SRB Retro-rockets for cars as an emergency braking system?

[Shuttle]

(Actually, in India, Airbags are still not that common)

Same thing with russian cars

[sanoj688]

Yes, let's fit our vehicles with bombs. That would be a good idea.

[DerekL1963]

Walk up to a stranger and ask that person what he/she thinks about an explosive charge stored in their steering wheel which will explode directly in their face in case of an accident. It's safe because there is a 0.08 mm membrane separating you from the explosion.

And.. yes, if it blows up spontaneously it can kill you or will break your arms. It will definitely break your thumbs.

Most of us do have cars with airbags.

Airbags aren't powered by explosive charges, they're gas generators and the difference is far more than semantic as (among other things) the former is uncontrolled while the latter is controlled. Asking people slanted and deliberately misleading questions is neither reasonable or useful.

[Seret]

Let's make one thing clear: A computer would have to detect imminent impact and this detection mechanism would have to be foolproof and 99.9% reliable.

That's actually not a biggy, we're at that point currently. Cars with autonomous emergency brakes are on the market now.

Retro-firing solid fuel rockets wouldn't be a great way to stop a car. The required impulse and the direction would vary widely. Obviously it'd be useless for impacts with pedestrians and dangerous in impacts with motorbikes and convertibles. Not to mention the fact that you'd be installing pyrotechnics into the crumple zone close to hot things like an ICE and its radiator, which would be a fire hazard. No thanks.

[LordOfTheOrbits]

A couple of reasons why this isn't going to happen:

1. If they fail, they could easily make your whole car explode in a crash

2. They would cause pretty high gee-forces which would often do the same or more damage to the driver than the actual crash

3. They would only work the way it was intended at a certain speed, for example if you would use the amount of solid fuel required to kill 200km/h at 100km/h you would go 100km/h backwards after firing them, which would almost certainly be much worse than the crash

4. When they fire they can easily set the car in front of you on fire

[Seret]

4. When they fire they can easily set the car in front of you on fire

Any cracks in the propellant and you'd set yours on fire too. Solid rocket motors are fairly delicate animals.

Personally I'd rather see the problem solved with much higher automation of cars, including short range comms between vehicles. I'd much rather have my car talk to the one I was about to hit and see if between them they could come up with a maneuver that avoided or lessened the severity of the crash.

[rasheed]

Why not use Water rockets instead?

[K^2]

"Finally, there is the question of velocity difference. Say I'm driving without ABS with someone who is driving with one. I drive only 2s behind at highway speeds. Say, 70mph."

2s at 70mph is 62 meters. Assuming 1 shock second until your foot is on the brake pedal you have less than half a football field to come to a stop or you crash. Fact of life, sorry.

I do not see the problem keeping a 45m distance to the car in front of me. Half a football field is not that much in a car.

If you actually pay attention, you'll notice that I do account for collision that follows. I do use .5s reaction time, because that's a typical reaction time in an accident. Some people might take up to a second to react. Some people might freeze all together. An alert person responds in approximately .2 - .3 seconds. But .5s is a good average. Why don't you go back and read why it's not as bad to collide in this scenario.

"I ride the guy's bumper just .5s behind."

Are we still at 70mph? Thats a distance of 15m. One would describe it as one car length. At this speed this is roadrage.

What sort of cars are you talking about? 15m is 3 car lengths.

And if you think that's too small, you've obviously never driven on a freeway outside a major city. 2 car lengths is not atypical, and you see less than 1 frequently when somebody changes lanes. In some places, such as vicinity of Miami, you'd be driving on a freeway with 4-5 lanes of traffic, ranging in speed from about 60mph to over 90mph as you go from right to left. All driving at 1-2 car lengths and constantly shifting from lane to lane.

Now, you might drive differently. But everyone else isn't going to suddenly change their driving style. The person behind you isn't going to know that your emergency braking system is going to stop you almost instantly, and he's still going to get behind you with less than 1 car length to spare in heavy traffic. The difference is that if you slam your brakes, you won't have time to slow much before he slams into you. And that means the impact won't be severe. If you use rockets, it gives you just enough time to stop before he hits you at full speed. Which I have already shown with numbers in my previous analysis which you haven't even bothered to look over, apparently.

Let's make one thing clear: A computer would have to detect imminent impact and this detection mechanism would have to be foolproof and 99.9% reliable.

<Video Link>

Do I really need to explain difference between operating an 18-wheeler and a passenger car? For starters, an 18-wheeler can't swerve in principle, because its trailer will keep going in the same direction, only helping the whole thing go into a flip.

[Kerbin Dallas Multipass]

If you actually pay attention, you'll notice that I do account for collision that follows. I do use .5s reaction time, because that's a typical reaction time in an accident. Some people might take up to a second to react. Some people might freeze all together. An alert person responds in approximately .2 - .3 seconds. But .5s is a good average. Why don't you go back and read why it's not as bad to collide in this scenario.

Where do you account for the collision that follows and why is a collision preferable to a braking maneuver? A collision is nothing different to braking, just that it happens within a shorter space and much higher g forces and all the other negative side effects of an actual crash.

