AIRCRAFT ACCIDENT REPORT AND EXECUTIVE SUMMARY

Section/division Occurrence Investigation Form Number: CA 12-12a AIRCRAFT ACCIDENT REPORT AND EXECUTIVE SUMMARY Reference: CA18/2/3/8321 Aircraft Registration ZS-OVZ Date of Accident 29 June 2007 Time of Accident 1342Z Type of Aircraft Boeing Stearman Type of Operation Private Pilot-in-command Licence Type Commercial Pilot Age 58 Licence Valid Yes Pilot-in-command Flying Experience Last point of departure Total Hours as on 29 June 2007 2796.7 Tzaneen Aerodrome (FATZ) (Limpopo Privince) Total Hours on Type as on 29 June 2007 563.5 Next point of intended landing Tzaneen Aerodrome (FATZ) (Limpopo Privince) Location of the accident site with reference to easily defined geographical points (GPS readings if possible) In bush-type area at the geographical position determined as S 23 49.824 E030 19.400. Meteorological Information Wind: 090 TN at 3 knots; Temperature: 22 C; Visibility: 10 km + Number of people on board 1 + 0 No. of people injured 0 No. of people killed 1 Synopsis On 29 June 2007, during an air show practice display by a commercial pilot in a Boeing Stearman aircraft, a loud bang was heard and witnesses observed that a large part of the aircraft separated and fell down from the aircraft. It was found that the engine separated from the aircraft during a display manoeuvre. The aircraft impacted with the ground at a 90 degree nose-down attitude and came to a complete stop in that position. During the investigation, it was determined that the primary causational factor for this accident was the type of washer utilized during the assembly of the engine mount bolts, combined with the operational envelope of the relevant aircraft. The low carbon washer steel contributed to an unfavourable reduction in thickness during operationally-induced compressive forces. These forces were exacerbated during aerobatic manoeuvres. The reduction in thickness of the washer caused movement of the bolts during operation. The movement of the bolts under strain led to fatigue crack initiation in the thread root areas on all four bolts. Probable Cause The aircraft s engine separated from the fuselage in flight, due to low carbon steel washers used during installation of the engine. IARC Date Release Date CA 12-12a 23 FEBRUARY 2006 Page 1 of 35

Section/division Occurrence Investigation Form Number: CA 12-12a Telephone number: 011-545-1000 E-mail address of originator: thwalag@caa.co.za AIRCRAFT ACCIDENT REPORT Name of Owner/Operator : Mark Beckley Studio (SA) (PTY) LTD Manufacturer : Boeing Model : E-75 Nationality : South African Registration Marks : ZS-OVZ Place : Tzaneen Date : 29 June 2007 Time : 1342 Z All times given in this report is Co-ordinated Universal Time (UTC) and will be denoted by (Z). South African Standard Time is UTC plus 2 hours. Purpose of the Investigation: In terms of Regulation 12.03.1 of the Civil Aviation Regulations (1997) this report was compiled in the interest of the promotion of aviation safety and the reduction of the risk of aviation accidents or incidents and not to establish legal liability. Disclaimer: This report is given without prejudice to the rights of the CAA, which are reserved. 1. FACTUAL INFORMATION 1.1 History of Flight 1.1.1 Shortly after take-off the pilot, call sign Stearman, changed the VHF frequency to 129.3 MHz, entered the display area south of the aerodrome, and commenced his display practice. 1.1.2 During the display there was a loud bang and witnesses observed that a large part of the aircraft separated and fell down from the aircraft. 1.1.3 No radio transmission was made from the pilot on the aircraft. The aircraft was observed by eyewitnesses to be falling out of the sky like a leaf. The aircraft started to swing forward and backwards with very little forward speed until it disappeared below the tree line. 1.1.4 The aircraft was found in a nose-down attitude on impact with the ground. 1.1.5 It was reported by an eyewitness that at approximately 13:30 Z on 29 June 2008, aircraft ZS-OVZ departed off Runway 06 at Tzaneen Aerodrome for a display practice for an airshow which was taking place on 30 June 2007. 1.1.6 Several eyewitnesses were interviewed after the accident and their statements all corroborated. CA 12-12a 23 FEBRUARY 2006 Page 2 of 35

