the online flight safety digest
the gremline digest — fatal stall or spin accidents
the online flight safety digest
the gremline digest — fatal stall or spin accidents
A Study of Fatal Stall or Spin Accidents to UK Registered Light Aircraft
1980 — 2008
The General Aviation Safety Council (GASCo) has recently published the results of a detailed study of all stall or spin fatal accidents suffered by UK registered light aircraft for the period 1980 to 2008. Copies of this report are being circulated to national and international safety organisations. Further copies are available from the GASCo office at Rochester Airport, Chatham, Kent ME5 9SD at a cost of £7 plus postage.
The report analyses 110 accidents to determine the factors affecting each accident. Among the findings were that:
· The percentage of fatal accidents due to stall/spin has remained almost unchanged during the period.
· There has been a major change in the pattern of accidents during the period. Early in the period there was a high percentage of accidents during low aerobatics/displays/beat-ups which were all but eliminated towards the end of the period. Conversely, in the 1980s there was a very low percentage of accidents following engine or airframe problems but since 2000 it has become the trigger for half of the accidents.
· There are marked differences in accident rates per 100,000 hours between aircraft types. Also, there are many types with a significant number on the UK register that have zero stall/spin accidents. For instance, the Piper PA28 has the greatest number of hours of all types and every one of the accidents was on the earlier constant chord wing version.
· Early in the period the stall/spin accident rate for aircraft under 600kg max gross weight was very much greater than that for heavier types. Since 2000 the figures have improved markedly, but are still considerably greater.
· The accident rate for the Slingsby T67 was throughout the period much greater than any other certified type and has been treated as a special case.
· With only one accident involving the Cessna 152 in 2.5 million flying hours, its record is similar to the tapered wing PA28, whereas the Cessna 150K, L and M models have had eleven accidents in one million hours. Investigating this, Brunel University, Uxbridge have carried out flight trials on several Cessna 152, 150L and 150M aeroplanes using calibrated data recording equipment to determine control loads etc. This showed significant differences, e.g. the stick force to stall the aeroplane was greater in the C152 thus providing a better alert to the pilot.
· Turning finals was long held to be a high risk point but the climb-out has now replaced this.
· It is a matter of concern to the GASCo study group that in 22% of the accidents there was an instructor on board as a crew member, although not always performing a training function. There has only been one fatal accident to a solo student since 1987.
Apart from the Slingsby T67, no significant problem has been revealed in spin recovery.
There were 359 fatal accidents to UK registered aeroplanes of 5,700kg maximum gross weight or less between 1980 and 2008. One hundred and thirty (36%) of these were found to be due to the pilot failing to maintain control resulting in a stall or a spin. During the 29 year period studied it was found that stalling, sometimes resulting in a spin, in visual flight conditions was the biggest single factor in fatal accidents, resulting in 216 deaths.
Accidents and Aircraft Excluded from the Analysis
Only fatal accidents were considered as these were able to be precisely defined and have been fully investigated by the UK Air Accidents Investigation Branch (AAIB) or by the foreign authority where the accident occurred outside the UK.
In order to concentrate the analysis on the sort of aeroplanes flown or owned by private pilots, whilst providing a reasonably sized data sample, of the 130 fatal stall/spin accidents the following have been excluded from the analysis:
· Twin-engine aeroplanes, (11 stall/spin accidents, most were loss of control after failure of one engine).
· Warbirds, including Harvards, (6 stall/spins, most frequently in an airshow/practice environment).
· Jet powered aeroplanes (3 stall/spin accidents).
· Microlights, as defined in the ANO, gliders (but motor gliders are included) helicopters and gyrocopters.
· The Slingsby T67, (8 fatal accidents) was excluded from the main numerical analysis but was studied as a special case. (See Appendix 2 of the full Report).
There were a few ‘unusual’ stall/spin accidents with unique circumstances including three cases of pilots under the influence of alcohol or drugs, a case of carbon monoxide poisoning, an unintended first flight and a pilot who had a heart attack in flight while suffering from a known major but undeclared medical condition. These are included in the analysis.
Analysis and Discussion
The meat of the report covers ten A4 pages that merit careful study by all those engaged in operating light aircraft as pilots, instructors or supervisors. This part of the report contains analysis and discussion of factors related to stall/spin accidents under no less than 14 headings. They are:
1. Annual Trend
2. Aeroplane Type
3. Weight Category
4. Activity and Circumstances
7. Stall Leading to Spin
8. Type of Stall Warning
10. Pilot Experience
11. Disorientation and Distraction/Overload
12. The Contribution of Engine and Airframe Problems
13. The Influence of Spectators
14. The Presence of an Instructor
The report contains Further Discussion under 7 headings. They are:
2. Aeroplane Types
3. Out of Balance
4. Airspeed Indicator, ASI Markings
6. Flying Training Organisations
There a several detailed tables of data and an appendix listing details of all the accidents reviewed.
The report makes nine recommendations. These are:
1. Based on the accident record, further test flights are necessary to verify the study’s initial flight test indication that the Cessna 150 L & M model aircraft may not comply with the criteria for stick force gradient in CS-23 and Federal Aviation Requirements, FAR 23 for light aircraft.