Typical reaction time for an unexpected event in traffic is about 1s. Reaction time for an expected event is obviously shorter, this goes down to 0.1s in martial arts. Consider that you do not only have to react mentally but also perform an action with a relatively clumsy and heavy limb aka your right leg. I will not argue about numbers you made up.

What sort of cars are you talking about? 15m is 3 car lengths.

Youre right. It's 3 car lengths. Its a gap that one car could squeeze in (we do not want the cars to touch).

And if you think that's too small, you've obviously never driven on a freeway outside a major city. 2 car lengths is not atypical, and you see less than 1 frequently when somebody changes lanes. In some places, such as vicinity of Miami, you'd be driving on a freeway with 4-5 lanes of traffic, ranging in speed from about 60mph to over 90mph as you go from right to left. All driving at 1-2 car lengths and constantly shifting from lane to lane.

Now, you might drive differently. But everyone else isn't going to suddenly change their driving style. The person behind you isn't going to know that your emergency braking system is going to stop you almost instantly, and he's still going to get behind you with less than 1 car length to spare in heavy traffic. The difference is that if you slam your brakes, you won't have time to slow much before he slams into you. And that means the impact won't be severe. If you use rockets, it gives you just enough time to stop before he hits you at full speed. Which I have already shown with numbers in my previous analysis which you haven't even bothered to look over, apparently.

Yes I do drive differently. Here you go 125mph in the left lane and you have to change lanes because you have to clear lane for someone who is coming up in the rearview going 30mph faster.

Fines for driving up too close to another vehicle are draconic in my country. It's easy to lose your license if they catch you a few times.

I admire your driving skills, but it's not safe and it's not a reason to categorically outrule improved braking mechanisms. (This thread being about a hypothetical, unrealistic one)

Do I really need to explain difference between operating an 18-wheeler and a passenger car? For starters, an 18-wheeler can't swerve in principle, because its trailer will keep going in the same direction, only helping the whole thing go into a flip.

The principle of emergency braking is the same for 18 wheelers and cars. I repeat we are talking about an unavoidable crash situation here and the mechanism we are debating would only use the remaining short distance to decelerate.

If I follow your logic you are also opposed to things like ABS and ESP because these concepts actively avoid swerving and are known to lengthen braking distance in optimal conditions.

[DaveofDefeat]

I would rather have an ejection seat in my car.

[Seret]

I would rather have an ejection seat in my car.

The sensible part of me thinks this is a terrible idea, but the ex-ejection seat technician part thinks it would be awesome. They're just the coolest machines built by man.

[K^2]

Where do you account for the collision that follows and why is a collision preferable to a braking maneuver? A collision is nothing different to braking, just that it happens within a shorter space and much higher g forces and all the other negative side effects of an actual crash.

Car in front of me has time to stop, and I slam into the back going 43mph. Given that it's an impact into a car of similar mass, my car is trashed, but I'm completely unharmed, save for a bruise from the airbag.

Now, lets suppose that instead of braking at .9g, the car in front suddenly pulls a 5G maneuver. In the .5s scenario above, even braking at .9g, I slam into car in front still going 63mph.

Typical reaction time for an unexpected event in traffic is about 1s. Reaction time for an expected event is obviously shorter, this goes down to 0.1s in martial arts. Consider that you do not only have to react mentally but also perform an action with a relatively clumsy and heavy limb aka your right leg. I will not argue about numbers you made up.

The eye-brain-hand reaction is about .15s, and can be close to .1s for some people. The eye-brain-foot pathway is much slower. You will never be able to respond in less than .2-.3 seconds.

As for 1s vs .5s for typical reaction time, 1s only makes my argument stronger, but feel free to run the numbers yourself.

Youre right. It's 3 car lengths. Its a gap that one car could squeeze in (we do not want the cars to touch).

That's still two cars that can squeeze in by my count.

Yes I do drive differently. Here you go 125mph in the left lane and you have to change lanes because you have to clear lane for someone who is coming up in the rearview going 30mph faster.

Fines for driving up too close to another vehicle are draconic in my country. It's easy to lose your license if they catch you a few times.

Again, how you drive on an autobahn in low density traffic is a separate issue. Most accidents in these conditions are single vehicle anyways, and no braking system helps. We are discussing mostly dense traffic at high speeds. If this never happens where you live, I'm very happy for you.

I admire your driving skills, but it's not safe and it's not a reason to categorically outrule improved braking mechanisms. (This thread being about a hypothetical, unrealistic one)

This isn't about my driving. I'm telling you what general traffic is like outside of many cities in US and Europe. I've personally driven in and near New York, Boston, Miami, and a number of smaller cities which are not quite as insane as these 3. LA is just as bad, but I've only been a passenger there. People don't slow down in these cities just because it's almost bumper-to-bumper traffic. And if you slam your brakes suddenly, you are almost guaranteed to be rear-ended.

But the difference is being rear-ended by someone as you are slowing down vs being rear-ended by someone after you've come to a complete stop. The retro-rockets system brings you to a stop in about the same time as it takes for a car following close behind to catch up. Meaning it's a full-speed-on-stationary collision, which is absolutely horrible for people in both cars.

