ANC94GA041
ANC94GA041

HISTORY OF FLIGHT

On March 21, 1994, at approximately 1115 Alaska standard time, an amphibian Grumman G21A airplane, N741, collided with glacial terrain under controlled conditions. The airplane is registered to the State of Alaska, Department of Public Safety, and was being operated as a public use aircraft. Departing Juneau at 1030, the airplane was on a VFR flight plan in visual meteorological conditions at the time of the accident. The location of the accident site was at the 9,650 foot level of the Margerie Glacier, located on the west side of Mount Fairweather, in British Columbia, Canada. A commercial pilot and two persons were on board at the time, two were uninjured and the third, a passenger in the right pilot's seat, received minor injuries. The airplane sustained structural damage consistent with substantial damage.

The Grumman amphibian airplane, owned and operated as a public-use aircraft was assigned to the Alaska Department of Public Safety, and hangared at the Juneau Airport. The manager responsible for the division's aviation program oversight in his position as, the Deputy Commissioner of Public Safety for the State of Alaska, was the pilot-in-command. On the day of the accident, the aircraft was being flown on a mission to transport the pilot and an Alaska State Trooper Lieutenant to Anchorage and the Director of the State's Highway Safety Planning Department to Yakutat. Additionally, eighteen numbered and boxed items were on board described as evidence to be delivered to Anchorage in conjunction with a state criminal proceeding.

Investigators found that the aircraft had impacted in controlled flight, with wings level, in a climb, on a snowfield above the Margerie Glacier at a location approximately one mile, 070 degrees (true) from the 15,300 foot peak of Mount Fairweather. Hull marks in the snow field indicated that the first touchdown was at approximately the 8900 foot elevation and the airplane came to a stop at approximately 9650 feet. Retractable wing floats were extended and the snow field showed wing float marks following the second hull skip mark at approximately the 9000 foot elevation. The retractable main wheels were extended equally approximately beyond flush with the hull. Snow was packed behind the extended main wheels. It could not be determined whether the retractable wing floats remained extended for the flight or were jarred down on impact with the snow field.

The saddle of the snow field above the impact point was measured by recovery aircraft altimeter to be approximately 11,000 feet in elevation extending south to north between Mount Fairweather and Mount Root (12,680'). Accurate elevation or contour lines are not shown on the Juneau Sectional Aeronautical Chart for the snow fields between the peaks. Both of these peaks are "boundary peaks" marking the U.S. and Canadian borders. (Refer to Wreckage and Impact Information.) The aircraft came to rest inside Canadian territory.

INJURIES TO PERSONS

The commercial pilot and the passenger in the rear of the airplane were uninjured. A second passenger in the airplane's copilot seat received back injuries not requiring hospitalization and as such were classified as minor injuries.

DAMAGE TO AIRCRAFT

The aircraft sustained little structural damage on the snow field during the gear-up landing. Investigators were unable to determine if any damage had occurred to the bottom of the amphibian's hull. Damage to the right horizontal stabilizer at the center of the right elevator hinge point involved the structure of that control surface and is classified as substantial damage. Subsequently the aircraft sustained further damage during salvage operations.

PERSONNEL INFORMATION

The pilot in command of the aircraft was Claude E. Swackhammer, who held a Commercial Pilot's Certificate and a Second Class Medical Certificate, with limitations of glasses for near and distant vision required. He said that he had flown "about 90 hours in the past year, in the Navajo and the Goose." (Piper PA-31-350 Navajo airplane and the Grumman G-21A).

Records of the department indicate the pilot flew 88.7 hours in the G-21A and 6.8 hours in the PA-31 during the 1993 and 1994 fiscal years. No flights were recorded after November 27, 1993 until the day of the accident. The Aircraft Manual of the Department of Public Safety requires all Department pilots to meet regulations of the Federal Aviation Administration for recent flight experience under FAR 61.57 and have flown within 90 days in order to act as pilot-in-command.

The pilot-in-command of the accident airplane served as the manager of the Department of Public Safety aircraft fleet.

Records provided by the State of Alaska indicated that the pilot had flown the accident aircraft most recently on November 27, 1993 on the occasion of a six month instrument check, and an annual multi-engine land and sea proficiency flight check. That check flight report indicated that 26 separate items were examined and were marked "satisfactory." No events or maneuvers were marked otherwise.

