Aviation Investigation Report A98H0003

4.3.10  Accident Investigation Issues

  1. 4.3.10.1 - Flight Recorder Duration and Power Supply
  2. 4.3.10.2 - Underwater Locator Beacon
  3. 4.3.10.3 - Non-volatile Memory

4.3.10.1  Flight Recorder Duration and Power Supply

TSB Recommendations A99-01 through A99-04 were sent to the Minister of Transport in Canada and to the JAA in the Netherlands. The TSB also sent copies of its recommendations to the Swiss Aircraft Accident Investigation Bureau, the United Kingdom Air Accidents Investigation Branch and the French Bureau d'Enquêtes et d'Analyse. Concurrently, the NTSB issued similar recommendations A99-16 through A99-18 to the FAA.

TC has agreed with the recommendations and advises that it is taking measures to amend its regulations by the dates stipulated by the TSB. However, TC also advises that because the FAA is dealing with similar issues raised by the NTSB, it intends to harmonize its actions with those of the FAA. In this regard, the NTSB has expressed its apprehension that, although the FAA has indicated the intention to implement the recommended actions, the dates for final action may not be met.

The Board recognizes that TC has started its consultative process with the Canadian aviation industry, and understands the value of a harmonized approach with US authorities. However, the Board is concerned that TC will also not meet its commitment to implement the required changes in a timely fashion.

4.3.10.2  Underwater Locator Beacon

The flight recorders were recovered from the ocean floor by tracking the acoustic waves emitted from their attached underwater locator beacon (ULB). Given the substantial fragmentation of the aircraft wreckage, and the low visibility water conditions, the ULBs minimized the time required to locate the recorders.

While the recorder's internal crash-protected memory modules were intact, the ULB brackets were damaged. The extent of the bracket damage suggests a high probability that one or both of the ULBs could have readily detached during the impact sequence. The issue of the adequacy of ULB attachments has been a concern of the international recorder community for years.

The Board recognizes that EUROCAE Working Group 50 is developing minimum operational performance specifications for crash-protection of airborne recorders. Currently, these specifications include requirements for the application of impact shock tests to ensure the integrity of a flight recorder's ULB attachment. As EUROCAE recommendations are advisory in nature, the Board is concerned that without adoption and harmonization by regulatory authorities worldwide, ULB attachment specifications may not be universally applied.

4.3.10.3  Non-volatile Memory

Modern aircraft are equipped with electronic systems that contain memory devices designed for data storage. Most commonly, such systems contain a type of volatile memory device whose data is lost when power is removed. Frequently, systems contain a memory device known as non-volatile memory (NVM), which is capable of retaining its stored data even though power has been removed. In the case of SR 111, the engines were controlled with the assistance of full-authority digital electronic control (FADEC) engine control units, which contained NVM devices. In the absence of FDR information, data retrieved from the engine FADEC 2 was helpful in providing some information about the final five-and-a-half minutes of the flight. However, as the FADEC memory was designed for engine maintenance troubleshooting purposes, and the "time stamp" indicating when faults occurred versus when they were written to memory, was of poor resolution for accident investigation purposes. In addition, many other NVM devices from other LRUs were extremely difficult and time-consuming to identify as there were no distinctive markings to facilitate identification.

The Board is concerned that manufacturers and designers of equipment containing memory devices may not consider the potential use of such devices for accident investigation purposes. These aspects are best considered at the design stage, when improvements in data quantity, quality, and ease of device recognition can generally be included for relatively low cost.

This report concludes the Transportation Safety Board's investigation into this occurrence. Consequently, the Board authorized release of this report on 27 March 2003.

Previous