HISTORY OF FLIGHT Use your browsers 'back' function to return to synopsisReturn to Query Page
On December 29, 2000, about 2115 Hawaiian standard time, a Lockheed L-1011-385-1 transport category airplane, N735D, operated by Delta Airlines as flight 219, experienced an electrical fire forward of the flight engineer's station while en route from San Francisco, California, to Honolulu, Hawaii. The flight was about 2 hours out of Honolulu when the event took place, and continued for an uneventful landing, arriving at the gate at 2330. There were no injuries to the airline transport rated captain, commercial rated first officer, airline transport rated flight engineer, 10 flight attendants, or the 287 passengers. The regularly scheduled domestic flight was operating under 14 CFR Part 121, and an instrument flight plan had been filed. The flight originated from San Francisco at 1910 Pacific standard time.
According to the flight crew, the airplane was in cruise flight at 32,000 feet mean sea level (msl) in clouds at the time of the incident. The captain reported the flight was encountering light and moderate turbulence in clouds. The flight was also experiencing a phenomenon called Saint Elmo's fire. The phenomenon is a discharge of static electricity often seen on airplanes flying through or near stormy weather. Static electricity from clouds collects on the airplane structure and discharges from sharp points in the form of visible light. This activity lasted for about 3 minutes, during which time loud popping noises were heard in the radios.
Subsequently, a loud electrical discharge occurred above the co-pilot's side window. Sparks, smoke, and a strong acrid smell of electrical burning accompanied the discharge from the panel. Oxygen masks and smoke goggles were donned and the smoke and fume checklist was performed. The smoke slowly dissipated.
After about 5 minutes the acrid smell increased and a glow was observed behind the same panel. Halon was discharged under the panel. All windshield heat switches were turned off and circuit breakers were opened and again the smoke dissipated.
After about 15 minutes the smell of burning increased and a glow was observed under the panel. Panel screws were removed for better access and view. Halon was again applied and additional bottles of halon were brought forward by a flight attendant as a precaution. The flight crew declared an emergency to air traffic control (ATC), and as a result ATC cleared the flight direct to Honolulu. The last 45 minutes of flight were conducted with no fire or smoke.
WRECKAGE AND IMPACT INFORMATION
Post incident examination of the cockpit's burned area revealed electrical arcing had occurred between the airplane structure, an Adel clamp, and a 30-wire bundle. Twenty of the 30 wires were observed burned. The burned wire bundle passed behind the flight engineer's station and overhead to the heated windshields.
Static dischargers, which are used to dissipate static electricity into the atmosphere, were examined after the incident and were found intact.
TESTS AND RESEARCH
The wire bundle and three clamps were removed and sent to the Safety Board's Materials Laboratory for examination, along with photographs of the wires taken before their removal. The wire bundle contained a burned area that nearly separated it into two segments (one segment being shorter than the other). The wire bundle consisted of four wires of 10-gage diameter, four wires of 16-gage diameter, and eight twisted pairs with red and blue markings. Heat-stamped numbers were found on most 10- and 16-gage wires, but the twisted wires had no numbers or marking sleeves.
Examination of the photographs depicted the shorter end of the burned bundle had been installed forward or furthest from the flight engineer station, and a "shadowed" area of the bundle was hanging beneath a molten clamp. The shadowed area was an area of less blackened material approximately equal in width as a sample clamp. The photographs revealed the burned portion of the wire bundle was aft of the molten clamp, toward the flight engineer station. Up to 4.2 inches of wires in the bundle were missing from aft of the molten clamped area.
Numerous wires were melted and resolidified. Some wires had been fused into solid copper for up to 0.6 inches in length, while others displayed resolidified beads of molten wire metal at their burned tips.
The wires appeared to contain silver to gray colored plating on copper cores with an aromatic poyimide-type of insulation (such as MIL-W-81381, or also known as Kapton). This was similar to the wire type Lockheed used as original equipment hook-up wire in L-1011 airplanes. The insulation style was tape wrapped with a yellowish-golden topcoat. A green vendor number found on many wires was 59 92507.
Microscopic examination (variable, at approximately 10-20 power) of undamaged wires revealed a general condition of wire that was smooth. The insulation reflected light evenly without a dry look. No microscopic radial cracks were found in the insulation. Impressed and deformed areas were found along the lengths of some individual wires.
As a test of insulation flexibility, the intact end of one undamaged 16-gage wire (wire number 3041 119 AN 2 16) was wrapped tightly around a 1-inch diameter form three times. Microscopic examination revealed no splits in the insulation. Wrapping a 10-gage wire around the same 1-inch form revealed two small breaches in the topcoat. The first was along the spiral feature of the insulation construction. The second was at the intersection of a heat-stamped character and another spiral. Neither breach appeared visually to extend to the conductor, and no further testing was done to establish the electrical integrity of the sample.