The principle of emergency braking is the same for 18 wheelers and cars. I repeat we are talking about an unavoidable crash situation here and the mechanism we are debating would only use the remaining short distance to decelerate.

By the time the crash becomes 100% unavoidable, it's too late to stop even with retro-rockets. That's what we're talking about. For an 18-wheeler it's different, because an unavoidable crash is determined much, much longer in advance.

If I follow your logic you are also opposed to things like ABS and ESP because these concepts actively avoid swerving and are known to lengthen braking distance in optimal conditions.

Personally, I would never drive a car with ABS or ESP, because these systems put me in more danger. But I go out and practice emergency braking on ice every winter, and I personally maintain my brakes and suspension, so that I know they work to their full potential.

For a typical driver, ABS does better emergency braking. And as I've pointed out, even if the person behind you skids, taking his friction coefficient from .9 to .3, collision is nowhere near as bad as full-speed-on-stationary collision. So ABS never creates a more dangerous situation than it is designed to avoid. Same deal with ESC and roll protection. But that's not the case with retro-rockets. They don't just reduce your braking distance while allowing for a safe impact from behind. They stop you dead. If they don't, then the distance from which you detect "imminent collision" has to be such that no system could ever predict it in real traffic. There is no way around these two limitations. You either end up with the system that ends up triggering in normal traffic without any emergency, or you end up with a system that stops so rapidly that it gives no chance for a safe collision from behind. There are no alternatives.

[rdfox]

For the record, frontal-impact airbags actually INCREASE the potential danger to properly restrained (i.e., seat belted) drivers; the only reason they're mandatory in so many countries is that nobody can be bothered to enforce seat belt laws and get people to wear their seat belts properly. If they were a safety benefit, they'd be mandatory in professional racing series instead of six-point safety harnesses; instead, virtually every racing series that uses production-based vehicles (either "showroom stock," sports car racing, or otherwise) mandates that any air bags be deactivated and/or removed, partly because of their potential danger to the driver, partly because they pose a genuine risk to rescue workers until they can kill the car's electrical system. I'd much prefer to NOT have an airbag in my car, because I've always worn my seat belt, and it's a personal rule that I've had since childhood that the car does not move when I'm not buckled up.

(Note that none of this applies to side-curtain air bags; those are beneficial to safety even with properly-restrained occupants, due to the fact that there's essentially no crumple zone possible in side impacts, and it helps protect from intrusion into the safety cell. The only thing that does apply is that they are a risk to rescue workers until the electrical system is deactivated by disconnecting the battery, which is why race cars prefer massive side bolsters on the seat to provide the same effect.)

[jwenting]

Best way to avoid a crash is to actually pay attention to other traffic, and keep some distance from other vehicles...

[K^2]

For the record, frontal-impact airbags actually INCREASE the potential danger to properly restrained (i.e., seat belted) drivers; the only reason they're mandatory in so many countries is that nobody can be bothered to enforce seat belt laws and get people to wear their seat belts properly. If they were a safety benefit, they'd be mandatory in professional racing series instead of six-point safety harnesses; instead, virtually every racing series that uses production-based vehicles (either "showroom stock," sports car racing, or otherwise) mandates that any air bags be deactivated and/or removed, partly because of their potential danger to the driver, partly because they pose a genuine risk to rescue workers until they can kill the car's electrical system. I'd much prefer to NOT have an airbag in my car, because I've always worn my seat belt, and it's a personal rule that I've had since childhood that the car does not move when I'm not buckled up.

And would you care to explain why this is in disagreement with every crash test ever conducted?

As for race cars, have you seen a typical race car collision? And a typical highway collision? These aren't the same thing. A highway collision involves driver continuing to move forward as the car very rapidly decelerates. Putting an airbag in the way helps. A typical race car collision involves car pirouetting through the air. So the best safety option is to have a very good harness that keeps you strapped to the sit and inside the roll cage. And yes, if passenger cars came with race car harnesses instead of the typical seat belt, you probably wouldn't need an air bag. But it certainly doesn't present an additional danger.

Now, for the reason why passenger cars typically have a basic belt instead of a full racing harness, that probably has a lot to do with convenience. Though, safety at low speed collisions and ease of exiting are also factors. But with such a belt instead of a harness, you are much better off with an airbag.

[Seret]

For the record, frontal-impact airbags actually INCREASE the potential danger to properly restrained (i.e., seat belted) drivers

No they don't. Studies have varied in the degree of positive impact from frontal airbags, but they've all agreed that they do improve safety. That's why they're mandatory in most countries.

the only reason they're mandatory in so many countries is that nobody can be bothered to enforce seat belt laws and get people to wear their seat belts properly.

Sorry, you've got this one backwards. One of the few cases where airbags are dangerous is if people aren't wearing their seatbelts. Effectively what happens in a frontal impact is the unbelted person accelerates towards the front of the car, and when the airbag deploys it slams them hard in the opposite direction. This very rapid change of direction results in a high jolt force (g s^-1) which is bad for head and neck injuries.

There are a few corner cases where airbags can be harmful. As mentioned above, unbelted passengers are at risk, as are babies in rear-facing car seats. But for properly belted adults airbags are a good thing.

