Takata’s Deadly Airbags: An Engineering Omnishambles

Engineers are, for the time being, only human. This applies even more so to executives, and all the other people that make up a modern organisation. Naturally, mistakes are made. Some are minor, while others are less so. It’s common knowledge that problems are best dealt with swift and early, and yet so often they are ignored in the hopes that they’ll go away.

You might have heard the name Takata in the news over the last few years. If that name doesn’t ring a bell you’ve likely heard that there was a major recall of airbag-equipped vehicles lately. The story behind it is one of a single decision leading to multiple deaths, scores of injuries, a $1 billion fine, and the collapse of a formerly massive automotive supplier.

Cost Savings Led to Airbag Failure, Injury, and Loss of Life

Takata Corporation began manufacturing airbags in 1988, initially using sodium azide as the propellant. This was difficult to handle safely, and so in the 1990s, production switched to tetrazole, a safer alternative. However, this came at a high price, so research was undertaken to find an alternative.

Takata made the decision to switch to using ammonium nitrate as a propellant. Prone to breaking down under hot and humid conditions, it can become unstable, leading to airbag modules exploding rather than safely inflating in a crash. In these cases, the airbag igniter case can rupture, spraying metal fragments throughout the vehicle. This can lead to serious injury or even death for occupants of the vehicle.

The properties of ammonium nitrate were well known; other airbag manufacturers had long ruled out its use due to its stability issues. Engineers warned in the late 1990s that the material was unsafe for the application, long before the parts in question were installed in vehicles. In spite of this, Takata forged ahead, winning contracts with major automakers. As the years went on, millions of airbags with ammonium nitrate propellant would be installed in vehicles worldwide, and the timebombs began ticking.

Event Zero Was Just The Beginning

The first rupture of a Takata airbag was in May 2004, occurring in a 2002 model year vehicle in Alabama. According to the 2016 NHTSA report, Takata was first notified of the event in 2005. A month later, a Takata engineer presented their conclusions based on the limited data available to them.

Photos from Event 0, the first noted case of a Takata airbag rupture. Source: NHTSA report

Based on photos of the aftermath, there was significant rust inside the inflator, which appeared to be present prior to the airbag firing. The cause of the problem appeared to be due to overpressure caused by the propellant, rather than substandard strength in the inflator casing material. After further research, Takata concluded the problem was likely due to improper sealing on the airbag components, and further action was not taken.

It would not be until 2007 that Takata learned of further dangerous airbag failures. By this time, millions more airbags had been installed in cars worldwide that included the dangerous inflators. Takata increased its research into the defects, searching for the root cause of the problem. This included a focus on investigating propellant density, as well as exploring whether there was a problem with poor seals on airbag components. In August 2009, the company had still not found a concrete cause of the fault, and enlisted the assistance of two independent outside companies, Fraunhofer and BakerRisk, to investigate the problem.

A Fault Tree Analysis document from within Takata shows that the investigation ruled out a fundamental chemistry issue early in the investigation. Source: NHTSA report

The investigation continued to go in circles for years. New failure events continued to crop up, with airbags from production lots outside the initial batches found to have problems. Key to the investigation was the “Stokes” press, which was suspected of producing low-density inflator materials which were more likely to cause ruptures. Engineers chased their tails trying to determine which airbags were produced on which press. Production lot numbers proved to be an unreliable record of which airbags were produced on which equipment, frustrating efforts to determine the cause of the problem.

Years went by, with teams investigating all manner of leads. In 2013, Takata were notified of the first case of an inflator failure from a batch manufactured outside the initial 2000-2002 period of interest. These were referred to as “Beta” failures, and began to highlight the broader issue at play. Fraunhofer’s continuing research began to show that instead of manufacturing issues, the problem was most likely caused by moisture and temperature cycling. At this point, the number of affected cars began to spiral out of control, as it became clear that few-to-none of Takata’s ammonium nitrate inflators were safe in the long term.

Early fault tree analysis from Takata shows that engineers were quick to rule out chemistry issues with their product. Instead, there was a strong focus on finding a manufacturing variability issue instead. While the investigation took its course, Takata factories continued to pump out ammonium nitrate inflators in their millions. This compounded the issue, as each new car sold with defective airbags would later become subject to a recall, requiring parts to be replaced.

A Two-Pronged Problem

Honda are the most closely associated manufacturer with the problem, however cars across over 10 manufacturers are now subject to recall.

Identifying the root cause of the issue was just the first, and easiest, step of the process. To solve the problem once and for all, millions upon millions of cars, from a huge number of manufacturers, would need to have their airbags replaced. The effort required to achieve this is truly Herculean in scope, and it has been far from smooth sailing thus far.

For cars with a single owner from new, it is a relatively simple task for dealers to contact the owner and send out a recall notice. However, with cars sold on the used market, records are poor to non-existent. Trying to educate the entire car-owning public that their car may be packing a dangerous explosive device is a difficult and expensive task. Car companies have taken out advertisements in newspapers, at stadiums, and gone so far as hiring private detectives to track down owners. Websites and other services have been created in various countries to allow motorists to check their Vehicle Identification Numbers against databases of recalled vehicles.  Despite this effort, and the fact that recalls have been ongoing since 2008, the message isn’t always getting through. And that’s only half the problem.

Parts Were Slow to Materialize for Recall Service

It’s one thing to know that your car could potentially have a defective and dangerous airbag installed. It’s another thing entirely to get the problem rectified. Parts availability has been a continual problem during the recall process. Owners have been told that they will be notified to bring their car in “when parts are available”, and told to avoid using the front seats in the meantime. Obviously, this is of cold comfort to drivers who rely on their cars for transport on a daily basis.

The unprecedented volume and scope of the industrywide Takata airbag inflator recalls have created unique replacement parts challenges for all vehicle manufacturers and global parts suppliers. At NHTSA’s request, Ford issued safety recalls for vehicles with certain Takata inflators even though replacement inflators with a different propellant would not be available for months or even years; in the meantime, new inflators are being redesigned and manufactured.

-Statement from Ford Motor Company, 2018

Given that cars were produced with defective inflators for well over a decade, it’s not a problem that has an easy fix. Airbag manufacturing is a niche industry, and production capacity exists to serve the demand for new vehicle production each year, and to maintain spare parts inventory. It simply isn’t possible to produce several years worth of new airbags for older vehicles overnight. This leaves many drivers with a difficult decision to make — to hope that their airbag is safe, or acquire another vehicle, often at their own expense.

