You ever take a physics class? There are extreme forces involved in a collision like this. The acceleration numbers alone are astounding. I'd say this is pretty darn good for 40 years of safety innovation driven by consumers who would rather pay for 1000 HP engines than safety belts.

If the car didn't have a scratch (ie. the energy wasn't dissipated by the crumple zones), a lot more of the impact would be felt by the passengers.

What's the point of having a car without a scratch of the driver is turned to jelly...

This is true even with roll cages in race cars. You'd think you want the strongest metal available, but that's not right. They actually use a softer steel so the cage deforms a on impact.

The guy who welded mine said the shock from an impact to a titanium cage would effectively shatter your body. He was far from a doctor, but it made sense.

Proper protection in a car crash means two things: you must be safe from being crushed directly, and deceleration must be slow enough to keep you from internal organs injuries and injures to your neck. The first is taken care by firm safety cage of the car, and the controlled crumpling/deformation of the front of the car slows down the crash taking care of the second case. More time = less deceleration = less stress on your body.

Sturdy cage only is not enough—you would not want to be dropped from sixth floor in metal safe, would you? The safe itself may survive the impact "without a scratch", but that would not help you much…

This is the sort of attitude that kills people. Years ago I saw on a Seattle news station an "investigative" story about how easy some cars to damage.

As part of their "investigation" they took a bunch of SUVs and backed them into poles (something they claim was easy to do by accident). They then asserted that the big dents in the bumpers were a sign of shoddy engineering and cheap products. Even being a young middle school whipper snapper I knew bullshit when I saw it.

This sort of thing saves your life.

Most of those investigations deal with the fact that many 8 and 5 mile an hour bumpers (meaning they shouldn't break under lower impacts) end up shattering at maybe 2mph and cost thousands of dollars to fix -- implying that the manufactuers are engineering them that way on purpose to sell replacement bumpers or that they are no effectively engineering the bumpers to meet the specifications.

Conservation of energy hasn't changed in 40 years. The force of the impact has to go somewhere, and consumers don't want to pay for a Titanium car. The driver looked like he was in pretty good condition.

Maybe in 40 more years...

I think what codyhilton was trying to say (correct me here) is that the 'no scratch on the car + no remote possibility of injury' claim is about accident prevention, not dealing with the actual impact. As the five or so replies show, safety innovation is seen in terms of 'what to do when you hit something or something hits you.' But wouldn't true innovation also look at how avoid getting into that tight spot in the first place?

Some car manufacturers clearly have been working on this. Better lights (showing dangers ahead), improved handling and braking (to avoid hitting stuff), and more innovative things like radar-based warning sounds if the car things you're about to change lanes and hit someone (the one in the BMW 5 series still has some issues, but it's not bad), or automated braking systems like the ones Mercedes-Benz and Volvo are promoting.

Those systems are still in their infancy. But safety innovation is definitely going in the direction of accident prevention. We have done a lot to protect the fragile human body inside the metal cage by cleverly absorbing impacts through deformation zones, airbags, etc. The next big step will be figuring out how to avoid needing to rely on them. One suggestion has been that the regulatory battle will be the hardest challenge: once you take some level of control out of a driver's hands, litigation issues and complicated fault assignment rear their ugly head. And of course, many drivers will dislike the idea of being babied by a 'nanny car' and refuse to give up any level of control over their vehicles, not unlike (tortured analogy ahead) low-level programmers who insist in coding to the bare metal.

You're better off if the front end of the car dramatically collapses during an accident. It may destroy the car, but the more force absorbed by the crumpling, the less is delivered to the passenger.

When two cars are going 40mph there is an immense amount of kinetic energy. That kinetic energy has to go somewhere in order for those cars to stop. It can go into the frame of the car, it can go into the passengers, and it will into heat and noise. But it has to go somewhere. Until they build cars out of space Nerf they found on a passing comet we can choose to have the car crumble and absorb the impact or have the passengers absorb it.

Imagine you built a wooden box just big enough to hold an egg. Drop it on the floor from 5-6 feet; the box will likely be fine, but the egg won't.