with Aromatic Polyimide Aircraft Wiring

INTRODUCTION

1. An RAF TriStar aircraft was undergoing maintenance with the APU running when suddenly circuit-breakers popped, aircraft strobe lights flashed and the ground warning horn operated. Finally there was a loud bang and flash from the Left Hand side of the aircraft. After the excitement was over, the investigation revealed that a wiring loom, consisting of 168 wires carrying a variety of supplies had 69 wires completely severed!
2. What could have caused this sequence of events? The investigation revealed that this highly dangerous situation was the result of a phenomenon known as carbon arc tracking. This devastating destruction of wiring is a fault that can occur with a type of cable insulation produced with a polyimide tape (trade name Kapton). How could a cable capable of this destruction be installed on an aircraft?

BACKGROUND TO THE INTRODUCTION OF KAPTON

3. Notwithstanding the advent of data-bus and fibre-optic links, the increased number of electronic and electrical equipments being fitted to modern aircraft, like the Tornado, has resulted in a proportionate increase in the number of interconnecting wires and cables. Whilst equipment manufacturers have been able to take advantage of microchip technology to reduce the size of their products, wire manufacturers have had to develop smaller lightweight wires having a performance equal to, or better than, previous wire types like Nyvin or Minyvin.
4. Wire manufacturers sought new insulation materials which would possess qualities commensurate with more stringent requirements. Some manufacturers turned to a polyimide material in a tape form made by Dupont, possessing exceptionally good dielectric properties and high physical strength. These qualities allowed the thickness of the insulating material to be greatly reduced, in some cases down to six thousandths of an inch, which in turn improved heat dissipation, thus permitting smaller conductors to be used to carry the same current. The resultant wires were therefore thinner and lighter than previous types and showed a vastly improved mechanical performance in laboratory tests.
5. Polyimide insulation with silver-plated copper conductors were approved for use on UK military aircraft in 1972 with the issue of Specification EL 2124, followed by EL 3001 for tin-plated conductors in 1981. Only the Lynx Helicopter and the (then) new-build Harriers had cabling to these specifications for airframe wire. Other polyimide types gained approval under PAN Standards as the main airframe wire for all marks of Tornado and Mil W 81381 for Harrier GR5. Polyimide has also been introduced on many older aircraft by the installation of modifications. Invariably, wire insulated with Polyimide has been hidden under an aircraft manufacturer’s specification, as opposed to the cable specification or NSN which made it unrecognizable to aircraft EA’s (Electrical Artificers?). As guidance was not sought from the equipment EA’s at the time, this type of insulation proliferated amongst mod kits issued over the last eight years as Design Authorities became pro-polyimide.

REALISATION OF THE PROBLEM WITH POLYIMIDE

6. The problem of fitting more wire into less space now appeared to be solved until the US Navy reported that they were experiencing excessive chafing, cracking and damage to the cable looms, especially in severe wind and moisture-prone areas (SWAMP). As these problems appeared to be attributable to a breakdown of the insulation material the US Navy conducted preliminary tests which revealed three potential problems which appeared to be unique to polyimide wire types:

a. Hydrolysis  Hydrolysis is a phenomenon characterised by cracking and breakdown of the insulation material through exposure to moisture; the speed of breakdown depending on both temperature and stress.

b. Wet Arc Tracking  Carbon arc tracking occurs when contaminating moisture or aircraft fluids create a short circuit between an exposed conductor and the aircraft structure or an adjacent exposed conductor at a different potential.

c. Dry Arc Tracking  Carbon arc tracking occurs in dry conditions when one or more conductors are shorted as a result of abrasion from the aircraft structure, wire to wire abrasion, installation error or battle damage.

