Kapton: Miracle Material With A Tragic History

On a balmy September evening in 1998, Swissair flight 111 was in big trouble. A fire in the cockpit ceiling had at first blinded the pilots with smoke, leaving them to rely on instruments to divert the plane, en route from New York to Geneva, to an emergency landing at Halifax Airport in the Canadian province of Nova Scotia. But the fire raging above and behind the pilots, intense enough to melt the aluminum of the flight deck, consumed wiring harness after wiring harness, cutting power to vital flight control systems. With no way to control the plane, the MD-11 hit the Atlantic ocean about six miles off the coast. All 229 souls were lost.

It would take months to recover and identify the victims. The 350-g crash broke the plane into two million pieces which would not reveal their secrets until much later. But eventually, the problem was traced to a cascade of failures caused by faulty wiring in the new in-flight entertainment system that spread into the cockpit and doomed the plane. A contributor to these failures was the type of insulation used on the plane’s wiring, blamed by some as the root cause of the issue: the space-age polymer Kapton.

No matter where we are, we’re surrounded by electrical wiring. Bundles of wires course with information and power, and the thing that protects us is the thin skin of insulation over the conductor. We trust these insulators, and in general our faith is rewarded. But like any other engineered system, failure is always an option. At the time, Kapton was still a relatively new wonder polymer, with an unfortunate Achilles’ heel that can turn the insulator into a conductor, and at least in the case of flight 111, set a fire that would bring a plane down out of the sky.

Space Age Stuff

Electronics hobbyists can recognize Kapton on sight. The familiar rich amber color shows up inside so many devices, from the flexible PCBs that link a laptop’s display to its motherboard to the seemingly miraculous adhesive tape that can withstand soldering temperatures or line a 3D printer’s bed for better adhesion. Kapton has become an indispensable material in electronics manufacturing.

Along with nylon, Teflon, Kevlar, Neoprene, and dozens of other polymers, Kapton was the product of the chemists at DuPont in Wilmington, Delaware. Kapton is an aromatic polyimide with outstanding thermal and dielectric properties. It can do things few other plastics can do, like withstand temperatures from nearly absolute zero to 400 °C. Formed into a film and aluminized on one side, Kapton made an early and dramatic public debut as the golden insulating blanket wrapping the Apollo lunar lander descent stage.

It didn’t take long for DuPont to discover myriad uses for Kapton, and by the early 1970s, the space age stuff was being used to create ultrathin, ultralight insulation for electrical wires. Aerospace companies latched onto the idea quickly; with hundreds of kilometers of wiring in a modern jetliner, shaving off a few grams of weight from each wire translates to huge weight reductions, leading to fuel savings, increased range, and increased payload. Driven by concern for the bottom line, Kapton made rapid inroads into passenger and military aircraft throughout the 1970s and 1980s, appearing in almost every major aircraft. Even the Space Shuttles were outfitted with the stuff.

Arcing and Charring

Degraded Kapton wires showing cracks and carbon charring due to arcing. Source: Lectromechanical Design Company, LLC

But the aerospace industry’s love affair with Kapton wiring would soon sour. As early as the 1980s, the US military was beginning to see aircraft fires and crashes related to electrical short circuits. Studies showed that Kapton was far from the ideal insulator everyone had hoped it would be — it tended to develop circumferential cracks from the slightest of nicks, exposing the conductors within. Kapton is also easily degraded by moisture, exacerbating the problem in humid environments or areas of an aircraft subject to moisture, like galleys and lavatories.

Once the insulation is compromised, arcing can occur, which leads to charring of the Kapton. This changes the insulation’s dielectric properties, turning it into a conductor. In some cases, this led to overloaded circuits that were not detected by circuit breakers, since the insulation was carrying the excess current rather than the conductor. Other times, the circuit breaker would trip to protect the circuit, but when the crew reset the breaker, the charred insulation would catch fire and burn like a fuse, flames traveling along the wires far from the original arc.

The US military mothballed many Kapton-equipped aircraft in the late 1980s. In 1999, a short-circuit during the launch of Columbia caused two of the engine control computers to fail, and NASA grounded the entire shuttle fleet to remediate Kapton insulation failures. Recommendations were made to develop methods to test all the wiring on the shuttles, but the fleet was retired with Kapton wiring still in place.

Boeing continued to manufacture planes with Kapton wiring until 1992; the Swissair MD-11 was one of the last Kapton-equipped planes to roll off the line, in fact. Planes built since then generally use TKT, which is Kapton with two thin layers of Teflon to provide better resistance to mechanical abrasion and water intrusion. Still, hundreds of planes take off and land every day with Kapton insulated wires routed through every nook and cranny, some almost impossible to reach for inspection. And even when wire bundles are accessible, inspection comes down to a flashlight and a Mark I eyeball, looking for that cracks that might be lurking.