2. The Cessna 150 and Cessna 152 should not be treated as the same type and in particular pilots transferring from the Cessna152 to the Cessna 150 should undertake formal Familiarisation Training.
3. Pilots of lighter weight aeroplanes are strongly recommended to obtain training with an instructor well experienced on the type or participate in the Light Aircraft Association Pilot Coaching Scheme.
4. The increased proportion of stall/spin accidents in the climb and during attempted forced landings following an engine or airframe problem should be publicised – for example in safety publications and posters, and within flying training environments. Thus, as a priority the CAA is strongly requested at the earliest opportunity to produce a new Safety Sense Leaflet on ‘Stall/Spin Awareness’ incorporating suitable elements of the Handling Sense Leaflets and the findings of this study. Ways should be sought to distribute the leaflet to all pilots.
5. Further research should be implemented into the suitability and use of angle of attack indicators in light aeroplanes.
6. The authorities should give consideration to mandating (as opposed to recommending) that flying instructors at any training facility may not undertake training flights with student pilots or passengers until after they have flown with and been checked for proficiency to instruct in the aeroplane type to be flown, by a Chief Flying Instructor, Examiner or Senior Instructor. This should not apply to the biennial one hour of flight instruction. Accordingly, in the absence of formal inspection of PPL training organisations, the flying training industry must be encouraged to formulate a ‘Best Practice Code’ and encourage all such organisations to use it.
7. A Code of ‘Best Practice’ for type conversion within the Single Engine Piston (SEP) class must be encouraged, including the need for thorough familiarity with the stall warning and characteristics for the aeroplane type they are to fly.
8. Further investigation should be conducted into the possible benefits of using the CS22 requirements for motor glider ASI markings in other aeroplane classes. In the meantime owners may wish to assess the usefulness by marking their own ASIs.
9. The authorities are recommended to review the PPL Skills Test tolerances that allow a wide margin in both climb speed and landing threshold speed and do not reflect differences between aeroplane types.
This GASCo study is a detailed examination of stall/spin accidents and deserves careful study. It should initiate more informed though on how to avoid these avoidable fatal accidents.
An fatal stall/spin accident that occurred later in the period covered by this study is a classic illustration of how things can go wrong on take off and how quickly this can have a fatal result if the pilot mishandles the situation.
A Commander 114TC with two pilots on board was departing Southampton Airport when the aircraft apparently suffered some problem that caused reduced performance but was not understood by the pilot. The take-off run was much longer than normal and the angle of climb was very shallow. The pilot announced he had a problem, but didn’t know what was wrong. He requested, and was given, clearance for an immediate turn to rejoin the circuit. Witnesses reported smoke or oil coming from the nose of the aircraft and the engine not sounding ‘normal.’ The aircraft was seen to enter a steeply banked turn to the right, lose height in a side-slip, apparently recover control at a very low altitude and speed, attempt to climb over some trees and then depart from controlled flight and impact the ground vertically. Both occupants were killed.
Investigations revealed a possible slight reduction in power because of a slight fault in one cylinder. More critically, in came to light that the aircraft had been refuelled, as requested, to full tanks. This, with the weight of the two pilots and the equipment carried in the cabin, resulted in the aircraft being 230 lb above its MTOW of 3305 lb. This excess weight would not only have degraded the aircraft’s takeoff and climb performance but would also have increased the stalling speed. The aircraft had the air conditioning selected on during takeoff, contrary to the Operating Manual.
Weight and Lift are equal in level flight so any increase in weight will demand an increase in lift. A rule of thumb is that the percentage increase in stalling speed is half the percentage increase in AUW. A Commander 114TC will stall at 68KIAS in level flight at 3305 lb (MTOW). If the aircraft is overloaded to 3535 lb the new level stalling speed becomes 72 KIAS. The Best Angle of Climb speed is 75 KIAS and it is probable that the pilot was trying to use this speed for the initial climb to clear rising ground in the direction of takeoff, so was operating very close to the stall.
Finally, we need to look at the effect of angle of bank on stalling speed. The stalling speed increases as the square root of the load factor. The load factor in a turn is proportional to the secant of the angle of bank. The table below shows the effect on stalling speed at different angles of bank. Interpolation allows one to see that the Commander would have stalled at 75 KIAS as soon as 22º of bank was applied in the attempt to turn back to the airport.
Angle of Bank Load Factor ‘n’ √’n’ Stall Speed KIAS
0 1.000 1.0000 72
10 1.015 1.0076 72.55
20 1.064 1.0315 74.27
30 1.154 1.0745 77.37
40 1.305 1.1425 82.26
45 1.414 1.1892 85.65
50 1.555 1.2427 89.80
60 2.000 1.4142 101.82
Try your own aircraft by inserting the basic stall speed (at the weight and configuration you are considering) in the space where ‘72’ is on the table. Now work down that column by multiplying your figure by the √’n’ figure for each angle of bank. Interesting? Even if your aircraft stalls at only 50 KIAS it will stall at 71KIAS with 60º of bank applied. Now you KNOW why you were taught that the preferred course of action after EFATO (Engine Failure After Takeoff) is to pick a spot ahead rather than to turn back.
Text and Photographs © 2010 Gremline & Hill House Publications, unless otherwise stated.