1.2 Injuries to Persons Injuries Pilot Crew Pass. Other Fatal 1 - - - Serious - - - - Minor - - - - None - - - - 1.3 Damage to Aircraft 1.3.1 The aircraft was substantially damaged. 1.4 Other Damage Figure 1: Indicating the damages that the aircraft sustained with no engine. 1.4.1 No other damages were caused. 1.5 Personnel Information Nationality South African Gender Male Age 58 Licence Number 0270200264 Licence Type Commercial Pilot Licence valid Yes Type Endorsed Yes Ratings Instructor rating grade 3, Instrument rating, Test pilot rating class 2, Night Rating Medical Expiry Date 2 October 2007 Restrictions Medical Restriction-To wear corrective lenses Previous Accidents None CA 12-12a 23 FEBRUARY 2006 Page 3 of 35

Flying Experience: Total Hours as on 29 June 2007 Total Past 90 Days as on 29 June 2007 Total on Type Past 90 Days as on 29 June 2007 Total on Type as on 29 June 2007 2796.7 29.7 4.1 563.5 1.6 Aircraft Information Airframe: Type Boeing Stearman E75 Serial Number 75-874 Manufacturer Boeing Year of Manufacture 1941 Total Airframe Hours (At time of Accident) 6316 Last MPI (Hours & Date) 6301 1 March 2007 Hours since Last MPI 15 C of A (Issue Date) 24 April 2002 C of A (Expiry Date) 23 April 2008 C of R (Issue Date) (Present owner) 19 March 2002 Operating Categories Standard Engine: Type Pratt & Whitney R985-AN3 Serial Number 21845 Hours since New 5876 Hours since Overhaul 415.19 Propeller: Type Hamilton Standard 2D 30-237 Serial Number N108260 Hours since New Unknown Hours since Overhaul 415.19 CA 12-12a 23 FEBRUARY 2006 Page 4 of 35

1.7 Meteorological Information 1.7.1 Official weather information obtained from the South African Weather Services: Wind direction 090 TN Wind speed 03 Knots Visibility 10 km + Temperature 22 C Cloud cover None Cloud base None Dew point Unknown 1.7.2 Weather conditions at the time of the accident: Surface Analysis A high pressure system was south-east of the country, extending into the northeastern part of the country. This high pressure caused partly cloudy conditions over Mozambique but fine weather with clear skies over the Limpopo Province. Upper Air Analysis At 500 hpa, a trough of low pressure was east of the country, causing southwesterly winds in the Tzaneen area. 1.8 Aids to Navigation 1.8.1 The aircraft was equipped with an Airpath Compass and a King KT 76A GPS. There were no recorded defects with the navigation equipment. 1.9 Communications 1.9.1 The communication equipment that was installed in the aircraft was a King KLX 135 VHF receiver. There were no recorded entries of defects with the communication equipment. 1.9.2 The pilot was operating in the vicinity of a manned aerodrome. He was practising an aerobatic manoeuvre and was broadcasting on VHF frequency 129.3 MHz prior to the accident. 1.10 Aerodrome Information Aerodrome Location Tzaneen Aerodrome ( FATZ) Aerodrome Co-ordinates S23 49.240 E030 19.360 Aerodrome Elevation 1914 Feet Runway Designations 06/24 Runway Dimensions 1420 x 20 Metres Runway Used 06 Runway Surface Asphalt Approach Facilities None Note: The above aerodrome information was the departure aerodrome and where the display was being held. CA 12-12a 23 FEBRUARY 2006 Page 5 of 35