Edited February 16, 2014 by Seret

[jwenting]

only reason (if at all) for higher injury and fatality ratings for airbag fitted cars is because there are a lot of idiots who think "it's got an airbag, I'm now invincible", go driving insanely, lose control, and hit a tree or light post.

[rdfox]

Gonna have to cobble this together, since the replies are on different pages...

And would you care to explain why this is in disagreement with every crash test ever conducted?

It's not. A properly restrained passenger never even reaches the airbag; one in an improper seating position or not wearing the belt properly (or in a crash beyond the design constraints of the car's safety cell, which is what IIHS has taken to doing to try and get more cars rated poor so that the insurance companies can charge higher premiums) will see the occupant strike the airbag and suffer blunt force injury to the face and chest. The only instance where a properly-seated, properly-restrained occupant will benefit from the airbag itself is in a crash so severe that the safety cell is compromised to the point that the steering wheel or dashboard are displaced sufficiently that they would strike them even with the seat belt in place. In *those* cases, the airbag reduces the acceleration experienced by the occupant by lengthening the time that the impact takes, but that's a situation where the safety cell itself has failed and severe injuries are to be expected regardless of what protection systems the car has due to cockpit intrusion. In cases where deformation of the safety cell is less severe, the airbag can cause injuries that would not happen if the occupant wasn't going to strike the car's interior.

As for race cars, have you seen a typical race car collision? And a typical highway collision? These aren't the same thing. A highway collision involves driver continuing to move forward as the car very rapidly decelerates. Putting an airbag in the way helps. A typical race car collision involves car pirouetting through the air. So the best safety option is to have a very good harness that keeps you strapped to the sit and inside the roll cage. And yes, if passenger cars came with race car harnesses instead of the typical seat belt, you probably wouldn't need an air bag. But it certainly doesn't present an additional danger.

I'm pretty sure I've seen more racing crashes than you have. Cars flying through the air are a very small percentage of racing crashes, as sanctioning bodies have put various aerodynamic measures into effect specifically to keep the cars on the ground and scrubbing off speed; the most obvious of these being the passive lift dumper "roof flaps" that NASCAR now mandates on all its cars, which use the same technique as the spoilers on an airliner to shed lift. The most typical racing crash sees a car either have a tire fail or lose control, slide off the racetrack, and into a barrier at high speed, usually in a single-car incident. In that way, they tend to be almost identical to highway accidents. (Indeed, the component of velocity perpendicular to the barrier is often surprisingly low--the crash that killed Dale Earnhardt at the 2001 Daytona 500, where his car was traveling at approximately 185 mph when it struck the wall, had a velocity component perpendicular to the wall of 41 mph.) And even in the rare incidents where a car DOES lift off and start pirouetting, it sheds almost no velocity while airborne. Virtually all of the delta-V comes from the multiple impacts with the *ground* as it tumbles along, and from friction as it slides to a stop.

The difference is essentially irrelevant, anyway. The actual behavior of the occupant is the same--the car decelerates abruptly, the occupant continues along under inertia, and the car's structure has enough time to get a significant change in its velocity before the occupant strikes it and is forced to make the same change in an immensely short time. In all situations, the ideal way to protect the occupant is A) to essentially "weld" them to the car's structure, so that they're a part of it and change velocity at the same rate, and prevent intrusion that would cause them to contact anything that's moving relative to the car's safety cell. (This is why street cars now use inertia-reel seat belts with explosive pretensioners, and race cars use belts fitted to the driver and then snugged down to the point of nearly cutting off circulation--the old theory that a little bit of slack in the belt to gradually slow the occupant down was shown to be faulty, instead causing blunt-force trauma impacts of the type mentioned above.) This is done by restraining the occupants in a way that doesn't allow them to move at all relative to the safety cage, and then by making the safety cage itself strong enough to remain intact in even the worst impact situations considered plausible. Air bags are an inferior solution, to be honest; a form of cushioning the impact akin to using a padded steering wheel or dashboard. If you *are* going to hit part of the vehicle's structure, then an airbag will help, but it's vastly preferable to *prevent* contact in the first place.

Now, for the reason why passenger cars typically have a basic belt instead of a full racing harness, that probably has a lot to do with convenience. Though, safety at low speed collisions and ease of exiting are also factors. But with such a belt instead of a harness, you are much better off with an airbag.

It has everything to do with convenience--the standard three-point belt is much easier to use than a five- or six-point harness, particularly if you're wearing a skirt (just try wearing the crotch strap then!), and it's more comfortable, too, since it has the slack of the inertia reel, plus it's less expensive for the automaker to install a single strap with a single buckle instead of five or six separate straps anchored to the car's frame. Ease of exit isn't really related to the seat belt (racing harnesses use a single-pull quick-release buckle similar to that used on airliners), but rather to the presence of the full roll cage that blocks a significant percentage of the door openings. As for low-speed collisions, there's no real difference between a three-point belt and a five/six-point harness there.

Sorry, you've got this one backwards. One of the few cases where airbags are dangerous is if people aren't wearing their seatbelts. Effectively what happens in a frontal impact is the unbelted person accelerates towards the front of the car, and when the airbag deploys it slams them hard in the opposite direction. This very rapid change of direction results in a high jolt force (g s-1) which is bad for head and neck injuries.

