超级军网F22因为电传而坠机？问题解决的如何？
来源：百度文库 编辑：超级军网 时间：2024/04/29 10:22:11
04年F22坠机的问题，据说是因为电传的问题？解决的如何？
现在都已经服役了？是否存在隐患？04年F22坠机的问题，据说是因为电传的问题？解决的如何？
现在都已经服役了？是否存在隐患？
现在都已经服役了？是否存在隐患？04年F22坠机的问题，据说是因为电传的问题？解决的如何？
现在都已经服役了？是否存在隐患？
当初J10不是以电传机里唯一一个没因为电传坠机做宣传重点之一的么，照这说法F22因为电传出问题不奇怪吧
我是说现在解决的如何？
我想服役的话，问题肯定解决的差不多了，不能带病上天吧。
我想服役的话，问题肯定解决的差不多了，不能带病上天吧。
YF23可否介绍以下情况
主要是飞控软件问题，解决方法就是检查对应模块的成千上万行代码，重写部分代码后进行模拟，没有问题后再重新装载进飞控系统中
飞控软件的严重缺陷一般在编写测试早期就解决了，但残留的一些bug不出事很难被发现，往往出现在某些特定情况下
这种电脑的程序控制的东西，总要出点状况才能解决的吧？
很想了解一下具体情况
很想了解一下具体情况
没解决怎么会服役呢？
F-22在试验时发生过空中死机问题（机载计算机死机），飞行员在空中重新启动的故障，还引起美国程序员之间关于Ada和C++的争论，F-22的飞控系统是由Ada和C++编写出来的，而以前的F-16、F-15是用Ada写的，在兼容性上多少都会有问题。
原帖由 frank01 于 2007-12-14 14:40 发表
F-22在试验时发生过空中死机问题（机载计算机死机），飞行员在空中重新启动的故障，还引起美国程序员之间关于Ada和C++的争论，F-22的飞控系统是由Ada和C++编写出来的，而以前的F-16、F-15是用Ada写的，在兼容性上多少 ...
晕，飞着飞着死机了，这麻烦就大了。应该由冗余的吧？
F-22在试验时发生过空中死机问题（机载计算机死机），飞行员在空中重新启动的故障，还引起美国程序员之间关于Ada和C++的争论，F-22的飞控系统是由Ada和C++编写出来的，而以前的F-16、F-15是用Ada写的，在兼容性上多少都会有问题。
＝＝＝＝＝＝＝＝＝＝＝＝
那是Ada和C++好，还是Ada好？
＝＝＝＝＝＝＝＝＝＝＝＝
那是Ada和C++好，还是Ada好？
记得当年有个笑话是ada粉挖苦C＋＋的，谁能回忆一下？我忘了小笑话里那三个键了，是弹射＋发动机＋fire?
我看过ngc放出来的出事的视频，感觉不是软件冲突的缘故。但是，C＋＋不支持异常处理的特性和编译器的毛病，语法的晦涩，确实不适合一些要求苛刻的场合
Ｊ１０是用什么写出来的，谁知道？
我看过ngc放出来的出事的视频，感觉不是软件冲突的缘故。
===============
根据你的分析，你认为是什么方面的原因？
===============
根据你的分析，你认为是什么方面的原因？
至少目前为止，在已经装备的来看，歼10坠毁数量比F22多
变稳机啊，变稳机
原帖由 tony123rql 于 2007-12-14 17:41 发表
我看过ngc放出来的出事的视频，感觉不是软件冲突的缘故。但是，C＋＋不支持异常处理的特性和编译器的毛病，语法的晦涩，确实不适合一些要求苛刻的场合
瞎扯淡啊。。。。。C++ 对异常有完善的支持。。。。。。
飞行员诱发震荡（PIO）
五花八门的计算机语言常常使我们程序员搞不清正在使用的是哪一种。下面的一次小型会议将有助于澄清你的疑惑。
任务：射你自己的脚
C：射你自己的脚。
C++：你不留神生成了一堆你自己的实例，所以只好挨个射他们的脚。紧急援救是不可能的，因为你不知道哪个是你的真拷贝，哪个只是指向你的指针。
fortran：你逐个射你的脚趾，一直循环到射没了所有的脚趾，然后你读入下一只脚并重复之。如果你没了子弹，你也得接着射，因为你没有意外处理机制。
pascal：编译器不允许你这么干。
ada：在你仔细地包装好了你的脚后，你试图以并行的方式上弹，扣扳机，尖叫，并射你自己的脚。然而，当你试了一下后，发现你的脚类型不对。
prolog：你告诉程序你想射你自己的脚。程序会自动找到具体的计划，不过语法上是不允许把这些计划告诉你的。
basic：你用水枪射你自己的脚。如果是在大系统中，重复直至你的下半身被水浸没。
Visual Basic：你其实只是装出好象是射了你的脚的样子。不过你觉得这么干更有趣所以也不在乎倒底射没射。
paradox：不但你可以射你自己的脚，你的用户也可以。
access：你用枪瞄准了你自己的脚，但子弹却把旁边所有标着borland字样的软盘打出了洞。
assembler：你试图射你自己的脚，结果发现你还得先自己来制造出枪支，
子弹，瞄准具，和你的脚。
modula2：当终于明白用这个语言什么也干不了时，你一枪射穿了你的脑门。
任务：射你自己的脚
C：射你自己的脚。
C++：你不留神生成了一堆你自己的实例，所以只好挨个射他们的脚。紧急援救是不可能的，因为你不知道哪个是你的真拷贝，哪个只是指向你的指针。
fortran：你逐个射你的脚趾，一直循环到射没了所有的脚趾，然后你读入下一只脚并重复之。如果你没了子弹，你也得接着射，因为你没有意外处理机制。
pascal：编译器不允许你这么干。
ada：在你仔细地包装好了你的脚后，你试图以并行的方式上弹，扣扳机，尖叫，并射你自己的脚。然而，当你试了一下后，发现你的脚类型不对。
prolog：你告诉程序你想射你自己的脚。程序会自动找到具体的计划，不过语法上是不允许把这些计划告诉你的。
basic：你用水枪射你自己的脚。如果是在大系统中，重复直至你的下半身被水浸没。
Visual Basic：你其实只是装出好象是射了你的脚的样子。不过你觉得这么干更有趣所以也不在乎倒底射没射。
paradox：不但你可以射你自己的脚，你的用户也可以。
access：你用枪瞄准了你自己的脚，但子弹却把旁边所有标着borland字样的软盘打出了洞。
assembler：你试图射你自己的脚，结果发现你还得先自己来制造出枪支，
子弹，瞄准具，和你的脚。
modula2：当终于明白用这个语言什么也干不了时，你一枪射穿了你的脑门。
Ada是一种表现能力很强的通用程序设计语言，它是美国国防部为克服软件开发危机，耗费巨资，历时近20年研制成功的。它被誉为 第四代计算机语言的成功代表。与其他流行的程序设计语言不同，它不仅体现了许多现代软件的开发原理，而且将这些原理付诸实现。因此，Ada语言的使用可大大改善软件系统的 清晰性， 可靠性， 有效性， 可维护性。Ada是现有的语言中无与伦比的一种 大型通用程序设计语言，它是现代计算机语言的成功代表，集中反映了 程序语言研究的成果。 Ada的出现，标志着软件工程成功地进入了国家和国际的规模。在一定意义上说，Ada还刺破了“冯.偌依曼思维模式” (Von Newman Mind-set) 的桎梏，连同Ada的 支持环境(APSE)一起，形成了新一派的所谓 Ada文化。它是迄今为止 最复杂，最完备的软件工具。 Ada语言是 美国国防部指定的 唯一的一种可用于 军用系统开发的语言，我国军方也将Ada做为军内开发标准（GJB 1383《程序设计语言Ada》）。
以上都是放狗搜的~
以上都是放狗搜的~
居然连ada和c++之争都扯出来了？
那次事故要说软件bug也应该算是飞控算法设计上的bug
不是所有软件系统都像通用商业软件那样补丁不可少的
航空航天领域确实存在无错的软件项目
那次事故要说软件bug也应该算是飞控算法设计上的bug
不是所有软件系统都像通用商业软件那样补丁不可少的
航空航天领域确实存在无错的软件项目
现在老美自己也开始在军事项目里用c++和商用设备了
多快好省是王道，ada 程序员太难找了
多快好省是王道，ada 程序员太难找了
这件事情确实跟飞控有关，但是不是PIO，原因主要是防错机制不够完善和飞行手册上表达模糊。
起飞前做完BIT后发现蓄能器压力不足，启动前用掉过多压缩空气就会有这个问题，按常规处理后发现蓄能器压力上升比较慢，于是飞行员把飞机滑行回去做人工充气。人工充气要求关闭发动机。发动机关闭了之后热管理系统关闭，辅助动力装置转入冷却模式，飞控的供电中断了，这时APU RUN灯亮指示APU冷却状态，飞行员于是重起了APU，这个间隔不到1秒，飞控的供电重新恢复，这个供电中断导致角加速度传感器自己闭锁了，此后输出为0。这种情况因为3个角加速度传感器输出一致，所以飞控的周期性BIT没有发现，必须要飞行员启动全机的BIT才能发现。但是飞行员认为他在第一次滑出时做了飞行员启动的BIT，所以没有必要重做，就在人工充气结束，蓄能器压力正常后重起发动机，滑出去直接起飞了。缺乏关键的角加速度数据导致飞机受到扰动后运动发散坠毁。
详情见于422中队对事故的总结
起飞前做完BIT后发现蓄能器压力不足，启动前用掉过多压缩空气就会有这个问题，按常规处理后发现蓄能器压力上升比较慢，于是飞行员把飞机滑行回去做人工充气。人工充气要求关闭发动机。发动机关闭了之后热管理系统关闭，辅助动力装置转入冷却模式，飞控的供电中断了，这时APU RUN灯亮指示APU冷却状态，飞行员于是重起了APU，这个间隔不到1秒，飞控的供电重新恢复，这个供电中断导致角加速度传感器自己闭锁了，此后输出为0。这种情况因为3个角加速度传感器输出一致，所以飞控的周期性BIT没有发现，必须要飞行员启动全机的BIT才能发现。但是飞行员认为他在第一次滑出时做了飞行员启动的BIT，所以没有必要重做，就在人工充气结束，蓄能器压力正常后重起发动机，滑出去直接起飞了。缺乏关键的角加速度数据导致飞机受到扰动后运动发散坠毁。
详情见于422中队对事故的总结
EXECUTIVE SUMMARY
AIRCRAFT ACCIDENT INVESTIGATION, F/A-22 S/N 00-4014
422nd TEST AND EVALUATION SQUADRON
NELLIS AIR FORCE BASE (AFB), NEVADA
20 DECEMBER 2004
On 20 December 2004, at 2340Z/1540 local time, the Mishap Aircraft (MA), F/A-22, Serial Number 00-
4014, crashed on initial takeoff from Nellis AFB. The Mishap Pilot (MP), assigned to the 422nd Test
and Evaluation Squadron, Nellis AFB, ejected safely and sustained only minor injuries. There were no
other casualties. The MA impacted the Nellis AFB airfield and was destroyed. The only other damage
was also to government property including an arresting cable, a runway sign, a runway light, and the
runway surface.
