the XB-70 that a collision was imminent.

Why was the F-104 close enough to touch the XB-70? The investigation board postulated numerous theories. Walker had a reputation as a fine test pilot and a levelheaded professional, dedicated, and safety conscious. He had nearly 5,000 hours of flight experience and had flown chase for the XB-70 nine times, eight in an F-104. It was hard to imagine that lack of formation proficiency or some lapse of judgement could have led to the disaster. In the end, the leading theory held that Joe had become distracted somehow and had caused the F-104 to move slightly, imperceptibly, toward the XB-70.

Photos of the F-104N crash site (click to enlarge)

During the course of the flight, ground controllers had alerted the formation to the presence of other air traffic in the area. An Air Force F-104D, a two-place aircraft with a photographer in the back seat had received permission to join the formation for a few pictures before returning to Edwards after a separate mission. Don Sorlie, the pilot of the F-104D later reported that the formation looked good, although the two aircraft on the left were not flying as close as the F-104 and the YF-5A on the right. This indicated that Joe was flying a close wing position. Shortly before the collision, Edwards reported a B-58 approaching the formation. It would pass high above, in the supersonic corridor, and wouldn't pose a hazard. Several pilots in the formation responded that they could see the B-58's contrail. Walker never made such a call. He may have been attempting to spot the B-58 at the time his aircraft collided with the XB-70.

The investigation board ultimately concluded that Walker's position relative to the XB-70 left him with no good visual reference points for judging his distance. Therefore, a gradual movement in any direction would not have been noticeable to him. The board blamed an "inadvertent movement" of the F-104 that placed it in a position such that "contact was inevitable."

The length of the precision formation mission may have also been a factor. Cloudy weather had extended the flight time and forced the formation to move to a different area than had originally been planned. Walker had been flying close to the bomber's wing, in a position that made it difficult to judge his distance and other air traffic in the area created distractions.

XB-70A tumbling out of control after collision with the F-104N (click to enlarge)

Others suffered the consequences for their role in organizing, planning, and approving the flight. John Fritz, chief test pilot for General Electric, set the chain of events in motion when he requested permission for the photo mission from a North American Aviation representative. Subsequently, Col. Joe Cotton, XB-70 Test Director, agreed to provide such an opportunity on a non-interference basis following a regularly scheduled test flight. Another North American official disapproved the photo flight, citing a tight schedule, but Cotton and Fritz lobbied to include it on an upcoming flight. Both North American and Cotton's immediate supervisors, Col. Albert Cate and John McCollom, finally approved the photo opportunity. No further approval was sought from higher headquarters. Cotton arranged to include an Air Force T-38, a chase aircraft that normally accompanied the XB-70. Mr. Fritz requested a Navy F-4 from Point Mugu Naval Air Station. It was authorized as a routine training flight in support of what was assumed to be an approved Air Force mission. Fritz himself piloted the YF-5A, bailed to General Electric by the Air Force. Although his officially stated purpose for the flight was to "perform engine airstart evaluations," he never actually did this. He tried to arrange for a B-58 to join the formation, and also for supplementary Air Force photo coverage, but was unsuccessful. General Electric contracted with Clay Lacy for use of the Learjet photo plane. Cotton supported Fritz in a request for the NASA F-104. As chief pilot at NASA FRC, Joe Walker was within his authority to schedule the chase operation, but his superiors were not aware of the photographic mission.

Photos of the XB-70A crash site (click to enlarge)

An Air Force public affairs officer in Los Angeles learned of the photo mission just two days before the flight through a call from a commercial photographer. He referred the caller to Col. James G. Smith, chief of public affairs at Edwards. Smith had also been unaware of the planned formation flight, but voiced no objections once he ascertained that Col. Cate had approved the mission. The accident investigation board later ruled that Col. Smith should have advised the responsible parties of proper procedures for approving such a mission through higher headquarters. A memorandum, dated 12 August 1966, from Air Force Secretary Harold Brown to the Secretary of Defense concluded: "The photographic mission would not have occurred if Col. Cotton had refused the General Electric request or at least not caused North American to reconsider its reluctance. It would not have occurred if Col. Cate had taken a more limited view of his own approval authority. It would not have occurred if Col. Smith had advised of the need for higher approval. It would not have occurred if Mr. McCollom had exercised the power he personally possessed to stop the flight. But it did occur."

Secretary Brown further stated that "these individuals acted in ignorance of prescribed procedures, rather than with intent to violate them." He noted that the commander of Air Force Systems Command, with the concurrence of the Air Force Chief of Staff, directed a number of disciplinary actions against the responsible parties. Col. Cate was relieved as Deputy for Systems test and reassigned. Col. Cotton and Col. Smith received written reprimands, as did John McCollom. The Air Force also made numerous administrative changes to improve operational procedures.