The check airman, as Aircraft Supervisor of the State of Alaska, Department of Public Safety, at Anchorage, is a civilian employee of the state and reports directly to Deputy Commissioner Swackhammer. State records showed that the same check airman had conducted satisfactory proficiency flight checks of the accident pilot in single and multi-engine airplanes on 11/27/93, 6/3/93, 5/27/93, 4/2/93, 11/4/92, and 10/8/92. No check flights were recorded between 1992 and 1985 where records of satisfactory check flights are once again recorded by the same check airman dating back to 1979. No unsatisfactory grades for maneuvers tested were recorded.

Investigators asked the accident pilot if he believed himself to be sufficiently current and proficient in the aircraft. He replied that he believed himself to be. He said that on the day of the accident that he did not use the aircraft checklists because he, "knew the airplane."

During an interview with the accident pilot, investigators learned that a preflight inspection of the airplane on the morning did not include a check of the weight and balance of the airplane or an examination or draining of the bilges for water. The pilot said that the airplane contained "about 300 gallons of fuel." He also said that he estimated the cargo to be "about 300 pounds, maybe a little more." No weight and balance calculations or forms were found in the aircraft.

Investigators were told that Federal Aviation Regulations applicable to preflight actions of a pilot in command (14 CFR Part 91.103 (b)(2)) regarding preflight duties apply to Department pilots under Section XIII of its own regulations.

The accident pilot told investigators that the flight, until the accident event was routine and conducted in visual meteorological conditions. He said that he flew the airplane so as to transit the glacier on Mount Fairweather, but that "a downdraft" prevented the airplane from climbing sufficiently to cross the mountain.

The accident pilot estimated the altitude needed to cross the glacier saddle to be 9000 feet, based on his knowledge of the sectional chart. (see Juneau Sectional Aeronautical Chart)

The pilot told investigators that he recognized from the airspeed indication and the passing terrain that the airplane was not climbing and as he neared the surface of the upsloping icefield he "decided to put it in straight ahead." He recalled the airspeed "around seventy (70 knots) and the VSI (vertical speed indicator) about 500 (feet per minute) down."

He was asked about engine power and settings. He said that the aircraft was performing normally, and that he had been cruising at about 10,000 feet enroute, using an engine RPM setting of 1800. The pilot told investigators that he "pushed the props up to about 1950 (RPM)" and that he had used "full throttle from around 7000 feet." The pilot said that he used a carburetor heat setting of "plus 20" (degrees centigrade).

During the interview the pilot was asked if he recalled that the tower (Juneau Air Traffic Control Tower) had reminded him, approximately five minutes after takeoff, that he had neglected to raise his landing gear. He said that he recalled that.

The pilot described raising the airplane's landing gear by the electrical switch, but did not recall finishing the retraction sequence (up until snug) by the hand crank.

In an interview of the check airman, following the accident, investigators asked if the time between the pilot's last flight and the accident flight was consistent with the department's currency requirements. To that question the check airman told investigators that he believed that it was not. The check airman said that unless otherwise noted that the "state followed the FAR's part 91" (14 CFR Part 91, Federal Aviation Regulations).

The check airman was asked who authorized the flight to be conducted, and the check airman said that, "he (the Deputy Commissioner) can authorize his own flights."

Following the accident flight, and prior to a department re-check proficiency flight or re-evaluation, the Safety Board learned that the accident pilot approved an additional flight, acting as pilot in command and sole occupant of the airplane, on a training flight from Juneau to Sitka and return. Following that flight, the check airman conducted a re-check flight on the accident pilot.

Investigators reviewed the check flight record of November 27, 1993 with the check airman. On that flight check sheet, subjects of aircraft operation, powerplants, performance and loading were checked "satisfactory." Under the heading of "AIR WORK," cockpit procedure, and engine controls were listed as "satisfactory." Comments on the November 27 check flight report indicated that the pilot had demonstrated "excellent knowledge, skill and judgement in all areas tested." The check airman was asked if the testing had included high altitude airwork or performance. He said that it did not.

