On Mythbusters, they frequently drop cars to simulate a crash. How accurate could such a simulation be? What effects (dropped vehicle vs. crashed vehicle) would be different, and how?
Please note, I’m assuming the drops and crashes happen at the same speeds and impact against the same type of barrier.
Pretty accurate. The difference is that the dropped car will essentially be in free fall (no internal gravitational acceleration) prior to the crash. But if the vehicle is going any sort of reasonable speed into a solid obstacle, the effects of gravity are going to be very minor compared to the effects of the crash.
The difference on things inside the car will be bigger. You couldn’t really do crash tests (with dummies) in this way because the dummies wouldn’t be in the right positions just before the crash, and that could change greatly how the restraints affect them, etc. But the car itself has relatively few freely moving parts that would be majorly affected in this way.
A dropped car may pirouette or twist unexpectedly on the way down, but if they do the math right, it will hit the ground at the appropriate MPH. I don’t think any of the Mythbusters crashes or drops have needed the amount of positional accuracy or repeatability of a “real” 35 MPH front offset crash or whatever. They’re more about smashing up cars without hurting anyone, and using a crane with a remote release serves the purpose well.
Mythbusters has tried doing crashes with the vehicles on the ground and pulling them into other vehicles or barriers, and the setup is challenging. They’ve had enough cables snap and pulleys and / or anchors break loose to make them realize that they aren’t the NHTSA with a “real” crash test facility.
Another difference is that after the impact, the car is still being pulled towards the object it impacted (the ground). In a regular crash, the car would be free to bounce away in any direction.
One problem is that energy won’t dissipate in the same directions. In a typical accident (or any rapid deceleration), the rear of the vehicle will rise, while the front dips. Some of the vehicle’s forward momentum is converted that way. That doesn’t happen if the vehicle is dropped, since there is no equivalent of the ground for the front wheels to rotate around.
My first job was writing software to model data taken from films of barrier tests (we never called them “crashes”) at Ford. I viewed a couple of tests live and saw a lot of film footage.
There is a lot that happens during a crash that is influenced by the downward vector of gravity. This is alluded to by **TerpBE **and Really Not All That Bright. The vehicle pitches front-down. We had to be very conscious of this type of motion. We did testing to satisfy government regulations that were not exactly precise by engineering standards, using language like “the steering column hub intruding into the passenger compartment parallel to the horizontal axis.” But there really is no fixed horizontal axis in a car undergoing a violent collision. So we used the rocker panel as an approximation, and it pitches front down.
The motion of the passengers would be much different in a drop scenario, possibly tumbling into the ceiling or over the top of the dashboard, depending on restraints. Further, when dropped the car could topple in any direction, such as landing on its roof. This will not happen in a barrier test.
If Mythbusters is trying to determine an order of magnitude for damage caused to the vehicle, then dropping it would be a pretty good approximation, especially at the point of initial impact. However, it would be a poor predictor of passenger injuries, the type of car damage, or “crashworthiness” of a car.
I will also mention that these barrier tests were conducted not by driving a car into the barrier but by towing a dry car with a cable. You don’t want any fluids flooding the test floor.
My biggest question is how much “give” the barrier would have. In a vertical drop, the ground is not going anywhere. In a horizontal collision, all but the strongest barriers will be moved by the car at least a little bit. Is a car crash test supposed to represent a collision with an invulnerable and immovable object?
At Ford the barrier was a massive concrete wall. Believe me, it didn’t move, not even a little bit.
It represents a head-on collision with an identical car.
NTHSA and Euro NCAP tests use deformable barriers, at least for frontal impact.