Beware the side curtain air bag deployment algorithm!

[mySSt]

Your SS side curtain air bags can deploy WITHOUT an impact and while all of your tires are still contacting the road. Don't know if the SIR computer's anti-roll algorithm is the same for the V6, LE or ZL1 models. As a minimum result and to replace with new, you are looking at 2 new air bags (there was a front seat passenger), 2 new seat belt assemblies, new SIR computer module, and a deformed head liner. GM list price for all the parts except the headliner total to about $1500. Add a new headliner and that adds about another $400.

I know the video has already been posted showing the silver Camaro drifting in the parking lot and his side curtains deploy. Here's the details of another account:

During some spirited driving and exiting a right hand corner on a fairly level asphalt road, the driver's side wheels hit some loose stuff while the steering wheel was not yet fully straight. The result was the car quickly rotated from left to right (about 60 degrees of yaw) and the car slid laterally down the road - all of this while the car is still on the road surface. Remember, less friction always "leads". To correct the slide, the the driver started to turn the steering wheel to the left and into the slide. At this moment, the 2 side curtain air bags were immediately deployed by the SIR computer. Needless to say, that shocking and unexpected event broke the driver's concentration and made a small issue a huge problem. By 'broke driver's concentration', I mean going from 'turn the wheel to correct this slide' to 'why the hell did my side curtain air bags just inflate!?'. Luckily, when the car then ended up leaving the right hand side of the roadway, it only spun around on some level ground (no trees, poles, ditches, etc) or this could have had a much different outcome.

Some of the tech details: stock suspension (other than the addition of a GM front strut tower brace), suspension in good shape, dry asphalt road, hot day, about 80 mph, 1.7 Gs, and up to 40 degrees of body roll when the computer anti-roll algorithm said "lock the seat belts and deploy the side curtain air bags". Even with a loading driver's side suspension, not sure the car can have 40 degrees of body roll without at least one wheel lifting off the road surface. Repeating, the wheels did NOT lift off the road - the car was sliding laterally down the road. Of course, the same device that deployed the air bags (based on an algorithm calculation and not any impact sensor) is the same module which was utilized to record some data and determine it "functioned normally" - hardly an independent verification.

It's interesting when you talk to others on and outside the forum which have had side curtain bags deploy in similar situations (no sudden jolts to the wheel/suspension, no body impact or crash, driving down a road during cornering or exiting a corner, ...), their story is very familiar and it sure leads you to the conclusion the anti-roll algorithm is set to deploy too sensitively for spirited driving. For pete's sake, this is a Camaro, not an SUV!

Just wanted to alert all to what can happen - better to be forewarned. Be safe!

[Inferno LSX]

Happen to me had stabilitrac turned off and was playing around. Car did a 180 and the side curtains deployed.

[cam45]

I've heard of these events and I was wondering if GM take care of the cost? That's a hefty price for deploying for no reason

[mySSt]

Originally Posted by cam45

I've heard of these events and I was wondering if GM take care of the cost? That's a hefty price for deploying for no reason

Nope, GM won't pay for any of it. They collect the data from the SIR module and then tell you "the system functioned normally."

[Number 3]

I've done some airbag deployment work in my career. A lonnnnngg time ago but it applies.

In order for the airbag to do anything it has to be deployed. If you don't make the decision to deploy soon enough it won't be there when you need it. So you have literally milliseconds to determine if the conditions are leading to crash. So for every move you make to avoid this condition you are likely finding more conditions where it should have deployed and did not.

I was at a different OEM at this time, but there is generally a "NO FIRE" event, a "MAY FIRE" range and a "MUST FIRE". So based on the accelerometers available there is a set of parameters under which the airbag should never go off and there is another range where it must ALWAYS go off. It's the range in between that gets very hard to calibrate.

Right or wrong, GM has chosen these parameters/conditions and they aren't likely changed because of the type of car.

So yes, the car likely performed as engineered in these types of events.

[its a syn]

So if you do some donuts they're going to deploy?

[oldfriend]

Originally Posted by its a syn

So if you do some donuts they're going to deploy?

Go ahead and do it , then let us know
But if they do deploy , don't send us the bill !

[snymat68]

Once they deploy, you don't have to worry about them any more. Just don't get into any wrecks after that!

[Synner]

Considering the loss of traction it seems like a valid deployment, 1.7G's is a crap ton. The majority of people out there are too poor of drivers to recover the vehicle themselves in such a situation. It just sucks for people who know how to correct significant oversteer.

[thahemp]

I have spun mine around a few times... I think I'll stop doing it after reading this! Number 3 makes excellent points. Tiny windows of data are sometimes impossible to draw accurate conclusions from. Just a case of better safe than sorry.

[mySSt]

Originally Posted by Number 3

I've done some airbag deployment work in my career. A lonnnnngg time ago but it applies.

In order for the airbag to do anything it has to be deployed. If you don't make the decision to deploy soon enough it won't be there when you need it. So you have literally milliseconds to determine if the conditions are leading to crash. So for every move you make to avoid this condition you are likely finding more conditions where it should have deployed and did not.

I was at a different OEM at this time, but there is generally a "NO FIRE" event, a "MAY FIRE" range and a "MUST FIRE". So based on the accelerometers available there is a set of parameters under which the airbag should never go off and there is another range where it must ALWAYS go off. It's the range in between that gets very hard to calibrate.

Right or wrong, GM has chosen these parameters/conditions and they aren't likely changed because of the type of car.

So yes, the car likely performed as engineered in these types of events.

Thanks for the input. Agree that its a decision they built into the computer algorithm to cover the "majority of drivers" - just like producing cars with built-in understeer (instead of neutral or oversteer) since it keeps the manufacturer's liability lower. However, its a shame that the type of vehicle doesn't make a difference in the firing point calculation for the deployment (lower center of gravity, car's ability to respond quickly to steering input, ...). Sure bites this is all about a roll over PREDICTION calculation versus an impact to a sensor to deploy front air bags for example.

[firedrop]

why would you want to do donuts in your car anyway? unless youre rich, and at that point who cares if the airbags deploy? $1,500 should be nothing to you if you're doing donuts on $350+ tires.

[thahemp]

Because donuts are fun? Why do we need 400+ HP cars? It's not like there is legal place to stretch our legs anyways.

::EDIT::

Besides on a closed course obviously.

[102SS]

At 1.7G's and 40* list and 80+mph it seems like a normal event to me,the computer thought the driver had lost it and a crash was inevitable and he probably had lost it.

If your going to drift something pick something with less technology.

BTW anyone see the vid of the Hennesey CTS-V catapult launch on the dragstrip where Onstar thought they had been rearended because of the launch G's?