This Isn’t a Small Fix

Unfortunately for Takata, and millions of drivers around the world, the problem wasn’t a short-lived manufacturing error. The choice to use ammonium nitrate was a fundamental design flaw, meaning entire product lines were affected.  This meant that the problem wasn’t as simple as correcting a production line error, and cranking out more parts. Airbags for hundreds of different models would have to be redesigned, tested, and put into production.

Ford is currently engaged in the complex and lengthy process of developing replacement nonammonium nitrate-based airbag inflators from alternative suppliers. Airbag inflators are highly engineered and unique to each model and model year. In addition to the lengthy design and revalidation process, the scarcity of global production capacity from inflator suppliers is contributing to a delay in global replacement parts availability.

-Statement from Ford Motor Company, 2018

Unsurprisingly, this caused headaches, if not migraines, for automotive companies the world over. Focus was placed on replacing parts in older cars first, in areas with high heat and humidity. The airbag industry is highly concentrated, however AutoLiv, TRW, and Daicel have stepped up to the plate, devoting their own resources to the effort to get replacement parts out in the field.

A statement from Ford’s website. Manufacturers are taking the recall seriously, but are struggling to effectively reach the public.

For many people, the fix still hasn’t come fast enough. As recently as 2016, vehicles were still being sold with defective inflators, with automakers admitting that these would have to be recalled by 2018. With the Takata airbags becoming more dangerous over time, the thinking is that these cars would likely be safe enough while resources are devoted to fixing riskier older vehicles. Owners of cars from certain manufacturers have been lucky enough to receive loan cars, while others continue to drive around, in the hopes that nothing goes wrong.

The Fallout

In 2016, Bloomberg reported over 13 people had died, and 100 had been injured, due to the defective Takata inflators. By 2017, Takata had filed for bankruptcy, with the cost of the recall and related liabilities exceeding $10 billion. The company was bought out by Key Safety Systems, with the aim being to keep the factories open and churning out replacement airbags with as minimal disruption as possible.

Millions of cars remain unrepaired in the US, with similarly worrying numbers elsewhere. Parts shortages continue to frustrate the repair effort, and many owners remain unaware of the dangers lurking behind their dashboard. Deaths continue to stack up, with the most recent case being a 2002 Honda Civic claiming the life of an Arizona man in June 2018. The vehicle in question had been under recall for over 3 years, with no repairs undertaken in that time.

Hindsight Is A Wonderful Thing

Producing airbag inflators with a dangerous propellant was a mistake, and one that would have dire consequences. However, the bigger mistake was ignoring the problem at every turn when it came to light. Thanks to the dedicated efforts of Takata to avoid facing the problem early on, millions more cars would ship with defective parts, and many more drivers and passengers would be placed in danger. This further complicated the efforts required to solve the issue.

Industry bodies and automakers continue to find themselves trapped between a rock and a hard place — trying to get the word out to customers, while simultaneously tell them they must wait until replacement parts are available.  It’s a painful communication problem with no easy solution. The major lesson going forward would be to stamp out spot fires before they become raging, uncontrollable conflagrations that threaten lives the world over. With Takata only pleading guilty in 2017, it’s clear that the problem could have been avoided entirely if early warnings had been heeded. It’s rare that a single engineering decision can lead to such a quagmire of pain and misery, but in concentrated industries with few suppliers, the stakes can be much higher.

Going Forward

The Takata airbag debacle will be an example case taught in business and engineering classes for decades to come. The very earliest warnings were ignored, and by the time the real problem became apparent 15 years later, the damage was already done. The biggest automotive recall in history is still in progress, and the after effects will likely be felt for some time yet. In the meantime, check your own vehicle, and advise your friends and family to do the same. And, as ever, if you’re warned by an engineer that you’re making the wrong decision, it might just pay to listen, in both dollars and lives.

116 thoughts on “Takata’s Deadly Airbags: An Engineering Omnishambles

  1. Not all cars have ever had airbags. If the faulty airbags are more dangerous than no airbag then would it not make sense to simply disconnect the potentially faulty airbag while awaiting replacements to be available?

    A waiver document might be required for the annual inspection if this leads to a fault warning, though.

    Something that isn’t 100% clear from this article is if the airbags spontaneously deploy (eek!) or deploy improperly only in a crash?

    1. Some vehicles will not allow the vehicle to be started if the driver’s airbag is disconnected.

      The inflator ruptures under ordinary deployment circumstances. There is no connection to unwarranted deployment.

      1. But you also have to consider the lives that are saved by working air bags. Far more people have been saved by even the defective air bags than killed by then. So even though they had a dangerous design flaw, disabling the air bags until a fix was ready would have resulted in even more deaths. It’s why OEMs continued to install and sell the defective inflators that they knew they’d have to recall with full knowledge and support of regulators. There simply was no easy fix and doing so was the best of the available options.

          1. Which is what they’re trying to do “…the thinking is that these cars would likely be safe enough while resources are devoted to fixing riskier older vehicles”

          2. Except they’re not removing the unsafe airbags. They’re replacing them, which isn’t happening fast enough because there aren’t enough made.

            If you can’t get a replacement soon enough, the defective airbag should be removed.

      2. This function can and will be overridden in case of an officially mandated removal. In the worst case it would have been replaced with a dummy electronics unit without the explosive charge.

    2. They don’t deploy spontaneously. The problem is that the compromised explosives burn faster than they were designed to and this causes the pressure to rise too rapidly. The inflator has a precisely sized opening that is covered with a burst disk calibrated to a specific pressure. When the pressure rises too fast the gas can’t exit through the opening fast enough to prevent the pressure from exceeding the limits of the inflator vessel. This can create a shrapnel bomb as the inflator vessel tears apart at probably 20k+ psi. Or if you’re lucky the vessel mushrooms open While staying in one piece.

