CV-22 Crash Caused by Flight Manual
On April 9, 2010, a new Air Force CV-22 tiltrotor crashed in Afghanistan. Four GIs were killed and 16 injured, most seriously. A controversial investigation concluded that the cause of the crash was unknown, mostly because the flight incident recorder disappeared, and the surviving copilot claimed amnesia. However, video recordings clearly show the aircraft dropped like rock just after its tiltrotors rotated upright. It is obvious that the CV-22 crashed simply because it attempted to hover when it could not.
The pilot, Major Randell B. Voas, was one of the Air Force's most experienced CV-22 pilots, but V-22s rarely hover with a heavy load. He encountered one of the tiltrotor's unique dangers. If a helicopter lacks the power to hover, it cannot take-off. However, a tiltrotor can perform a long, rolling take-off and gain lift with its wings to fly with a higher gross weight than it can hover land. Major Voas assumed that he could land based on aircraft performance charts and CV-22 simulators programmed with those capabilities.
The official V-22 hover limit is 5400 ft. at a typical gross weight (fuel and payload) of 47,000 lbs. That is listed by both Boeing and Bell as a spec on their websites, and is embarrassing because similar-sized modern helicopters with longer, flat rotors can hover twice as high with the same payload. The V-22 Hover Out of Ground Effect (HOGE) "hover" chart is based on old performance goals, even after OPEVAL testing showed much lower performance. Nevertheless, these "goals" were published in the actual flight manuals and used to program simulators, lest critics obtain copies and use them to expose the failure of the tiltrotor concept. Aviation experts who evaluated the V-22 during testing were appalled that the pilot manual was filled with inflated performance data, which was also used to program V-22 simulators.
Proving the Critics Wrong
Tiltrotor limitations are known in U.S. Army Aviation. The V-22 program began with the U.S. Army, which dropped the idea after learning that flat helicopter rotors and twisted aircraft propellers are quite different. A tiltrotor uses a compromise "proprotor" that provides half the efficiency in either mode. Just before the CV-22s arrived in Afghanistan, the Army dominated Special Operations Command voiced its displeasure with the CV-22's performance. An article in the March 28, 2010 issue of "Aviation Week", quoted Army Special Operations Colonel Clay Hutmacher, of the 160th Special Operations Squadron, explaining why no more CV-22s were desired:
“Above 4,000 ft.,
there’s a significant [hovering] limitation on the V-22,” he said. Tiltrotor
engineers concede that while the V-22 hovers well in many situations, the
special twist and size of its “proprotors” leave it unable to carry as much
useful load pound-for-pound as most helicopters hovering in similar conditions.
Certainly, everyone in the small CV-22 community read that. Since Army helicopters were able to land at mid-altitude in Afghanistan, newly arrived CV-22 crews with their new, ultra-expensive tiltrotors were under pressure to prove their value. One week after Col. Hutmacher criticized the CV-22's poor hover ability above 4000 ft., someone sent three fully-loaded CV-22s to hover land at 5226 ft. on their first Afghan combat mission. The first to land crashed because it couldn't terminate in a hover.
Let's See If the Major Can Land
Landing in an area where an enemy may appear is dangerous. The basic tactic is to swoop in, drop off passengers, and quickly depart. There were three CV-22s on this mission, so the best tactic was for all three to land simultaneously. In this mission, the CV-22 with the squadron's most experienced pilot went first, while the other CV-22s circled overhead. They knew this was a dangerous landing near the CV-22's hover limit, so they wisely stayed airborne to watch events unfold.
After the first CV-22 piloted by Major Voas crashed and caught fire, the two CV-22s overhead did not swoop down and land nearby to assist. They circled overhead burning off more fuel, then decided to land at different spots a mile from the crash site, with a short, rolling landing. This is hard proof that the CV-22 pilots knew they couldn't hover land.
The Official Mishap Report
The lengthy official report can be read online. The conclusion is remarkable, since it was inconclusive. Despite the availability of the report, the media never bothered to decipher the true cause of the accident, which becomes evident by reading a few key passages. These are quoted below in blue, with this author's comments inserted in black. They are not in sequence but grouped by issue for clarity:
"The MA impacted the ground at approximately 2009:15Z on 8 April 2010 at N 32° 4.7479’ E 066° 47.1487’ at 5,226 feet MSL. (Tab M-7, Tab Z-27) This was 0.23 nm short of the intended LZ. The MA had all three landing gear down and locked, the nacelles nearly vertical, and the speed was approximately 75 KGS. (Tab Z-27, Tab JJ-3 thru JJ-4)"
"The MA’s take off gross weight was 45,485 lbs, which resulted in out of ground effect (OGE) plus five percent. (Tab K-6) This would allow a 50 foot hover with an extra five percent power margin."