Featured Image Credit: By Aero Icarus from Zürich, Switzerland (28as – Swissair MD-11; HB-IWF@ZRH;14.07.1998) [CC BY-SA 2.0], via Wikimedia Commons

63 thoughts on “Kapton: Miracle Material With A Tragic History

      1. Even the Chinese know thier products are crap and products with English looking names sell better OVER THERE, but most of the serfs/peasents over there can’t read English so they just notice the Roman style characters and assume the product is better. Local merchants caught on an started naming producs in jibberish Roman letters because it then looked imported/quality because the merchant’s couldn’t read English either. There are lots of truthful/non-nondeceptive people over there but they are islands in a sea of trickery some of which is an amazingly low-grade lie, kind of like the products. ….10A SPST switch, yea right, try 3A wihout it melting.

  1. “At the time, Kapton was still a relatively new wonder polymer, with an unfortunate Achilles’ heel that can turn the insulator into a conductor, and at least in the case of flight 111, set a fire that would bring a plane down out of the sky.”

    Wonder technology is kind of like that. Just look at the history of radioactive substances.

    1. Low-level IS/LFN (low-frequency noise) from slow-running fans in poorly designed ventilation systems and heat exchanger units triggering serious CNS-pathologies in (office)workers and residents. Well documented since the early 80s, still being ignored by many manufacturers…

      1. Right… and whomever is savvy in electronic surveillance (ES) or technical surveillance (TS) and electronic warfare (EW) or even just audio injecting signals into any device that makes a noise using either wired or wireless methods.

        I’ve blogged my concerns on: http://dewdetectionprojects.blogspot.com/2017/08/

        Consider the human body resonance frequencies like EEG, ECG, EMG, etc. that are publicly disclosed. Then there are AC and DC analog and digital components if I understand correctly that are not well disclosed.

        We know bi-neural beats have an effect and Rasmussens work goes into the subsonic and sonic details a little: http://www.donmar.com/Tech/VIBRATION-BODY.pdf

        Also, I am concerned with microwaves more and the combinations of phased synchronized signals that combine to the 95GHz and I think other GHz ranges are more directly connected with the human body though not well disclosed. Like permanent damage can occur even with the 95Ghz system if over exposed or used incorrectly. I only found on GBPPR noting 95Ghz and one other site referencing a German ( “Bembenek P: Akupunktur und bio-resonanz (in German), CO’MED Nr. 6:50, 1998” article of reference. ) paper I haven’t been able to find yet: http://ieminstitute.org/_Resources/Articles/HEFmesurementRubik.html

        That’s not only the carrier frequency effects… there can be pulse train duty cycle effects also. Like say a 95GHz carrier with a modulation in the sonic or ultrasonic range.

        1. I worked for a military contractor in 1972, and environmental tests were already standard then. Of course, new or more stringent tests come about when failures show that existing tests are inadequate.

  2. I see three technical problems in this story:

    First, why the cracks form? Is there some sort of mechanical stress induced by environmental factors (chemistry, thermal swelling of metallic conductors, something else)?

    Second, how could the existence and location of cracks be detected? Obviously, there are spots on the wire where local capacitance between the wire and neighbouring wires is lower, because there is a column of lower epsilon-r above that spot. In theory, that should disturb impedance of the conductor and a Time Domain Reflectometer should see something (although faint, but always at the same spot, so integrating over many firings should bring it out from the noise).

    Third, how would you mend those cracks? Perhaps a very long capillary tube could be inserted into the wiring harness and very low viscosity adhesive could be delivered at the spot, where it would wet all the surfaces, including the sides of the cracks, and then it would solidify (hopefully into an elastic, rubbery substance) …

    1. Rewind the video to the beginning and you’ll see a good explanation about the test. The test intentionally nicks the wires with what looks to be about 1 cm spacing between adjacent nicks, drips 1% saline solution on it, and wires are powered by 120V 400 Hz large generator. It simulates what’s been seen in old aircraft wiring such as having metal shavings on wires, and nicked wires near the lavatory dripped with lavatory “blue water”.

    2. Aircraft vibrate, a lot. Also, moisture is a problem due to lower temperatures at high altitude. Bleed air from the engines help heat the cabin/cockpit and this means moisture trapped inside the frame areas will turn into condensate. We have to spot tie every so many inches to keep wires from running due to compounded vibration over time. You would be amazed to see how much damage can occur. A lot of the inner areas where wiring is run will have plenty of carbon and oil residue from jet exhaust leaking in. All of these factors are what play into testing for a real disaster. That’s not even including the potential for foreign object debris (FOD) from things like aluminum chips, bucked out rivets and an occasional screw/nut/bolt bouncing around. This can have disastrous consequences.

      1. Yes, TDR is used to find wiring faults all the time. It is very easy in coax and twisted pair. The US Navy uses it extensively to find opens/shorts, etc in shipboard wiring. I would imagine it also gets used on board aircraft. I have used it to find bad coax in the runs up my ham radio antenna tower.

  3. It isn’t just kapton, The drive to shave weight or material costs goes to much of the wiring industry. The insulation on regular household wire is actually foamed to reduced material costs. It still meets all the design specifications for insulating properties,degradation over time, etc. It just has worse mechanical properties. If it is installed correctly and carefully, no problem. But how many cookie cutter construction companies ensure that every worker knows how to handle incidental interaction with wiring or PEX or many other modern construction materials?