    3. There has been one case where deployment occurred while the man was detailing his vehicle. Also, and this is not in the news, a technician was injured when an airbag deployed during removal. Honda has had the most incidents and is tirelessly working to locate all vehicles and owners and repairing the vehicles onsite, no dealer appointment necessary, even on weekends. The older the vehicle the more dangerous and likely to explode they are. 2001 and 2002 Vehicles are the worst and have a 50% failure risk. Do you want to trust your life and well being to a flip of a coin? If a Honda rep shows up to your house and you have the car, get it fixed, even if it doesn’t run, it’s dangerous. If you no longer own it, be as honest and give as much info to the rep as you can, it could save a life.

      1. Accidental deployment while working on a vehicle is a completely separate issue. Improperly disconnecting sensors or inflators while working on a vehicle can cause deployment of any kind of airbag, not just the Takata ones. The problem is that because collisions can cause severing of wiring harnesses, the airbag systems are designed to operate even after power has been removed. In the case of my car (1997 Mountaineer), the service manual specifies that the battery must be disconnected for at least a full minute before disconnecting any airbag components. Airbag deployment when working on a vehicle is very dangerous because the person doing the work is generally not in a position relative to the airbag, that the airbag system is designed for. Or to put it more simply, you CAN get injured or even killed by an airbag that is in perfect working order.

        1. OK, that’s the reason for the big cap in an Airbag control unit somebody did a teardown of.
          Like an anti-tamper fuze in some kind of ammunition. “If something could be wrong or sombody fiddles with the device and you are in doubt, just detonate it.” I did not know, airbags are designed like this.

  2. Then, preventively, why not just remove the offendind parts and leave the car without them while waiting for the replacement parts ?

    Driving safely is something that car drivers can manage by themselves, since the failure of the part is out of anybody´s control.

    1. I am not sure I agree with this analysis, but it’s the official line:

      “I’m concerned that my airbag isn’t safe. Should I disable my airbag?
      No. NHTSA estimates that frontal airbags saved 2,400 lives in 2014 alone. According to NHTSA, it is far more likely that your airbag will perform properly and protect you than rupture and cause harm should you be involved in a crash. An airbag that is purposely disabled has a 100 percent chance of failing to provide any protection in a crash.”

      (from https://owner.ford.com/service/frequently-asked-questions-regarding-takata-airbag-inflator-recalls.html )

      I guess that the real message here is “consider not crashing”

      1. IF your airbag is compromised it’s best to disable, if you can. But chances are it isn’t compromised. If you live in a low humidity area you’re probably ok, but if you live near the sea and your car is old… you’re throwing the dice.
        Nissan tried to avoid replacing the Takata airbags in low-humidity areas (whole countries/regions), but was finally forced to, at least in most of the EU some time between 2014 and 2018 I think.
        Mine was replaced, supposedly (god knows what they did… inspect it for moisture damage? swap it for a newer but equally prone to failure model? -these crooks know no boundaries). Fortunately it’s very dry here so even if they swapped it for another ammonium nitrate airbag I’m most probably ok.

        1. If Takata only has 1% of the market then the answer would be 24, twice as many as have been allegedly killed by the airbags. And I suspect that the actual market share is nearer 25% (based on three other manufacturers being mentioned)
          So despite the quote not specifically saying so, it does appear that a Takata airbag is more likely to save you than to kill you. Not that that is much of a comfort when you are picking shrapnel out of your body parts.

          1. That’s a faulty appeal to statistics, because the old airbags are only getting worse and the probability of killing the driver instead is creeping up as the years go by.

            Most accidents are survivable without the airbag, so the 2,400 estimate is also questionable. It’s based on accidents without airbags vs. accidents with airbags and comparing the difference, but, this is not a fair comparison because cars without airbags are older and/or cheaper, and driven by different sorts of people – such as your teenage son crashing his cheap 90’s Honda Civic vs. some pensioner buying a new Nissan to drive to bingo night.

    2. On August 7th 2018 I was in a fender bender and the air bag exploded and knocked me out and thru shrapnel in my arms and my head, and fracktured the femur Bone in my chest, this totally devastated my whole world, I have reported this incident to Honda,tekati, and nobody has offered me any relief, just bring your car in and we’ll see if your on the recall list, well of course I am ,I have a 2005 Honda Accord. What’s the problem?

        1. Yes, but was the femur bone in his chest before the accident as well? And was it “his” femur bone, or his parasitic twin? Was it even a human femur bone? This begs so many questions.

          1. What the heck are you even trying to say,is that what you like to do argue, distract and confuse people who are seriously try to deal with something that has put a effect on a persons life and turn it into a big joke? Really?

    3. Yes. good question.
      Does it come down to probability ?
      ie: % chance of having an accident where airbag deploys and stop injury vs % risk of it going off on it’s own randomly / ticking time “bomb”.

      And the cost of removing them (typically there is a method to do this of course for after they deploy or end of vehicle life) verses leaving them in situ till a fix comes.
      At least for the UK, modifying the airbag system is considered an MOT fail ** so you wont be allowed to drive the car on the road without the aribag fitted and working – no fault light lit
      Of course there are plenty of ways to fool that, but you’ll run the problem of insuring a now modified car too ***
      The manufactuers I suppose may need to get type approval on board as technically they wouldn’t have the same EuroNCAP rating wiithout the airbag.

      But would a manufacturer want to put the driver at risk?
      Is it just the driver air bag or all of them in the car including in some models the curtain deployments?

      Also in the UK we tend not to get the same level of recalls that you might see in the US for the same models. For example the old Accord which was for the UK market built in the US and essentially the same never got the dealer level recalls that the US did.

      But as ever, it’s the company which is guilty. A company is not a person, they are not a being which makes concious decisions. They are made up of many parts and frankly the management should all be going to jail for this.
      But that’s rarely the case. The company gets wound up, a few people get token jail sentences, but nothing really gets fixed. No execs are held liable, the board just quieltly takes their payment and moves on. Everyone claims they never knew – but funny isn’t it, how when things are on the up everyone wants to know about it / be associated within corporate world.
      And so this will continue to happen like it’s happened for the last hundred years or so – corporate manslaughter / murder without real punishment of the exec/board that is responsible – is why they get paid so well.

      ** unless a bonefide rally car which requires an MOT for road stages
      *** people do it anyway and technically they are not insured because they dont declare it

      1. In both of the cars I owned during the recall period, it was only the passenger side airbag that was recalled: the driver side was apparently okay. I just didn’t take passengers during the recall period. (And I live in a desert, so the chances were lower that it would deploy from humidity-caused degradation.)