These figures are based on the textbook HOGE limit of 5400 ft at 47,000 lbs, but that is ISA (in still air), and a mild wind existed at the landing site. In addition, the passengers were not weighed, and were assumed to be 240 lbs. each, including their gear. Finally, jet engines produce slightly less power as their blades wear down, and don't produce 100% of the expected power. However, lead investigator and author of the report, Brigadier General Donald Harvel never expressed concern about the thin 5% margin of error that winds, payload underestimates, and engine wear can erase. Moreover, he never considered that the CV-22 HOGE chart was inaccurate, even after Colonel Hutmacher publicly warned that a V-22 can't hover well above 4000 ft. General Harvel wrote:
"After a thorough, careful and complete investigation, I ruled out multiple causes including: enemy action, brownout, vortex ring state, mid-air collision, loss of hydraulic system, electrical failure, drive shaft failure, swashplate actuator mount failure, flight control failure, thrust control lever (TCL) rigging, avionics failure, and crew physiological events."
"I was unable to determine, by clear and convincing evidence, the cause of this mishap."
General Harvel considered many factors, but ignored the HOGE performance issue, even after the CV-22 crashed because it couldn't hover.
"At 0.5 nm from the LZ, the MA had slowed to 128 KGS. (Tab JJ-4) The normal speed should have been 60 to 70 KGS. (Tab BB-31) A very excited discussion occurred in the cockpit seconds prior to impact. (Tab V-60.28, V-60.34) A member of the MCR counted down "10, 9, 8, 7" rapidly and at "7" the aircraft impacted the ground. (Tab V-20.4)"
General Harvel noted that once the tiltrotors moved to the full upright position, the crew was distraught that the aircraft sank quickly and they could not halt its descent, hitting the ground seven seconds earlier than the planned glide path (pictured) and before it cleared a small drainage channel, which collapsed its nose wheel. It is obvious to any pilot that it was too heavy to safely land. It could not hover, and couldn't even stay airborne with its rotors upright and the help of 100 knots of air flowing under its wings.
"Unanticipated high rate of descent
Video analysis indicated the MA maintained a relatively steady, but high rate of descent, beginning at 0.4 nm from the LZ and 150 feet AGL, until the main landing gear (MLG) impacted the ground. (Tab Z-3, Z-5, Z-27, Tab JJ-4) At one point during the final seven seconds, the descent rate was approximately 1,800 feet per minute. (Tab JJ-4) The normal descent rate during this phase of the approach should have been 200 feet per minute. (Tab V-24.7 thru V-24.8, Tab BB-31)
The high altitude environment, temperature, winds, and the MA’s weight, airspeed, and nacelle configuration affected the descent rate. (Tab K-8, Tab R-77, Tab V-13.6, V-60.35, Tab II-6, Tab JJ-11, JJ-16) Simulator flights confirmed that a go-around would have been possible if the descent rate was caught early enough, the pilot performed appropriate control inputs, and the engines were performing normally. (Tab JJ-11 thru JJ-17) The preponderance of the evidence supports the conclusion that the MA’s high rate of descent was abnormal for this phase of the LVA."
General Harvel didn't question the data used by the simulator, even though an actual flight showed it faulty. The CV-22 began to descend nine times faster than a CV-22 simulator told him would occur. An obvious cause would be that the HOGE performance chart was wrong, so the simulators were wrong.
"The excited conversation and the countdown show they were aware of their descent rate, attempting to correct it, but unable to do so, because of abnormal engine response.
It is also possible that the MCR [crew] made the decision to go-around at this point, but did not have the altitude or power necessary to accomplish this maneuver with the descent rate established. In this case, the MCR would have selected the correct course of action, but were unable to accomplish it."
"The touchdown was smooth enough to leave the ramp, MLG tires and doors, and the underside intact. (Tab R-23, Tab V-7.35) The absence of large impressions in the sand, at the initial impact point, indicates an intentional, perfectly executed, and straight, roll-on landing. (Tab Z-3, Z-5) The greater weight of credible evidence shows the MP would have only executed a roll-on landing if he believed the MA did not have sufficient power to execute a go-around.