1.10.1 The accident occurred in bushes at the geographical position determined as S 23 49.824 E030 19.400. 1.11 Flight Recorders 1.11.1 The aircraft was not fitted with a flight data recorder and a cockpit voice recorder as these were not required in terms of the Civil Aviation Regulations. 1.12 Wreckage and Impact Information 1.12.1 The aircraft was flying at a height of approximately 2500 feet above ground level. The pilot was performing a stall turn manoeuvre, when the engine of the aircraft separated from the fuselage. The aerobatic display box was measured from ground level up to and including 3000 feet above ground level. 1.12.2 The aircraft started to swing backwards and forwards until the aircraft impacted with the ground. 1.12.3 The aircraft impacted with the ground at a 90 angle nose-down and came to a complete stop in that position at a geographical position of S23 49,818 E030 19 395 and an elevation of 1838 Feet. 1.12.4 On impact with the ground, all the right-hand wing support struts broke and on the left-hand wing only one support strut broke. 1.12.5 The engine of the aircraft was located approximately 500 metres away from the main wreckage in the direction of 355 M. Figure 2: Indicating the angle and the impact of the aircraft. CA 12-12a 23 FEBRUARY 2006 Page 6 of 35

Figure 3: Indicating the engine and the damages sustained. 1.13 Medical and Pathological Information 1.13.1 A post-mortem examination was performed on the deceased pilot after the accident. 1.13.2 The results of the post-mortem report and toxicology tests were not available at the time when the report was compiled. Should any of the results be positive, an attachment will be added to this report to include the results. 1.13.3 The investigator was informed by the forensic pathologist who performed the autopsy on the deceased pilot, that the cause of death was attributable to head injuries only. 1.14 Fire 1.14.1 There was no evidence of a pre- or post-impact fire. 1.15 Survival Aspects 1.15.1 This accident was considered survivable. The cabin area was intact. The seat did not collapse. The safety harness was found intact and did not fail. 1.15.2 The fatal injuries sustained by the occupant might have been prevented had he worn a full-face helmet. 1.16 Tests and Research 1.16.1 The four bolts which attach the engine mount to the airframe were sent for metallurgical analysis. The metallurgical report concluded the following: CA 12-12a 23 FEBRUARY 2006 Page 7 of 35

Engine Cradle Connecting Points. The engine cradle connecting points showed no clear signs of rack formation. The mount shock rubbers also seem to be in good condition. The bulging damages on the right-hand top and bottom points can be attributed to the impact forces. The engine cradle connecting points cannot be attributed as a contributing factor to this accident. Airframe Connecting Points. No pre-impact crack formation could be detected and all damages can be attributed to the impact and the initial parting of the engine cradle from the airframe. Both the remainder of the right-hand top and left-hand top mount bolts were still in position. The right and left-hand bottom airframe connecting points, together with the remainder of the respective mount bolts, were not supplied for this investigation. No clear evidence was uncovered to identify the airframe connecting points as contributing to this accident. Mounting Bolts. The bolts retrieved showed the NAS-148 designation marks. It is an internal wrenching type, cadmium plated, high tensile bolt with a minimum tensile strength of 160 000 PSI. The test results proved that all four bolts conform to this specification. The thread TPI number is approximately 10, placing it under the 'coarse thread' designation for torque purposes. All four of the bolt fracture surfaces displayed fatigue crack profiles to varying depths and propagation rates. No clear signs of possible over-torque of the bolts during assembly could be detected. The right-hand top bolt is exposed to severe gyroscopic forces during operation, particularly during aerobatics and in the pitching plane. The fracture surface investigation of the right hand top bolt suggests that it was most probably the first of the four bolts to fail. Deposits and other impact damages suggest that this bolt probably failed some period of time before the fatal flight. The bolt failure was induced by fatigue crack formation. The result from its final fracture is the overloading of the remaining three bolts during operation. The fatigue profiles from the remaining bolts indicate that although some pre-existing fatigue crack initiation is evident, the primary segment of the fatigue cracks was formed during the phase between the right top bolt failure and the right-hand bottom bolt failure that followed. The right-hand bottom bolt failed during the last moments of flight, leading to the engine being supported by only the two remaining left-hand bolts. The fatigue profiles from the two left-hand bolts confirm this fact. Both left-hand bolts failed within a short period of time after the right-hand bolt failures leading to the separation of the engine cradle from the airframe in flight. The engine assembly veered to the left in relation to the fuselage, resulting in the propeller impacting with the left main wheel and spat. The investigation revealed no clear discrepancies with regard to the integrity of the four mounting bolts and can therefore not be attributed as a contributing factor to this accident. Mounting Washers. Only three of the four washers were available for this investigation. No clear identification marks were detected on any of the washers. Spectrometry results indicated that the base material is of the low carbon (0.041%C0) steel type conforming to C1005 specification. The hardness results indicate an average of 137 Brinell Hardness Number (BHN). The hardness of the C1005 washer is approximately 33 BHN lower than the AN960/970 series, aircraft type washer manufactured from C1020 (MIL-S-7952) steel at a specified 170 BHN. The recommended torque forces supplied by the NAS-148 bolt manufacturer are based on the NAS-148 and AN960/970 washer combinations. The carbon content of the C1005 washer is approximately 0.041% while the aircraft type C1020 steel contains 0.18% to 0.23% carbon. The resultant effect of this variation in carbon content and initial hardness is a variation in the strain hardening exponent, n. This CA 12-12a 23 FEBRUARY 2006 Page 8 of 35