THE THEORY OF CARBON ARC TRACKING

7. The phenomenon of carbon arc tracking has been known for decades in the Electrical Distribution Industry and many test methods have been devised to classify the resistance of installation materials to failure by this mechanism. Fig 1 shows a strip of insulation material with an electrode attached to its surface at each end; if there is a suitable potential difference between the electrodes and if they are now bridged by a film of wet conducting contaminant, a current of a few milliamps will flow through the moist layer and cause slight heating.
8. This heating will lead to the formation of an occasional very narrow "dry band" in the film (see Fig 2). When one of these bands is formed, most of the voltage between the electrodes is concentrated across the tiny dry gap and a small flashover will occur. This tiny arc cannot be sustained because of the high resistance of the moisture but in a typical situation such dry bands and sparks will occur continually and randomly over the surface in an effect known as scintillation. (ticking fault?)
9. The micro arcs have a temperature around 1000 degrees Celsius and so cause intense heating of the insulation surface on a micro area basis sufficient to pyrolyse (chemically decompose by the action of heat) any organic polymer. The pyrolysis products of the particular polymer will determine its tracking behaviour.
10. In the case of a "tracking" polymer (i.e. Kapton insulation) each scintillation will deposit a micro-spot of carbon char (see Fig 3) with a thermally stable conducting graphitic structure. There is no change in leakage current at this stage and the formation of spots will continue, often with a characteristic ‘tree’ pattern until a sufficiently complete path has formed to enable the next flashover to be sustained as a ‘power arc’ (see Fig 4) through the newly formed low resistance graphitic carbon track. At this point there is an electrical and thermal avalanche effect which will have a magnitude governed by circuit and power source impedance and by any circuit protection devices.
11. In the case of a "non-tracking" polymer i.e. Ethylene Tetrafluorethylene (ETFE) and Polytetrafluorethylene (PTFE) insulation, under the same conditions the intense micro heating of the surface gives gaseous pyrolysis products so that a minute quantity of polymer evaporates away leaving the composition of the insulation surface unchanged. There is therefore no char or track formation and no thermal or electrical runaway. The loss of tiny quantities of polymer (as gas) gives rise to erosion (see Fig 5) and the rate of erosion will depend upon the type of polymer. In the case of a non-tracking fluoropolymer the fluorinated gases given off have the additional property of suppressing electrical arcing.
12. In summary, many aromatic polymers (compounds with carbon rings) are literally converted from insulators to conductors when subjected to very high temperatures (as in pyrolysis). It is this feature which appears to control the susceptibility of aromatic wire insulations to tracking at relatively low voltages (e.g. 16 volts).
13. Examples of predominantly aromatic ‘tracking’ polymers include Kapton*, Peek* and Ultern*. Examples of predominantly aliphatic ‘non-tracking’ polymers (compounds with carbon chains) include Polyethylene (PE), Polyvinylidene (PVDF), Polyvinylchloride (PVC), Polytetrafluorethylene (PTFE) and Tefzel*, a form of ETFE.

*Trade names

DRY ARC TRACKING

14. More recent work shows that bundles of wire insulated with an aromatic ‘severe tracker’ can exhibit total bundle destruction via carbon arc tracking under completely dry conditions such as impact damage or when vibration leads to a conductor making direct contact with a metallic structure or a wire at a different voltage. The small arcs involved emanating from intermittent contact again convert adjacent wire installations to graphitic conductors and lead to catastrophic failure of the wire bundles. Where the wires are insulated with a non-tracking polymer, in an otherwise identical setup, there is no avalanche effect and no extension of damage beyond the initial fault.

TYPE OF DAMAGE BEING EXPERIENCED

15. Polyimide insulated wires can be found in many guises with a variety of part numbers and specifications; a selection of the most common is as follows:

a.	Polyimide insulation, liquid H lacquer topcoat	Mil W 81381
b.	Polyimide insulation. Fluorinated Ethylene Propylene (FEP) lacquer topcoat	EL 2124 EL 3001 PAN 6423
c.	Polyimide insulation. PTFE tape lacquer topcoat	PAN 6411 ACT 150 ACT 260
d.	PTFE tape. Polyimide insulation. FEP or PTFE topcoat	EFA 200 MDS 4480

16. Polyimide insulation can be recognised by its bright translucent copper colour. This is often misinterpreted as the conductor being exposed when the topcoat cosmetic layer has been removed due to damage. Although this has not degraded the cables insulating properties, this mottling of the cable does lead to actual chafes being harder to detect amongst flaking lacquer. Complacency can also creep in with tradesmen assuming it is actually flaking lacquer. A considerable amount of flaking is now being exhibited on the Tornado and BAE 146 and many man-hours are being expended by electrical tradesmen re-evaluating damage that has already been examined on a previous occasion.
17. Cracking and splitting of the polyimide insulation is normally found where cable bend radii have been exceeded or excessive flexing of the wires allowed. This form of damage has been found on numerous occasions within Tornado weapons pylons. This damage has been accelerated by hydrolysis action within the SWAMP areas. Once this type of damage has occurred and the conductors are exposed, the looms are in a primed condition for a wet or dry carbon arc track event. There are already occurrences of inadvertent weapons release from the Tornado due to this type of damage.
18. Polyimide is a tape-wrapped insulation with, in some cases, a PTFE tape-wrap as a topcoat. In manufacture these tapes are sintered together to seal them in place. Experience has shown that flexing and damp conditions allow these tapes to unwrap, once again exposing the conductors.
19. Polyimide insulated wires are stiffer than other types of wire and this has proved to be its weakness in fighter aircraft. It is reluctant to move with areas of vibration and so chafe damage is inflicted on the insulation. Due to the volume of wiring on current fighter aircraft, where space is at a premium, cable is susceptible to inadvertent damage by tradesmen removing or fitting equipments. If this minor scuffing is ignored then catastrophic consequences could result. If we are to avoid a high maintenance bill then husbandry of wiring installations must be improved.

