Stall at the Approach Bottom . . .

By Richard N. Aarons
May 01 , 2012

Over the years we’ve seen a relatively large number of approach accidents involving aerodynamic stalls during the last moments of instrument approaches. The airplanes involved — often turboprops — are typically flown by experienced, professionally qualified pilots who somehow get behind the situation as they transition from instruments to visual reference. More often than not, the operation is single pilot.

This month, we’ll look at the investigation into the loss of Mitsubishi MU-2B-60 N80HH. The pilot and three passengers were killed when the aircraft crashed during an instrument approach to Runway 07 at Lorain County Regional Airport (LPR) in Elyria, Ohio. This accident is in no way extraordinary, and that is what makes it worth reviewing, I think. Perhaps in examining the investigator’s findings, you’ll come away with the sense that the last moments of an instrument approach are always the most critical.

The surface weather observation at LPR at 1353 on Jan. 18, 2010, indicated the following conditions: winds, 240 deg. at 9 kt.; visibility, 2 mi. in mist; overcast at 500 ft.; temperature, -1C; dew point, -3C; altimeter, 29.93. Weather Depiction Charts for 1100 and 1400 depicted an extensive area of IFR conditions over the region. The closest VFR conditions were over 200 mi. south of Elyria. The freezing level was at the surface with freezing temperatures at all levels aloft.

Pilots reported an extensive overcast layer extending over Ohio with bases from 100 to 1,200 ft. and tops at 2,200 to 3,800 ft. There were 12 reports of light rime to mixed type icing, and four reports of light to moderate intensity icing conditions in clouds below 3,000 ft.

The accident flight departed Gainesville (Fla.) Regional Airport at 1100 on an IFR flight plan. The 30-year-old ATP-rated pilot was in the left seat and was conducting the flight as a single-pilot operation. A private-pilot-rated passenger occupied the right seat, and two passengers were seated in the cabin. This was an FAR Part 91 operation. The en route flight phase was without incident. Investigators used ATC recordings, radar track analysis and eyewitness testimony to put together the details of the approach phase.

The airplane approached LPR on a heading of 325 deg. At 1335:51, ATC informed the pilot that he could expect radar vectors for the ILS Runway 7 approach. Runway 7 is 5,002 ft. long; airport elevation is 794 ft. The ILS Runway 7 approach course is 070 deg.; glideslope/glidepath intercept altitude is 2,400 ft.; straight-in landing minimums were 994 ft. MSL DH with one-half mile visibility. Circling approach minimums were 1,240 ft. MSL MDA with 1-mi. visibility.

At 1345:53, ATC informed the pilot that he was 4.5 miles from RAWLS, the final approach fix for the approach, and instructed him to turn right to a heading of 050 deg. and maintain 2,600 ft. until established on the localizer. The flight was cleared for the ILS Runway 07 approach. The pilot acknowledged the clearance. Radar track data indicated that the airplane flew through the inbound course of 070 deg. and continued on a 055-deg. heading.

At 1347:03, ATC instructed the pilot to turn to 090 deg. to intercept the inbound course. The ATC controller also stated, “I didn’t adjust for the wind there.”

At 1347:19, ATC instructed the pilot to turn to 100 deg. and asked the pilot if he wanted to continue the approach or accept radar vectors to get reestablished on the inbound course, since he would be intercepting the inbound course near or at RAWLS. The pilot elected to continue the approach.

At 1348:27, ATC instructed the pilot to change radio frequency to LPR’s advisory frequency. The pilot acknowledged the frequency change.

At 1349:33, the pilot advised ATC that he was executing a missed approach. Radar track data indicated that the airplane’s altitude during the approach was never lower than 1,500 ft. MSL. The decision height for the ILS Runway 7 approach was 994 ft. MSL.

At 1350:29, ATC instructed the pilot to climb to 2,500 ft. MSL and turn left to a heading of 280 deg. for radar vectors for the ILS Runway 07 final approach course. The pilot requested that the controller extend the outbound leg to provide more time to get established on the inbound course. The radar track data indicated that the airplane was about 11 mi. from the airport before it turned inbound to intercept the inbound localizer course.

At 1358:18, ATC instructed the pilot to turn left to 100 deg. and maintain 2,600 ft. MSL until established on the localizer. The flight was cleared for the ILS Runway 7 approach. The pilot acknowledged the clearance.

At 1401:12, ATC instructed the pilot to change to the advisory frequency. The pilot acknowledged the frequency change. Radar track data indicated that the airplane’s altitude increased to about 3,000 ft. MSL when it turned inbound and intercepted the localizer. The altitude was about 2,200 ft. MSL when it crossed RAWLS. (The crossing altitude at RAWLS is depicted as 2,263 ft. MSL).

The airplane continued inbound and the altitude continued to decrease. The radar track data indicated that the airplane had descended to 1,300 ft. (506 ft. AGL) about 1 mi. from the runway. The airplane’s heading had started to go left of the centerline when it was at about 1,400 ft. MSL, and it continued to drift left until the last radar return. The last radar return indicated that the airplane was about 1,000 ft. left of the centerline and 206 ft. AGL. The linear distance from the last recorded radar return to the initial impact point was about 750 ft.