Actually, this is the place where airbags provide the most benefit. Because the airbag is not a rigid structure, but instead is an air spring that also is designed to DEflate as quickly as it inflates, the unbelted occupant strikes a (relatively) soft cushion that extends the length of the acceleration compared to striking the steering wheel, the dashboard, or the windshield. Even if the bag is still inflating when they strike it--which is unlikely; it's designed to deploy completely before any occupant could reach it in the worst credible crash--it will still deform much more than the car's structure, allowing a more gradual acceleration.

A good comparison to the airbag in terms of "softness" would be the SAFER Barrier developed to protect oval-track racing drivers in crashes by putting an energy-absorbing barrier between them and the concrete retaining wall on the edge of the track, which will deform as much as 18 inches (half a meter) during a heavy impact by a heavy vehicle. When it was first being introduced, people referred to the barrier as "soft walls," because the concept was that they were softer than the concrete retaining walls. That said, they're not actually SOFT; the front surface that the cars hit is 1/8" steel plate, with the energy absorbance coming from the styrofoam backing behind it. It still smashes hell out of the car, and if you kick it, you'll still break your toe, but it's a lot softer than bare concrete. It's a similar thing with airbags--they're not big soft pillowy things; they can easily cause severe bruises and even broken noses and orbital bones, but they're a hell of a lot softer than the car's structure.

Edited February 17, 2014 by rdfox

[K^2]

It's not. A properly restrained passenger never even reaches the airbag

I've seen properly restrained passengers with facial bruises that say otherwise. Crash test data concurs.

(or in a crash beyond the design constraints of the car's safety cell, which is what IIHS has taken to doing to try and get more cars rated poor so that the insurance companies can charge higher premiums) will see the occupant strike the airbag and suffer blunt force injury to the face and chest.

Because we all know that real life crashes never exceed any safety limits. That's why everyone always walks away from them perfectly fine.

Your argument is absolutely absurd. What matters is that we take a range of collision speeds from a bump in the parking lot to a head-on collision with oncoming at highway speeds. And we look at range of injuries and survivability with and without air bags. That's what matters in real life situation. That's what crash ratings look at. And yes, they go past what the car is meant to withstand, because guess what, that's the scenario that's going to kill you. If your crash was mild enough that your seat belt was enough and you never hit airbag, good for you. But if your crash was severe enough for you to plummet face first into the airbag, it might have just saved your life.

which use the same technique as the spoilers on an airliner to shed lift.

Airliner nver does anything to reduce lift. And it's not how spoilers work either.

The most typical racing crash sees a car either have a tire fail or lose control, slide off the racetrack, and into a barrier at high speed, usually in a single-car incident. In that way, they tend to be almost identical to highway accidents.

And both are fatality free in majority situations. So they are kind of irrelevant. How many racing crashes do you know that are head-on either into an on-coming car or into a fixed barrier? (Not one of these soft crash barriers.)

A typical racing incident either has the car brake into a specially designed barrier, or is a glancing collision, which sends the car spinning or rolling. (Yes, "flying" was an exaggeration.) You don't see extreme forward deceleration as much as lateral in racing accidents.

Road accidents are by far dominated with linear deceleration.

Indeed, the component of velocity perpendicular to the barrier is often surprisingly low

That's kind of the point, yes.

The difference is essentially irrelevant, anyway. The actual behavior of the occupant is the same--the car decelerates abruptly, the occupant continues along under inertia, and the car's structure has enough time to get a significant change in its velocity before the occupant strikes it and is forced to make the same change in an immensely short time.

Just so long as we keep in mind that direction of acceleration in racing and road accidents tends to be very different, moving on...

In all situations, the ideal way to protect the occupant is A) to essentially "weld" them to the car's structure, so that they're a part of it and change velocity at the same rate, and prevent intrusion that would cause them to contact anything that's moving relative to the car's safety cell.

This is true only while you can guarantee that the acceleration of the safety cell proper is within survivable envelope. I can strap you to the inside of a solid steel structure and use it as a wrecking ball. You aren't going to make it, simply because the very harness to which you are attached is going to undergo accelerations that will brake your rib cage and ruin your internal organs.

This is a limitation for real cars as well. Which results in there being severe limitations on how strongly the restraints can hold you in case of rapid, linear deceleration. These are the cases in which we rely on the air bag. The safety cell proper decelerates too fast for you to survive with restraints that can hold you. A typical belt will simply break your ribs if it's strong enough to hold you to the seat.

This is why such a thing as airbag exists. You reduce the holding power of the seat belt, you allow passenger to travel through the cell under extreme deceleration, and you mitigate the problem with an almost-instantly deployed airbag.

And the reason this isn't a factor in racing is

Indeed, the component of velocity perpendicular to the barrier is often surprisingly low

Good. Now we're learning something.

