d/d 31jan57, AFFTC (jan57-may57) w/o 27may57

Note: The build of 55-2970 was finished on 18 September 1956. After acceptance flights the aircraft was finally accepted by the USAF on 28 September and delivered on 31 January 1957. It was decided that this 55-2970 stayed with Lockheed for specific airframe and system testing during the phase the USAF tested the F-104A also with the Air Force (Edwards AFB). First of all the 55-2970 was used for BLC system testing but around May the aircraft was used to test modifications to prevent gun debris damaging the aircraft sides and entering the inlet. During the tests at Edwards this problem was issued.

Accident Reports:

27 May 1957 (YF-104A)
This F-104A crashed 12:20 hours PST at San Bernardino, 20 N Miles NW of Camp Irwin on Leach Gunnery Range, in California during an gun shooting test mission after encountering a compressor failure. Pilot, Lockheed Engineering Test Pilot, Arthur Joseph Peterson tried several airstarts which were not successful. Then he ejected safely at 2500 feet. Aircraft took off from AF Plant 42, Lockheed, California, a flight which lasted 21 minutes before crash. After the accident the engine was shipped to Evandale for study. Results are unknown but very likely cause was debris from the gun which entered the inlet and damaged compressorblades inside the engine.
The official report spent a lot of time to the project in which this F-104 was involved. It also showed how many effort was spent to get the J79 engine running normally again.

USAF REPORT (briefly):
Previous testflights have shown that gun debris (links and shell cases) from the T-17 E-3 gun when fired in flight from the F-104A strike the fuselage and ventral fin in a random manner from an area in line with the engine inlet air ducts aft. A fix to this problem consisting of installing an 8”x16”vertical steal deflector plate in front of the shell case and link ejection chutes had completed wind tunnel tests and was being flight tested. Tests were not complete but were not progressing satisfactorily as the fuselage and ventral fin continued to receive random strikes from the gun debris.
The accident occurred during the fifth flight (fourth flight that the guns were fired) in a planned series of flights to test the deflector plate. The test aircraft was F-104A 55-2970 on LAC s/n 1016, on bailment to Lockheed Aircraft Corp (LAC) flown by Peterson, an LAC Engineering test pilot.
The chase aircraft was an USAF F-100 from Edwards Air Force Base flown by Capt Carson, an ARDC Engineering test pilot.
The purpose of the flight was to fire a two second burst in a two “G”pull out at 400K. then fire the remainder of a full ammunition load (760 rounds) in a two second bursts at one “G” and 350K.
Peterson took off in the testaircraft from Palmdale, California at 12:59 hours PDT. Take off and climb were normal. Radio contact was established with Herbert 3, Lockheed Flight Test Engineering, immediately after becoming airborne. Rendevous with the chase aircraft was effected at 19,000 feet over Edwards AFB. The two aircraft proceded to the Leach Lake Gunnery Range at 19,000 feet and 350K. The pilot of the test aircraft notified Herbert 3 (Lockheed Engineering) of his position and flight card number. He then increased power to 100% and transmitted his engine instrument readings to Herbert 3.The first firing pass was made at military power (100% RPM no Afterburner), 400K in a two “G” pull up on a West heading. The chase aircraft flew line abreast 50 feet out on the left side.
A two second fire burst was made and the pilot noted all engine instruments were normal. The chase pilot noted a wide dispersal of gun debris.
Power was retarded to 92% and a rectangular pattern flown at 19000 feet for the second firing pass. The pilot of the test aircraft advised Herbert 3 of a cockpit pressure fluctuation of 2000 feet and that he had 2900 lbs of fuel remaining. Speed brakes were used to bring air speed down to 340K. Power was increased to 100%. Dive brakes were brought up and the pilot of the test aircraft advised Herbert 3 of his engine instrument readings all of which were normal. At 350K the pilot of the testaircraft fired a two second burst. He noticed no unusual vibration, instrument readings, or warninglights on the warning light panel commonly refered to by pilots as the “peek and panic” panel. The chase pilot was in the same position as on the first firing run.
It should be noted that the deflector plate installed on the aircraft caused continuous buffeting at all speeds and gave the effect of partial dive brakes. Also the aircraft was instrumented to study the effects of gun firing which was recorded by an oscillograph. The oscillograph is set in action by pressing the gun switch to the first detant 10 seconds before firing and turns itself off after a short delay when the gun switch is released. The oscillograph recorded both firing runs and was recovered.
Film recordings were also made of the lower aft portion of the aircraft but were not recovered. Engine thrust calibrations indicate 16% RPM is required to sustain flight at 19,000 feet in a clean condition. With the deflector plate installed at least 86% RPM is required with the engine producing reated thrust for the RPM.