Immediately upon leaving the ground, the MA began a series of un-commanded and progressively
more violent yaw, roll, and pitch transients. Unable to control the aircraft, the MP ejected seconds
before the MA impacted the ground.
The Accident Investigation Board President determined the cause of the mishap, supported by clear
and convincing evidence, was an inoperative Flight Control System, resulting from a power
interruption, which made the MA uncontrollable. The MP was , unaware of this condition because he
did not perform an Initiate Built in Test (IBIT), the only means of detecting the problem. Failure to
perform the IBIT was the result of ambiguous Technical Orders and a mistaken belief in continuous
RSA power availability.
During the mishap sequence, the MP started engines, performed an IBIT, and had a fully functioning
Flight Control System. Subsequently, the MP shut down engines to allow maintenance personnel to
service the Stored Energy System. During engine shut down, the MA’s Auxiliary Power System (APU)
was running. The MP believed the APU provided continuous power to the Flight Control System, and
therefore another IBIT after engine restart was unnecessary. This belief was based on academic
training, technical data system description, and was shared by most F/A-22 personnel interviewed
during the investigation.
In fact, the MA’s Flight Control System did experience a brief power interruption during the engine
shut down sequence. The interruption produced an unforeseen catastrophic Flight Control System
failure that rendered the MA un-flyable.
Under 10 U.S.C. 2254(d), any opinion of the accident investigators as to the cause of, or the factors
contributing to, the accident set forth in the accident investigation report may not be considered as
evidence in any civil or criminal proceeding arising from an aircraft accident, nor may such information
be considered an admission of liability by the United States or by any person referred to in these
conclusions or statements.
AIRCRAFT ACCIDENT INVESTIGATION, F/A-22 S/N 00-4014
422nd TEST AND EVALUATION SQUADRON
NELLIS AIR FORCE BASE (AFB), NEVADA
20 DECEMBER 2004
On 20 December 2004, at 2340Z/1540 local time, the Mishap Aircraft (MA), F/A-22, Serial Number 00-
4014, crashed on initial takeoff from Nellis AFB. The Mishap Pilot (MP), assigned to the 422nd Test
and Evaluation Squadron, Nellis AFB, ejected safely and sustained only minor injuries. There were no
other casualties. The MA impacted the Nellis AFB airfield and was destroyed. The only other damage
was also to government property including an arresting cable, a runway sign, a runway light, and the
runway surface.
Immediately upon leaving the ground, the MA began a series of un-commanded and progressively
more violent yaw, roll, and pitch transients. Unable to control the aircraft, the MP ejected seconds
before the MA impacted the ground.
The Accident Investigation Board President determined the cause of the mishap, supported by clear
and convincing evidence, was an inoperative Flight Control System, resulting from a power
interruption, which made the MA uncontrollable. The MP was , unaware of this condition because he
did not perform an Initiate Built in Test (IBIT), the only means of detecting the problem. Failure to
perform the IBIT was the result of ambiguous Technical Orders and a mistaken belief in continuous
RSA power availability.
During the mishap sequence, the MP started engines, performed an IBIT, and had a fully functioning
Flight Control System. Subsequently, the MP shut down engines to allow maintenance personnel to
service the Stored Energy System. During engine shut down, the MA’s Auxiliary Power System (APU)
was running. The MP believed the APU provided continuous power to the Flight Control System, and
therefore another IBIT after engine restart was unnecessary. This belief was based on academic
training, technical data system description, and was shared by most F/A-22 personnel interviewed
during the investigation.
In fact, the MA’s Flight Control System did experience a brief power interruption during the engine
shut down sequence. The interruption produced an unforeseen catastrophic Flight Control System
failure that rendered the MA un-flyable.
Under 10 U.S.C. 2254(d), any opinion of the accident investigators as to the cause of, or the factors
contributing to, the accident set forth in the accident investigation report may not be considered as
evidence in any civil or criminal proceeding arising from an aircraft accident, nor may such information
be considered an admission of liability by the United States or by any person referred to in these
conclusions or statements.
]]
Without RSA data, the FLCS computer wasn’t receiving angular acceleration feedback telling the
FLCS computers how the aircraft was reacting to the pilot’s stick/rudder inputs. The only information
fed to the FLCS computers was the position of the flight control surfaces. This situation was recreated
in the handling qualities simulator at Edwards AFB using the highly experienced F/A-22 test pilots.
Without RSA data, all pilots crashed in the first few of seconds of flight. High stick sensitivity/overcontrol
was the norm. The test pilots’ conclusion was “the aircraft is completely uncontrollable without
any angular acceleration/rate feedback to the FLCS” (Tab J-9).
Immediately after the MA became airborne, it yawed into the wind (nose right) with approximately 3
degrees of sideslip. The climb continued to be relatively stable varying between five and eight degrees
of nose-high pitch. Through timely pilot input, the nose right yaw rate was arrested and began to yaw
aggressively left. At the same time, power was reduced to approximately 45% for approximately two
seconds. The MP testified that this was in reaction to the jet not “flying like an airplane.” The MP did
not consider a high speed abort because he believed he could fly the MA out of these yaw transients.
The aircraft continued to yaw left to a sideslip angle 15 degrees. Due to his low altitude, the MP
selected MAX power in an attempi to climb. After another large yaw oscillation to right (22 degrees of
sideslip) MP reduced power back to military (MIL) for the remainder of the flight. The MP still felt he
could fly his way out of this. The MP had not reached his ejection decision-yet (Tab CC-7) (Tab V-5.14).
The MP testified that his ejection decision is based on three criteria: 1) aircraft is not flying in
coordinated flight, 2) the aircraft is not correctly responding to control inputs, and 3) the aircraft is
progressing to a position that is unrecoverable either in pitch or roll (Tab V-5.15). The MP’s well
thought-out ejection decision prevented a probable fatality.
The landing gear was still down. Because the airspeed was continuing to climb and the fact that the
aircraft was not flying well in its current configuration, the MP raised the landing gear. Immediately
after landing gear retraction, the jet started to pitch-up, roll right, and yaw left. Sideslip peaked at 31
degree nose left. The right roll rate could not be arrested by stick and rudder inputs--the MA was no
longer responding to the MP’s flight control inputs. At 1540:09, 11 seconds after becoming airborne, the
CANOPY UNLOCK ICAW asserted indicating ejection (Tab CC-8) (Tab V-5.15).
