Episode 195: Cannonball Chemistry

Air Date: November 11, 2012

A three-story fall into shallow water can be made more survivable by cushioning the fall with a mattress floating on top of the water.


This myth came from a scene in the TV series Burn Notice, in which the hero throws a hotel mattress from a balcony into a 4.5 ft (1.4 m) deep swimming pool and safely jumps onto it from 35 ft (10.7 m). Adam and Jamie began testing this scenario using a dummy with an accelerometer placed in its chest. When Buster, the dummy, was dropped butt-first into the plain water, he experienced 50g on impact with the water and another 29g impact with the bottom of the pool. When he was dropped onto an innerspring mattress, he experienced a deceleration of 86g. With 50g as their benchmark for lethality, neither drop appeared to be survivable.

To get even better results from their tests, Adam and Jamie repeated the drops using a sophisticated simulated cadaver. An orthopedic surgeon was also brought in to evaluate the damage using a portable X-ray machine. In the plain water fall, the simulated cadaver suffered a shattered pelvis and a paralyzing spine injury. When falling onto the mattress, it had far more damage all over, and as Jamie described, was “even more dead”.

At this point the myth was busted, but the duo wanted to see if skilled human jumper could possibly survive the jump. They brought in a stuntman who demonstrated that a normal, feet-first jump from 35 ft carried him to a depth of about 12.5 ft (3.8 m). The stuntman began training Jamie with techniques to reduce the depth of his entry. After a day of practicing jumps from increasing heights, Jamie deferred the full-height jumps to the stuntman due to soreness and concern for further injury. Ultimately the stuntman was able to jump from the full height while only dipping a few inches below the 4.5 ft mark, which Adam and Jamie reasoned would be a survivable impact had the pool only been that deep.

Cannonballs made of stone could be as effective as metal ones while also disintegrating on impact and not leaving the enemy anything to fire back.


The Build Team began by firing an iron cannon ball from Civil War-era cannon loaded with 1 lb (0.45 kg) of black powder. They measured its velocity at 1,200 ft/s (366 m/s) and used that for a benchmark when calibrating their other shots. When the team began tests with a self-made cannon that they had previously used for other myths, a misfire sent a cannonball into a residential neighborhood in California, an event that made national news and postponed their filming.

The team resumed their testing several months later. They made cannonballs out of three different types of stone: sandstone, limestone, and granite. They changed their testing location to a remote rock quarry and decided to use the Civil War cannon for all of their shots. They calculated how much black powder each type of ball needed to match the velocity of the iron ball and fired one test shot for each. To simulate castle ramparts, they used pallets of large concrete bricks. The iron ball penetrated 2 layers or bricks and damaged 4 layers. The sandstone ball penetrated 1 layer, damaged 2 layers, and disintegrated. The limestone ball penetrated 1 layer, damaged 3 layers, and disintegrated. The granite ball caused the same amount of damage as the iron ball and also mostly disintegrated, making this myth plausible.


  1. Jon Schwalm says:

    for the jumping into shallow water myth, I noticed that it didn’t look exactly like the footage. I was just curious if this might be more plausible with a different kind of mattress. also, you only test butt first sitting position. would the results differ if the person landed on their back? the pelvis definitely wouldn’t be nearly as injured and the spine would most likely be in better condition too. I liked the new medical dummies, but it this myth seemed to have been done just to get it over with; the testing wasn’t very thorough. also to add as a reason this could be possible. people have jumped into shallower water from this height and survived by belly flopping so I think the change in position would help a lot; the belly flopping world record is 36 feet into a kiddie pool. obviously when they filmed it they did it differently but this myth was no very conclusive, and it was not tested nearly as thoroughly as past myths.

  2. Dirk Bruere says:

    A belly flop from 36 feet into a child’s inflatable pool with 2 feet of water is survivable and has been done. HOWEVER, it may depend on being able to displace water due to the container.

  3. Jon says:

    Well, you try the belly flop. The accelerometer results are enough of a deterrent for me.

  4. Anon Y Mouse says:

    It seems intuitive to me that a human in control of their fall can do much better with the mattress than a dummy that cannot orient itself, deflect/absorb impact energy in a variety of ways, etc. Of course, a dummy human may share results.

  5. Stephen says:

    I should think the stone cannon ball “myth” should be confirmed.
    The presenters said something about “it was too time consuming make the cannon balls out of stone”.
    If you look up slave labour, peasantry and serfdom that existed in the middle ages, it explains how they made those cannon balls out of stone. They had the “human resources” to do it.

  6. Ian says:

    If you jumped into the water holding the mattress so that it hit the water first, couldn’t that also make the jump survivable? The mattress would take the bulk of what is almost a survivable impact (50g) and cushion against hitting the bottom. The mattress may also increase air resistance and (slightly) slow the fall.

Leave a Reply