entails that the softer, lower n value material will be more prone to compacting during extreme compression stresses (see par. 1.5.). Although it proved to be ideal for the washer to be of a softer material than the bolt head to enhance 'bedding' during torque, the strain hardening effect will cause the washer to 'harden' allowing compaction only to a certain percentage of the initial thickness. In this case the softer C1005 steel washer will allow compaction to a higher percentage than the C1020 steel, when exposed to the same compression forces during operation. A percentage of the reduction in thickness will be due to the initial torque placed on the bolt during assembly. Ideally the washer must absorb this reduction without cracking but should not allow any further reduction in thickness, as it will lead to movement of the bolt during operation. Movement of the high tensile bolt, particularly in the weaker shear plane, will result in crack initiation over time under severe stresses. The right-hand top washer revealed multiple impact marks leading plastic flow. This is an indication that the washer had been exposed to severe and repeated compression forces over a period of time. The multiple impact marks were most probably induced by the mount bolt moving during operation. The washer material and the performance thereof during operation can be considered as a contributing factor to this accident. The following most probable sequence of events leading to the accident are based on the results from this investigation on the supplied parts only. The primary causational factors for this accident are the type of washer utilized during the assembly of the engine mount bolts, combined with the operational envelope of the relevant aircraft. The low carbon washer steel led to an un favorable reduction in thickness during operationally-induced compressive forces. These forces were enhanced during aerobatic maneuvers. The reduction in thickness of the washer led to movement of the bolts during operation. The movement of the bolts under strain led to fatigue crack initiation in the thread root areas on all four bolts. The right-hand top bolt was exposed to amongst others, severe, alternating gyroscopic induced tension and compression forces during operation. The right-hand top bolt finally fractured after a period of time under these extreme conditions, leading to the increase in crack propagation rates of the remaining three bolts. The right-hand top bolt was probably finally fractured prior to the fatal flight. During the fatal flight and the final maneuver, the right-hand bottom bolt finally fractured, leading to the separation of the engine cradle from the airframe connecting points. The left-hand bolts, now the only support, failed almost immediately after the right-hand set of bolts in flight. This resulted in the propeller blades impacting with the left main wheel and spat before impacting with the ground. Although the resulting effect would have been comparable, the possible over-torque of the bolts during assembly could not be conclusively determined. 1.16.2 The engine was subjected to an engine teardown examination after the accident at an approved maintenance facility, to determine if there was a catastrophic failure in the engine which could have resulted in the engine separating from the airframe, whilst in flight. The report on the engine teardown examination concluded the following: 1. In general the engine sustained extensive accident-related damage. 2. There was no indication of malfunction in the operation of the engine that could have led to engine failure. CA 12-12a 23 FEBRUARY 2006 Page 9 of 35