The MP stated that he was still trying to fly the aircraft with his right hand and pulled the ejection
handle with his left. The MP’s recollection of the events leading up to his ejection decision was
accurate-almost an exact reiteration of CSMU data. Aircraft parameters at the time of the CANOPY
UNLOCK ICAW are as follows (Tab CC-8):
• Angle of Attack: -4.4 degrees
• Sideslip: -39.0 degrees (nose left) • Pitch : 25 degrees nose up
• Roll: 56 degrees ring wing down (continuing to roll more right wing down)
• Magnetic heading: 002
• Airspeed: 210 Knots Calibrated Airspeed (KCAS)
• Altitude: 2360’ (approximately 500 feet above the ground)
The MP suffered minor abrasions to the neck during the ejection (Tab B-3). See section 4f for a
discussion of the ejection seat performance.
Following the MP’s ejection, the MA continued to simultaneously roll right and pitch further nose up.
The MA continued to a near inverted attitude, and pitched nose up, essentially performing a “Split-S”
and impacted the runway going backwards. The last frame of CSMU data recorded the following
parameters (Tab J-23):
• Angle of Attack: 90 degrees
• Sideslip: -45.0 degrees (nose left)
• Pitch : 35 degrees nose up
• Roll: 71 degrees left wing down
• Magnetic heading: 223
FLCS computers how the aircraft was reacting to the pilot’s stick/rudder inputs. The only information
fed to the FLCS computers was the position of the flight control surfaces. This situation was recreated
in the handling qualities simulator at Edwards AFB using the highly experienced F/A-22 test pilots.
Without RSA data, all pilots crashed in the first few of seconds of flight. High stick sensitivity/overcontrol
was the norm. The test pilots’ conclusion was “the aircraft is completely uncontrollable without
any angular acceleration/rate feedback to the FLCS” (Tab J-9).
Immediately after the MA became airborne, it yawed into the wind (nose right) with approximately 3
degrees of sideslip. The climb continued to be relatively stable varying between five and eight degrees
of nose-high pitch. Through timely pilot input, the nose right yaw rate was arrested and began to yaw
aggressively left. At the same time, power was reduced to approximately 45% for approximately two
seconds. The MP testified that this was in reaction to the jet not “flying like an airplane.” The MP did
not consider a high speed abort because he believed he could fly the MA out of these yaw transients.
The aircraft continued to yaw left to a sideslip angle 15 degrees. Due to his low altitude, the MP
selected MAX power in an attempi to climb. After another large yaw oscillation to right (22 degrees of
sideslip) MP reduced power back to military (MIL) for the remainder of the flight. The MP still felt he
could fly his way out of this. The MP had not reached his ejection decision-yet (Tab CC-7) (Tab V-5.14).
The MP testified that his ejection decision is based on three criteria: 1) aircraft is not flying in
coordinated flight, 2) the aircraft is not correctly responding to control inputs, and 3) the aircraft is
progressing to a position that is unrecoverable either in pitch or roll (Tab V-5.15). The MP’s well
thought-out ejection decision prevented a probable fatality.
The landing gear was still down. Because the airspeed was continuing to climb and the fact that the
aircraft was not flying well in its current configuration, the MP raised the landing gear. Immediately
after landing gear retraction, the jet started to pitch-up, roll right, and yaw left. Sideslip peaked at 31
degree nose left. The right roll rate could not be arrested by stick and rudder inputs--the MA was no
longer responding to the MP’s flight control inputs. At 1540:09, 11 seconds after becoming airborne, the
CANOPY UNLOCK ICAW asserted indicating ejection (Tab CC-8) (Tab V-5.15).
The MP stated that he was still trying to fly the aircraft with his right hand and pulled the ejection
handle with his left. The MP’s recollection of the events leading up to his ejection decision was
accurate-almost an exact reiteration of CSMU data. Aircraft parameters at the time of the CANOPY
UNLOCK ICAW are as follows (Tab CC-8):
• Angle of Attack: -4.4 degrees
• Sideslip: -39.0 degrees (nose left) • Pitch : 25 degrees nose up
• Roll: 56 degrees ring wing down (continuing to roll more right wing down)
• Magnetic heading: 002
• Airspeed: 210 Knots Calibrated Airspeed (KCAS)
• Altitude: 2360’ (approximately 500 feet above the ground)
The MP suffered minor abrasions to the neck during the ejection (Tab B-3). See section 4f for a
discussion of the ejection seat performance.
Following the MP’s ejection, the MA continued to simultaneously roll right and pitch further nose up.
The MA continued to a near inverted attitude, and pitched nose up, essentially performing a “Split-S”
and impacted the runway going backwards. The last frame of CSMU data recorded the following
parameters (Tab J-23):
• Angle of Attack: 90 degrees
• Sideslip: -45.0 degrees (nose left)
• Pitch : 35 degrees nose up
• Roll: 71 degrees left wing down
• Magnetic heading: 223
e. Impact.
Aircraft F/A-22A, S/N 00-4014, impacted the ground on initial takeoff at 2340:12Z/1540:12L (Tab J-
23) at 36 degrees and 14.79 minutes north and 115 degrees and 01.41 minutes west, Nellis AFB,
Nevada (Tab C-3). In addition to the loss of the aircraft, other government property that was damaged
included a BAK-12 arresting cable, a runway guide sign, a runway light and runway 03R. MA broke
into several pieces leaving a debris field scattered over the departure end of runway 03R, the runway
overrun and ground area beyond the departure end.
f. Life Support Equipment, Egress and Survival.
The MP initiated ejection at 2340:09Z/1540:09L during an uncommanded right roll at approximately
2,360 feet MSL, 500 feet AGL (Tab CC-8). The parachute deployed normally (Tab V-5.16).
The Air Force Safety Center at Kirtland AFB, NM evaluated the following Emergency Escape
Sequencing System (EESS): modified ACES II ejection seat (S/N 225014), two Thermal Battery
Ejection Seat Initiators (S/Ns 50 & 36), two Electro-Explosive Devices (EED) (S/Ns 1678 & 1670), the
Canopy Jettison Rocket Motor (S/N 19), Inertia Reel Gas Generator (S/N 11134), Inertia Reel EED (S/N
125), Primary Parachute Mortar Cartridge (S/N 9269), Back-up Parachute Mortar Cartridge (S/N
9276), Drogue Deployment Mortar Cartridge (S/N 31563), the CKU-5B/A Ejection Seat Rocket
Catapult, the Gyro Spin-up Cartridge (S/N 16443), the Vernier Rocket Motor (S/N 11490), the Harness
Release Cartridge (SIN 24443), two Drogue Severance Assemblies (S/Ns 3736 & 3701), two 1.15 SecondDelay Reefing Line Cutters, the 4.0 Second Delay Locking Cord Cutter, the Recovery Sequencer (S/N
1199-1389), and two Universal Water Activated Release System (UWARS) parachute releases (Tab H-3
through H-5).
In the opinion of the evaluators, the EESS and all pyrotechnic components functioned properly during
the ejection in the Mode I range (Tab H-6).
g. Search and Rescue.
The MP ejected safely and sustained only minor injuries. Mongoose 01, an HH-60G transitioning in
the traffic pattern, landed on runway 03R to ensure the MP had not sustained significant injury. The
MP was transported by ambulance to Mike O’Callahan Federal Hospital where he received evaluation
and treatment. The MA impacted runway 3R and was completely destroyed. The crash and resulting
debris field were confined to Nellis AFB property. Fire response was swift, and the fire was
extinguished.
h. Recovery of Remains.
There were no fatalities as a result of this mishap.
5. MAINTENANCE
a. Forms Documentation.
The 422 Aircraft Maintenance Unit (AMU) maintained the aircraft forms for the F/A-22A Raptor on
an Integrated Maintenance Information System (IMIS) database server which is accessed through a
portable computer called a Portable Maintenance Aide (PMA). The PMA interfaces with the aircraft for
maintenance, servicing, and diagnostics. The Pratt & Whitney F119-PW-100 engine is maintained by
Pratt & Whitney under a contract support system in which maintenance is performed in a technical
support concept. The data can also be accessed via the Core Automated Maintenance System (CAMS)
which is a computer system used for maintenance management and trend analysis. The following
documents (which are used in lieu of 781 series forms) were thoroughly reviewed and provided
pertinent information:
• Electronic IMIS aircraft records
• CAMS print out
• Aircraft Jacket File
• Electronic IMIS engine records
• Time Change Records
• Weight and Balance Handbook
• AFT O Form 95
• Support Equipment Form 244
• Avionics and Weapons Log Books
Time Compliance Technical Orders (TCTO) status was reviewed. Of relevance to the mishap is TCTO
1F/A-22A-654, which would have replaced all three RSAs, Part Number 553500-O1, installed on MA
with Part Number 553500-02 which is an upgraded RSA. The TCTO was not yet accomplished and
parts were on order. The grounding date for this TCTO was 15 May 2006. RSA serial numbers and part
numbers were extracted from the IMIS Hardware and Software Configuration pages from the PMA and
the CAMS print out. The MA’s RSA part numbers were 553500-01, serial numbers 002, 009, 026 (Tab
U-3, U-4).
