Advanced Airbag Restraint Systems

In 1996 reports were getting stronger that airbags were not the magic protective restraint system some had hoped. In several moderate speed (i.e. 12 to 20 mph) collisions the airbag deployment caused the death of a number of short stature drivers and several of children that were seated in the front passenger seat. [ Rigid Barrier Test ] The first generation of airbags was designed to catch unbelted adult male driver and passenger in a 30mph frontal crash (straight on and at a 30 degree angle). [Man falling from three story building ] If a speed of 30 mph doesn't sound too serious, realize that it is the equivalent of falling from a three story building onto the pavement. There aren't too many people that have fallen from a three story building and lived to tell about it. The airbag power needed to cushion this blow proved to be too much for short statue, closely seated drivers and small children. The government responded by first changing the requirements such that the industry could de-power the airbags. Then, bit by bit, new technologies were introduced: Dual threshold crash sensors; Dual stage inflators; Occupant classification sensors; Seat position sensors; Belt usage sensors; As well as pedal positioning and steering column adjustments. All that eventually lead to the "Advanced Airbag Rule" in which the government mandated new certification testing with different speeds, belted and unbelted dummies, and with dummies positioned right on top of the airbags during deployment. The advanced airbags were born.

While simulation analysis was well established for occupant protection analysis, back in 1996, the tools to analyze injury risk to occupants in close proximity to the airbag were in their early development stage. Testing was therefore the only viable solution to investigate the issue. While some research had shown that for a person, who was right on top of the airbag module, only the strength or aggressivity of the inflator determined the injury risk, this was less clear for a person sitting a mere 6" (150 mm) away from the airbag module. Laboratory testing with the 5th percentile female dummy (height 5'0") showed that module design and airbag fold could also play a role.

It was found that airbag modules that have a single piece cover increase the chance that the airbag gets caught under the chin of a closely seated occupant. Once caught there, the force of deployment will snap the head back with great force, increasing the risk of neck injury. Single piece covers are often desired by vehicle stylists. Optimizing the airbag fold such that the airbag deploys free from interaction with the occupant is an engineering challenge. It is more difficult than with module designs that have a cover that splits through the center which therefore have the engineer's preference.
[ OOP deployment ] [ OOP deployment ] If the bag is folded incorrectly, it may lock-up under the chin of a closely seated occupant, potentially causing severe neck injuries.

An other challenge engineers face is the ongoing cost cutting to remain competitive. At less than $2 a set, airbag tethers are frequently targeted by the bean counters. Testing showed, however, that they are a worthwhile expense to improve airbag safety. Tethers are strips of fabric, hidden inside the airbag, that span from the back to the front of the cushion. They reduce the throw of the airbag during deployment and determine the final thickness of the cushion. Without such thickness control the airbag will expand into a spherical shape. The extra thickness increases the load on the head of a closely seated, short stature occupant.
[ OOP deployment ] [ OOP deployment ] The head now has to pull the body backwards which increases the neck loads. Without tethers an airbag will balloon into a spherical shape.

The Magic Returns

Step by step the airbag systems were improved. [ RFIS Warning ] As an intermediate step the National Highway Traffic Safety Administration sanctioned the de-powering of airbags by changing the certification requirements. The resulting airbags were some 30% softer than the first generation. Furthermore, the NHTSA allowed the passenger airbag to be manually turned off in case a rearward facing infant seat had to be placed in the front passenger seat. You will find either of these modifications in most vehicles between model years 1998 and 2002. While depowered airbags reduce the chance of airbag induced injury to the short stature drivers and young front seat passengers, it also reduces the amount of protection to larger occupants and occupants in more severe crashes. This affects the area airbags have been proven to be very successful. The government and industry set to work to regain what was lost.

One of the first technologies to be introduced were dual-output inflators. Those have the ability to deploy the airbag with either a part or a full charge of gas. They were first combined with dual-threshold crash sensors that would signal if the crash was of moderate or high severity. The full gas charge would only be released if the crash sensor signalled a high severity crash.

That may sound easier than it is. The crash of a vehicle driving at 30mph into a rigid barrier [ Crash Event Timing ] only lasts about 100 milliseconds, or 0.1 seconds, or about the time it takes to blink your eyes. Out of this time the crash sensor only has about 15 to 20 milliseconds to determine that a crash is happening and whether or not it is a severe one. By that time the vehicle may have only slowed down 2 mph (from 30mph to 28mph).
No two crashes are alike. The vehicle has a totally different response whether one drives into a tree or into the back of a truck. Both can be of similar severity, however. Often multiple sensors and sophisticated software algorithms are needed to make the distinction.