(This is why street cars now use inertia-reel seat belts with explosive pretensioners, and race cars use belts fitted to the driver and then snugged down to the point of nearly cutting off circulation--the old theory that a little bit of slack in the belt to gradually slow the occupant down was shown to be faulty, instead causing blunt-force trauma impacts of the type mentioned above.) This is done by restraining the occupants in a way that doesn't allow them to move at all relative to the safety cage, and then by making the safety cage itself strong enough to remain intact in even the worst impact situations considered plausible. Air bags are an inferior solution, to be honest; a form of cushioning the impact akin to using a padded steering wheel or dashboard. If you *are* going to hit part of the vehicle's structure, then an airbag will help, but it's vastly preferable to *prevent* contact in the first place.

Now, I'll give you a few minutes to compare what you just said here, focusing on difference in design of passenger car belts and racing harnesses, with everything going on above.

It has everything to do with convenience--the standard three-point belt is much easier to use than a five- or six-point harness, particularly if you're wearing a skirt (just try wearing the crotch strap then!), and it's more comfortable, too, since it has the slack of the inertia reel, plus it's less expensive for the automaker to install a single strap with a single buckle instead of five or six separate straps anchored to the car's frame. Ease of exit isn't really related to the seat belt (racing harnesses use a single-pull quick-release buckle similar to that used on airliners), but rather to the presence of the full roll cage that blocks a significant percentage of the door openings. As for low-speed collisions, there's no real difference between a three-point belt and a five/six-point harness there.

I agree that it's a major factor. Thank you for spelling out the differences for people who aren't aware of them. But you have to understand that it's not only a matter of convenience. There are factors here that have to do with differences between typical road and racing collision. And this goes for everything from the durability of the cell, to amount of crumple space allowed by a vehicle, and by far not least, the way the collisions actually happen. Some of these limitations are due to cost, yes. You can make a car that's half less likely to kill you in a catastrophic collision, but if it costs twice as much for all the same features, nobody is going to buy it. And that's a fact you have to deal with. Saying that this is why airbags are useless is absolutely wrong. They provide similar levels of protection at a significantly lower costs not just to the restraint system, but to the entire vehicle. From frame, to engine design, to the seats themselves.

Actually, this is the place where airbags provide the most benefit. Because the airbag is not a rigid structure, but instead is an air spring that also is designed to DEflate as quickly as it inflates, the unbelted occupant strikes a (relatively) soft cushion that extends the length of the acceleration compared to striking the steering wheel, the dashboard, or the windshield. Even if the bag is still inflating when they strike it--which is unlikely; it's designed to deploy completely before any occupant could reach it in the worst credible crash--it will still deform much more than the car's structure, allowing a more gradual acceleration.

You seriously need to look at some of the actual reports, both from crash tests, and actual accidents.

Airbag dramatically increases risk of neck injuries to unrestrained passengers in low speed collisions. In fact, there have been cases of airbags killing an unrestrained passenger. You seem to have zero understanding of how these systems work or timing involved. A properly restrained passenger strikes airbag after it has deployed. An unrestrained passenger is hit by deploying airbag, typically, at an upward angle, which causes the head to be pushed backwards while torso continues down and forward. This is how people brake their necks.

[Nibb31]

The best safety device would actually be a huge metallic spike mounted on every steering wheel and pointed at the drivers chest. If this was mounted on every vehicle, I think everybody would keep a nice safe braking distance from the vehicle in front.

[Seret]

You seriously need to look at some of the actual reports, both from crash tests, and actual accidents.

I agree. While much of what you're saying sounds perfectly reasonable rdfox, it doesn't agree with the actual studies and data. Stuff like this:

The results of this study suggest that seatbelts alone or in combination with an airbag increased the incidence of AIS1 spinal injuries, but provide protection against more severe injury to all regions of the spine. Airbag deployment without seatbelt use did not show increased protection relative to unrestrained occupants.

and this:

Passenger air bags may be a hazard to unrestrained children and of little benefit to unrestrained adults

At most you find studies that show air bags don't make much difference compared to seat belts (like this):

If the associations are causal the average risk of driver death was reduced 8% (95% confidence interval 4% to 12%) by an air bag. Benefit was similar for belted and unbelted drivers

What you don't find is anything saying that airbags help unrestrained passengers. It's interesting to note that most of the research dates back over 10 years and that there's been very little done lately. This shows how uncontroversial the current body of evidence is IMO. I still would have liked to see something like a Cochrane study done to round everything up, but from a quick scrounge I couldn't find anything.

[spudman2]

Why don't we mount SRBs for acceleration instead of deceleration? If you use them while trying to accelerate there's a much lower chance that you would hit someone in front of you. They'd have to be in front of you, which is the direction you're looking and acceleration, and you would be much more inclined to check if there was someone in front of you before pushing the button.

[Seret]

Why don't we mount SRBs for acceleration instead of deceleration?

Well, we have.

[rdfox]

Because we all know that real life crashes never exceed any safety limits. That's why everyone always walks away from them perfectly fine.

Your argument is absolutely absurd. What matters is that we take a range of collision speeds from a bump in the parking lot to a head-on collision with oncoming at highway speeds. And we look at range of injuries and survivability with and without air bags. That's what matters in real life situation. That's what crash ratings look at. And yes, they go past what the car is meant to withstand, because guess what, that's the scenario that's going to kill you. If your crash was mild enough that your seat belt was enough and you never hit airbag, good for you. But if your crash was severe enough for you to plummet face first into the airbag, it might have just saved your life.