TCTO 1F/A-22-574 installed Operational Flight Program (OFP) 3.1.2 in June 2004. This OFP load
incorporated an APU restart feature. APU restart overrides the APU shutdown in order to help the
SES charge and Thermal Management System (TMS) timeline. The change was tested in Lockheed
Martin Aero (LMA) Vehicle System Simulator (VSS). RSA units were not installed during testing of the
APU re-start mode.
In addition, a new OFP, version 3.1.3, was loaded on the aircraft on 14 December 2004 under TCTO
1F/A-22-747. The changes integrated on this version did not affect the flight control system (Tab BB-9,
BB-10). The OFP was loaded and tested by LMA personnel and included regression testing with FLCS
IBIT. As of 25 December 2004, other F/A-22s with this OFP have flown 72 sorties and 104 flight hours
without any FLCS problems (Tab CC-9 through CC-11). OFP load 3.1.3 was not a factor in the mishap.
Engine S/N 730052, position #1 had 190 total operating hours and was installed on the aircraft on 22
November 2004. Engine S/N 730036, position #2 had 260 total operating hours and was installed on the
aircraft on 10 November 2004. Prior to installation on aircraft 4014, Pratt & Whitney performed a
reliability centered maintenance retrofit to engines S/N 730052 and 730036. This retrofit consisted of
18 TCTOs which required disassembly of the engines. Engine records were reviewed and were
unremarkable (Tab D-5).
b. Inspections.
All scheduled inspections were satisfactorily completed.
c. Maintenance Procedures.
With the exception of not to verifying/re-accomplishing a FLCS IBTT after engine restart,
maintenance procedures were sound and in IAW USAF directives. See 13c(4) for discussion of FLCS
IBIT.
d. Maintenance Personnel and Supervision.
Maintenance training records consisting of AF Form 623As, Career Field Education and Training
Plans, AF Form 797s, and Special Certification Rosters were reviewed for the ground crew involved
with the MA.
Preflight servicing of the aircraft, supervision, and performance of all personnel was within current
directives. All training and certifications were current. Upgrade training and testing was progressing
well for all airmen concerned. Personnel assigned to the MA were trained and had the necessary skill
level and qualifications to perform assigned duties. Inadequate technical data regarding power modes
during APU operation was a factor and is addressed in 13c(4).
e. Fuel, Hydraulic and Oil Inspection Analysis.
A review of fluid analysis records and accident fluid analyses on the aircraft, engines, and support
equipment showed normal results (Tab D-7 through D-17).
f. Unscheduled Maintenance.
The aircraft had experienced FLCS discrepancies, not including RSA failures, which were repaired.
These discrepancies were not a factor in the mishap. SES problems appeared throughout the recent
history of the aircraft and are a fleet-wide occurrence.
6. AIRCRAFT AND AIRFRAME, MISSILE, OR SPACE VEHICLE SYSTEMS
a. Condition of Systems.
The MA had flown 146 sorties of 176 sorties attempted and 150.4 hours since 29 May 2003. 95 sorties
were Code 1 (no discrepancies), 35 sorties were code 2 (flyable discrepancies), and 16 sorties were code
3 (grounding discrepancies). Reported discrepancies were recorded and all discrepancies were corrected
and properly documented (Tab CC-10, CC-11).
With the exception of the canopy and ejection seat, the MA was intact at impact. CSMU data shows
all relevant systems (control surfaces, engines, navigational systems, instruments, warning systems,
fuel, lubrication, electrical, hydraulic, pneumatic, avionics, communications, and environmental
control) were normal with the exception of the three latched RSAs (Tab CC-4 through CC-9).
RSA serial numbers and part numbers were extracted from the IMIS Hardware and Software
Configuration pages in the PMA and the CAMS print out. All three RSAs were initial production model
-101 versions with part numbers 553500-01, serial # 2, 9, and 26. LMA mishap analysis revealed that
two of the three RSAs installed on aircraft 4014 were returned to BAE systems for repair: # 9 due to
latching and # 26 due to shock sensor. They were repaired and returned to service in the -101
configuration and installed on the MA on 6 May 2003.
LMA was aware of occasional RSA latching at startup since 2002. This latching condition is due to a
pin used by the manufacture to get the unit into the test mode. This pin connects to a Universal
Asynchronous Receiver Transmitter (UART) chip. The UART is looking for a leading edge 5 VDC
square-wave trigger to get the unit into test mode. With this pin left “floating” the RSA can latch in the
test mode on power-up or cycled power.
Aircraft F/A-22A, S/N 00-4014, impacted the ground on initial takeoff at 2340:12Z/1540:12L (Tab J-
23) at 36 degrees and 14.79 minutes north and 115 degrees and 01.41 minutes west, Nellis AFB,
Nevada (Tab C-3). In addition to the loss of the aircraft, other government property that was damaged
included a BAK-12 arresting cable, a runway guide sign, a runway light and runway 03R. MA broke
into several pieces leaving a debris field scattered over the departure end of runway 03R, the runway
overrun and ground area beyond the departure end.
f. Life Support Equipment, Egress and Survival.
The MP initiated ejection at 2340:09Z/1540:09L during an uncommanded right roll at approximately
2,360 feet MSL, 500 feet AGL (Tab CC-8). The parachute deployed normally (Tab V-5.16).
The Air Force Safety Center at Kirtland AFB, NM evaluated the following Emergency Escape
Sequencing System (EESS): modified ACES II ejection seat (S/N 225014), two Thermal Battery
Ejection Seat Initiators (S/Ns 50 & 36), two Electro-Explosive Devices (EED) (S/Ns 1678 & 1670), the
Canopy Jettison Rocket Motor (S/N 19), Inertia Reel Gas Generator (S/N 11134), Inertia Reel EED (S/N
125), Primary Parachute Mortar Cartridge (S/N 9269), Back-up Parachute Mortar Cartridge (S/N
9276), Drogue Deployment Mortar Cartridge (S/N 31563), the CKU-5B/A Ejection Seat Rocket
Catapult, the Gyro Spin-up Cartridge (S/N 16443), the Vernier Rocket Motor (S/N 11490), the Harness
Release Cartridge (SIN 24443), two Drogue Severance Assemblies (S/Ns 3736 & 3701), two 1.15 SecondDelay Reefing Line Cutters, the 4.0 Second Delay Locking Cord Cutter, the Recovery Sequencer (S/N
1199-1389), and two Universal Water Activated Release System (UWARS) parachute releases (Tab H-3
through H-5).
In the opinion of the evaluators, the EESS and all pyrotechnic components functioned properly during
the ejection in the Mode I range (Tab H-6).
g. Search and Rescue.
The MP ejected safely and sustained only minor injuries. Mongoose 01, an HH-60G transitioning in
the traffic pattern, landed on runway 03R to ensure the MP had not sustained significant injury. The
MP was transported by ambulance to Mike O’Callahan Federal Hospital where he received evaluation
and treatment. The MA impacted runway 3R and was completely destroyed. The crash and resulting
debris field were confined to Nellis AFB property. Fire response was swift, and the fire was
extinguished.
h. Recovery of Remains.
There were no fatalities as a result of this mishap.
5. MAINTENANCE
a. Forms Documentation.
The 422 Aircraft Maintenance Unit (AMU) maintained the aircraft forms for the F/A-22A Raptor on
an Integrated Maintenance Information System (IMIS) database server which is accessed through a
portable computer called a Portable Maintenance Aide (PMA). The PMA interfaces with the aircraft for
maintenance, servicing, and diagnostics. The Pratt & Whitney F119-PW-100 engine is maintained by
Pratt & Whitney under a contract support system in which maintenance is performed in a technical
support concept. The data can also be accessed via the Core Automated Maintenance System (CAMS)
which is a computer system used for maintenance management and trend analysis. The following
documents (which are used in lieu of 781 series forms) were thoroughly reviewed and provided
pertinent information:
• Electronic IMIS aircraft records
• CAMS print out
• Aircraft Jacket File
• Electronic IMIS engine records
• Time Change Records
• Weight and Balance Handbook
• AFT O Form 95
• Support Equipment Form 244
• Avionics and Weapons Log Books
Time Compliance Technical Orders (TCTO) status was reviewed. Of relevance to the mishap is TCTO
1F/A-22A-654, which would have replaced all three RSAs, Part Number 553500-O1, installed on MA
with Part Number 553500-02 which is an upgraded RSA. The TCTO was not yet accomplished and
parts were on order. The grounding date for this TCTO was 15 May 2006. RSA serial numbers and part
numbers were extracted from the IMIS Hardware and Software Configuration pages from the PMA and
the CAMS print out. The MA’s RSA part numbers were 553500-01, serial numbers 002, 009, 026 (Tab
U-3, U-4).
