[RomP]

This is simply unkind to the competition (and to the poles alike): "Tesla achieved this outcome by nesting multiple deep aluminum extrusions in the side rail of the car that absorb the impact energy (a similar approach was used by the Apollo Lunar Lander) and transfer load to the rest of the vehicle. This causes the pole to be either sheared off or to stop the car before the pole hits an occupant."

There's winning and there's crushing the opponents. One shouldn't gloat in the latter case, but this case deserves an exception.

Congratulations Elon and the team on tremendously nice engineering!

[axylone]

Here's the video of the Tesla pole test: http://youtu.be/_Al3IUHt9Wc?t=32s

Honda Accord for comparison: http://youtu.be/ycntFFUfGkU?t=30s

[IgorPartola]

Anyone else notice Tesla's flying rear view mirror that nearly hit the dummy in the face? Maybe they forgot to apply the little alcohol wipe before gluing it on the windshield :).

[superfest]

If Tesla can bring this maintain this level of vehicle safety to its lower priced vehicles (gen iii vehicles: http://insideevs.com/tesla-gen-iii-to-eliminate-price-premiu...), then their dedication to quality, and not just economics, will be even more applaudable.

[mgolawala]

Tesla makes an impressive vehicle, but it is unfair to compare a ~$80K machine to a ~$30K one.

A fairer comparison would be to a BMW 5xx or Mercedes E class. Tesla may still fair better but not by as huge a margin.

Having said all that... wow, that is a surprisingly huge difference in results!

[Recoil42]

Same test with a Jeep Liberty:

http://www.youtube.com/watch?v=z4T1BimfUf4

Lexus RX350:

https://www.youtube.com/watch?v=mAz6ZMeIC9s

Lexus ES350:

https://www.youtube.com/watch?v=4l0o9MTY0JA

BMW 5-Series:

https://www.youtube.com/watch?v=-r20YfGYhvc

Mercedes C-Class:

https://www.youtube.com/watch?v=zTkbG64V4tQ

https://www.youtube.com/watch?v=sg0ubBqhZjg

[ericHosick]

It also looks like, in an accident, the lights on the Tesla start to flash.

[pilom]

The lights wouldn't appear to be flashing, they are blinking at a faster rate than the human eye can see and thus just to a person appear dimmer than they would if they were on full power.

Most dimming of LED's is done with this super fast blinking technique because LED's can be either on or off unlike incandesant bulbs. https://en.wikipedia.org/wiki/Pulse-width_modulation

[JVIDEL]

That's an Accord coupe, the B pillar is more thick than a 4-door version, wouldn't that mean the sedan would score even worse?

[nkoren]

Okay, that's impressive. The Accord isn't a bad car per se, but its dummy is clearly having a much worse day.

[Florin_Andrei]

> There's winning and there's crushing the opponents.

Crushing the opponents while your car remains uncrushed in the test (or the least crushed).

That's what you get when you ask the rocket scientist to make you a car.

[ISL]

If you ask a rocket scientist to make you a car that passes the crash test ratings alone, they'll build you a tungsten dump truck.

If you're going to win, you might as well reach the destination without noticing that you've had a collision.

[Someone]

The goal of a crash test is not to keep the car intact, it is to keep the driver intact. On "keep the car intact", today's cars do way more badly than those of the 1950's.

[jacquesm]

> On "keep the car intact", today's cars do way more badly than those of the 1950's.

I don't think that is true.

http://www.youtube.com/watch?v=joMK1WZjP7g

[Someone]

Nice video, deserves my upvote, but I am going to call it just anecdata ;-), so that I can add some partial counterarguments.

The video is a commercial, so they will have chosen an impact angle and speed that makes the newer car look best). For example, you can see that the newer car is heavier than the old one from the video; its front wheels still move forward when the wheels of the old one already go back. Because of that, the new car has a much longer braking distance than the old one. Things would have looked relatively better for the old car if they had chosen a collision with a concrete wall.

Also, Reading http://en.wikipedia.org/wiki/Crumple_zone, the crumple zone was patented in 1952; chances are this 1959 car doesn't have a good one.

If its crumple zone is bad or absent, chances are that the older car would have survived way better in a frontal collision, where the beam carrying the engine would be elastically compressed without deforming permanently.

That would be just the car, though; those old cars could be lethal at incredibly low speeds, for example by impaling them on their non-collapsible steering column.

[tempestn]

Great video. Pretty clear that both occupant and vehicle fared worse in the 1950s. (Although after a serious collision I doubt many people care much how the car looks.)

