PDA

View Full Version : SOT: how do airbags work



rob101
02-19-2007, 05:19 PM
I've just read a post on another forum about airbags not deploying in the car and the subsequent discussion of how people think they work, and there is alot of misinformation out there about it.

I was under the impression that there were only accelerometers, speed, roll over angle sensors and maybe yaw speed sensors involved. I have heard people talk about things like impact sensors that sense increases in pressure in the cabin but as an engineer my gut feeling says thats a load of BS. As i thought the main way they sensed an impact was from the rapid deceleration of the car (through accelerometers) would anyone care to shed light on the subject that is in the know?

t.wak
02-19-2007, 07:50 PM
I'm also on Eurocca, and was looking forward to finding out for good, too.

t.wak
02-19-2007, 07:52 PM
http://auto.howstuffworks.com/airbag1.htm

Accelerometres is all they mention here, makes sense.

LunatiC
02-19-2007, 08:04 PM
I think you are right Rob. I vaguely remember a Wheels magazine article on how airbags work when airbags first came to cars that were sold in Australia.

Accelerometers and a bunch of other sensors would trigger the airbag to propel. Pressure sensors is a load of BS, I don't see airbags being deployed whenever you open or close your doors ;)

When all else fails, consult Wikipedia's article on Airbag triggering conditions (http://en.wikipedia.org/wiki/Airbag#Triggering_conditions) :D

Eurocca is full of BS anyway, but you already knew that :P

rob101
02-19-2007, 08:27 PM
Eurocca is full of BS anyway, but you already knew that :P
remember dude, whatever you do don't change your lifetime oil. If you do you will be cursed forever muhahahahaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.
~strokes his money bags in the back of his 330ci.
(private joke between myself and Luny if you're freaked out)

rob101
02-19-2007, 08:37 PM
one of the other things i don't get was the dude saying words to the effect of because the crash wasn't front on the angle would reduce the impact that the sensor sees,(because they are directional) assuming of course that you only have accelerometers that feel forces in the front to rear direction.
Wouldn't you put an extra sensor which detects left to right forces, i mean what is the point of having side airbags if they don't go off in a side impact? Just seems really obvious to me and therefore any designer that you would have an extra set of accelerometers for that axis.

GJPinAU
02-20-2007, 04:25 AM
My bro-in-law used 2 other commodores to stop his commodore (WAY TO GO!) and the airbags didn't deploy, which I thought was f#$%ing scary.
Typical of good quality Aussie cars.
Aren't they sending commodores to the US now?

LunatiC
02-20-2007, 06:55 AM
Aren't they sending commodores to the US now?

They have been for some time -- Monaro (which is just a 2 door commodore in my eyes, no matter how hard Holden tries to convince me otherwise) been in the states for some time, and now the "new" Pontiac G8 making its debut soonish.

VentoGT
02-20-2007, 08:14 AM
The Commodore was called the GTO here and Pontiac is no longer making it. I believe the new GTO will be more 'Americanized' than the last GTO, which apparently people didn't like...I think it had something to do with the $40,000-ish price more than the styling or lack of 'Americana' personally...

Hal
02-20-2007, 10:05 AM
I designed accelerometers for air bag deployment in the old days, and some millions of our parts are out there somewhere waiting for an accident to happen. There is a big difference, however, between deployment of the front air bags (driver and passenger) and the side airbags. The front airbag deployment is described well in the Wikapedia article.

For a side impact, however, there is not enough time for an accelerometer in the front or cabin of the car to respond when the SUV is trying to come through the driver's door. One scheme is to monitor the air pressure in the door cavity itself. As it is being crushed, the pressure inside the door spikes, which is detected and the side airbag deployed. This can happen well before the main body of the car starts to move (accelerate). There can be issues, however, particularly with some door mounted speaker systems. When a door is slammed, there is not quite the same pressure rise in the door cavity as when it is being crushed.

I have not been in this business lately, so I am not sure what the most common way of side impact detection is these days, but I think there are many cars on the road with this pressure-based system.

Hal

1994 530iT 162,000 mi Nikasil powered
1999 528 iT 93,000 mi non-Nikasil

rob101
02-20-2007, 03:01 PM
I designed accelerometers for air bag deployment in the old days.......
Hal

1994 530iT 162,000 mi Nikasil powered
1999 528 iT 93,000 mi non-Nikasil
Interesting read Hal, makes sense about the accelerometers not being as effective in the left right axis, there is alot less stroke.
In the post on the other forum the guy was hit from the side in the front fender area. if the sole detection of the side impact is through a pressure increase in the doors it would explain why the system didn't pick it up. apparently the guy's engine ended up a few meters to the side of the car.

Hal
02-20-2007, 04:47 PM
The following is a link to a Siemens 2003 press release on their pressure-based side impact detection system:
Siemens link (http://www.car-data.com/xpage.news/news.template.asp?article=siemens.tech.paper.addre sses.side.impact.txt)

As the price of accelerometers came down, I think that more systems started to use accelerometers, usually mounted in the B-pillar, to be as close as possible to the driver.

Both front and side air bags do not deploy for every crash of every type. Most will not deploy if you are hit from behind, or even from the side if the impact is well away from the B-pillar. Historically, the systems also have been different in the US vs. the rest of the world, due to problems with the US legal system. For example, at one time (probably still today) the air bags for the US systems were sized assuming that the driver and passenger were NOT using their seat belts, while the European systems were based on the assumption that seat belts were used. It is, of course, much easier to design an effective air bag system if you know that the passenger is going to be reasonably well restrained by the seat belt.

Hal

indierthanthou
02-20-2007, 04:48 PM
it wasn't a gto, just a sunfire with a few styling changes and rebadged.





oh, and a 6 liter 400hp ls2 engine, pushing it to a screaming 5 second 0-60 time

LunatiC
02-20-2007, 05:23 PM
it wasn't a gto, just a sunfire with a few styling changes and rebadged.


I am not sure if you were being serious or not, but in case you were I think you're wrong. The Sunfire was based on the GM J platform, which the only car that ever was developed in Australia sharing this platform was the Holden Camira, not the Commodore.

The Camira was a real lemon here Down Under. You know you have hit the pits and are completely desperate when you need a car and all you can afford is a used Camira.

Also it should be noted that the Camira was the only car produced in Australia on the J platform (which ended production in 1989), while the US used the J for many years after.

rob101
02-20-2007, 10:16 PM
The following is a link to a Siemens 2003 press release on their pressure-based side impact detection system:
Siemens link (http://www.car-data.com/xpage.news/news.template.asp?article=siemens.tech.paper.addre sses.side.impact.txt)
.......Hal
Just got my copy of edition 6 of bentley publishing's Bosch Automotive handbook.

And it has a 2 sections one on front airbags and another on side. now they did mention the side airbag problem and how they must inflate in 5-10 ms. now they have peripheral accelerometers mounted on the B and C pillars as well as in the seat belt tensioners and side airbags themselves.

"Nowadays those acceleration sensors are surface-micromechanical sensors consisting of fixed and moving finger structures and spring pins. A special process is used to incorporate the "spring mass system" on the suface of a silicon water."
it is also mentioned that
"Peripheral Pressure Sensors inside the door cavity are to be used in the future to measure the adiabatic pressure changes caused by deformation of the door"