TCTO 1F/A-22-574 installed Operational Flight Program (OFP) 3.1.2 in June 2004. This OFP load
incorporated an APU restart feature. APU restart overrides the APU shutdown in order to help the
SES charge and Thermal Management System (TMS) timeline. The change was tested in Lockheed
Martin Aero (LMA) Vehicle System Simulator (VSS). RSA units were not installed during testing of the
APU re-start mode.
In addition, a new OFP, version 3.1.3, was loaded on the aircraft on 14 December 2004 under TCTO
1F/A-22-747. The changes integrated on this version did not affect the flight control system (Tab BB-9,
BB-10). The OFP was loaded and tested by LMA personnel and included regression testing with FLCS
IBIT. As of 25 December 2004, other F/A-22s with this OFP have flown 72 sorties and 104 flight hours
without any FLCS problems (Tab CC-9 through CC-11). OFP load 3.1.3 was not a factor in the mishap.
Engine S/N 730052, position #1 had 190 total operating hours and was installed on the aircraft on 22
November 2004. Engine S/N 730036, position #2 had 260 total operating hours and was installed on the
aircraft on 10 November 2004. Prior to installation on aircraft 4014, Pratt & Whitney performed a
reliability centered maintenance retrofit to engines S/N 730052 and 730036. This retrofit consisted of
18 TCTOs which required disassembly of the engines. Engine records were reviewed and were
unremarkable (Tab D-5).
b. Inspections.
All scheduled inspections were satisfactorily completed.
c. Maintenance Procedures.
With the exception of not to verifying/re-accomplishing a FLCS IBTT after engine restart,
maintenance procedures were sound and in IAW USAF directives. See 13c(4) for discussion of FLCS
IBIT.
d. Maintenance Personnel and Supervision.
Maintenance training records consisting of AF Form 623As, Career Field Education and Training
Plans, AF Form 797s, and Special Certification Rosters were reviewed for the ground crew involved
with the MA.
Preflight servicing of the aircraft, supervision, and performance of all personnel was within current
directives. All training and certifications were current. Upgrade training and testing was progressing
well for all airmen concerned. Personnel assigned to the MA were trained and had the necessary skill
level and qualifications to perform assigned duties. Inadequate technical data regarding power modes
during APU operation was a factor and is addressed in 13c(4).
e. Fuel, Hydraulic and Oil Inspection Analysis.
A review of fluid analysis records and accident fluid analyses on the aircraft, engines, and support
equipment showed normal results (Tab D-7 through D-17).
f. Unscheduled Maintenance.
The aircraft had experienced FLCS discrepancies, not including RSA failures, which were repaired.
These discrepancies were not a factor in the mishap. SES problems appeared throughout the recent
history of the aircraft and are a fleet-wide occurrence.
6. AIRCRAFT AND AIRFRAME, MISSILE, OR SPACE VEHICLE SYSTEMS
a. Condition of Systems.
The MA had flown 146 sorties of 176 sorties attempted and 150.4 hours since 29 May 2003. 95 sorties
were Code 1 (no discrepancies), 35 sorties were code 2 (flyable discrepancies), and 16 sorties were code
3 (grounding discrepancies). Reported discrepancies were recorded and all discrepancies were corrected
and properly documented (Tab CC-10, CC-11).
With the exception of the canopy and ejection seat, the MA was intact at impact. CSMU data shows
all relevant systems (control surfaces, engines, navigational systems, instruments, warning systems,
fuel, lubrication, electrical, hydraulic, pneumatic, avionics, communications, and environmental
control) were normal with the exception of the three latched RSAs (Tab CC-4 through CC-9).
RSA serial numbers and part numbers were extracted from the IMIS Hardware and Software
Configuration pages in the PMA and the CAMS print out. All three RSAs were initial production model
-101 versions with part numbers 553500-01, serial # 2, 9, and 26. LMA mishap analysis revealed that
two of the three RSAs installed on aircraft 4014 were returned to BAE systems for repair: # 9 due to
latching and # 26 due to shock sensor. They were repaired and returned to service in the -101
configuration and installed on the MA on 6 May 2003.
LMA was aware of occasional RSA latching at startup since 2002. This latching condition is due to a
pin used by the manufacture to get the unit into the test mode. This pin connects to a Universal
Asynchronous Receiver Transmitter (UART) chip. The UART is looking for a leading edge 5 VDC
square-wave trigger to get the unit into test mode. With this pin left “floating” the RSA can latch in the
test mode on power-up or cycled power.
Approximately 20 -101 RSAs had been returned to the supplier for suspected latching. Most had been
single failures with one double RSA failure. All occurred on the ground and had been detected by builtin-
test. LMA determined the probability of a catastrophic airborne triple RSA failure to be extremely
low and developed a latching fix. This involved adding a pull-up resister to give a constant Sv signal to
the UART, ensuring the leading edge trigger will never be seen by the UART. In June 2004, Lockheed
Martin submitted a charge request which added the latching fix to an ongoing RSA upgrade plan. The
improved RSA, designated the -103, was to be gradually retrofitted in all F/A-22s through TCTO 1F/A-
22A-654 (Tab J-11, J-26).
The F/A-22 Systems Program Office (SPO) did not conduct an internal safety review of the -101
latching issue or proposed fix (Tab J-11). The SPO’s technical lead for flight controls, which include
RSAs, had no knowledge of either latching problems or the fix until after the mishap (Tab V-6.1). The
SPO has since implemented procedures to ensure a proper review of change requests (Tab V-16.5).
b. Testing.
Post-mishap testing revealed the potential for triple RSA latching. After MA’s CSMU data showed
zero output from all three RSAs, LMA Air Vehicle Systems engineers conducted tests to better
characterize what had previously been described as a random tendency to latch at startup. Extensive
power cycling tests were performed in the VSS at LMA Fort Worth. The VSS had been used in the past
to duplicate the latching characteristic of the -101 RSAs.
The -101 RSAs were installed in an F/A-22 flight control system simulator and the power switched
(on-off-on) over a range from a few milliseconds to 16 seconds. Testing determined that the probably of -
101 RSA latching depends on the length of time power is lost: less than 1.6 seconds the failure almost
always (99%) occurs, from 1.6 to about 3.5 seconds the chances are 50%, and from 3.5 seconds on the
latching do not occur. -103 RSAs did not latch during the testing (Tab J-10).
7. WEATHER
a. Forecast Weather.
Forecast weather for Nellis AFB on 20 December 2004 called for variable wind direction at six knots.
Sky conditions were to be clear with a visibility of seven statute miles. The altimeter setting was 30.05
inches. Between 20 and 21 December 2004 the only significant change in the forecast indicated winds
would be out of the north at six knots, and a scattered layer of clouds would be present at 20,000 feet
(Tab F).
b. Observed Weather.
The weather prior to Raptor O l’s scheduled takeoff revealed winds variable at three knots, visibility
40 statute miles, a scattered cloud layer at 20,000 feet and an altimeter setting of 29.90 inches. A
special observation less than two minutes after the mishap indicated the same report with the only
difference being the winds from 070 at four knots (Tab F).
c. Space Environment.
Not Applicable.
d. Conclusions.
Weather was not a factor in this mishap.
8. CREW QUALIFICATIONS
The MP is an experienced instructor pilot and dual qualified in the F-15C/D and the F/A-22. He had
2,001.2 hours in the F-15C/D, 1,111.6 of which were logged as an instructor. His total time in the F/A-
22 was 54.6 hours, and 39.6 of those were as an instructor. The MP’s possessed 2150.2 total military
flying time. He was current and qualified to perform the duties planned for the mishap sortie (Tab G-7).
Recent flight time is as follows (Tab G-4):
F/A-22 Hours Sorties
30 days 6.9 4
60 days 17.0 11
90 days 24.1 15
F-15C Hours Sorties
30 days 0.5 1
60 days 3.0 3
90 days 12.3 9
F- 15D Hours Sorties
30 days 0.0 0
60 days 0.8 1
90 days 2.1 2
9. MEDICAL
a. Qualifications.
The medical and dental records of the MP were reviewed. He was medically qualified for flying class
II duties at the time of the mishap. He was not on a waiver. The medical and dental records of
maintainers involved with the MA were reviewed with no findings (Tab X-3).
b. Health.