  3. So… For cars on the road to be legal, you *register* with the state for a plate. It requires a VIN.

    If this recall is as critical as “someone set you up the bomb. airbag detonation ON”… Why aren’t we using govt records to contact people?

      1. You may want to contact Honda directly at this link if you haven’t already.
        http://hondaairbaginfo.com/

        I bet Honda will make completing the recall on your 2003 Accord a high priority. I’ve been very impressed with the effort Honda is putting forth to get the PSAN inflators off the road. When they completed the recall on my Pilot, they did not replace the recalled inflator with another Takata PSAN inflator that had desiccant, which Takata claims is a remedy to the defect. Honda doesn’t buy that, and neither do I. Honda is putting non-Takata inflators back on the road.

  4. My 2K4 Honda CR-V has had airbags replaced/recalled twice…
    I guess the first replacements were still using ammonium nitrate.
    (that’s the stuff Timothy McVeigh and Terry Nichols used to bomb the Alfred P. Murrah Federal Building in Oklahoma City on April 19, 1995)

    1. How the hell did you get your hands on a Honda CR-V, from the year 2400!?

      (If I could travel through time and bring things back from the future, I can honestly say that a Honda CR-V would be one of the last things that I’d be bringing back…)

    2. That’s also the stuff that devastated a chemical fertilizer plant in the early 20 century and it’s used in many mining explosives.
      So it is just a very common material for civilian use. I think the military does not use it very much, perhaps they have higher stability requirements.

      1. It’s not a stability issue, ANFO is used as an improvisation so it’s a short term thing, it’s not used much because it has a fairly low detonation rate of 2500 metres a second whereas plastics typically detonate at 5000 metres a second.

        1. ANFO of course can be an improvised thing.
          But there are several civilian/mining explosives of similar chemical composition – AN and some plastiziers/gelatinizing agents or something. AFAIK in mining you often want a slower material. Not pulverizing the rock, but breaking into larger chunks. But I do not work in this field.

        2. Here in Australia, ANFO is regularly used in ton lots in mining. Drive up the Great Northern Highway out of Perth, & chances are you’ll encounter (& be stuck for miles behind) a “road train” – 120 tons of multiple trailers behind a prime mover. Get closer than you should, & you’ll see the safety placards “Ammonium Nitrate, Oxidizing Agent”, etc.
          ANFO is about the cheapest bulk explosive available, hence its use.

    1. Simple. If they went with someone new. The car manuf. would have to pay them for each new airbag replacement. Keeping Takata means they have to supply the replacement at no cost.

  5. My assumption has been, since ammonium nitrate becomes less explosive as it absorbs moisture that the actual failure mechanism is contaminated material plugs the vent holes rather than becoming vapor and this leads to the explosion. The US military has done research into how long a fertilizer based IED remains dangerous until the ammonium nitrate becomes saturated with water. Heat and humidity hasten the process of degradation.

    In some videos showing the explosions the bright light of combustion from the vent holes is blocked off by debris before the rupture. Had the gas generators been built with a rupture groove that would vent without causing shrapnel then the airbag deployment might be compromised, but the flying debris would not be exploded in a way to injure people.

    It is possible that cars will not operate without the driver’s side airbag. It makes sense that the ecu/ecm would shut off the engine if the airbags deployed and therefore prevent the motor from being started.

    1. To amplify, looking at the Takata chart, they don’t include the possibility that that the propellant is too inactive – that the material closest to the holes would gain the most moisture and be less reactive and therefore act to block the holes they removed hole blockage as a possible reason, probably because they only considered fabrication flaws, not dynamic ones.

    2. If the car maker officially removes the airbag for safety reasons, it will reprogram the ECU. Shut off in case of deployment is a different thing from preventing the car to start. The first is only applicable if the engine is running and the latter if it is off.

  6. As someone familiar with the properties of ammonium nitrate explosive this story is incredibly surprising. Not only is ammonium nitrate very hard to detonate in small containers (you need at least a few cm of radial volume to get the deflagration to detonation wave going). But also, water content from absorbed by the hydroscopic ammonium nitrate will make detonation nearly impossible. Water *does not* increase the sensitivity. It massively decreases it.

    It’s like this story is backwards. I find it hard to believe. I’d expect the airbags to simply fail to deploy and cause problems that way.

    The only thing I can imagine going on here is the dense ammonium nitrate aborbing water, increasing in volume and then evaporating off decreasing in density making it every so slightly more likely to detonate.

    1. I think the summary is that the moisture degraded the ammonium nitrate near the exhaust holes making it less/slower to explode. The dry unaffected ammonium nitrate when triggered explodes normally, but the less reactive ammonium nitrate effectively plugs the exhaust hole which causes internal pressure to build up exceeding the design pressure of the case. This then causes the case to fragment creating shrapnel and it is the shrapnel from the case of the inflator that causes the injuries.

      Here’s a tool for US readers:
      https://www.nhtsa.gov/recalls

    2. My understanding from the other comments and some follow up reading is that the de-activated agent plugged the vent holes, creating a pressure chamber. The metal face then split, creating the shrapnel, and was propelled forward by the higher than expected pressure wave through the airbag (and into people).

      I expect there was actually less total pressure produced as a result of some dead agent in there, but since the metal damaged the airbag I don’t know how much that pressure reduction mattered. And the accidental creation of an additional pressure chamber means that the initial wave was somewhat more powerful than expected.

  7. As someone familiar with the properties of ammonium nitrate explosive this story is incredibly surprising. Not only is ammonium nitrate very hard to detonate in small containers (you need at least a few cm of radial volume to get the deflagration to detonation wave going). But also, water content from absorbed by the hydroscopic ammonium nitrate will make detonation nearly impossible. Water *does not* increase the sensitivity. It massively decreases it.

    It’s like this story is backwards. I find it hard to believe. I’d expect the airbags to simply fail to deploy and cause problems that way.

    The only thing I can imagine going on here is the dense ammonium nitrate absorbing water, increasing in volume and then evaporating off. The fluffiness resulting decreasing the density making it every so slightly more likely to detonate in small volumes.