MAINTENANCE OF POLYIMIDE INSULATED WIRING

20. The following paragraphs set out guidelines for the correct maintenance of polyimide insulated wiring.
21. Routing and Tying  Harnesses should be routed and supported well so that they will stay in position and not contact installed equipments or structure. Looms should not be readjusted during servicing and the build standard maintained by ensuring correct lengths between cleating and equipments. Ties should be spaced close enough to hold the harness together, preventing splaying of individual wires where the harness bends to avoid snagging when maintenance is carried out in the area.
22. Harness Twist  Harness flexibility is greatly improved if the wire are twisted (one to two turns per foot) prior to tying. Twisting allows the harness to bend by rotational motion of the wires, rather than trying to stretch the wires. This is carried out in initial build and should be maintained during rewire and modification action.
23. Hinge areas  Harnesses which must cross areas of relative motion (eg wing fold mechanisms, landing gear and hinged access panels) should be mounted so that the loom will twist rather than bend when the joint moves. Care must be taken to ensure that the loom does not bind or pinch during motion.
24. Connector Strain Relief   It is important that strain relief supports be reinstated after maintenance and that the wire is properly routed and supported by them. This is especially relevant with the single tie-wrap post back shells which do not have the same grip on the cables as the older saddle clamps. Accurate cutting to length of the loom should be carried out especially on 90 degree clamps, in order to avoid single wires being placed under stress.
25. Bend radius  Harnesses should be installed with bends as generous as possible to avoid strain on the wires as the structure flexes. The minimum bend radius allowed is six times the diameter of the loom or ten times the diameter of the original wire – whichever is the greater. The minimum bend radii have been exceeded on some fly leads at the rear of connectors even during aircraft build. This practice should not be repeated on re-wires.
26. Use of Test Prods  Piercing of the insulation for testing purposes is STRICTLY FORBIDDEN.
27. Stripping and Crimping  Great care must be taken to avoid any insulation damage with stripper and crimping tools. Any scuffing of the insulation from blades or grips is to be considered unserviceable and the cable re-prepared.
28. Wire Marking  All marking must be carried out using the tape dwell time, temperature and pressure recommended by the wire manufacturer. Where hot stamp identification marking is involved, a mandatory high voltage spark test must be carried out after marking. A badly printed cable, which burnt through to the conductor, was the cause of six feet of loom being destroyed in a carbon arc on a civilian airliner. ‘Inter-connect’ and ‘Equipment Wires’ should not be hot stamp printed.
29. Inspections   Every opportunity should be taken to inspect installed wiring for signs of damage or chafing. This should be done when equipments or panels are removed and access can be gained to normally obscure areas. Inspection is especially important after servicing has been carried out and is essential prior to panelling up. All trades should be aware of the dangers of even minor damage to Polyimide and an electrical tradesman with 10X magnification should be called to inspect all instances of suspected damage. Cable looms should be maintained clean and dry and contamination of cables by toilet and galley waste should be rigorously prevented.
30. Damaged wire   If any flats are detected on the translucent copper polyimide insulation, apart from missing lacquer, then it should be assumed that some of the insulation has been removed and the cable should be repaired or replaced. Repairs to the cable should be carried out in accordance with AP100B-01 Order 4901 using the environmentally sealed in-line splices.
31. Wire Installation  The practice of ‘pulling through’ cables during replacement should be avoided. If cables are ‘laid in’ then damage to the top-coat lacquers would be avoided and the snagging of PTFE topcoats, causing the tapes to separate, would also be prevented.

to Tell-Tale Docs