The MP’s and maintainers’ health were not contributing factors to this mishap (Tab X-3). Postejection
evaluation of the MP revealed only minor injuries that required no additional medical treatment (Tab
X-2).
c. Toxicology.
Post-mishap analysis of carbon monoxide, blood ethanol and drug urine screens revealed no
abnormalities. This was accomplished on the MP and the key maintainers (Tab X-3).
d. Lifestyle.
There is no evidence that unusual habits, behavior, or stress contributed to the accident (Tab V-5.3).
e. Crew Rest/Crew Duty Time.
The MP met all crew rest requirements (Tab V-5.3).
10. OPERATIONS AND SUPERVISION
a. Operations.
The 422 TES is comprised of all highly experienced Instructor Pilots. The unit requires this level of
experience to execute complex operational test scenarios. There were no indications that either
experience or operations tempo were a factor in the mishap.
b. Supervision.
Operations supervision complied with all governing directives. Supervision was not a factor.
11. HUMAN FACTORS ANALYSIS
All human factors listed in AFPAM 91-211 were reviewed. No human factors played a role in this
accident (Tab X-3).
12. ADDITIONAL AREAS OF CONCERN
There are no additional areas of concern.
13. GOVERNING DIRECTIVES AND PUBLICATIONS
a. Primary Operations Directives and Publications.
(1) T.O. 1F/A-22A-1
(2) T.O. 1F/A-22A-1CL
b. Maintenance Directives and Publications.
(1) AFI 21-101-Aerospace Equipment Maintenance Management
(2) AFI 21-101 NAFB SUP 1 dated 26 Mar 2004- Red Ball Procedures
(3) AFI 21-124--Air Force Oil Analysis Program
(4) PMA T.O. 1F/A-22A-- Launch Checklist
c. Known or Suspected Deviations from Directives or Publications.
(1) Mishap Crew.
T.O. 1F/A-22A-1 (dash-1) was the relevant governing directive for the MP. It provides both general
knowledge of the aircraft and specific operating procedures. With regards to the requirement for a
FLCS IBIT and Flight Control System power, the dash-1 is ambiguous and incorrect (Tab BB-2).
Page 2-20 contains the following warning under Normal Procedures in the Before Taxi Checklist:
“Do not take off without completing a successful FLCS IBIT. If the aircraft is
shutdown during ground operations a FLCS IBIT must be reaccomplished.”
The MP completed a successful FLCS IBIT after the original engine start. The need for FLCS IBIT
after engine restart is unclear because definition of “shutdown” is ambiguous. Eleven of 18 F/A-22
pilots surveyed believed that “shutdown” means all power is removed from the aircraft (Tab CC-3).
Because the APU remained on, the MP believed the FLCS never lost power, and he had not shutdown.
This belief is based, in part, on the following dash-1 passage and is reinforced in pilot systems
academics (Tab BB-3).
“The FLCS is operational any time 28 volt DC power is available on the
aircraft, except while on the ground when the battery is the only source of
power.”
This is incorrect. This investigation identified several aircraft power states that are not mentioned in
the dash-1. The most important is the SES power state. During the SES power state 28 volt DC power
is available from the APU, but the FLCS is not powered. The only time the aircraft enters this SES
power state on the ground is during SES start and APU cooldown. The dash-1 describes cooldown as
follows:
“During cooldown the APU continues to run, the APU RUN light flashes, the
auxiliary generator supplies power, the auxiliary hydraulic pump is
depressurized, the APU bleed air to the ECS is terminated. Cooldown only
occurs during ground operation.”
This statement further reinforces the belief that the FLCS remains powered. The statement that “the
auxiliary generator supplies power” is only partially true. While the auxiliary generator does continue
to supply power, it is only 5 kw instead of the normal 22 kw. This reduced power output is not
addressed in the dash-1. It is during this period (the SES power state) of reduced APU power output
that the FLCS is not powered. This contradicts previously cited dash-1 information.
This misunderstanding was widespread. F/A-22 pilots, maintenance personnel, Air Force Engineering
and Technical Services (AFETS) experts, and avionics/electrics technicians interviewed during this
investigation were unaware of the SES power state and the FLCS power ramifications (Tab CC-3, CC-
13, V-9.1, V-11.2, V-15.1). This investigation finally found the information in the Software
Requirements Specification for the Electrical Power System Software of the F/A-22 Weapons System
provided by LMA through the SPO.
(2) Lead Crew/Others.
Not applicable.
(3) Operations Supervision.
Operations supervision complied with all directives and publications.
(4) Maintenance.
Compliance with the F/A-22 launch checklist under normal launch conditions was evident during the
first launch attempt. The crew chief’s suspected failure to verify a FLCS IBIT after engine restart
warrants review. At this point in the operation, the crew chief is expected to return to the launch
checklist. However, for this scenario, no directives specify a starting point on the checklist. According to
the Chief of Quality Assurance and four Crew Chief trainers, the crew chief uses judgment and aircraft
system knowledge to determine which events need to be reaccomplished. The belief that a system was
previously checked and had not changed would explain the failure to re-accomplish the check (Tab CC-
12).
The testimony of four crew chief trainers, four electrical/avionics specialists, five Air Force
Engineering and Technical Services representatives, and two Field Training Detachment instructors
with F/A-22 experience demonstrates a common belief in continuous FLCS power during engine
shutdown with APU running. Technical data, checklists, and maintenance training courseware do not
address potential for power interruption (Tab CC-13).
14. NEWS MEDIA INVOLVEMENT
On the day of and the day after the mishap, local media interest was high. All of the local news
channels ran stories, and articles were written in two local newspapers. The Associated Press also
printed reports not only covering the mishap, but also the F/A-22 fleet’s return to the air. Public affairs
released two statements. The first simply reported the incident, and the second advertised a public
address to be accomplished by Major General Stephen Goldfein, Commander, Air Force Warfare Center
(AFWC). At approximately 1430L on 22 December 2004, Maj Gen Goldfein briefed members of the
media on the mishap and the investigation process. Since then, media interest has been minimal (Tab
DD).
3 February 2005 STANLEY T. KRESGE, Colonel, USAF
President, Accident Investigation Board
single failures with one double RSA failure. All occurred on the ground and had been detected by builtin-
test. LMA determined the probability of a catastrophic airborne triple RSA failure to be extremely
low and developed a latching fix. This involved adding a pull-up resister to give a constant Sv signal to
the UART, ensuring the leading edge trigger will never be seen by the UART. In June 2004, Lockheed
Martin submitted a charge request which added the latching fix to an ongoing RSA upgrade plan. The
improved RSA, designated the -103, was to be gradually retrofitted in all F/A-22s through TCTO 1F/A-
22A-654 (Tab J-11, J-26).
The F/A-22 Systems Program Office (SPO) did not conduct an internal safety review of the -101
latching issue or proposed fix (Tab J-11). The SPO’s technical lead for flight controls, which include
RSAs, had no knowledge of either latching problems or the fix until after the mishap (Tab V-6.1). The
SPO has since implemented procedures to ensure a proper review of change requests (Tab V-16.5).
b. Testing.
Post-mishap testing revealed the potential for triple RSA latching. After MA’s CSMU data showed
zero output from all three RSAs, LMA Air Vehicle Systems engineers conducted tests to better
characterize what had previously been described as a random tendency to latch at startup. Extensive
power cycling tests were performed in the VSS at LMA Fort Worth. The VSS had been used in the past
to duplicate the latching characteristic of the -101 RSAs.
The -101 RSAs were installed in an F/A-22 flight control system simulator and the power switched
(on-off-on) over a range from a few milliseconds to 16 seconds. Testing determined that the probably of -
101 RSA latching depends on the length of time power is lost: less than 1.6 seconds the failure almost
always (99%) occurs, from 1.6 to about 3.5 seconds the chances are 50%, and from 3.5 seconds on the
latching do not occur. -103 RSAs did not latch during the testing (Tab J-10).
7. WEATHER
a. Forecast Weather.
Forecast weather for Nellis AFB on 20 December 2004 called for variable wind direction at six knots.
Sky conditions were to be clear with a visibility of seven statute miles. The altimeter setting was 30.05
inches. Between 20 and 21 December 2004 the only significant change in the forecast indicated winds
would be out of the north at six knots, and a scattered layer of clouds would be present at 20,000 feet
(Tab F).
b. Observed Weather.
The weather prior to Raptor O l’s scheduled takeoff revealed winds variable at three knots, visibility
40 statute miles, a scattered cloud layer at 20,000 feet and an altimeter setting of 29.90 inches. A
special observation less than two minutes after the mishap indicated the same report with the only
difference being the winds from 070 at four knots (Tab F).
c. Space Environment.
Not Applicable.
d. Conclusions.
