Aviation Accident Report: United Air Lines Flight 823/Attachment 2

From Wikisource
Jump to navigation Jump to search

CIVIL AERONAUTICS BOARD
BUREAU OF SAFETY

October 9, 1964

Mr. George S. Moore
Director
Flight Standards Service
Washington, D. C. 20553

Dear Mr. Moore:

During the Board's investigation of the recent accident near Parrottsville, Tennessee, on July 9, 1964, involving a United Air Lines Vickers Viscount, N7405, several static and inflight tests were conducted on the fire extinguishing systems for the underfloor cargo compartment using similar aircraft. The results of these tests pointed out certain discrepancies which could seriously affect the safety of the aircraft and passengers.

At least fifteen fire extinguishers, Pyrene Duo Head Model DCD-10 were discharged during the static and inflight tests. This extinguisher is located behind the first officer on the right hand side of the flight deck companionway. Connection of the extinguisher to the underfloor cargo compartment spray tube assembly is accomplished by means of a flexible hose between the bottle discharge head and a bayonet floor fitting. This arrangement allows the bottle to be utilized as a portable extinguisher should the need arise. All of the static tests were conducted with the flexible hose configuration. United Air Lines replaced the flexible hose with a rigid tube for the inflight tests.

The most serious discrepancy noted during the tests concerned spraying of CO2 into the cockpit when extinguishers were discharged. At least five of the fifteen tests resulted in the gas escaping into the cockpit where CO2 concentrations in the atmosphere at head level were measured at maximum values of 12 percent. This leakage occurred at the bayonet floor fitting on the flexible hose installations and at the valve in the bottle discharge head of the extinguishers with the rigid tube installation. In each of these cases very little of the CO2 was discharged into the under-floor compartment.

A second discrepancy concerned improper installation of the metal seal diaphragm, Pyrene P/N 49/90140, which is installed in the discharge head of the extinguisher. This seal diaphragm covers the outlet port of the discharge valve and is installed prior to charging the extinguisher. Brass terminal blocks attached to each end of the seal are connected to electrical wiring in the discharge head for normal electrical discharge. When the discharge button is pressed in the cockpit current flows through the seal diaphragm which acts as a resistance and results in heating the seal. When the temperature rises sufficiently, the metal is softened and the 200 psi pressure in the extinguisher causes the seal to rupture and the CO2 flows into the underfloor compartment.

A number of these seal diaphragms were removed from the extinguisher discharge heads during the tests and examined. Several were found to have been installed off center resulting in improper or incomplete rupture. Photographs of representative seals are enclosed for your information. Comparison of the seal terminal blocks to their locating slots in the discharge heads indicates the reason for the off-center installation. The slots are three-fourths of an inch wide while the terminal blocks are five-eights of an inch wide. This one-eighth of an inch tolerance allows the seal to be mispositioned a critical one-sixteenth of an inch off center. The fact that mispositioning of the seal by such a small amount can result in questionable operation of the extinguisher system would indicate marginal, if not poor, design.

The Viscount Maintenance Manual contains explicit instructions for ensuring that the seal diaphragm is centered over the discharge valve opening. However, from the number of mispositioned seals noted during the tests, it is apparent that the manual instructions are not being followed.

Based on the results of the static and inflight tests as detailed in the preceding paragraphs, the Board believes that the Pyrene fire extinguishing system for the underfloor cargo compartment as installed on Viscount aircraft is not only inadequate for its intended purpose but also poses a danger to the flight crew.

Therefore, the Board recommends that the Federal Aviation Agency evaluate the design of the Pyrene Duo Head Model DCD-10 fire extinguisher system and take such corrective action as will increase its reliability and prevent release of CO2 into the habitable portions of the aircraft.

Flight Standards Service personnel assisting in the investigation of this accident are aware of these discrepancies and will continue to receive information relative to the problem. Personnel of our Engineering Division will also be available for further consultation on this matter.

Sincerely yours,

/s/ B. R. Allen

for Leon H. Tanguay
Director, Bureau of Safety

FEDERAL AVIATION AGENCY
Washington, D. C. 20553

In reply
refer to: FS-120

July 26, 1965

Mr. B. R. Allen
Director, Bureau of Safety
Civil Aeronautics Board
Washington, D. C. 20428

Dear Mr. Allen:

This is in answer to your letter of October 9, 1964, reference B-80-96, concerning the CO2 fire extinguishing system on the United Air Lines Vickers Viscount aircraft. It supplements our acknowledgment letter of October 19, 1964. Your letter of November 17, 1964, and our answer of December 15, 1964, also bear, in part (as will be explained), on this same subject.

We have evaluated the design of the Pyrene Duo Head Model DCD-10 fire extinguisher system. During this evaluation particular attention was paid to the two discrepancies mentioned in your October 9, 1964, letter. The first of these, spraying CO2 into the cockpit when extinguishers were discharged, was the subject of your letter of November 17, 1964, and our answer of December 15, 1964. The second discrepancy was in regard to the improper installation of the metal diaphragm, Pyrene P/N 49/90140, installed in the discharge head of the extinguisher.

As a result of our investigation, we have published a proposed airworthiness directive, applicable to the Vickers Viscount Model 744, 745D, and 810 series aircraft equipped with Pyrene fire extinguishers. The proposal appeared in the Federal Register of June 29, 1965, as Docket No. 6734.