3. There was no indication of any metal shavings in the engine. 4. The engine oil was clean. 1.17 Organisational and Management Information 1.17.1 This was a private flight. 1.17.2 The pilot was the owner of the aircraft. 1.17.3 According to available records, the Aircraft Maintenance Organisation (AMO) that had certified the last MPI on the aircraft prior to the incident, was in possession of a valid AMO Approval, No.104 with an expiry date of 31October 2007. 1.18 Additional Information 1.18.1 On-site investigation Inspection of the wreckage after the accident revealed that the right-hand wing support struts failed and on the left-hand wing only one support strut failed on impact. It was found that all 4 bolts which attached the engine mount to the airframe had sheared. The left-hand tyre was found cut. One of the propeller blades went through the tyre as there was tyre smear marks found on the one propeller blade. On the inspection of the propeller, nothing abnormal was found and all damages were sustained due to impact. The counter weights on the propeller were still attached. One propeller blade was found broken, but this was due to impact with the ground and the piece of blade was found next to the engine. 1.18.2 Evidence gathered during the investigation indicates that the aircraft was used for aerobatic manoeuvres on a frequent basis. The fatal accident occurred during a practice session for an aerobatic display for an airshow. The pilot was performing a stall turn manoeuvre when the engine separated from the airframe in flight. 1.18.3 The Pilot Operating Handbook (POH), Section II, on page 8, paragraph 18, dated 30 January 1944, states that the following manoeuvres are prohibited: Inverted flight Inverted spins Outside loops Snap rolls, at more than 106mph indicated airspeed Slow rolls at more than 124mph. Do not exceed an indicated airspeed of 141mph. 1.18.4 Evidence found in the pilot logbook indicates that the pilot obtained aerobatic rating on the aircraft type on 21 July 2002. 1.19 Useful or Effective Investigation Techniques 1.19.1 None. CA 12-12a 23 FEBRUARY 2006 Page 10 of 35

2. ANALYSIS 2.1 On 29 June 2007 at approximately 13:30Z, aircraft ZS-OVZ took off from Runway 06 at Tzaneen Aerodrome for a display practice for an airshow the following day on 30 June 2007. During the practice,, several eyewitnesses heard a loud bang and then observed a large part of the aircraft detach and fall from the aircraft whilst in flight. 2.2 The accident occurred in daylight conditions. Fine weather conditions prevailed at the time and place of the accident and the weather therefore was considered not to have contributed to the accident. 2.3 Although the aircraft had a valid Certificate of Airworthiness, evidence gathered indicated that the aircraft had not been maintained in compliance with the Civil Aviation Regulations, because of the type of washers that was used on the engine mountings. 2.4 The pilot was medically fit and was in possession of a valid licence with an expiry date of 2 October 2007. The aircraft type was endorsed on the pilot s licence at the time of the accident. 2.5 During the investigation, it was found that the engine had separated from the aircraft s fuselage in flight. It was further found that the four mounting bolts that attach the engine to the airframe had failed and as a result were sent for metallurgical analysis. The metallurgical report concluded the following: The primary causational factor for this accident is the type of washer utilized during the assembly of the engine mount bolts, combined with the operational envelope of the relevant aircraft. The low carbon washer steel led to an unfavorable reduction in thickness during operationally-induced compressive forces. These forces were enhanced during aerobatic maneuvers. The reduction in thickness of the washer led to movement of the bolts during operation. The movement of the bolts under strain led to fatigue crack initiation in the thread root areas on all four bolts. The right-hand top bolt was exposed to amongst others, severe, alternating gyroscopic induced tension and compression forces during operation. The right-hand top bolt finally fractured after a period of time under these extreme conditions, leading to the increase in crack propagation rates of the remaining three bolts. The right-hand top bolt was most probably finally fractured prior to the fatal flight. During the fatal flight and the final maneuver, the right-hand bottom bolt finally fractured, leading to the separation of the engine cradle from the airframe connecting points. The left-hand bolts, now the only support, failed almost immediately after the right-hand set of bolts in flight. This CA 12-12a 23 FEBRUARY 2006 Page 11 of 35

