Would a bullet made from ice be capable of killing somebody? Does it would melt because of the friction with the air?

By ice I mean any kind of Ice, including Ice II and all of its variations

I'm fully aware of that question: Would an ice bullet work?

  • $\begingroup$ Provided you get it moving quick enough, of course, it can. $\endgroup$ May 22, 2017 at 18:09
  • $\begingroup$ Ice has a tendency to shatter quite easily, which is non-ideal for bullets. However, there are things you can do that sort of "temper" the ice. For example, to give it strength and some plasticity you can mix in sawdust to get pykrete. This still wouldn't make a great bullet but it's a step in the right direction. $\endgroup$ May 22, 2017 at 18:16
  • $\begingroup$ This has been tried on a couple science shows, Mythbusters was one example. They all concluded it doesn't work. $\endgroup$
    – cobaltduck
    May 22, 2017 at 18:27
  • 3
    $\begingroup$ You are aware of that other question, how does this one difer? Is it only in removing the distance constraint? $\endgroup$
    – Theraot
    May 22, 2017 at 19:11
  • 2
    $\begingroup$ Mythbusters also showed that falling icicles can kill. So a sufficintly large calibre would “work” by some definition. $\endgroup$
    – JDługosz
    May 22, 2017 at 20:24

3 Answers 3


Mythbusters TV series attempted this on two different episodes.

In both, they were unable to create an ice bullet. They originally tried in episode 1, then in episode 14.

Magic Bullet: This myth tested the feasibility of magic bullets that can be used to assassinate without leaving evidence, used as a plot device or otherwise mentioned in many movies, such as Most Wanted or Three Days of the Condor.

An ice bullet can kill someone without leaving a trace. Busted The ice bullet evaporated before it could leave the barrel. This myth was retested in Myths Revisited, and remained busted with slow-frozen ice.


  • $\begingroup$ This is what I was coming here to say +1! $\endgroup$ May 22, 2017 at 20:48

Ice is pretty low density, 1 versus 11 for lead. It is also pretty brittle and will tend to fragment on impact so wound channels will be short. It will also tend to melt under the heat of the combusting powder. Finally, a normal barrel will spin it at a very rapid rate. (37,000+ rpm for .45 ACP) Which will cause it to shatter on barrel exit.

On the other hand, get something going fast enough and it gets dangerous.

So suppose we go for something like a smooth bore musket using compressed gas instead of combusting powder. It looks like modern air rifles can go 1,500 FPS. We could probably go better than that with a really high pressure supply and fancy ultrafast open valves. Maybe go 25mm bore, a length of 50mm and thus a projectile weight of 22 grams or so. Kinetic energy looks to be north of 400 Joules. This should be enough kinetic energy to do some real damage even if the projectile doesn't penetrate very well.

  • $\begingroup$ Colonel Moran can help you with your air rifle experiments! $\endgroup$
    – Willk
    May 23, 2017 at 16:07

Not practically

Bullets in guns get a lot of heat from the gunpowder exploding and propelling them through the barrel, and especially in threaded barrels the resistance the bullet experiences along with the gas is sufficient to not only melt the ice, but evaporate the water.

Then there is this case where the water is held by a container. Still, the water has only 1/9th the density of lead and air currents in supersonic flight make the projectile very inaccurate even when held in a container that keeps it from disintegrating right off the bat.

  • $\begingroup$ You ignore the effect of thermal conductivity. Applying heat to the exterior of a block of ice does not instantly heat the core. From a purely thermal perspective, the heat of ignition and barrel friction will simply melt the very outer layers of the ice, but they are applied for such a short time that the main body of the bullet will survive. The same cannot be said for mechanical forces. $\endgroup$ May 22, 2017 at 19:56
  • $\begingroup$ @WhatRoughBeast I took mechanical friction in account in my answer as well. But you're right. $\endgroup$
    – Hyfnae
    May 22, 2017 at 20:09

Not the answer you're looking for? Browse other questions tagged .