I determined that only an aircraft performance issue could completely account for the MP’s decision to execute a roll-on landing. During a rapid descent, it is unlikely that this very experienced and competent MP would have chosen to execute a roll-on landing on rough terrain if he had power available to go-around and set up for another approach."
Exactly, an aircraft performance issue! The only two possibilities were the engines not providing full power, or that the aircraft was outside its performance envelope attempting to hover at 5226 ft. with the weight of 16 combat equipped passengers aboard. General Harvel carefully examined the possibility of an engine problem, yet little evidence existed. For unknown reasons, he refused to consider the alternative, despite Colonel Hutmacher's recent warning that a CV-22 can't hover well above 4000 ft. The crew discussed the possibility of "abnormal engine response" because instruments showed both engines were at full power yet their CV-22 was dropping like a rock, and the HOGE chart and simulator had told them that would not happen. General Harvel concludes his report:
"The absence of the Flight Incident Recorder, the Vibration Structural Life and Engine Diagnostics control unit, and the right engine prevented the board from obtaining clear and convincing evidence of the cause of this mishap."
The failure to recover the Flight Incident Recorder (FIR) was blamed on two comical excuses. It was announced that no one knew where they were located in the CV-22, or that they didn't exist, even though CV-22 squadron personnel were present with the rescue force. It was also claimed that rescue personnel didn't have time, although they spent two hours at the crash site with no signs of enemy activity. Crews returned the next day from the nearby airbase, but it was claimed the FIR was "missing", stolen by Afghans. The CV-22 wreckage was promptly destroyed, lest the Taliban reverse engineer the design and build their own V-22s.
Although General Harvel spent several days in Afghanistan at that CV-22 base, he was unable to collect the needed facts. He could have flown the exact same mission in a CV-22 with a crew to see if it could land safely. Without the 16 passengers, the CV-22 would be some 4000 lbs. lighter and should be able to hover land at that 5225 ft. site. If his CV-22 was unable to hover land, or did so with a hard landing, his investigation would be complete. The report never mentions why this common sense test was not performed at the site or back in the USA. The report never explains why the investigating General was flown to the crash site in a Russian built helicopter flown by contractors, rather than a CV-22.
1. The CV-22 crews were under pressure to prove the value of their aircraft from an airbase well above sea level.
2. The very experienced CV-22 pilot was surprised to lose altitude so quickly during final approach and couldn't abort. This is not as simple as a fixed wing aircraft that can instantly apply more power to abort. The tiltrotor was at maximum power and couldn't hover, so it would have to tilt its rotors forward to gain airspeed, which requires several seconds. Major Voas attempted a tricky rolling land, which was partially successful.
3. The other two CV-22s circled overhead to see if his risky landing was successful. The crash confirmed that a hover landing was not possible, so they dumped fuel (i.e. weight) and found other spots to perform a rolling landing.
4. All the evidence from the crash site was destroyed before the investigation began, including the flight recorder.
5. The pilot and flight engineer died. The co-pilot, Captain Brian Luce, survived but provided few details, claiming amnesia. Tab Q of the report notes that what he did say is not for public release. Luce quickly recovered from his leg and back injury, returned to flight status, and was promoted to major, only to suffer injuries during another CV-22 crash in June 2012.
6. The investigation did not consider that the CV-22 was unable to hover land at that altitude since the charts and simulators indicated that it could be done, albeit with only a 5% margin of safety, which can be less with degraded engine power, winds, and if the passenger weight is greater than estimated. No doubts about the CV-22's hover performance ability were pursued except the possibility of an engine was not fully operating.
7. General Harvel's report was delayed, questioned by higher officers, and he left the Air Force shortly after it was released. He was ordered not to speak to the press or with family members of the deceased.
8. There is some evidence that a loss of power in one engine contributed to the crash. This would explain why the experienced pilot did not abort the landing.
9. The mishap cause was limited to either blaming the pilots, or an engine malfunction. The possibility that others are to blame for inserting inflated performance data in the flight manual was never addressed.
A CV-22 was sent to land where it could not. The pilots were fooled by exaggerated performance data in the aircraft manual and simulator. This needs to be corrected immediately, and an investigation launched to see who was responsible. This may have already occurred, but like much of the official report, it may be unfit for public release.
Carlton Meyer editor@G2mil.com