resulted in the propeller blades impacting with the left main wheel and spat before impacting with the ground. Although the resulting effect would have been comparable, the possible over-torque of the bolts during assembly could not be conclusively determined. 2.6 It was further found that the aircraft was used for aerobatic displays and that aerobatic manoeuvres were practised with the aircraft. It can therefore be assumed that the aerobatic manoeuvres over the years took their toll on the bolts and washers which attach the engine to the airframe. 2.7 The investigator was informed by the forensic pathologist who performed the autopsy on the deceased pilot, that the cause of death was attributable to head injuries only. 3. CONCLUSION 3.1 Findings 3.1.1 The pilot was the holder of a valid commercial pilot s licence and had the aircraft type endorsed on his licence. 3.1.2 This was a practice flight for an aerobatic display at an airshow. 3.1.3 The aircraft held a valid Certificate of Airworthiness. 3.1.4 Aircraft examination revealed that the engine separated from the airframe whilst in flight. 3.1.5 It was found that the four bolts that attach the engine to the airframe had failed. These were sent for metallurgical analysis, and the conclusion was that the primary causational factor for this accident was the type of washer utilized during the assembly of the engine mount bolts.(see 1.16.1) 3.1.6 The aircraft was substantially damaged on impact. 3.1.7 The AMO that certified the last Mandatory Periodic Inspection prior to the accident was in possession of a valid AMO Approval certificate from the CAA. 3.1.8 Weather conditions at the time of the accident were not considered to have had a bearing on the accident. 3.1.9 It was found that if the pilot had worn a full-face helmet, he could have survived the accident. 3.2 Probable Cause/s 3.2.1 The aircraft s engine separated from the fuselage in flight, due to low carbon steel washers used during installation of the engine. CA 12-12a 23 FEBRUARY 2006 Page 12 of 35

4. SAFETY RECOMMENDATIONS 4.1 It was recommended that the Commissioner for Civil Aviation ground all Boeing Stearman Aircraft fitted with a Pratt & Whitney 985 series engine, which allows for a take-off power setting of 450 horse power for the purpose of an inspection. (This safety recommendation has already been implemented.) 4.2 It is recommended that the Commissioner for Civil Aviation issue an Aeronautical Information Circular advising all pilots involved in aerobatic manoeuvres to wear a helmet, as this accident could have been regarded as a survivable accident if the pilot had worn a helmet. 4.3 It is recommended that the Commissioner for Civil Aviation issues a MAN requiring the correct specifications are used when low-carbon steel washers are utilized in engine mount assemblies for this aircraft type. 5. APPENDICES 5.1 Appendix 1: Metallurgical report from Crash Lab. Report reviewed and amended by Advisory Safety Panel: 28 July 2009. -END- CA 12-12a 23 FEBRUARY 2006 Page 13 of 35

Appendix 1 CA 12-12a 23 FEBRUARY 2006 Page 14 of 35

CA 12-12a 23 FEBRUARY 2006 Page 15 of 35

CA 12-12a 23 FEBRUARY 2006 Page 16 of 35

CA 12-12a 23 FEBRUARY 2006 Page 17 of 35

CA 12-12a 23 FEBRUARY 2006 Page 18 of 35

CA 12-12a 23 FEBRUARY 2006 Page 19 of 35

CA 12-12a 23 FEBRUARY 2006 Page 20 of 35

CA 12-12a 23 FEBRUARY 2006 Page 21 of 35

CA 12-12a 23 FEBRUARY 2006 Page 22 of 35

CA 12-12a 23 FEBRUARY 2006 Page 23 of 35

CA 12-12a 23 FEBRUARY 2006 Page 24 of 35

CA 12-12a 23 FEBRUARY 2006 Page 25 of 35

CA 12-12a 23 FEBRUARY 2006 Page 26 of 35

CA 12-12a 23 FEBRUARY 2006 Page 27 of 35

CA 12-12a 23 FEBRUARY 2006 Page 28 of 35

CA 12-12a 23 FEBRUARY 2006 Page 29 of 35

CA 12-12a 23 FEBRUARY 2006 Page 30 of 35

CA 12-12a 23 FEBRUARY 2006 Page 31 of 35

CA 12-12a 23 FEBRUARY 2006 Page 32 of 35

CA 12-12a 23 FEBRUARY 2006 Page 33 of 35

CA 12-12a 23 FEBRUARY 2006 Page 34 of 35

CA 12-12a 23 FEBRUARY 2006 Page 35 of 35