A witness — waiting at the airport for the airplane to arrive — heard the Unicom transmissions and looked toward the approach end of the runway to watch the arrival. He spotted the airplane as it descended out of the clouds. It was in a nose-low attitude, rolling to the right into a steep right turn with the wings at almost a 90-deg. position relative to the ground. The airplane was “definitely out of control” when he saw it.

The witness saw a “huge cloud of snow” billow up as the airplane struck the ground. When the snow cloud cleared, he observed the airplane wreckage at the west end of the airport property. It all happened very fast, he told investigators — “in the blink of an eye.” He reported that the airplane’s landing light was off as it descended.

Another witness told investigators he heard the airplane as it was approaching his house near the airport. He looked out a window and saw the airplane about 150 ft. above the ground. Using an airplane model to describe the airplane’s flight profile, he indicated that he observed the airplane in a 60-deg., nose-low attitude with an 80-deg. right bank.

The airplane had impacted a field within the airport’s boundary about 2,150 ft. west of the Runway 07 threshold and about 720 ft. north of the extended centerline of Runway 07. The wreckage path was 194 ft. long and was oriented on a heading of 100 deg. magnetic. The wings and landing gear separated from the fuselage. The cockpit cabin had partially separated from the rest of the fuselage during the impact sequence, but the flight control cables were not severed. The empennage remained attached to the fuselage. There was no post-impact fire.

The pilot and passengers were dead on the scene of multiple blunt force trauma. There was no indication of pre-impact physical impairments and the FAA’s Civil Aeromedical Institute tests were negative for the list of drugs and pharmaceuticals normally checked.

Inspection of the wreckage revealed that the landing gear had been extended and the flaps were set to 5 deg. The pitch trim indicator was about 15 deg. nose-up, the rudder trim indicator was about 1-2 deg. right, and the aileron trim was neutral. The trim surfaces corresponded to the indicated trim settings.

The flight control cables were examined for continuity. The elevator and rudder push-pull rods and cables exhibited continuity from the flight controls to the control surfaces. The wing spoiler cables had continuity from the control yoke to the mixer box located in the wing center section. The push-pull tubes from the mixer box to the spoilers were broken and exhibited impact damage. The attach points of the push-pull tubes to the spoiler bell cranks exhibited continuity.

The airplane’s two Honeywell TPE331-10-511M turboprop engines were disassembled and examined at the factory. Both exhibited rotational scoring of the first-stage compressor impeller shroud. The leading edges of the first-stage impeller blades were bent opposite the direction of rotation. There was rotational scoring through 360 deg. on the second-stage compressor housing impeller shroud and rotational scoring on the shroud line of all second-stage compressor impeller blades. The engines also exhibited metal spray deposits adhering to the suction side of the second-stage turbine stator vanes and to the suction side of the second-stage turbine blades. Both had rotational scoring damage to the sun gear and propeller shaft. In other words, they were operating at impact. The propellers and gearboxes showed no pre-impact damage. Special studies of the autopilot turned up no pre-impact studies.

A Cessna Citation landed at LPR 10 to 15 min. prior to the MU-2’s approach. The Citation crew reported they flew the ILS Runway 07 approach, then circled to Runway 25. The pilot said they entered the clouds at 3,000 ft. on the descent and received radar vectors for the approach. Visibility was good above the clouds and there was no turbulence in the clouds.

The Citation pilots leveled off at the MDA and flew the right-hand circling pattern at 1,300 to 1,350 ft. MSL (about 500 to 550 ft. AGL). He said they had 3 mi. visibility and remained clear of clouds during the circle-to-land maneuver. The airplane’s anti-icing and deicing equipment were on during the approach. He observed about one-eighth inch of ice on the nose of the airplane when they pulled it into the hangar. He reported that the airplane was in the clouds for about 2 to 2.5 min. The weather was essentially as forecast.

The MU-2 pilot held an ATP certificate for single-engine and multiengine land airplanes, and helicopters. He was a certified flight instructor with single-engine airplane, multi-engine airplane and helicopter ratings; and he was an instrument instructor in airplanes and helicopters. The pilot’s latest first-class medical certificate had been issued on Nov. 29, 2007. He had accumulated over 2,000 flight hours, 1,285 of which were in multiengine aircraft with 1,250 hr. in the MU-2. He had 231 hr. in helicopters. He recorded 290 hr. of flight in actual instrument conditions and had flown 180 hr. in the MU-2 within the preceding 12 months, and had flown 30 hr. of instrument flying within the preceding 12 months.

The pilot’s training records — obtained from the SimCom Training Center — indicated that he obtained initial MU-2 simulator training in October 2002. He returned to SimCom for recurrent MU-2 simulator training on a yearly basis. On Jan. 28, 2009, the pilot attended the SimCom Training Center and received a certificate signifying that he had satisfactorily completed a Special Federal Aviation Regulation (SFAR) 108 compliant MU-2 Recurrent course for the MU-2B-60 model. The pilot was scheduled to return to SimCom for recurrent MU-2 simulator training on Jan. 25-27, 2010.