During our investigation of this subject, we worked with representatives from UAL and the Air Registration Board, as well as your own technical personnel. Several independent flight tests were conducted, and also ground (using an installed system), and laboratory tests. Reports on these tests were submitted as exhibits at the public hearing at Knoxville, Tennessee, January 11–16, 1965.

We believe that the proposed AD contains adequate instruction for eliminating the noted discrepancies. If the overhaul, inspection, and tests provisions contained therein are carried out, the leakage of CO2 into the cockpit from the underfloor cargo compartment fire extinguishing system will be a minimum consistent with approved flight crew emergency procedures for fire protection. These procedures were listed, in part, in our letter of December 15, 1964. Subsequent to compliance with the noted proposed AD, manual discharge of the CO2 system will not be necessary, although this mode of operation will still be available if desired.

All communications received on or before July 27, 1965, will be considered by the Administrator before taking action upon the proposed rule. Since few operators use the affected airplanes, we do not anticipate many comments. In any event, we feel assured that our corrective action will eliminate the problems uncovered during your investigation, as stated in your subject letter.

Sincerely yours,

/s/ G. S. Moore

George S. Moore
Director
Flight Standards Service

CIVIL AERONAUTICS BOARD
BUREAU OF SAFETY

November 17, 1964

Mr. George S. Moore
Director
Flight Standards Service
Federal Aviation Agency
Washington, D. C. 20553

Dear Mr. Moore:

During the course of our continuing investigation of the accident involving a United Air Lines Vickers Viscount aircraft near Parrottsville, Tennessee, on July 9, 1964, at was revealed, during static and flight tests, that certain discrepancies existed in the Pyrene fire extinguisher. The most serious of these discrepancies was the spraying of CO2 into the cockpit area when extinguishers are discharged. This was determined to have been due to leakage at either the bayonet floor fitting or at the valve in the head of the discharge bottle.

This discrepancy was previously brought to your attention in our correspondence of October 9, 1964, wherein we recommended that the Administrator evaluate the design of the Pyrene Duo Head Model DCD-10 fire extinguisher system. Your letter of October 19, 1964, indicated that our recommendation was being studied.

We believe that during the period required for your study the potential hazards outlined in our previous recommendation continue to exist. Therefore, it is further recommended that all operators of Viscount aircraft having a Pyrene cylinder installation within the general cockpit area, be requested to re-emphasize the need for pilots to don smoke masks before discharging the lower cargo bin CO2 cylinder.

In addition the operators should be informed that a CO2 spray directed into the cockpit area could result in considerable distraction if the pilots are unaware that such an incident could occur.

Sincerely yours,

/s/ B. R. Allen

B. R. Allen
Director, Bureau of Safety

FEDERAL AVIATION AGENCY
WASHINGTON, D. C. 20553

December 15, 1964

In repy
refer to: FS-431

Mr. B. R. Allen
Director, Bureau of Safety
Civil Aeronautics Board
Washington, D. C. 20428

Dear Mr. Allen:

This is in reply to your letter of November 17, 1964, regarding discrepancies in the Pyrene fire extinguisher system on the United Air Lines Vickers Viscount aircraft.

Our studies concerning the Viscount CO2 system of which you were advised in our letter of October 19, 1962, have not been completed. For your information, Vickers Armstrong has notified all Viscount operators with cockpit CO2 systems to manually discharge the extinguishing agent into the lower cargo bin

Pending completion of our investigation, as an interim precautionary measure, we are instructing our field personnel to request all Viscount operators with this same system to take the following action:

  1. Re-emphasize to their pilots the need to don smoke masks before discharging the lower cargo bin CO2 cylinder.
  1. Instruct their pilots to manually discharge CO2 for the lower cargo compartment
  1. Notify their pilots that it is possible for CO2 to leak into the cockpit area.

We certainly appreciate your recommendations.

Sincerely yours,

/s/ C. W. Walker

for
George S. Moore
Director
Flight Standards Service

FAA AIRWORTHINESS DIRECTIVE

65-21-6 Vickers Amdt. 39-131 Part 39 Federal Register September 3, 1965. Applies to Viscount Models 744, 745D, and 810 Series Airplanes That Are Equipped With Pyrene Fire Extinguishers, Type DCD 2, DCD 2-1/4, DCD 10 or DCD 11.

Compliance required at first airplane overhaul or within six months after the effective date of the AD, whichever occurs first, unless already accomplished.

To prevent further failures of Pyrene fire extinguishers of the subject models, accomplish the following.

(a) Overhaul, inspect, and test the fire extinguishing systems that are equipped with electric discharge facility, in accordance with revised requirements in British Aircraft Corporation (Weybridge Division) Preliminary Technical Leaflets No. 256 (700 Series), No. 120 (800/810 Series), (Amendment TR 25 to Viscount Maintenance and Instruction Manuals and Amendment TR3 to the Accessories Manual cover the same subject). Subsequent overhaul of the fire extinguishing system must be carried out in accordance with these specified periods in the FAA-approved maintenance schedule.

(b) The spray ring system must be blown through with warm dry air to insure that the piping and discharge holes are free of obstructions during the overhaul of the airplane at periods specified in the FAA-approved maintenance schedule.

This directive effective October 3, 1965.