Just as the test aircraft completed the second firing run the pilot of the chase aircraft, Capt Carson, saw a long yellow flame of about one and one half second duration come out of the tail pipe of the test aircraft and was typical of a compressor stall. Capt Carson called Peterson, pilot of the test aircraft, and told Peterson he had a compressor stall. Peterson immediately reduced power. EGT did not show overheat condition and RPM had dropped to approximately 80%. Peterson could not positively state whether RPM drop was due to retarding the throttle of existed prior to reducing throttle. The chase aircraft fell in trail in such a position that he could see up the tail pipe of the test aircraft.
Peterson initially attempted to clear the stall by retarding the throttle and adding power several times. He could not get over 78% to 80% RMP and vibration increased as he reached this point. Peterson brought the throttle back to idle and hit both air start switches then advanced the throttle but the engine would not follow beyond 78% and 80%. Airspeed was approximately 250K. During these attempts to clear the stall Capt Carson observed yellow flame through the turbine blades in the lower half of the engine but could see no flame in the upper half. Peterson then stopcocked and when the flame went out Capt Carson pulled up again on the left side in a line abreast position and stayed there until Peterson left the aircraft. Capt Carson observed the attempt to clear the engine but did not observe the subsequent air starts from the “in trail” position.
When repositioning the throttle and hitting the air start switches proved unsuccessful (this is normal procedure and will usually clear a stall or “hung” engine as indicated by interviews with other F-104 pilots) Peterson stopcocked and let RPM fall to 50%. He made an air start and the engine started normally and ran smoothly up to 78%-80% RPM but became rough and would not rise above that point. EGT was normal for the RPM. Overheat EGT was never noted. Peterson again stopcocked the engine and let RPM fall to 25%. He dove the aircraft to 270K, and made another air start. The start was normal and the engine run smooth to 78%-80% when it again started vibrating and the RPM stabilized. Capt Carson observed the test aircraft from a line abreast position during both air starts. He looked the aircraft over very carefully but could see no damage. The deflector plate was in position. Between each air start fuel came out of the overboard drain in a normal manner and after each start he observed heat waves from the tail pipe.
Capt Carson asked Peterson what were his intentions. Peterson advised he would eject at 6,000 feet (sea level). Capt Carson advised he was then at 6,000 feet and to eject. The tarrain in this area is 3,500 feet. Peterson trimmed the test aircraft slightly nose high with power at 80% RPM and ejected.
The ejection was made with no apparent difficulty. Peterson made two forward summersaults and separated from the seat in an upright position. He reported the automatic features of releasing the seat and opening the parachute worked faster then he could think and manually follow.
When the parachute opened the shroud lines were twisted above the harness for about three feet.
Peterson was rotating slowly during descent. As the shroud lines unwound he hit the ground drifting sideways and fell backwards. He was dragged a short distance and released himself by unsnapping his leg and breast strape. The parachute blew a short distance away and hung on some brush. Peterson crawled to his parachute and awaited rescue.
Capt Carson observed the aircraft, after Peterson ejected, continue straight ahead and crash in a level position, explode and burn briefly. He followed Peterson down to get his position then returned  to Edwards AFB and accompanied the rescue helicopter back to the crash scene where Peterson was picked up and returned to Edwards AFB.  Peterson was on the ground for 1:20 hours after landing.