    1. Aluminum wires, too. In the 1970s, it was for cost reasons, because the price of copper had spiked even higher than usual above aluminum. They adjusted the wire thickness for aluminum’s higher resistance, but didn’t change the way it was terminated. After quite a lot of house fires, the naysayers were finally taken seriously and new standards were enacted to account for aluminum’s very different behavior under screws, inside wire nuts, and elsewhere.

      You also see aluminum wire sneaking into cheap knockoff goods: Network cables, jumper cables, pretty much anything that doesn’t have to be UL listed. Copper-clad aluminum looks pretty similar, too, unless you cleave the end and look carefully.

      Recently, there’s a push to use aluminum in automotive wiring, principally for weight reasons. The stuff can’t be soldered, and it works out of regular crimps, so there are exotic ultrasonic-weld machines being used for new termination styles. The tech is neat — the vibration breaks up the oxide layer during the operation — but there are still widespread doubts about its long-term reliability.

      1. And aluminum bring bigger resistance, so more heating in cables.

        Also, for the non-copper cables in non-regulated uses, I bought some thin electric wire from a common electronics shop for a quick fix in a project, and was disappointed to see that what I expected to be copper cables would get attached to even weak magnets ….

        1. I had a similar experience, though instead of being copper plated iron wire, I found 3 strands of copper wire inside a HEAVY insulation layer.

          It “looked” like 18 AWG, until you actually removed the connector from either end and stripped back the insulation.

          1. Also many electronic parts have copper clad iron terminals or contacts. Like electrolytic capacitors, resistors, dip chips, almost everywhere… Not only leaded parts and not all of them either, some just have iron/nickel clad contacts or similar as a barrier. Get a magnet a check your parts stash.

      2. The difference between copper-clad aluminum and full core copper wire is made quite obvious once you torch the bare wires. With stranded wire a lighter is just enough. The aluminium melts almost right away and the copper gets orange red hot before melting.

        1. I’m curious about this as well. Soldering aluminum requires getting through the oxide layer, which has an incredibly high melting point. Special blends of flux would be required unless rosin based flux can dissolve the oxide. I hadn’t heard that it can.

          1. I haven’t tried it myself, but I’ve been told (and seen in a YouTube video) that the oxide layer can be removed by scraping the surface of the aluminum under oil. The oil is necessary to prevent a new oxide layer from forming immediately when heat is applied. This was using an iron – not sure how it would be done with a torch.

  4. Another good reason to use Kapton (polyimide) insulation in space applications is its excellent resistance to radiation. Teflon (PTFE) has fairly low radiation resistance. Tefzel (ETFE) is quite good (and more flexible but Kapton), but gets damaged easily.
    In my application (low voltage, low power satellite components), Kapton is king. However, the wire we use does have an innermost layer of PTFE around the conductors, then 2 layers Kapton.

  5. Kapton used to be used at the outermost layer of thermal-insulation blankets used on various spacecraft, including satellites (giving that gold look when layered over reflective metalized mylar), but when I worked for the Space Dynamics Lab/USU in the 1980s we stopped using the stuff because of “ablation” –the wearing-away of the polyimide molecules by the extremely-high UV radiation found even in the upper atmosphere. We switched to white Beta cloth, layered with metalized mylar and polyester netting–about twelve layers of each.

  6. “In some cases, this led to overloaded circuits that were not detected by circuit breakers, since the insulation was carrying the excess current rather than the conductor.”

    Much more likely that circuit breakers didn’t trip because the overload wasn’t a direct short – the compromised insulation formed a medium to high resistance path that didn’t draw enough current to trip the breaker. Insulation doesn’t give you some magic end-run around the breaker.

    1. Also, circuit breakers are sized according to the current carrying capacity of the wires, not the current carrying capacity of bits of carbon that aren’t supposed to be there, and it’s the carbon that got hot enough to start fires, even though the current was below the rating of he wires and therefore the trip rating of the breakers. This is also what led to the many house fires in the 1970s involving aluminum wiring. The aluminum wire under screw terminals would cold-flow, reducing the pressure on connections. This, combined with the slightest amount of oxidation, would cause local heating at the contact point, even when the current was well within the current capacity of the wires.

  7. I worked for the company that designed and built the entertainment system for Swissair. Things were a little tense there for a while, when the fault was initially traced to the entertainment system itself.

  8. If memory serves right this was the stuff that caused the TWA800 downing too. It was supposedly pulled from Air Force One and a number of military aircraft almost immediately after the NTSB report came out. The rest of us? Keep flyin’ and hopin’

  9. New materials are great, 30 years after discovery. I fear to see what was missed in the transition from lab-to real-world in the “new” composite air-liners. I will stick with the well understood 100 year old aluminum aircraft when given the choice. I hope history does not repeat it self, but it tends to..

Leave a Reply to Hiro ProtagonistCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.