The passenger in the right seat of the airplane during the accident flight was a state official who was being transported on official business to Yakutat. He received back injuries not requiring hospitalization beyond 24 hours. A second passenger, an Alaska State Trooper Lieutenant, in the rear of the airplane was on official business enroute to Anchorage. He was not injured.

AIRCRAFT INFORMATION

The Grumman G-21 (Goose) amphibian, S/N B-97, was manufactured for the U.S. Navy as a JRF-5 utility amphibian transport airplane. A weighing record of the airplane, dated June 27, 1991, showed that the airplane had a weight of 6717 pounds which included a 41 pound life raft. The maximum gross weight of the airplane is 9200 pounds, 9000 and 8800 pounds, the limitations being different, depending upon several locations in the flight manual.

Section XVIII of the Departments regulations, p.23, limits the Grumman Goose to a maximum weight of 8800 for takeoff and 8000 for landing. The regulations state, in part, "These limits shall not be exceeded. . . "

The flight handbook notebook for the accident airplane, provided investigators by the State of Alaska, listed 9000 pounds as the maximum weight for the G-21A airplane in a supplemental data sheet at the front of the book. On a weight and balance sheet dated 6/27/91, the aircraft gross weight was listed as 9200 pounds. In the approved flight manual portion of the handbook the gross weight limit is listed as 8800 pounds. Several unindexed supplements with varying limitations and instructions, along with a photocopy of a Pilot's Handbook of Flight Operating Instructions (15 June 1945) were provided to investigators as the operating instructions. Weight and balance calculations provided performance data through weights limited to a maximum of 8800 pounds.

Investigators calculated the accident airplane to have weighed 10,202.8 pounds at takeoff from Juneau on the day of the accident, based upon the aircraft empty weight, the weight of cargo, passengers and full fuel (336 gallons.) That weight exceeded the maximum gross weight of the aircraft authorized by Department regulation by 1403 pounds.

Officials of Aero Services, Inc., contracted to provide aviation fuel to the state's airplane, provided a record to the NTSB that indicated that the airplane, prior to the accident flight, was "topped off."

Investigators estimated the weight of the airplane at the time of the collision with terrain to be 9,802.8 pounds (estimating 400 pounds of fuel burned enroute.) Investigators also estimated the landing weight of the airplane, as planned at the interim stop at Yakutat, would have been approximately 9603 pounds or 1603 pounds over the maximum allowable weight by Department regulation. (Section XVIII)

The handbook provided other airplane flight characteristics and limitations of performance at 8800 pounds germane to the conduct of the flight and the accident performance calculations. As set forth in the handbook, these include:

The single-engine service ceiling is 6000 feet. Best multi-engine rate of climb is 105 knots. Best multi-engine angle of climb is 95 knots. Normal climb power is 33.5 MP (inches of manifold pressure) and 2200 RPM. This is also the maximum climb power and maximum cruise power.

Recommended safe single-engine speed is 86 knots. This is the velocity of minimum control (VMC). This speed provides the minimum speed at which full rudder deflection will provide controlled flight with maximum power on the critical engine (left). Flight at speeds below VMC would result in loss of aircraft control as a result of either engine failure.

The handbook does not provide "L over D" curve charts (lift over drag performance charts) indicating at what airspeed "back side of the curve" flight power requirement occur. The "best glide" speed of 90 knots calibrated airspeed, flaps up (clean), approximates the least power required for a given airspeed and 66 knots is listed as the velocity of stall (clean) at 8800 pounds gross weight. Investigators calculated that 70 knots seen before impact was approximately 25 knots slower than "best angle of climb" and that the airplane was settling ("mushing") prior to full aerodynamic stall. Since the Grumman G-21A aircraft wing is designed (See Flying Characteristics, JRF-5 Flight Manual, p.17), so that the onset of boundary layer separation (aerodynamic stall of the wing) originates at the root, the ailerons and thus roll control about the longitudinal axis may be maintained while the aircraft approaches full aerodynamic stall while settling (mush). No numerical flight test data has been recorded at the weight of airplane (1000 pounds over gross) at the time of the accident.

The flight manual's "estimated operating limits - engine power curve, Figure 7, indicated that the maximum engine horsepower at an RPM setting of 1950 is limited to approximately 320 HP at 5000 feet. With the throttles full forward, manifold pressure and horsepower reduces during the climb, to about 280 horsepower at 9000 feet, the manifold pressure reducing to about 26 inches.

