Pulsar, G-MCMS, 25 July 2004

56 AAIB Bulletin No: 5/2005 Ref: EW/G2004/07/21 Category: 1.3 Aircraft Type and Registration: Pulsar, G-MCMS No & Type of Engines: 1 Rotax 582 piston engine Year of Manufacture: 1993 Date & Time (UTC): 25 July 2004 at 1305 hrs Location: Near Taynuilt, Scotland Type of Flight: Private Persons on Board: Crew - 1 Passengers - None Injuries: Crew - 1 (Serious) Passengers - N/A Nature of Damage: Damage to landing gear and fuselage belly Commander's Licence: Private Pilot's Licence Commander's Age: 52 years Commander's Flying Experience: 408 hours (of which 50 were on type) Last 90 days - 24 hours Last 28 days - 4 hours Information Source: Aircraft Accident Report Form submitted by the pilot and metallurgical examination of radiator failure History of the flight The Pulsar, pictured at right, is a low-wing composite single-seat kitplane operated under a Permit to Fly. The aircraft was on a cross-country flight from Perth to Oban. Ten miles away from Oban the pilot called Oban Radio and requested the airfield information. He did not receive a reply but he heard the wind direction and speed being passed to another aircraft. While at 4,000 feet on the Oban QFE the pilot detected a slight burning smell. He advanced the throttle but the engine did not respond. The pilot immediately declared a MAYDAY, stating his position and the nature of his emergency to Oban Radio, but the Oban radio operator was unable to decipher the message. Due to a strong westerly wind the pilot decided that he would be unable to reach Oban. There were no suitable fields nearby for a forced landing so the pilot selected a field on flat ground beside a river and planned a 57 circuit while repeating his MAYDAY transmission several times but with no response. At approximately 1,000 feet agl the engine seized. The aircraft reached the chosen field but the touchdown was hard and the field was rough with cows grazing at the eastern (near) end. The landing gear separated during the ground roll and the aircraft decelerated rapidly to a rest. The propeller had stopped in a horizontal position and so was undamaged. The pilot was able to vacate the aircraft by opening the canopy as normal and then telephoned '999' from his mobile phone for assistance. It was later determined that the pilot had suffered from a crushed vertebra. Aircraft's engine cooling system The aircraft's Rotax 582 engine is liquid cooled using two small radiators that are mounted within the forward cowling (see Figure 1). The primary hoses A and D between the engine and each radiator have an inside diameter of 1 inch. The cross-over hoses B and C between the radiators have an inside diameter of 5/8 inch. There are rectangular slots within the upper and lower cowlings to hold the radiators in place. The grey adhesive 'gaffer' tape that is visible on the port radiator in Figure 1 was, according to the owner, applied to ensure a snug fit within the upper cowling slot. The aircraft's instrument panel was fitted with a water coolant temperature gauge but not with a temperature exceedance warning light or aural warning. Because the Rotax 582 is a two-stroke engine there were no oil temperature or oil pressure gauges. The Pulsar construction manual warns: 'if the water ever leaks out of the system in flight, you'll find yourself in a glider in probably less than a minute.' Engine examination A Popular Flying Association (PFA) inspector examined the engine following the accident. He discovered that all the coolant had drained from the radiators. There was a small pool of coolant in the lower cowling and the port radiator's outboard hose coupling had detached from the radiator - circled area in Figure 1 and close-up in Figure 2. The failed coupling, the port radiator and the port radiator hoses were sent to the AAIB for a more detailed examination. An engine strip was subsequently performed by the PFA inspector which revealed severe over heating damage of piston number two. Examination of radiator hoses There was no evidence of any leaks within radiator hoses C or D. Hose D had a slight bend in it but it was a straight hose that had retained some of its bend from having been bent into position for a long period. The Pulsar construction manual states that hose D should be constructed from two pieces of straight hose that are then joined together with the 45° copper fitting supplied with the kit 58 (the alternative is to use a curved moulded hose). This should ensure that no side load is imparted to the radiator coupling. Examination of radiator and failed coupling The port radiator and failed coupling were examined by an independent organisation with metallurgy and fracture expertise. The following is a summary of their findings. A visual examination of the brass radiator showed that the whole circumference of the detached area exhibited a fracture surface and that the corresponding coupling still had a section of the brass radiator attached to it (see Figures 3 and 4). This indicated that the failure occurred in the radiator material rather than in the brazed joint between the coupling and the radiator. A cross section of the failed coupling and the intact coupling on the other end of the radiator revealed their composition (see Figures 5 and 6). A brass connecting tube had been inserted into an opening within the radiator and then brazed with a silver based filler metal. A copper coupling had then been soldered to the brass connecting tube with a lead-tin based solder. The evidence of lead-tin solder on the inner surface of the brass connecting tube suggested that the copper coupling had been dip soldered. Lead-tin solder also coated the outer surface of the brass connecting tube and radiator joint. There were no differences in the construction of the failed and intact couplings although the connecting tube of the failed coupling appeared to be offset downwards at an angle of approximately 5°. The majority of both the fracture surfaces on the radiator and on the coupling were masked by contaminants apart from a small area of the fracture in the upper left position of the radiator (see Figure 3) which was clean. The fracture surface of the detached section was examined under a Scanning Electron Microscope (SEM). The SEM examination did not reveal conclusive evidence of the failure mechanism but it was likely that the clean area of the fracture was the last section to detach and that it occurred due to overload. The lower half of the fracture (with reference to Figure 3) was more obscured by damage and surface contamination; however, small patches were observed which showed a transgranular fracture that appeared to have corroded. This evidence suggested that this part of the crack had been present for a longer period than the cleaner area hence the crack appeared to be progressive. The fracture surface in this area also showed some faint parallel markings that may be evidence of fatigue striations, although they were not clear enough to confirm this positively. There was a build up of solder around the lower part of the fracture (see Figure 3) and there was solder contamination on the lower right area of the fracture – the area from which the crack is likely to have propagated. Lead-tin solder usually melts in the range of 180°C to 320°C depending on its exact composition. The solder contamination on the fracture surface may have been a result of local overheating during the engine failure which resulted in some of the solder melting and being drawn 59 into the crack. However, the radiator was not in direct contact with the engine and therefore the engine bay itself would have had to reach a temperature of 180°C to 320°C. It is also possible that the solder contamination is evidence of a repair. Because it is likely that the crack growth was progressive, it is possible that a coolant leak was observed at some point before final failure oc

Accidents Database

Pulsar, G-MCMS, 25 July 2004

Report Overview

Pulsar, G-MCMS

Summary:

Official Documents

Document Content Preview

G-MCMS-5-05.pdf

Change History

Related Entities