The owner of the airplane told investigators the pilot was competent and qualified to fly the MU-2 single pilot. He and the accident pilot routinely flew together, and they would switch pilot and copilot responsibilities. They routinely flew in instrument conditions and had often flown IFR approaches in actual instrument conditions. He stated that the accident pilot was a good instrument pilot and that there were no issues with his flying or his technique. The pilot had worked for the owner of the airplane for about 13 years.

The owner reported that neither he nor the accident pilot used the autopilot while flying instrument approaches for landing. He stated that typically, during the approach, the flaps were set to 5 deg. at 175 kt. and then to 20 deg. at 155 kt. He stated that the approach is normally flown at 150 kt. with 20 deg. of flaps. At the bottom of the approach the airplane would be slowed to 115 kt., then the flaps would be moved to 40 deg. The owner reported that he had never seen the accident pilot use 5 deg. of flaps below 500 ft. in visual flight conditions. He stated that the accident pilot always used 20 deg. of flaps from the initial approach point (IAP) to the missed approach point (MAP). The owner stated that it was a “mystery” to him why the flaps had been set at 5 deg. during the accident approach.

The pilot-rated passenger, employed by the owner to maintain the accident airplane and a helicopter, held an A&P rating and a private pilot certificate with a single-engine land rating. During his third-class medical examination on Oct. 10, 2008, the pilot reported that his total flight time was 190 hr. The airplane owner reported that the pilot-rated passenger was not performing the duties of copilot during the flight.

NTSB performance specialists plotted radar tracks and ATC recordings in an effort to define the aircraft flight path, ground track, ground speed, rate of climb and ATC communication event-time history. They determined that the calibrated airspeed was about 130 ±10 kt. on the final approach, but subsequently decreased to about 95 to 100 kt. during the 20-sec. period prior to loss of radar contact. During the final approach and descent, but prior to the airspeed decay, the calculated flight path angle was about -3 ±1 deg., the calculated bank angle was about 0 ±10 deg. and the estimated angle of attack ranged from 4 to 6 deg.

During the airspeed decay period, the estimated AOA increased by 6 to 8 deg., ending up at 10 to 14 deg. (depending on the assumed engine power setting). According to the Mitsubishi MU-2B-60 AFM, the flaps -5 minimum control speed was 99 kt. and the wings-level power-off stall speed at the accident aircraft weight was about 91 kt.

The MU-2B-60 AFM Approach Checklist calls for flaps lowered to 5 deg. as speed drops below 175 kt., at which time the gear can be lowered. (The AFM also recommends that the operator set the flap switch to 5 deg. and wait for an indication before going to the 20-deg. position.)

The Before Landing Checklist states that the flaps are set to 20 deg. (below 155 KCAS) or 40 deg. (below 120 KCAS). The AFM contains performance charts for “Landing Approach Speed — Flaps 20 Deg.” and “Landing Approach Speed — Flaps 40 Deg.,” but it does offer a chart for a landing approach speed using 5 deg. of flaps.

The AFM also states, “Use of 40-deg. flaps for landing considerably restricts the go-around capability should an engine failure occur in the approach or landing phase. During landing, do not select 40-deg. flaps when operating in icing conditions. The FAA has determined that ice accumulations on the tail plane of many aircraft may result in a reduced down force on the horizontal stabilizer when full flaps are used. This reduced down force may result in the aircraft pitching nose down.”

The general instructions for the MU-2 approach procedures state: The minimum airspeed when using 5 deg. of flaps is 140 kt. (25-30% torque). Check gear down, flaps 20 deg. when approaching glideslope intercept. The minimum airspeed with 20 deg. of flaps is 120 kt. Perform landing check (approximately 25% torque). When landing is assured, check flaps 20 deg. (or 40 deg. of flaps below 120 kt.). At the threshold, fly Vref airspeed (20% torque). At touchdown, retard the power levers to flight idle stop.

Just why the pilot flew the approach with flaps set at 5 deg. — intentionally or unintentionally — is not known. Ultimately, the NTSB determined the probable cause of the accident was simply “The pilot’s failure to maintain adequate airspeed during the instrument approach, which resulted in an aerodynamic stall and impact with terrain.”

Unfortunately, there is little unique about this accident. Single-pilot minimums approaches always require maximum situational awareness on the part of the pilot flying. While it is unknown whether the rated-pilot passenger was participating in any way in this phase of the flight, there have been many cases in which distractions developed when a single pilot was being “assisted” or even offering training to a passenger in the right seat.

Single-pilot minimums approaches are just about the most-demanding non-emergency maneuvers required of high-performance aircraft pilots. These operations require planning, use of practiced routine procedures and broad situational awareness. As primary flight instructors are fond of remarking to their students, “You’ve got to stay on your game and fly the airplane all the way to the chocks.”


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: