SH36, vicinity East Midlands UK, 1986
From SKYbrary Wiki
|On 31 January 1986, at night during an instrument approach, a Shorts SD3-60 operated by Aer Lingus Commuter experienced a loss of control attributed to airframe ice accretion. When fully established on the Instrument Landing System (ILS), the aircraft began a series of divergent rolling oscillations which were accompanied by a very high rate of descent. The crew was able to regain control of the aircraft just before contact with power cables and subsequent impact with terrain near East Midlands Airport.|
|Actual or Potential
|Loss of Control, Weather|
|Type of Flight||Public Transport (Passenger)|
|Intended Destination||East Midlands Airport|
|Take off Commenced||Yes|
|ENR / APR|
|Location - Airport|
|Airport vicinity||East Midlands Airport|
|Tag(s)||Approach not stabilised|
|Tag(s)||In Flight Airframe Icing|
|Damage or injury||Yes|
|Aircraft damage||Hull loss|
|Causal Factor Group(s)|
On 31 January 1986, at night during an instrument approach, a Shorts SD3-60 operated by Aer Lingus Commuter experienced a loss of control attributed to airframe ice accretion. When fully established on the Instrument Landing System (ILS), the aircraft began a series of divergent rolling oscillations which were accompanied by a very high rate of descent. The crew was able to regain control of the aircraft just before contact with power cables and subsequent impact with terrain near East Midlands Airport.
During the descent and before entering the cloud tops at about FL60, the crew switched on the aircraft’s anti-ice system, which heats the windscreen, engine air intakes, propellers, static air vents and pitot probes, but in accordance with normal operating procedure, they did not use the wing and tail de-icing system. At this time the freezing level (0°C isotherm) was at 1000 feet, the temperature at FL60 was minus 6°C and the air was saturated. Whilst in cloud, which according to the non-handling pilot was particularly dense, ice thrown from the propellers was heard striking the side of the aircraft fuselage, and it was suggested that the propeller rpm be increased to expedite the removal of the ice. Both pilots state that, at some time during the descent, they visually checked the aircraft for ice but saw none. Nevertheless, several other flights during that evening have since reported the occurrence of severe icing.
Having leveled at 3000 feet, still in and out of cloud, the aircraft was directed by radar to intercept the ILS and was fully established on the glideslope and centerline by 10 NM, at which point the final descent was initiated. A normal approach was established and continued, past the outer marker beacon situated at 3.9 NM from touchdown, down to around 1000 feet above the runway threshold height. The crew state that up to this point they had neither experienced any significant turbulence nor observed any ice forming on the aircraft.
As the aircraft descended through 1000 feet, it suddenly rolled very sharply to the left without any apparent cause. With the application of corrective aileron and rudder the aircraft rolled rapidly right, well beyond wings level position. This alternate left and right rolling motion continued with the angles of bank increasing for some 30 seconds, causing the commander to believe that the aircraft might roll right over onto its back. The angles of bank then gradually decreased. During this period and subsequent few seconds the aircraft established a very high descent rate approaching 3000 feet/min. Subsequently, with the aid of full engine power, the airspeed increased and the rate of descent was arrested just as the aircraft struck an 11KV power cable.
In Section 220.127.116.11 dedicated to the Aircraft De-icing System, the Report states:
The leading edges of both wings outboard of the engines, both horizontal stabilisers and the fin are provided with pneumatically inflating de-icing “boots”. These function by selection in 1 minute or 3 minute cycles and are normally, apart from test purposes, selected ON only when a moderate build-up of ice is visually detected. These were not selected ON in the accident aircraft, and were subsequently found in the OFF position.
Through the flight, the de-icing boots were managed according to the standard operating procedure. In the situation which the crew found themselves during the approach there was apparently no requirement to operate the boots.
From experience gained over many years, de-icing boots are not used until there is a visible build-up of ice on the leading edges of the wings and tail. The reason for not doing so is that a thin layer of rime ice may not be brittle enough to fracture and may not shed effectively.
As to Cause of the accident, the Report states:
- The accident most probably occurred as a result of the effects of a significant accumulation of airframe ice degrading the aircraft’s stability and control characteristics, such that the crew were unable to maintain control…Turbulence and or downdraught may have contributed to the accident.
- Other contributory factors were the difficulty in detecting clear ice at night on the SD3-60 which resulted in the airframe de-icing system not being used.
- The delay in application of go-around power may have also contributed to the accident.
The Safety Recommendations in the report included the recommendation that:
- Pneumatically inflated wing and tail de-icing systems be exercised during the final approach to land, when an aircraft is flying, or has recently flown in conditions conductive to the accretion of ice.
For the complete list of the Safety Recommendations also addressing institutional and manufacturer’s issues see Further Reading.
- Accident and Serious Incident Reports: WX
- Ice Formation on Aircraft
- Aircraft and In Flight Icing Risks
- Piston Engine Induction Icing.
- AT73, en route, Roselawn IN USA, 1994
- ATP, en-route, Oxford UK, 1991
- DH8D, en-route, South West of Glasgow UK, 2006
- AT43, en-route, Folgefonna Norway, 2005
For further information see the full accident Report published by AAIB UK.