Next on the scene were weight sensors that would determine the presence of a passenger and determine whether it was an adult or a small child. It was without any doubt that an infant in a rearward facing child seat had no benefit whatsoever from a deploying airbag. But what about a six year old child, involved in a high speed crash? Well, a 12-month old in an infant seat, covered with a blanked weighs about as much as a 3-year old child. So if we turn the airbag off for the infant, then the weight sensor will also cause the airbag to be turned off for the 3-year old child. A 3-year old, sitting in a child seat can weigh as much as a 6-year old child sitting on the seat by itself. So, the weight sensor might decide that the airbag will be switched-off for them too. Only sophisticated, pattern recognition software could tell the difference, but not in 100% of the cases.

Henk Helleman has been on the forefront of the development of the next generation airbag systems. This culminated in his co-authorship of the BREED Technologies' TOPSTM strategy for airbag deployment. (fellow authors are Dr. Russel Brantman and Dr. Said Nakhla)

TOPS™ is an acronym for Tailorable Occupant Protection System. It combines several of the new and improved sensors with the dual-output gas generators to tailor the airbag for the (crash) situation at hand. TOPS™ was unique in the industry in that it allowed a phased introduction of third generation airbag systems with increasing levels of sophistication. This is important to allow cost effective development and a rapid return to full protection airbag systems.

Cars of model years 2003 and newer may have airbag systems that typically comprise of the following:

  • A crash sensor that can discriminate multiple levels of crash severity.
  • A gas generator with a low and high output level, known as a dual stage inflator.
  • A belt usage sensor.
  • A driver seat position sensor.
  • A passenger seat weight sensor.

Furthermore there is an increasing chance that your car of choice will be equipped with seat belt pretensioners. These will take the slack out of the belt in the early stages of a crash, making the belts more effective (only if you wear them, of course). In addition to this the seat belts might incorporate a load limiter. This device will cap the seat belt load at a certain maximum, which will reduce the force of the belt on the chest and the shoulder. This reduces the chance of injuries caused by the seat belt, such as cuts and broken ribs (yes, belts can cause injuries too...)
If you are lucky enough to have a car with telescoping steering wheel adjustment and adjustable pedals, use them to create as much distance between you and the airbag module as possible.

With all this in place the restraint system can be tailored to the needs of the occupant. The airbag deployment strategy on the driver side might look like this:

Belt UseCrash Severity1 Seat Position2->Airbag Deployment
yeshighbackward->low level deployment
nomediumforward->low level deployment
nomediumbackward->low level deployment
nohighforward->low level deployment
nohighbackward->high level deployment
1) A medium crash severity is equivalent to a vehicle driving into a rigid wall at 12 mph or more. A high crash severity is equivalent to a vehicle driving into a rigid wall at 20 mph or more. 2) Seat forward position typically means the seat is in the forward most 1/4 of the seat track. Consequently, seat backward position means the seat is further back than 1/4 from the front of the seat track.

The airbag deployment strategy for the front passenger side might look like this:

Belt UseCrash Severity Occupant Weight3->Airbag Deployment
yeshighhigh->low level deployment
nomediumhigh->low level deployment
nohighlow->low level deployment
nohighhigh->high level deployment
3) The occupant weight is typically considered low if it is less than 35 kg (75 lb). Above that it is considered high.

These systems will deal with the two most pressing problems:

  • Avoid deployment of the passenger airbag if a child seat is placed there.
  • Reduce the airbag deployment force for closely seated drivers.

Sensor accuracy (gray area's) and the realization that everything man makes can ultimately fail, make that the real deployment strategy is somewhat more complicated then the one described above.
Sensors that can dynamically (i.e. during the crash) gauge the distance of the occupant to the airbag are still under development. They could affect the deployment decision in case the occupants are being thrown out of position prior to the impact.

Fortunately at long last the simulation analysis techniques have caught up too, so that the interaction of the occupant with the deploying airbag can be studied. This helps the automakers balance the airbag strength between providing sufficient protection and minimizing injury risk.

[ Driver Airbag Single Stage Deployment ] [ Driver Airbag Dual Stage Deployment ]
Driver Airbag Single Stage Deployment Simulation Driver Airbag Dual Stage Deployment Simulation

If you have questions about the contents of this page, please send email to

|Home |Services |Gallery |About Us |Contact |External Links |Site Info|

Pages last modified on January 15. Copyright © 2017 Hork Enterprises (