You look at a range of collision delta-Vs (not speeds, but the net change in velocity, covering both speed *and* angle of impact), and you design the car to protect against what's considered the Worst Credible Case. It's a basic principle of engineering that you do NOT design to withstand the absolute worst case possible; you design to withstand the worst case *likely*. It's a subtle, but key difference, and explains why, in ALL designs, there are situations where the answer to the question, "What if THIS goes wrong?" end up being, "Well, it's pretty much gonna be a bad day." The reason for this is that there's ALWAYS a way that you could be putting it through worse conditions; the old phrase "cost-benefit analysis" comes into play. It's certainly *possible* to design a car that would protect its occupants against collisions from any angle at any closing speed that would be theoretically possible. The problem is that it would be far too large to meet with legal limitations (wouldn't fit in a lane, too long for a standard driver's license, etc.), too heavy, completely impractical due to the sort of structure that you'd have to navigate to enter or exit the vehicle, and, most of all, so expensive that nobody could afford it, not even Bill Gates.

Indeed, the primary standard that cars are designed to is, depending on where it's meant to be sold, either the NHTSA or Euro NCAP crashworthiness standards, because those are the ones that you *must* pass for it to be legal to sell. (I can't remember all of them, but the frontal-impact standard for NHTSA NCAP is no fatal injuries to unrestrained occupants in a 30 mph crash into a full-width fixed barrier perpendicular to the velocity vector--yes, that particular standard was set in the late 60s, why do you ask?) Cars aren't designed to withstand a head-on collision with both cars going at freeway speeds for the simple reason that such crashes are seen as not credible cases, for various reasons.

Airliner nver does anything to reduce lift. And it's not how spoilers work either.

Very wrong. Have you ever ridden on an airliner in a position where you could see back half of the wing out your window? Immediately after touching down, the aircraft's spoilers (which can be deployed to a small extent in flight for aerodynamic braking effect) are deployed fully, to approximately an 85 degree angle. This disrupts the airflow over the wing, both shedding lift (to place the aircraft firmly on the runway to make the wheel brakes effective) and directly generating a great deal of drag to help slow the airplane down; it also completely destroys the lift being generated by the airplane's flaps, causing them to provide additional drag to slow the airplane down. (Technically, airliners also reduce lift significantly as they climb and accelerate out from the airport, too, by gradually retracting their flaps; once airspeed is high enough, you don't need the additional lift that extended flaps and slats generate, and retracting them reduces drag.) Basically, you want all the lift you can get right up to the moment that you've planted the mains on the runway, and then you want *no* lift, to help make sure you can stop before going off the end of the runway.

And both are fatality free in majority situations. So they are kind of irrelevant. How many racing crashes do you know that are head-on either into an on-coming car or into a fixed barrier? (Not one of these soft crash barriers.)

Well, let's see. Oncoming car, not often, though I've seen plenty in all sorts of racing series where someone plowed head-on into a *stopped* car at full racing speed, but for a fixed barrier? Right off the top of my head, I can name Michael Waltrip in the NASCAR Busch Grand National Series at Bristol International Raceway, Tennessee, in 1991, when he blew a tire, shot straight into the wall, hit a crossover gate, which failed, and plowed head-on into the fixed concrete barrier beyond the now-open gate. Mike Harmon, same series (now the "NASCAR Grand National Series"), same place on the same track (now Bristol Motor Speedway) in 2002, almost identical crash, except that the wreckage of his car was then hit by ANOTHER car after coming to rest. Mark Martin, NASCAR Sprint Cup Series, Michigan International Speedway, 2012, when he spun down pit road, managing to get to the inside pit wall just at an opening where the cars can enter the garage area, and hit the concrete wall SIDEWAYS at over 100 mph, just aft of the driver's seat. J.R. McDuffie, NASCAR Winston Cup Series, Watkins Glen International Raceway, 1991, when he suffered complete brake failure, shot off the track at full speed, crossed the grass runoff area, and plowed head-on into the wall, killing him. Tony Roper, NASCAR Craftsman Truck Series, Texas Motor Speedway, 2000, when a freak accident caused his truck to turn (not slide) hard right to hit the wall head-on at nearly full speed, killing him. Adam Petty and Kenny Irwin, Jr., both killed testing for the Grand National series at New Hampshire International Speedway in 2000, in two separate but nearly-identical crashes where their throttle stuck wide open, and the configuration of the corner caused them to hit the wall with a velocity vector perpendicular to it. I'm sure that if I put my mind to it, I could come up with dozens more, in dozens of different series--those are just the ones that popped into my head within a minute of reading that question.

A typical racing incident either has the car brake into a specially designed barrier, or is a glancing collision, which sends the car spinning or rolling. (Yes, "flying" was an exaggeration.) You don't see extreme forward deceleration as much as lateral in racing accidents.

Road accidents are by far dominated with linear deceleration.