Weather was not a factor in this mishap.
8. CREW QUALIFICATIONS
The MP is an experienced instructor pilot and dual qualified in the F-15C/D and the F/A-22. He had
2,001.2 hours in the F-15C/D, 1,111.6 of which were logged as an instructor. His total time in the F/A-
22 was 54.6 hours, and 39.6 of those were as an instructor. The MP’s possessed 2150.2 total military
flying time. He was current and qualified to perform the duties planned for the mishap sortie (Tab G-7).
Recent flight time is as follows (Tab G-4):
F/A-22 Hours Sorties
30 days 6.9 4
60 days 17.0 11
90 days 24.1 15
F-15C Hours Sorties
30 days 0.5 1
60 days 3.0 3
90 days 12.3 9
F- 15D Hours Sorties
30 days 0.0 0
60 days 0.8 1
90 days 2.1 2
9. MEDICAL
a. Qualifications.
The medical and dental records of the MP were reviewed. He was medically qualified for flying class
II duties at the time of the mishap. He was not on a waiver. The medical and dental records of
maintainers involved with the MA were reviewed with no findings (Tab X-3).
b. Health.
The MP’s and maintainers’ health were not contributing factors to this mishap (Tab X-3). Postejection
evaluation of the MP revealed only minor injuries that required no additional medical treatment (Tab
X-2).
c. Toxicology.
Post-mishap analysis of carbon monoxide, blood ethanol and drug urine screens revealed no
abnormalities. This was accomplished on the MP and the key maintainers (Tab X-3).
d. Lifestyle.
There is no evidence that unusual habits, behavior, or stress contributed to the accident (Tab V-5.3).
e. Crew Rest/Crew Duty Time.
The MP met all crew rest requirements (Tab V-5.3).
10. OPERATIONS AND SUPERVISION
a. Operations.
The 422 TES is comprised of all highly experienced Instructor Pilots. The unit requires this level of
experience to execute complex operational test scenarios. There were no indications that either
experience or operations tempo were a factor in the mishap.
b. Supervision.
Operations supervision complied with all governing directives. Supervision was not a factor.
11. HUMAN FACTORS ANALYSIS
All human factors listed in AFPAM 91-211 were reviewed. No human factors played a role in this
accident (Tab X-3).
12. ADDITIONAL AREAS OF CONCERN
There are no additional areas of concern.
13. GOVERNING DIRECTIVES AND PUBLICATIONS
a. Primary Operations Directives and Publications.
(1) T.O. 1F/A-22A-1
(2) T.O. 1F/A-22A-1CL
b. Maintenance Directives and Publications.
(1) AFI 21-101-Aerospace Equipment Maintenance Management
(2) AFI 21-101 NAFB SUP 1 dated 26 Mar 2004- Red Ball Procedures
(3) AFI 21-124--Air Force Oil Analysis Program
(4) PMA T.O. 1F/A-22A-- Launch Checklist
c. Known or Suspected Deviations from Directives or Publications.
(1) Mishap Crew.
T.O. 1F/A-22A-1 (dash-1) was the relevant governing directive for the MP. It provides both general
knowledge of the aircraft and specific operating procedures. With regards to the requirement for a
FLCS IBIT and Flight Control System power, the dash-1 is ambiguous and incorrect (Tab BB-2).
Page 2-20 contains the following warning under Normal Procedures in the Before Taxi Checklist:
“Do not take off without completing a successful FLCS IBIT. If the aircraft is
shutdown during ground operations a FLCS IBIT must be reaccomplished.”
The MP completed a successful FLCS IBIT after the original engine start. The need for FLCS IBIT
after engine restart is unclear because definition of “shutdown” is ambiguous. Eleven of 18 F/A-22
pilots surveyed believed that “shutdown” means all power is removed from the aircraft (Tab CC-3).
Because the APU remained on, the MP believed the FLCS never lost power, and he had not shutdown.
This belief is based, in part, on the following dash-1 passage and is reinforced in pilot systems
academics (Tab BB-3).
“The FLCS is operational any time 28 volt DC power is available on the
aircraft, except while on the ground when the battery is the only source of
power.”
This is incorrect. This investigation identified several aircraft power states that are not mentioned in
the dash-1. The most important is the SES power state. During the SES power state 28 volt DC power
is available from the APU, but the FLCS is not powered. The only time the aircraft enters this SES
power state on the ground is during SES start and APU cooldown. The dash-1 describes cooldown as
follows:
“During cooldown the APU continues to run, the APU RUN light flashes, the
auxiliary generator supplies power, the auxiliary hydraulic pump is
depressurized, the APU bleed air to the ECS is terminated. Cooldown only
occurs during ground operation.”
This statement further reinforces the belief that the FLCS remains powered. The statement that “the
auxiliary generator supplies power” is only partially true. While the auxiliary generator does continue
to supply power, it is only 5 kw instead of the normal 22 kw. This reduced power output is not
addressed in the dash-1. It is during this period (the SES power state) of reduced APU power output
that the FLCS is not powered. This contradicts previously cited dash-1 information.
This misunderstanding was widespread. F/A-22 pilots, maintenance personnel, Air Force Engineering
and Technical Services (AFETS) experts, and avionics/electrics technicians interviewed during this
investigation were unaware of the SES power state and the FLCS power ramifications (Tab CC-3, CC-
13, V-9.1, V-11.2, V-15.1). This investigation finally found the information in the Software
Requirements Specification for the Electrical Power System Software of the F/A-22 Weapons System
provided by LMA through the SPO.
(2) Lead Crew/Others.
Not applicable.
(3) Operations Supervision.
Operations supervision complied with all directives and publications.
(4) Maintenance.
Compliance with the F/A-22 launch checklist under normal launch conditions was evident during the
first launch attempt. The crew chief’s suspected failure to verify a FLCS IBIT after engine restart
warrants review. At this point in the operation, the crew chief is expected to return to the launch
checklist. However, for this scenario, no directives specify a starting point on the checklist. According to
the Chief of Quality Assurance and four Crew Chief trainers, the crew chief uses judgment and aircraft
system knowledge to determine which events need to be reaccomplished. The belief that a system was
previously checked and had not changed would explain the failure to re-accomplish the check (Tab CC-
12).
The testimony of four crew chief trainers, four electrical/avionics specialists, five Air Force
Engineering and Technical Services representatives, and two Field Training Detachment instructors
with F/A-22 experience demonstrates a common belief in continuous FLCS power during engine
shutdown with APU running. Technical data, checklists, and maintenance training courseware do not
address potential for power interruption (Tab CC-13).
14. NEWS MEDIA INVOLVEMENT
On the day of and the day after the mishap, local media interest was high. All of the local news
channels ran stories, and articles were written in two local newspapers. The Associated Press also
printed reports not only covering the mishap, but also the F/A-22 fleet’s return to the air. Public affairs
released two statements. The first simply reported the incident, and the second advertised a public
address to be accomplished by Major General Stephen Goldfein, Commander, Air Force Warfare Center
(AFWC). At approximately 1430L on 22 December 2004, Maj Gen Goldfein briefed members of the
media on the mishap and the investigation process. Since then, media interest has been minimal (Tab
DD).
3 February 2005 STANLEY T. KRESGE, Colonel, USAF
President, Accident Investigation Board
STATEMENT OF OPINION
F/A-22 ACCIDENT
20 December 2004
Under 10 U.S.C. 2254(d) any opinion of the accident investigators as to the cause of, or the factors
contributing to, the accident set forth in the accident investigation report may not be considered as
evidence in any civil or criminal proceeding arising from an aircraft accident, nor may such information
be considered an admission of liability of the United States or by any person referred to in those
conclusions or statements.
1. OPINION SUMMARY:
The Accident Investigation Board conducted a complete and thorough investigation of this mishap. In
addition to interviews with the MP, maintenance personnel, and SPO personnel, the investigation
reviewed technical evaluations, CSMU data, technical orders, and applicable academic material.
The cause of this mishap, supported by clear and convincing evidence, was an inoperative FLCS,
resulting from a power interruption to the three RSAs, which rendered the MA uncontrollable. The MP
was unaware of this condition because he did not perform an IBIT, the only means of identifying the
problem. Failure to perform the IBIT was the result of ambiguous and incorrect Technical Orders and a
mistaken belief in continuous RSA power availability.
2. DISCUSSION OF OPINION:
RSAs are a critical component of the F/A-22 flight control system. Their function is to provide angular
acceleration (yaw, pitch, and roll) feedback so the flight control computer knows how the aircraft is
responding to flight control inputs. During the mishap, the MA’s three RSAs were “latched” and
therefore not providing the necessary feedback. Simulator testing showed that in this condition, the
aircraft is not flyable.