    1. This is not about sensitivity, it’s about what happens after detonation (airbags are not being set off randomly, they continue to function “correctly”, but they become so powerful that they send shrapnel towards the occupants). Apparently some moisture can somehow make the explosion more powerful.

    2. Ignore the 9news stuff, it’s sensationalist and wildly off the mark. No one credible appears to be suggesting the mixture is detonating. The suggestion is that the mixture is cracking or changing morphologically and that is causing increased burn rate.

    3. I won’t claim huge expertise… in this case the Ammonium Nitrate is involved in a Deflagration reaction, and not a Detonation. The propellant is being burned to make gas to deploy the airbag. The hole where the gas is meant to escape becomes clogged, and the metal casing bursts, sending shrapnel out. As another poster above has pointed out, if the casing contained features to allow it to burst in a safe direction, this could improve things. Still sounds like it was a poor choice of oxidant for this application.

  8. I own a Mazda/Ford pickup with this airbag on the passenger side, and in Sweden they are apparently not recalled?

    The problem with the propellant is that it is made into pellets, pellets that gives off gas in a predictible way, but with thermal cycling and moisture, the pellets decay into powder, a powder that will burn a lot faster, causing an much faster explosion that is big enough to shoot pieces of the aluminium frame in the face of the passenger.

  9. ” the most recent case being a 2002 Honda Civic claiming the life of an Arizona man in June 2018″

    Well, I drive a 2003 Honda Civic and am currently in the waiting section of the recall, so I feel safe. :\

  10. Somehow, this all could have been averted by not being cheap to begin with… (sadly, a lot of problems comes from being cheap, and/or lazy.)

    Another solution is to have more then one propellant of choice, so if it is something wrong with the chemistry, then the effect is lessened due to not all one’s products being defective. (Ie, don’t put all eggs in the same basket.)

    1. The motor industry pays a lot of attention to being cheap. Spend 10 cents less per vehicle, sell 1 million vehicles, save 100k, pay for 2 more years of development engineer and get a better product overall. There are not many fields where such tiny savings on such an expensive product can be so worthwhile.
      The car manufacturers probably had no idea what the propellant was, and care even less. They trust the suppliers to provide a functional product, test a few and place the order.
      This is a lot like the PC manufacturers who bought duff capacitors. Nobody seems to be blaming the motherboard makers for not checking capacitor chemistry there, so why is so much blame point aimed at the car manufacturers here? I wonder how many people have been killed by dodgy capacitors? I suspect it’s not none.

      1. This is business everywhere.
        It started when engineers were no longer allowed to make decisions.

        I used to be involved in conversations about cutting 10 cent parts from a $80 bom that would go on for weeks about a trade off on useability verse savings over several thousand made.
        Add up 10 of those 10cent parts and you’ve saved $10k.+ over a production run = profit = margin = good numbers to report to bosses.

        Having external suppliers also helps shift blame and cost.
        Some supplier agreements I’ve been party too for large corps I would never consider signing. But the far east mfrs will happily do it to get the contract. Punative damages that put those firms under when they inevitably screw up.

        Another frightening thing is the push for “agile developent” which in reality means ship the product with bugs and fix them later. And has so many shortcuts onces it’s been bastardised by management looking to improve cycle time (testing, surely our customers can do that for us and report bugs, so lets skip it).
        Sure…. That never happens in transportation huh

      2. “why is so much blame point aimed at the car manufacturers here?”

        Well, I were actually talking about the airbag manufacturer….

        Though, it wouldn’t be unreasonable for the car manufacturer to ask what propellant is used, and search for information in regards to how good it is from third parties. After all the article do point out that the propellant in question were a well known fact that it weren’t reliable.

        “The properties of ammonium nitrate were well known; other airbag manufacturers had long ruled out its use due to its stability issues. Engineers warned in the late 1990s that the material was unsafe for the application, long before the parts in question were installed in vehicles. In spite of this, Takata forged ahead, winning contracts with major automakers.”

        So the facts were on the table, though simply ignored.

        Unless Takata decided to keep the propellant a secret, and not tell the car manufacturers about it.

        1. Why not point the blame at the auto manufactures? they still have a responsibility for all parts THEY put in the cars. Just because you outsource it does not release you from liability. Every company exec. knows that the low bidder has somehow cut something to get the lowest price. Like buying a real Rolex watch off a street vendor for $100. Gosh was I supposed to know it was stolen?

    2. That’s just a “postcode lottery”, and doesn’t actually make consumers happier. It also means you split your R&D and manufacturing resources, likely reducing the quality or increasing the price of both types.

  11. “It simply isn’t possible to produce several years worth of new airbags for older vehicles overnight.”

    Well, not without cutting into the supply chain and thus decreasing profits.

    Which, to any corporation, is the opposite what they want.

    1. No, it simply isn’t possible. Profits aside, the financial impossibility is proven by Takata’s bankruptcy. Even more importantly, it’s physically impossible. The airbag manufacturers size their production facilities to match the most optimistic estimates of demand and that sets their production capability limit. To replace more than a decade of Takata’s production, the whole airbag industry would have to produce at full production for over a year, not including the time it takes to retool for different airbag models.

      In the long term, this is all the consumer’s fault. If the lazy, superficial jerks that refused to use their seatbelts had used them, the auto collision death and injury totals would have tumbled and there would have been no pressure to develop “passive” restraints, i.e. airbags.

      Ideally, each seat in a passenger vehicle would be fitted with a racing harness and everyone would use them. But people are fools, thus the mess we have today.

      1. A racing harness these days includes a Hans device or equivalent, which requires a helmet, which reduces peripheral vision substantially. I drove that way once to get a race car to a repair shop, and I hope to never do it again. Lane changes were really awkward and might have been dangerous without friends escorting me around in their own cars.

        1. If you are driving on the public road you could have easily just disconnected the HANS device and left the helmet off. And if you’re not driving on a circuit and not wearing a helmet your unlikely to get into a crash severe enough to need the HANS device (Which is mostly because of the added weight of the helmet which the human body is not built to handle). A simple 4 or 5 point harness is perfectly fine for normal road use, but anybody that’s ever used one on a regular basis know they are a PAIN to use compared to the normal 3 point seatbelt.