The three clamps submitted were "P" shaped clams with rubber padding seated in the inside circumference. A normal clamp of this type is made of a metal band that has a straight side, which extends from a bolt-hole, then gradually loops around into a "U" shaped cup, which holds the wires. The clamp band then bends 90 degrees at the other end, which contains another bolt-hole. This allows the mating of the bolt-holes on each end of the clamp. The clamp is then lined (insulated) with a rubber piece, which rests on the inside of the metal band. The rubber insulator has a flat end and a wedged end near the mating ends of the clamp.
The three clamps were examined, and it was noted that one of the clamps was separated into two segments and the other two clamps were intact. The separated clamp was missing nearly half of the circumference that had been padded with rubber. The remaining padded portion on the damaged clamp contained the end of the pad without the wedge. The damaged clamp was compared to the other two, and it was noted that the clamp's normally straight side was bent 90 degrees, which with the two bolt-holes mated would have created a tennis racquet shape. The section of the damaged clamp, which usually contains the part number, was destroyed.
The two intact clamps were marked with different part numbers; one being LS9345-8, and the other LS9345-7. Review of the photographs revealed the -8 clamp was located aft of the damaged clamp and the -7 clamp was positioned forward of the damage clamp. The -8 clamp was larger in diameter, but the overall shape was distorted, so that when the clamp was closed, it resembled the shape of a tennis racquet, rather than a "P." The wire bundle was placed into the -8 clamp and the fit was snug, but the wires could be moved longitudinally. The bolt could be inserted through the bolt-holes of the -8 clamp. The wire bundle was then placed into the -7 clamp and the fit was tight, but the bundle could be installed and the bolt could be inserted. When the wire bundle was in the smaller -7 clamp, smaller wires tended to get between the wedge of the rubber pad and the ends of the metal clamp band.
Review of Advisory Circular (AC) 25-16 "Electrical Fault and Fire Prevention and Protection," revealed it was created to provide information on electrically caused faults, overheat, smoke, and fire in transport category airplanes. The AC defines two situations called "arc tracking" and "insulation flashover." Arc tracking is defined as a "phenomenon in which a conductive carbon path is formed across an insulating surface. This carbon path provides a short circuit path through which current can flow. Normally a result of electrical arcing." Insulation flashover is a "result of arc tracking" and is an "instantaneous burn-through of the insulated wire with the possibility of continuing the burn into surrounding wires. This failure mode, which is a result of the high temperature degradation of the insulation experienced during arcing, can propagate through a complete wire bundle severing the entire grouping."
AC 25-16 indicates, "Care should be taken to ensure that the clamps and ties around wire bundles do not present a rough surface that may damage the wire insulation. These clamps and ties should be tight enough to hold the wires in place, but not so tight that insulation damage would occur during fabrication or installation, or later in service." The clamps should also be oriented such that abrasion of the wires or the clamp insulation does not occur.
The AC warns its readers to realize that circuit protective devices (circuit breakers and fuses) are considered to be "slow-acting" devices and may not offer sufficient disconnect protection from events such as arc-tracking or insulation flashover. The AC also indicates the "effects of electrical faults can include component overheating; toxic fumes; smoke; fire; damage to wires, wire bundles, or parts...; electromagnetic interference (EMI) with equipment. These effects can be caused by the following conditions or failure modes for which circuit protective devices often do not provide timely, if any automatic protection:
...(viii) Arcing on wire insulation, or the resulting insulation damage. Service experience documents only a relatively small number of incidents of arcing damage for all types of insulation. However, service experience of aromatic polyimide insulation, as previously constructed, documents a failure mode called 'insulation flashover' where conduction at insulation breakdown areas has damaged or destroyed the wire or wire bundle in which it occurs. Also, other adverse effects have sometimes occurred as a result of this failure mode. Arcing on wire insulation, or 'arc tracking,' can result from electrolytic contamination of wire having insulation cracks or cuts that expose the conductor. It can also result from chafing damage that reduces the dielectric strength of dry insulation..."
According to the USAF Aircraft Mishap Investigation Handbook for Electronic Hardware (WL-TR-95-4004, January 1995), the short-circuiting events that continue for periods of time without tripping a circuit breaker are a characteristic of wire insulated with aromatic polyimide (Kapton). The handbook states, "the carbonized insulation is conductive and provides a current path between the conductor and other live electrical components." Once the wires separate, the insulation of the portions that remain connected to electrical power can continue to carbonize, so that the damage progresses from the area of separation toward the source of power.
The FAA's Aging Transport Systems Rulemaking Advisory Committee Report of Intrusive Wire Inspections documented an average of more than 3.5 breaches of Kapton wire insulation per 1,000 feet in an L-1011 and that clamps and other hardware act as focal points for insulation defects.