Airbag / Passive Restraint Design Standard

[Sparweb]

Hi,
While considering the reliability statistics of various devices on a project here, the subject came up about the reliability of airbags. There are two questions we can ask:

"Can the bag fail to inflate when the controller commands it", or
"Can the bag inflate even though it was not commanded to?"

They are probably very rare occurrences. In fact I can't find any statistics about these kinds of failure rates at all. Perhaps all drowned out by the other kinds of failures (bad sensors, too much pressure, flying debris, occupants who are much smaller than the 50% man, etc.)

Personally, I've never heard of one going off without a reason, aside from the "nuisance reasons" like impacts that barely even damaged the car. I have found no auto recalls, for example that say "faulty airbag inflation device", though there are plenty of complaints about airbags that do not deploy!

Can anyone think of any data regarding the failure rate of airbag inflation systems? Either in terms of inflating when they're not supposed to, or not inflating when they should?

Thanks!


Steven Fahey, CET

 

[MacGyverS2000]

There have been recalls for bags that inflated with no provocation... as early as a few years ago, too.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Sparweb]

Well, I keep finding more examples - but they keep coming up as faulty wiring, damaged sensors, improperly programmed control units, and so on. Things that go wrong at the assembly plant or during the service life. What about the components themselves?

Maybe my real question is: What are the quality control criteria for the procurement of airbags from an OEM?

Steven Fahey, CET

 

[Sparweb]

Yes, lots of recalls: There have been inflators that shed metallic debris, burst at too high a pressure, ouch - turning the safety device into a hazard itself.

But I'm still not getting to statistics on airbag failures.
Fatalities per million vehicle years is not the same as device failures per million vehicle years. The first gets printed in the newspaper, the second should be designed into the hardware at the factory.


Steven Fahey, CET

 

[IRstuff]

I think you'd have to go to DOT to get anything along those lines, or possibly the Insurance Institute. They should have accident reports that list the accident factors.

Regardless of the reason, there are between 700 and 1400 deaths supposedly attributed to airbag failures over about a decade, which supposedly include both failure to deploy of any kind as well as inadvertent deployment. As this is out of supposedly 170 million vehicles so equipped, that makes for very thin statistics. If you assume a conservativ 10K miles per year per vehicle, that works out to one death per 1.2 billion vehicle miles.

As for your two questions:
"Can the bag fail to inflate when the controller commands it"
> Clearly, yes. You might argue that that majority of deployment failures are actually latent failures; I wouldn't disagree with that, but that simply puts the actual component failure rate quite low, as it should be. The entire system is quite simple, overall, and only the sensor and processor are powered during car operation. That puts the nominal MTBF in the vicinity of 300,000 hrs or higher, based on complexity. This is presumably coupled with a fail-safe hardware and software design. Your question already pre-selects the failures, since the "controller commands" basically means that everything is functional except for the possibility of the initiator and explosive, and possibly the airbag and its cover.

"Can the bag inflate even though it was not commanded to?"
> Probably, but extremely unlikely. There's the actual explosive and some sort of initiator. The only reason such a system might initiate on its own would be due to electromagnetic interference (EMI), but of an extremely high level, since the engine compartment is not without some high levels of EMI. I would think that it would take a lightning strike or EMP to set off the initiator, since it has to require a certain level of supplied energy to initiate.

So both questions are talking, at most, about the initiator, the explosive, the airbag, and the cover. All these are relatively inert, and not subject to any disturbances except for temp and vibe. The airbag and its cover, ostensibly, could only fail due to assembly and fabrication problems, so they're also latent failures. That leaves the initiator, which is a heater, and the explosive itself. Likewise, short of latent failures, failures could only be introduced through some sort of moisture related effects, like corrosion. This would require subjecting the car to more moisture than normally possible, like taking a convertible into a carwash with the roof down.

TTFN

FAQ731-376

 

[drwebb]

Along those lines, my '01 Volvo calls for a service for the SRS airbags next year. Anyone know or speculate if something is actually involved with such a 'service' (as this seems peculiar to Volvos)?

 

[patprimmer]

Maybe they set it off to see if it works ;-)

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[BrianPetersen]

I know that VW calls for an inspection after 10 years and every 2 years thereafter.

The inspection consists of verifying that the "airbag" warning lamp comes on when the ignition key is switched on, and goes out a couple seconds afterward. (I'm serious - that's what it says to do in the factory service manual.)

 

[Sparweb]

IRstuff:
"Can the bag inflate even though it was not commanded to?"> Probably, but extremely unlikely.

My conclusion, too. Funny I just did the same math as you did this morning before seeing your reply. Our numbers match up quite well. Light trucks have airbags, now, too, so that brings the vehicle count up even more than you assumed. When getting down to the component level, you can consider at least two initiators per vehicle (4 with side-bags now).

The overal airbag failure rate comes to about 1 for every billion vehicle trips per year, and all I'm asking for is one very unusual failure mode in the entire complex airbag system.

Some of the most useful information in my search has come from the IHS, not just the NHTSA. This isn't helping my love/hate relationship with insurance adjusters and lawyers.

Brian,
It sounds like VW has a diagnostic built into the controller, but the reason they make the fuss about having it in the shop is so that the mechanic can send a message back to the factory about as many vehicles as possible. An information collecting exercise.


Steven Fahey, CET

 

[drwebb]

"An information collecting exercise."- subsidized by the customer (unless there is N/C for this particular maintenance) - I may decline to participate in the 'survey', but I'll keep an eye on the light.

 

[AlexKRacing]

Early model Mercedes -Benz model 203 chassis (c class) had a fairly large recall on the driver airbag unit itself. I never really looked into the specifics on the issue but I can look it up next week and perhaps get some proprietary information for you. I know this doesn't give you the statistics your asking for but if you have information on the exact failure of component perhaps you can modify your search to something more conclusive.

 

[MacGyverS2000]

I can look it up next week and perhaps get some proprietary information for you.

Wouldn't this be a bad idea?

Dan - Owner

http://www.Hi-TecDesigns.com