      2. >refused to use their seatbelts

        Modern airbags become deadly if you don’t use the belts. Airbags were initially developed as a substitute for seat belts, but those systems used compressed air and could not inflate rapidly – which also made them useless, and rejected by manufacturers in the 70’s. The government however did try to make them mandatory, despite expert opinions that they’re not safe.

        The airbag later became a necessary “supplemental restraint system” because cars were made smaller and more rigid to withstand a crash without deforming the passenger area, which subjected the driver to a higher deceleration in a crash. The belts would injure the person by the sudden jerk, so the airbag found use to provide additional cushioning and improve the survival rate of car crashes.

    2. Also, I have to question the use of the word “overnight”. If the first incident was in 2004, TWELVE YEARS later, the industry was still shipping cars with this known time bomb. This is inexcusable. Note that Takata was only supplying a fraction of the airbag inflators for the industry – we’re not talking about a single source. Surely the airbag inflator manufacturers could have ramped up a little quicker than this, had the NTSA done its job and flat-out banned this over a much shorter timetable.

      I think this really comes down to class warfare. If it takes three or four years for an inflator to become dangerous, why would anybody who buys a new car every three years even care?

        1. I can’t dispute the possibility that I need medications. But just how does bigotry come into this? If I may reiterate what you clearly don’t get, they were shipping vehicles with known problems that will occur most likely not with the original owners of the vehicles, but with 2nd or 3rd owners, who won’t be tracked by the owner information that the manufacturers have. Therefore there is a high probability that these WILL fail, and the people affected at the time of failure will never have been notified of the recalls. This saves the manufacturers a lot of money, because a customer who doesn’t know about the recall doesn’t cost them any money. Shipping product that WILL fail later, knowing that this WILL continue to injure and kill people, is callous in the extreme.

  12. Interesting analysis of the root cause of the problem. Having worked at Takata in the early 1990s this material was proposed, tested and the results even then, it was not viable to use. All new products go though a failure mode analysis, and what occurs is management will override the data and facts to improve the financial condition. Takata had some management in place at one point that had no experience in the safety area not automotive in general. This is their failure. Sad to say I always liked working at Takata, but due to thier frequent uppper management turing over early on, I left. O well.

    1. Awesome, someone who thinks they’ve solved the problem Takata thought they had already solved. I don’t see a description of what the method is, but I wonder if ANST stands for Ammonium Nitrate STrontium.

    2. And given the Tanaka bankruptcy I see _zero_ chance of any manufacturer going for this. (I would like to think that the deaths would be the deciding factor, but…

  13. if it is profits instead of making faulty airbags why not committ collusion and price gouging and price fixing along with working their employees off the clock (oh we need you to clean the bathroom) as soon as they punch out and when they are ready to be upgraded to the next pay level then let them off and rehire them back at the bottom.

    1. Yes, there should be. The argument that you’re more likely to be killed by not having a functioning airbag, than by a faulty airbag, becomes invalid if you plug “I choose to always wear a seat belt” into the equation.

      1. That is completely and utterly false. Airbags ADD to the safety on top of a seat belt. Without a seat belt the airbag is also highly unlikely to work as intended as you’ll likely get launched over the top of the wheel into the windscreen instead of face first into the airbag. An airbag is not a replacement or substitute for a seat belt. It’s an additional safety feature that reduced forces on the neck and head by cushioning the blow before the seat belt reaches full extension. If you have only a seat belt to take the blow then the deceleration of the head happens completely by the neck and spine which are not made to take those loads, leading to neck and spinal injuries, as well as brain issues (concussion and the like). The airbag spreads the deceleration and “catches” the head, taking some of the forces from the neck.

    1. But in that case you do not have the “A” and “B” pillars of the car. And you hopefully do not really suggest to wear a helmet in the car? Except for high-risk motor sports events.

      1. I think he tried to reply to NSFW above, on his comment about driving with a helmet and HANS device. In which case I already replied it’s a poor idea to be wearing a helmet on the public road inside a car. If you ARE wearing a helmet and HANS device then turning your head is not really an option, you have very limited range of motion when wearing those.

  14. The business lesson here is that sandbagging an investigation can lead to fifteen additional years of executive wages. Even in bankruptcy it’s not clear anyone responsible has sufficient disincentive to have done differently.

  15. Ok so besides all this bullshit you guys are talking back and forth, what is the fucking problem , it is what it is, I’m sure the number of deaths and injury’s is higher than what’s said.,I know because I actually am a victim, nobody notified me or made sure my cars airbag was safe and a little fender bender broke my femur in my chest and,put shrapnel in my arms and head, and knocked me out with a major hit on my face,where’s my justice” the accident wSnt even my fault.

      1. Who are you? You are nobody to me ,I’m not trying to impress you or get your approval, so lay off obviously you have no life” what can you do for me? Nothing not a thing ,but what I can do for you is give you the little bit of importance ,to to make you feel special like your someone ,take a step back and look at what’s truly wrong with your self,and before you come back on this site think about if what your commenting on or saying to people, will it help them , or just make you feel good about thinking your, right,wrong, or smarter”..and before you try to pick my comments or opinions aparts.look at yourself.what are you doing in this world to help others. Hhhmmm because if you think making sarcastic comments helps anyone , your wrong, stop wasting time. And space”

    1. The injuries these defective inflators induce are horrific, I wish you quick recovery. A US District Court established an Individual Restitution Fund, as part of Takata’s criminal plea. There’s a process for victims to make a claim. Here’s a link to the website http://www.takataspecialmaster.com/

      That must have been terrifying. I’m curious, what make and model vehicle were you in? And were you the driver of the vehicle or a passenger? Passenger side inflators have more much propellant than the driver side inflators. It’s thunderous when they explode.
      Most of the recalled Takata inflators have been replaced with other Takata inflators that have the same Ammonium Nitrate propellant, and an additional desiccant to absorb moisture. Desiccant is just moisture adsorbing material. It works like dried rice in the plastic bag with a wet cell phone if you’ve ever googled how to remedy a cell phone that quit working after being dropped in water. You sometimes find a desiccant pouch in pill bottles. Takata claims that moisture will absorb into the desiccant instead of the propellant. That may be true for some time, but eventually the desiccant will saturate, then moisture will go into the propellant. NHTSA must decide whether or not to recall these additional 100 million PSAN inflators by the end of this year (in 6 months). In 2017 I met with lawyers from the senate commerce committee. They asked me if I thought the desiccated inflators would be safe. I told them that relying on a desiccant to fix the problem is like relying on desiccant in a pill bottle to extend the expiration date of a drug that injures or kills people when it expires instead of loosing potency. Prescription Medications lose potency when they expire. The FDA wouldn’t deliberately approve a drug that caused injury or death once it expires, and rely on desiccant to extend the expiration date. NHTSA should take the same position. I’m concerned NHTSA will not require the 100 million desiccated PSAN inflators to be recalled. You have an important voice as a victim. I hope NHTSA listens to you and the hundreds of other victims like you that have paid the human price for this.