Incorrect. The vast majority of road accidents are, technically, glancing blows where the two vehicles do not hit each other square, and end up spinning away from the point of impact. (This is why Euro NCAP and the IIHS use offset frangible barrier tests rather than the NHTSA fixed-barrier test; this represents a more realistic collision where the cars aren't lined up perfectly square.) The difference is that it's more dramatically visible in racing accidents due to the higher speeds involved, making them slide further and thus be able to spin more before stopping.

This is true only while you can guarantee that the acceleration of the safety cell proper is within survivable envelope. I can strap you to the inside of a solid steel structure and use it as a wrecking ball. You aren't going to make it, simply because the very harness to which you are attached is going to undergo accelerations that will brake your rib cage and ruin your internal organs.

This is a limitation for real cars as well. Which results in there being severe limitations on how strongly the restraints can hold you in case of rapid, linear deceleration. These are the cases in which we rely on the air bag. The safety cell proper decelerates too fast for you to survive with restraints that can hold you. A typical belt will simply break your ribs if it's strong enough to hold you to the seat.

The entire job of the design engineers is to guarantee that the acceleration of the safety cell proper is within the survivable envelope, using crumple zones and other energy-management features, in any crash up to and including the Worst Credible Case. If the car goes through a crash within those design constraints, then the safety cell *will* remain within a survivable acceleration envelope.

Also, have you ever been in a car crash? I have. I've been in a side-impact crash well beyond the car's design criteria (closing speed of 60 mph, directly into the passenger door), with my younger brother sitting in the passenger seat. He underwent far worse acceleration loadings than in any credible head-on collision, but because he was wearing his seat belt properly, he suffered relatively mild injuries--none of which came from the seat belt itself. (He broke three ribs and his collarbone due to the other car intruding into the safety cell, and one rib punctured his lung when he was writhing in pain as the paramedics examined him before they extracted him from the car; even so, he came home from the hospital four days later.) Seat belts, be they three-point belts in production cars or five/six-point belts in race cars, inherently have some "stretch" to them that allows a total of about a quarter-inch of additional passenger forward movement in a crash (which is why it's important to replace your seat belts if your car is fixable after a major crash--the stretch is permanent and the belts are thus compromised and no longer safe after the crash).

If your seat belts can decelerate you at a rate that would be fatal, and the safety cell is decelerating at that rate, then you WILL decelerate at that rate, regardless of whether you have airbags or not. Airbags provide *supplemental* deceleration to try and prevent you from impacting the car's structure; if anything, the acceleration loading when you hit the airbag is *higher* than when you're just being restrained by the belt. However, the force needed to generate this greater acceleration is spread over a much larger area, spreading the load out over a larger area of the body and thereby reducing the force on each individual part of the body.

This is why such a thing as airbag exists. You reduce the holding power of the seat belt, you allow passenger to travel through the cell under extreme deceleration, and you mitigate the problem with an almost-instantly deployed airbag.

No, no, no, completely wrong. The seat belt is not reduced in holding power in any way; this would be insanely dangerous in the approximately 10% of crashes where one or more of the airbags fail, because then the occupants would collide with the car's structure as though they weren't effectively restrained, and we're down to eating steering wheel again.

You seriously need to look at some of the actual reports, both from crash tests, and actual accidents.

Airbag dramatically increases risk of neck injuries to unrestrained passengers in low speed collisions. In fact, there have been cases of airbags killing an unrestrained passenger. You seem to have zero understanding of how these systems work or timing involved. A properly restrained passenger strikes airbag after it has deployed. An unrestrained passenger is hit by deploying airbag, typically, at an upward angle, which causes the head to be pushed backwards while torso continues down and forward. This is how people brake their necks.

I've been paying attention to crash test results since the late Reagan administration, before airbags were mandatory. I also recall the Big Three spending huge amounts of money on airbag research in the late 60s in hopes of avoiding having to pay license fees to Volvo for their patented three-point seat belt design, figuring that it could replace the shoulder belt, or possibly even the entire seat belt. I grant you, based on further research quoted later, that it has now been shown to be of minimal benefit to the unrestrained driver, but the original design was based entirely on Head-Injury Criteria (HIC) and Total Acceleration numbers; it was intended to keep the unrestrained occupant from striking the vehicle structure hard enough to be instantly fatal. The whiplashing effect causing basal skull fracture and internal decapitation was unforeseen, because it wasn't something the engineers were told to consider.

Now, I will acknowledge, my opinions on this may be colored somewhat by my only personal experience with airbags being a crash my grandparents were in back in the early 90s, where they rear-ended another car at about 20 mph, and the front-seat occupants (with three-point belts and frontal air bags) suffered severe chest and facial bruising and broken noses, while the rear-seat occupant (with only a three-point belt, and the same motion clearance before they would strike anything) being left uninjured. Airbag technology has advanced greatly since then, but I still contend that *frontal* airbags are of limited value for a properly-restrained occupant; I'd much prefer a good seat belt system combined with a well-designed bucket seat (deep side and anti-submarine bolsters) and proper seating position for frontal crashes.

[Forte]

I know this post has sort of gone off topic, but another problem is that the retro-rockets would have to be balanced perfectly. If they were too low on the car, the back would most likely go flipping over the front, and that would make it a bad safety feature