The MP was unaware of the flight control problem. The MA’s flight control surfaces operated
normally on the ground. The FLCS automatic PBIT is designed to compare the output of the three
RSAs to determine if the output of one differs from the other two. If so, it warns the pilot. PBIT cannot
identify that all three RSAs are latched because their output (zero) is the same. The only way to
discover the problem is to perform a FLCS IBIT.
The MP performed a FLCS IBIT after engine start during the first attempt of the mishap sortie. The
FLCS was operating normally until the MP shutdown both engines to allow maintenance personnel to
service the Stored Energy System (SES). The MP ensured the APU remained running while the
engines were off. During engine shutdown, all three RSAs latched. The MP did not perform a FLCS
IBIT after restarting engines.
The F/A-22 Pilot Checklist requires a FLCS IBIT after engine start. T.O. 1F/A-22A-1 contains the
following warning: “Do not take off without completing a successful FLCS IBIT. If the aircraft is shut
down during ground operations a FLCS IBIT must be reaccomplished.” The MP was aware of this
guidance. However, because the APU was running while the engines were off, he believed that the
FLCS received continuous power and did not consider the aircraft to be “shutdown”. These beliefs were
based on F/A-22 academic training, T.O. systems description, and are shared by most F/A-22 personnel
we surveyed.
The F/A-22 maintenance T.O. requires the crew chief to verify FLCS IBIT after engine start. There is
clear and convincing evidence that the crew chief did so after the first attempt engine start. It is
uncertain what was said after the engine restart. However, there is no clear consensus on what the
T.O. requires following the engines shutdown-APU running-engines restart sequence. Most
maintenance personnel interviewed considered the FLCS IBIT optional because of continuous APU
power.
Unfortunately, FLCS power is interrupted during APU cooldown when engines are below 53% RPM.
There is clear and convincing evidence of the following events. The MP selected TMM in an attempt to
correct the SES problem. With TMM selected, the APU enters cooldown mode when engines are shut
down. OFP 3.1.2, installed in the MA in June 2004, gave the pilot the ability to override APU cooldown.
The MP shut down the engines, experienced APU cooldown, then selected override. During the brief
period when the APU was in cooldown and the engines were both below 53%, power to the RSAs was
interrupted.
The MA contained 5VV50035-101 (-101) configuration RSAs. Post-mishap testing determined the
probability of -101 latching depends on the length of time power is lost. If less than 1.6 seconds,
latching almost always (99%) occurs. Between 1.6 to 3.5 seconds, the probability is 50%. Greater than
3.5 seconds, latching does not occur. There is substantial evidence that the MA’s RSAs experienced a
less than one second power loss, which almost guaranteed all three RSAs would latch.
Prior to the mishap, approximately 20 -101 RSAs had been returned to the supplier (BAE Systems)
for suspected latching during ground power up. The failures have mainly been single RSAs, with one
occurrence of dual failure. All occurred on the ground and detected by built in tests. A catastrophic
triple RSA failure was considered nearly impossible. In June 2004, Lockheed Martin Aeronautics
incorporated a fix to the latching problem into the new -103 RSA, to be eventually retrofitted in all F/A-
22s. Based on the information available at the time, the retrofit plan was reasonable.
3 February 2005 STANLEY T. KRESGE, Colonel, USAF
President, Accident Investigation Board
F/A-22 ACCIDENT
20 December 2004
Under 10 U.S.C. 2254(d) any opinion of the accident investigators as to the cause of, or the factors
contributing to, the accident set forth in the accident investigation report may not be considered as
evidence in any civil or criminal proceeding arising from an aircraft accident, nor may such information
be considered an admission of liability of the United States or by any person referred to in those
conclusions or statements.
1. OPINION SUMMARY:
The Accident Investigation Board conducted a complete and thorough investigation of this mishap. In
addition to interviews with the MP, maintenance personnel, and SPO personnel, the investigation
reviewed technical evaluations, CSMU data, technical orders, and applicable academic material.
The cause of this mishap, supported by clear and convincing evidence, was an inoperative FLCS,
resulting from a power interruption to the three RSAs, which rendered the MA uncontrollable. The MP
was unaware of this condition because he did not perform an IBIT, the only means of identifying the
problem. Failure to perform the IBIT was the result of ambiguous and incorrect Technical Orders and a
mistaken belief in continuous RSA power availability.
2. DISCUSSION OF OPINION:
RSAs are a critical component of the F/A-22 flight control system. Their function is to provide angular
acceleration (yaw, pitch, and roll) feedback so the flight control computer knows how the aircraft is
responding to flight control inputs. During the mishap, the MA’s three RSAs were “latched” and
therefore not providing the necessary feedback. Simulator testing showed that in this condition, the
aircraft is not flyable.
The MP was unaware of the flight control problem. The MA’s flight control surfaces operated
normally on the ground. The FLCS automatic PBIT is designed to compare the output of the three
RSAs to determine if the output of one differs from the other two. If so, it warns the pilot. PBIT cannot
identify that all three RSAs are latched because their output (zero) is the same. The only way to
discover the problem is to perform a FLCS IBIT.
The MP performed a FLCS IBIT after engine start during the first attempt of the mishap sortie. The
FLCS was operating normally until the MP shutdown both engines to allow maintenance personnel to
service the Stored Energy System (SES). The MP ensured the APU remained running while the
engines were off. During engine shutdown, all three RSAs latched. The MP did not perform a FLCS
IBIT after restarting engines.
The F/A-22 Pilot Checklist requires a FLCS IBIT after engine start. T.O. 1F/A-22A-1 contains the
following warning: “Do not take off without completing a successful FLCS IBIT. If the aircraft is shut
down during ground operations a FLCS IBIT must be reaccomplished.” The MP was aware of this
guidance. However, because the APU was running while the engines were off, he believed that the
FLCS received continuous power and did not consider the aircraft to be “shutdown”. These beliefs were
based on F/A-22 academic training, T.O. systems description, and are shared by most F/A-22 personnel
we surveyed.
The F/A-22 maintenance T.O. requires the crew chief to verify FLCS IBIT after engine start. There is
clear and convincing evidence that the crew chief did so after the first attempt engine start. It is
uncertain what was said after the engine restart. However, there is no clear consensus on what the
T.O. requires following the engines shutdown-APU running-engines restart sequence. Most
maintenance personnel interviewed considered the FLCS IBIT optional because of continuous APU
power.
Unfortunately, FLCS power is interrupted during APU cooldown when engines are below 53% RPM.
There is clear and convincing evidence of the following events. The MP selected TMM in an attempt to
correct the SES problem. With TMM selected, the APU enters cooldown mode when engines are shut
down. OFP 3.1.2, installed in the MA in June 2004, gave the pilot the ability to override APU cooldown.
The MP shut down the engines, experienced APU cooldown, then selected override. During the brief
period when the APU was in cooldown and the engines were both below 53%, power to the RSAs was
interrupted.
The MA contained 5VV50035-101 (-101) configuration RSAs. Post-mishap testing determined the
probability of -101 latching depends on the length of time power is lost. If less than 1.6 seconds,
latching almost always (99%) occurs. Between 1.6 to 3.5 seconds, the probability is 50%. Greater than
3.5 seconds, latching does not occur. There is substantial evidence that the MA’s RSAs experienced a
less than one second power loss, which almost guaranteed all three RSAs would latch.
Prior to the mishap, approximately 20 -101 RSAs had been returned to the supplier (BAE Systems)
for suspected latching during ground power up. The failures have mainly been single RSAs, with one
occurrence of dual failure. All occurred on the ground and detected by built in tests. A catastrophic
triple RSA failure was considered nearly impossible. In June 2004, Lockheed Martin Aeronautics
incorporated a fix to the latching problem into the new -103 RSA, to be eventually retrofitted in all F/A-
22s. Based on the information available at the time, the retrofit plan was reasonable.
3 February 2005 STANLEY T. KRESGE, Colonel, USAF
President, Accident Investigation Board
:L 楼上还不睡？:D
原帖由 死亡之翼 于 2007-12-15 04:15 发表
:L 楼上还不睡？:D
:D 周末，周末
那个总结看到我头爆啊~
]]
原帖由 yf23 于 2007-12-15 04:01 发表
这个供电中断导致角加速度传感器自己闭锁了，此后输出为0。这种情况因为3个角加速度传感器输出一致，所以飞控的周期性BIT没有发现
这个bug比较ft……
那现在是否解决了这个BUG？
]]
那就是说，这其实也不是一个真正BUG，只是飞行员的操作不当，或者说设计的不太方便而已？
事先没有想到会发生这种情况，所以飞行手册里对于这种情况要不要再次作全系统自检表达模糊，以后应该在这方面会有明确的规定
非常感谢
PS：23兄是北航的还是沈航的？
PS：23兄是北航的还是沈航的？