  16. Great write up Lewin Day. Your closing statement really resonates. Sometimes life as an engineer can be frustrating, but your perspective makes me grateful for the calling. It’s agonizing to watch the aftermath when warnings are ignored. Even more agonizing to see subtle statements designed to downplay the specific root cause, like Takata PSAN inflators can “explode with too much force”. Imagine if NASA had tried to get away with that in 1986? The Space Shuttle Challenger “exploded with too much force”. As if all space shuttles explode; sometimes they just do it with too much force. I would add that the longer a problem festers, the more new problems it creates. When I saw news of the recall in May 2016 I ran some quick calculations and realized the dangers associated shipping 70 million of these defective inflators in cardboard boxes was very real. Most people don’t know that recalled Takata inflators are very dangerous in a fire. Fire is the most likely initiation scenario during shipping. I warned government officials of these shipping dangers 3 months before the Takata truck explosion in Texas on August 22, 2016. Those letters set into motion a chain of events that culminated with subjecting more than 150 recalled inflators to the UN 6(c) bonfire test at an explosives lab, exactly one year ago this week. I purchased the recalled inflators specifically for the test from vendors that sold them as replacement parts from salvaged vehicles. It was a lot of work. I needed a corporation with D-U-N-S number just to comply with certain aspects of the DOT’s formal process for submitting the bonfire test reports. I didn’t even know what a DUNS number was before the DOT told me that I needed one. Interestingly, five short months after that bonfire test, the EPA made a formal ruling on ‘Safe Management of Recalled Airbags’ that acknowledged the recalled airbags are hazardous waste, and should be shipped in a container that addresses the shipping hazard. The bonfire test results can be found in this link to comments on EPA’s Ruling. https://www.regulations.gov/document?D=EPA-HQ-OLEM-2018-0646-0034

      1. Thanks Lewin. Respect for your coverage of this. To understand the danger, people need to understand precisely what’s been covered up, the depth and the breadth of it. This rabbit hole is much deeper than most people realize. I worked as an Engineer for Takata for just under 3 years. It was my first job out of school and great place to start practicing problem solving skills gained from a fresh engineering degree, but I stayed there less than 3 years. In late 2000 I was asked to investigate excessive variability of the new PSAN inflators. Inflator combustion pressures are modeled with mathematical equations. The equations that had reliably predicted operating pressures of other inflators I was familiar with, didn’t work on PSAN inflators. Dr. Blomquist gives a great explanation of how combustion pressures are balanced in paragraph ‘c’ on page 7, of the root cause report, commissioned by NHTSA, in this link.
        https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/documents/expert_report-hrblomquist.pdf
        The mathematical models in my big nasty excel spreadsheet only made sense if the assumptions about propellant surface area were wrong. It was clear I needed to quench the propellant in the PSAN inflators to see if it was getting fractured. Quench testing is tricky, the idea is to rapidly depressurize the main propellant chamber (PSAN in this case) immediately after the ignition train hits it so that the propellant never burns. The goal is to see if the ignition energy is breaking the propellant. After a few failed attempts, I came up with a unique approach that worked. It proved the ignition system was fracturing every single piece of the propellant wafers like fine china. The quench test explained precisely why the ballistic performance was so variable. It also proved the PSAN inflators were grossly overloaded with propellant, 30-40% more propellant than was needed for a given inflator output. It also proved precisely why the carbon monoxide levels were about twice what they should have been. I worked tirelessly to summarize the findings in a report titled “Propellant Integrity Evaluation” and recommended that an official project # be opened with focus on resolving the problems discovered. I had the support of a great engineering manager, and an excellent test lab. When I submitted the report, the feedback from a very high executive level was that the report needed to go away. Shortly after that, Takata announce it was eliminating the engineering department I worked at. Takata offered to move me to another facility. I said thanks – but no thanks, and opted for lay-off instead. Many of us at that facility chose to leave Takata around that time. I wasn’t the only engineer there who’s findings were being ignored.
        If people started connecting some of the dots they’ll see Takata’s PSAN inflators have multiple design flaws besides propellant chemistry. They’re grossly overloaded with PSAN propellant, and the ballistic output is variable because the propellant surface area is not properly controlled. It gets fractured by the ignition system, resulting in excessive variability. It’s a design flaw.
        Here’s a link to the indictment. https://www.justice.gov/opa/press-release/file/926066/download
        Notice, item #27 on page 11 of the indictment states that most often the data Takata falsified was “related to ballistics or effluent gas, Takata’s PSAN inflators had difficulty meeting the OEM’s specifications relating to ballistics and effluent gas”. We know that these PSAN inflators do bad things after moisture ingress and years of exposure to thermal cycles, but that’s not what caused effluent gas (carbon monoxide) levels to be so high in test reports that were falsified and submitted to OEMs, referenced in the indictment. Excessive propellant load and uncontrolled propellant surface area is the root cause of that. And look back at that paragraph on P7 of the root cause report in the first link again. The 2nd sentence of that paragraph states “The designer of an inflator creates this balance by manipulating the propellant surface area”. That’s true and well put by Dr. Blombquist, the surface area is typically manipulated by using a specific propellant geometric structure called a “wafer” in this case, but other grain geometries such as tablets are also common in the industry. If the propellant surface area needs to increase you use smaller tables and more of them. If surface area needs to decrease, larger tablets and fewer. If a given propellant load has too much surface area, for a given vent area, chamber pressure will be too high. It’s why pistol powder shouldn’t be used in rifle cartridges when reloading ammo. So how did the designers of Takata’s PSAN inflators manipulate the propellant surface area (geometry) to maintain chamber pressure balance with an ignition system that fractures the propellant? How did they know exactly how the propellant would fracture every time? Did they have some method established to calculate the resultant surface area after the ignition train fractured all the propellant wafers? And what prompted the inflator designers to come up with this controlled ignition fracturing approach that was presumably an intentional feature?
        Also, notice paragraph 18(b)(i), on page 6 states: “The main propellant…..burns layer by layer as designed” But check out footnote 5… What?? “controlled ignition fracturing” How is the fracturing “controlled” exactly, and what are the specific controls. And more importantly, precisely why would PSAN inflators have a design feature that intentionally fractures propellant, if fractured propellant is indicated as a root cause of failure in the early fault the tree analysis. The box “Crush” below “Poor physical properties” refers to the propellant’s crush strength. Crush strength is in the fault tree because people that understand how inflators work, know that propellant wafer crush strength is important because low crush strength results in fractured propellant, which is bad. Also notice the box titled “Propellant Damage” under the “Ignition too Aggressive” box that is X’d out. Why would Takata have an inflator design that utilized controlled ignition fracturing if they listed propellant damage due to an ignition system that was too aggressive in an early fault tree analysis?
        Why would Takata want Dr. Blombquist, and the world to believe that an ignition system which fractures propellant is an intentional feature of their PSAN Inflator designs?
        To understand the danger, people need to understand precisely what’s been covered up, the depth and the breadth of it.