[at-fates-hands]

You also have to factor in a lot cars in those days didn't have, and weren't required to have seat belts. This safety feature alone has saved thousands of lives regardless of how safe the car is.

[Too]

Another interesting crash test is the Smart car. Youtube some videos of it. Due to its small size it hardly has any deformation zones and has to be built as a hard cage. It looks quite intact after a crash but i don't think the passengers inside would be.

[cenhyperion]

Rather my car take the impact than myself.

[notaddicted]

I believe one of the tests is crashing into a wall that yields very little, so the tungsten dumptruck would fare poorly, exciting as it would be to drive.

EDIT: this one: http://youtu.be/V5R80yUUVNk

[ISL]

The wall at left in that video is, at most, four dump trucks in volume and perhaps constructed of reinforced concrete (density ~2,500 kg/m^3)?

A single solid tungsten dump truck would have almost twice the mass of the wall. Even with a completely inelastic collision, the wall's going to shatter and move. For a 35 mph collision, I think you're right, that wall would ultimately bring the truck to a halt from friction with the ground. At higher speeds, I think the truck might make it through.

Tungsten is quite dense (19,300 kg/m^3). Such a truck would have a mass of >400 metric tons (2 x 3 x 4 m x 19 tons/m^3), or >247 Toyota Tacomas (2013 extended cab, curb weight 3560 lb --> 1618 kg).

The truck would be terrifying to drive, once you got it going. I don't know how you'd turn.

Alas, Youtube is short on tungsten dumptrucks, but this may suffice. In a demonstration, a truck uses its brakes to stop after obliterating a few cars. The truck is driven by a real person.

http://www.youtube.com/watch?v=h6TRqjjnO58

[jlgreco]

>400 metric tons

So around the max takeoff weight of a 747, with approximately as heavy duty a wheel setup I assume. That would be quite a truck.

[Florin_Andrei]

Given a heavy enough tungsten truck...

[mikeash]

I imagine NASA's crawler transporters would score exceedingly well on most of these tests, assuming the technicians survived.

[gchpaco]

Well considering they had a max speed of two miles an hour unloaded, you could probably jump off and get a beer before the crash happened.

[Ntrails]

"We're gonna crash, noooooooooooooooooooooo"

Flashbacks of Austin Powers and a steam roller...

[agumonkey]

Is it too far fetched to say he benefited from SpaceX research ? if so, this kind of gene transfer reminds me of Apple years when the iPhone benefited from iPod and Mac hard and software knowledge to leapfrog the competition, and then later helped the MBA do it again.

[mncolinlee]

I would imagine this is less the case than Subaru and Saab benefiting from their own aeronautical research. Unlike Subaru and Saab, SpaceX is a different company with different engineers than Tesla.

Rather, I think this speaks more to Elon Musk's uncanny ability to attract and hire good engineers.

[cheapsteak]

Similarly: "during validation of Model S roof crush protection at an independent commercial facility, the testing machine failed at just above 4 g's."

[bobbles]

Is there some kind of diagram that explains how this works? I can't really see how this works just by imagining it.

[curiousDog]

Not exactly a diagram but the a video of the test: http://www.youtube.com/watch?v=dz2FMfv-CSc

A bmw 3 series (not full frontal): http://www.youtube.com/watch?v=4XG8BQt4sHY

[ISL]

Modern crash engineering is really impressive (that under-car view is really neat!)

This 2009 vs. 1959 offset crash shows how much things have changed:

http://www.youtube.com/watch?v=joMK1WZjP7g

[bradleyland]

It's really unfair to compare offset and frontal crash tests. You can't draw any conclusions by comparing the two. For years, auto manufacturers were able to perform competently in the frontal crash test, but when the offset test was introduced, the failure rate went through the roof.

I don't mean to diminish the performance of the Tesla Model S. It is, without qualification, an incredible car. I just don't think comparing it to the offset video of a BMW are relevant.

[trafficlight]

Wow. The windshield doesn't even crack.

[kmfrk]

You have to consider that there is no engine in the front - just empty space for storage. That changes the game of frontal collisions completely.

[klinquist]

0:58 - 1:07 of the first video link... the car continues to attempt to propel itself into the wall.

[_mulder_]

Only for 0.3 seconds. The first camera shot at normal speed show's how quick this is all happening.

[sliverstorm]

It sounds like it's a special band running around the car, made of multiple layers of aluminum with space in between each layer. The layers crumple and transfer impact across the side of the car like a ski resting on soft snow.