U.S. Navy: Record Breaking Rail Gun Shot
December 13th, 2010When I was at Brooks Institute in the 1990s, I used to be fascinated by the work done by the people who specialized in high speed photography (although, the food photography freaks were the weirdest kids on campus). I learned a bit about how speed stuff works, but that was a long time ago. Speeds have increased a few orders of magnitude since I last paid any attention to this kind of thing:
And then there’s the DRS Imacon 200: 200 million frames per second.
Now, with this rail gun video release from the U.S. Navy, it’s not the FPS that’s particularly interesting, but I’m quite baffled by the scene that starts at around 30 seconds.
How was the pan accomplished?
Via: Washington Post:
Inside a cavernous building at the Naval Surface Warfare Center in Dahlgren, Va., on Friday, a gigantic pulse of electricity hurled a 20-pound slug of aluminum out the barrel of an experimental gun at seven times the speed of sound.
The slug trailed a pillar of fire as it left the weapon and the building, illuminating the surrounding woods like a giant flashbulb. It streaked down range, generating a small sonic boom, and traveled about 5,500 feet before tumbling to the ground harmlessly.
In an adjacent building, there was a round of applause from observing scientists.
It was the latest test of the Navy’s electromagnetic railgun – a futuristic weapon that is right out of the latest video war game and could one day change the face of Naval warfare.
Roger Ellis, the railgun program manager, said people “see these things in the video games, but this is real. This is what is very historical.”
The gun is fired with a huge jolt of electricity that can propel a round more than 100 miles and at such velocity that it does not need an explosive warhead.
Two tests were conducted Friday – the first of which the Navy said generated a world record 33 megajoules of force out of the barrel. The second shot, witnessed by reporters, produced 32 megajoules.
Yeah that seemed totally fake even. What the hell??? I’m no expert in photography but that just looks fake.
I imagine one or several cameras mounted on a spinning platform, perhaps with a telescopic lens enabling it to be placed sufficiently distant from the bullet’s path that a very small amount of arc covers a good deal of the flight. The camera’s position and speed could even initiate the test sequence.
i would think that since this is a research video, and the unit itself was not designed by the military, but a contractor, that they would also have access to a variety of targeting and tracking gear provided by the parent company/affiliates and would not be likely to be using a sony handicam for this work, in these circumstances.
using a larger assembly to collect the entire run in a single frame, then using software boxing to track and zoom into the appropriate area, while cutting off the impact phase on the other end gives a clear and debris free area of view.
it wouldn’t surprise me to learn that this was a potential test for high speed object tracking within unmoving frames for fast track targeting such as rpg and other high speed projectiles.
using a larger assembly to collect the entire run in a single frame, then using software boxing to track and zoom into the appropriate area, while cutting off the impact phase on the other end gives a clear and debris free area of view.
That’s definitely not what they did.
The camera is panning with the projectile, causing the foreground and background to blur. If it was a static shot, with object zoom added in post, the foreground and background wouldn’t show the motion blur.
Cryptogon reader, P, sent this in. He neither has nor wants an account here, so I’ll just post the message:
anyway
the way the projectile may have been imaged is with a lens system that
incorporates a revolving prism or mirror.
See how the imaging systems of satelites work that image the earth
that are used by google/landsat etc. The scanning of the image plane is
not done by moving the entire lens and sensor, but rather by moving a
mirror that reflects the image into the lens and then the sensor.
The satelite images are made from lines of pixel data similar to the
way the image is built up on a TV.
The prism/mirror can rotate at any speed you want.
8X speed of sound = approx 8x 1,230 km/h (wikepedia sez it’s about 1,230
km/h) = 9840 km/h or 2744 metres/sec
Lets say the camera is about 100 metres from the projectile path.
And the path the projectile travels is about 20 meters.
Let’s work out the angular speed.
20 metres moving past the 100m distant camera (maybe perpendicular from
travel path for this example) takes about
20/2744 seconds or 7.28 milliseconds.
so, for the projectile to go 20 metres approximately 100 meters distant
from the camera it will take 7.28 milliseconds.
The angle the camera (and lens) needs to move is about 11.4 degrees.
(2 x atan(10/100) if you are interested)
As I said the camera need not move, only the moving mirror/prism needs
to move.
11.4 degrees per 7.28 msec
11.4 / (1/7.28e-3) = 1566 degrees per second.
which is about 4.3 complete rotations per second.
or maybe 258 revs per minute.
Most of the satelite imaging systems have prisms with 3 or more faces,
thereby reducing the speed that the mirror assembly needs to rotate
by this factor of 3 – or more.
Using a long focal length lens and timing the mirror rotation
I reckon it’s possible to get the footage they got.
The US govt can also afford really sharp lenses of very long focal
length so that seeing 20 kg mass move 100 metres away is easy.
I wrote back to P:
That panning shot starts out static and then pans with the projectile as it emerges from the muzzle blast area. Could the spinning prism theory somehow account for the shot starting out in a fixed position?
And P responded:
I think so, I estimated only 350rpm. Fast servo control is a reality.
See some of the pick and place robots that are available now.
http://www.youtube.com/watch?v=4A_2wMNZWEw
and
http://www.youtube.com/watch?v=tgVzYASPlPg&feature=related
gives you some sense of the speed of accurate motion.
I would think it’s pretty simple to get a mirror to rotate at this speed,
starting from rest and accelerating to 350 rpm almost immediately.
I notice the projectile is not tracked exactly by the camera setup.
The projectile starts at the left side of the frame, gradually gets into
the middle and
finishes out of frame at the right hand edge.
The camera is panning (mirror rotating) at increasing speeds too.
After looking at the film again, I suspect the camera/mirror rotation is
started at a period of time after the test firing. I think the
projectile went a little faster
than expected since it raced out of frame near the end.