          1. Thanks for that. People do need to understand what’s been covered up. The people that have more recently needed to see that propellant integrity investigation from 19 years ago have seen it. Maybe at some point everyone else should too.

  17. I’ve been a validation and test engineer for many years working with/for many different automotive organizations and products. There is a common theme at all of these organizations to deny that a product is defective or has a design flaw. It is a constant struggle to get any test failure acknowledged as anything other than a test setup error. because it costs money and time to address the real root issue of a failure.

  18. Airbags were foisted upon by safety Nazis in the US government. Since I own the car, I should be able to have these things removed. Of course, lots of dumb people here who think any regulation the government creates is for protection and makes sense.

    There are some hand made classic Mustangs being built that have the newer emission controls but no air bags, yet us peons are not entitled to the same option.

    1. Did you know you can’t disengage your seat belt while upside down and why you should ALWAYS have a belt cutter and window break tool. Without it you can’t get out of a burning car or underwater…

      Personally, seat belt has saved my life from going through the front windshield.

      That said the air bags are good because of other damned people. Aka using the turn signal for a lane change should be law (if fact it is) but Phuket they cut you of anyway. Don’t they.

  19. It is not clear in this article if changes were made in current manufacturing of the airbag to make it safer and remove the issues causing problems, did this take place or are the current airbags just as dangerous?

    1. All Takata inflators with “PSAN” propellant will eventually become dangerous. The 100+ million Takata inflators that have been recalled have what Takata calls “Phase Stabilized” Ammonium Nitrate (PSAN) propellant in them. Most airbag propellants dud out (looses energy) when exposed to moisture, but Takata’s PSAN propellant becomes unstable, causing death or injury. Many of Takata’s recalled PSAN inflators were replaced with other Takata inflators that had the same defective PSAN propellant but Takata claims they fixed the problem by adding a desiccant to absorb any moisture that would otherwise go into the propellant. However, all desiccant will eventually saturate. Once that happens moisture will go into the propellant and cause the same problem. There are another 100 million of these Takata inflators that have desiccant but have not been recalled yet. These inflators with desiccant should also be recalled because they will eventually become dangerous too.

      Prescription medications are a good analogy. Most prescription medications loose potency after they reach the expiration date. For Takata to convince NHTSA that desiccant fixes the problem is like a pharmaceutical drug manufacturer releasing a drug that causes death or injury once it expires, then convincing the FDA that drug is safe because they added a desiccant to the pill bottle. And BTW, you sometimes find desiccant pouches in pill bottles to extend the expiration date. The FDA would never approve a drug that was proven to cause death or injury once it expired no matter how much desiccant was added. NHTSA should take the same approach.

      The change Takata made was to add desiccant, and the Takata inflators with desiccant have not been recalled yet, but NHTSA must decide by the end of this year if the inflators with desiccant will be recalled also. If NHTSA decides not to recall the other 100 million Takata inflators with PSAN propellant + desiccant, we wont know which vehicles have these other 100 million Takata inflators in them until the problem starts manifesting itself like it did with the first 100 million.

      1. Prescription meds statement. FDA just did a test regarding the stockpile for the DOJ & military.

        Aka Vacuum Sealed and light proof container.

        All medications were good up to 10 years.

        Only one that dropped in efficacy was Amoxicillin. That one was at 50% after 7 years.

        Read the report yourself.

        Don’t spread mis-information

        1. I’ve not spread misinformation. Medications are a good analogy of the approach NHTSA should take regarding desiccant as Takata’s solution to the problems with PSAN propellant.

          As you point out, with the case of Amoxicillin, medications loose efficacy when they age. Some medications have a longer shelf life than others, but they don’t change at the molecular level in ways that cause them to become categorically harmful like the PSAN propellant in Takata’s inflators.

          Other airbag propellants dud out after environmental exposure. Takata claims they fixed the problem by adding desiccant, but it’s not a real fix, because the desiccant will eventually saturate. The FDA is the regulatory body of the pharmaceutical industry, like the DOT (NHTSA) is for the Auto Industry. NHTSA should not accept desiccant as a solution to the defect. The FDA wouldn’t approve a drug that caused death or injury after it expired, even if adding desiccant to the pill bottle could extend the expiration date by some margin. NHTSA, should take the same approach with Takata’s PSAN inflators.

  20. Between insurance companies and motor vehicle registration you can find every single one of these cars then you dismantle them dismember them shut them off whatever mechanism needs to be done you protected the people

  21. Oh I forgot inspection stations and Emission Inspection Stations as a source to find all these cars not hard not rocket science and they’re all government agencies they can be tapped to do what they need to do inotify everybody

  22. In Australia manufacturers are buying back older cars that do not have replacement airbags. Honda paid a AUD $3400 for our well worn 1999 CRV which seemed OK. Some other companies have upset people with low prices offered for good cars.
    Are buybacks happening in the rest of the world?

Leave a 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.