The flight manual indicates the use of climb power in a schedule which calls for 2200 RPM, horsepower of approximately 335 per engine and a resultant manifold pressure of about 27.5 inches.

These calculations provide for this nominal performance if the carburetor air temperature was followed as per the schedule (minus 3 degrees centigrade at 9000 feet). (Standard temperature lapse rate, ISO).

The use of 20 degrees of carburetor heat (no icing conditions present) provides an engine-performance density altitude of approximately 12,000 feet at 9000 foot pressure altitude. Increased carburetor air temperature is proscribed for nominal operation of the engine unless in icing conditions as it degrades engine performance.

At the power setting and carburetor air temperature setting used by the pilot on the day of the accident (full available manifold pressure and 1950 RPM), the aircraft reached maximum performance at approximately 5000 feet mean sea level. A climb beyond that altitude was accomplished at the expense of airspeed and increased angle of attack.

The Preflight Checklist (Before Start) indicate that the first item to be accomplished is: "1. Drain Hull."

The After Takeoff check list showed the following first three items to be accomplished:

1. Full power to 300 feet (36.5 HG 2300) 2. Climb power (33.5 HG 2000) 3. Gear rest of the way up additional items

Investigators found that the aircraft's main wheels were not stowed in flight but were extended to the position remaining after electrical retraction was accomplished. The G-21A Flight Manual requires that following electrical retraction, the landing gear be cranked up until snug with the hull and visually checked up. The main landing gear, when extended out of its wheel well, can be seen to be not stowed. The performance charts do not quantify the amount of parasitic aerodynamic drag produced by the wheels in the unstowed position, nor can the amount of degradation in climb performance be determined.

The contents of the airplane, not listed on the airplane's weighing record, were inventoried, transported and weighed at the Juneau airport on the day following the accident. The contents of the airplane weighed 830.8 pounds.

The scale used for the weighing of these items were calibrated by an FAA certificated A&P and AI and observed by the NTSB IIC. The scale was calibrated to within .4 pounds throughout its range.

Investigators estimated the bilges to have contained a minimum of 80 pounds of water. The water was frozen and could not be removed at the accident site. Only that portion of the bilge that was amidship in the hull was examined. Fore and aft portions of the hull were not examined at the site.

METEOROLOGY

The nearest surface observation to the accident scene were made at Yakutat, Alaska, approximately 50 miles north northwest of the site. At 1047 Alaska standard time that weather observation showed Yakutat to have broken ceilings of 12000 feet and winds from the southeast at 8 knots. Approximately one hour later, and 35 minutes after the accident time, the weather was essentially the same, and the wind had dropped off to 4 knots.

Upper air winds at Yakutat and Juneau at 9000 feet were from the same direction (140 degrees true) at 10 and 14 knots respectively.

Winds from that direction provided a relative tail wind, blowing "up glacier" as the aircraft navigated the Margerie Glacier to the west north west. The air mass encountered prior to the impact at the 9000 foot level was consistent with lifting energy rather than falling energy. In order that the aircraft to have encounted a "down draft," wind directions of southwest through northwest would have been present. Information from meteorological observations do not support weather criteria to be factors in this accident scenario.

WRECKAGE AND IMPACT INFORMATION

Based on snow field scars, the airplane touched down in a wings-level, nose up, attitude, consistent with a aerodynamic stall-mush. After approximately four skips the airplane slid to a stop 40 degrees left of the original heading after approximately 1000 feet of ground run. The airplane rested on its left wing float.

ADDITIONAL DATA/INFORMATION

Under Alaska statute and interagency agreement, the State of Alaska requested that the National Transportation Safety Board assume investigative lead in the accident. The Alaska Department of Public Safety was granted party status and assisted in the investigation.

Canadian authorities (Transportation Safety Board of Canada, Vancouver) delegated the investigative authority for the accident to the U.S. NTSB Investigator-in-Charge, Northwest Field Office, Anchorage, at 2030 on 21 March, 1994. They declined on-scene participation.

Use your browsers 'back' function to return to synopsis
Return to Query Page