[X] My Assistant

[Comga]

You can also look at the JPL site

http://deepimpact.jpl.nasa.gov/tech/instruments.html

There is a wealth of information on the instruments, and lots of links.

You can see that with a pixel projection of 10 microradians, (10 meters at 1,000,000 meters distance) if it snaps a picture ten seconds out, at a distance of 100 km, the resolution would be one meter. If it gets one five seconds out, its half a meter. The 20 cm resolution mentioned as a limit would require taking the picture at 2 seconds out, from 20km, and sending it to the fly-by spececraft before being vaporized. That assumes that the impactor doesn't get hit with a dust particle, which could cause it to lose stability. The crater would be uneffected, but any images would be smeared.

[edstrick]

The Ranger 7-9 moon probes were each transmitting images when they went ZIP-CRUNCH. In each case, a variable portion of a full-frame image <either A or B frame> and one of the 200 line partial scan <P1 through P4> images was "in transmission" at impact. Since the data was analog slow-scan raster, there was no problem making an image of the fragment up to the last millisecond of image.

Are the impactor's data compressed, so that partial frame data will be partially corrupted, or are they raw bit stream, so that every last pixel up to the one "in transmission" at the moment of impact will be recoverable?

[djellison]

It's going to be a bit like NEAR in a way isnt it - image image image wow - didnt expect them this low - image image - crunch

I would imagine there isnt enough grunt or time to compress, it's not like they can re-transmit to get back some lost bits - and compressed data suffers a lot with a single lost bit, so they'll wizz straight back to the fly-by uncompressed I'd imagine - and thus, the last image will almost certainly be partial
Doug

[dilo]

Thanks again, Comga!
About compression, I made a little small calculation: based on Impactor Technology description (http://deepimpact.jpl.nasa.gov/tech/impactor.html), it will transmit at nominal data rate of 64 Kbps, so this translates into a mere 16 KBytes every 2 seconds... a full res image from ITS, however, should be 1 MB uncompressed (assuming 8bit/pixel), so they should strongly compress it before transmission (more than 60:1 ratio!); this would require heavy/fast number crunching but, probably, board computer is not so sophisticated and, in addition, Doug is right about compression side effects.
I tend also to exclude pixel re-bin (subsampling) because this would compromise nominal resolution which is declared to be 20cm @ 20 Km distance. The only possible conclusion is that, at least for very last pictures, only a (central) small portion of image will be transmitted, probably something like 256x256 or even less, considering that shortly before impact the pace of imaging will increase until it reaches a maximum of one picture every 0.7 second...
So, under the best conditions, let's prepare to see only a small portion (50m?) of the comet nucleus with maximum resolution...

[Bob Shaw]

Judging from the Giotto dust impact experience (it *really* needed the Whipple bumpers!) I'd expect the smart bullet to get some pretty hefty whacks, so there's every chance it's tumbling whern it hits. Assuming that the transmitter antenna geometry isn't critical then that could actually give us an interesting spread of images - or none at all!

[Guest_BruceMoomaw_*]

Keep in mind also that that camera is going to get seriously sandblasted as it plows through the central regions of the coma -- I wouldn't count on its photos being much better than those from the main spacecraft's HRI.

[Comga]

The images will be sub-frame towards the end.

Note that Tempel 1 is a much less active comet than 1P/Halley, and seems to be less active this orbit than on its last orbit. Deep Space 1 flew by Borelley without ever recording an impact that registered on the attitude control system. However, 20 km is a lot closer than either Giotto or DS1 got.

If memory serves, Giotto got hit on the camera steering mirror that was sticking out around the edge of the Whipple shield. A probe can either be safe and blind, or stick its "neck" out, get the data, and take the risk. The same goes for the fly-by.

[abalone]

If memory serves, Giotto got hit on the camera steering mirror that was sticking out around the edge of the Whipple shield.  A probe can either be safe and blind, or stick its "neck" out, get the data, and take the risk.  The same goes for the fly-by.

There is also a very big difference in encounter velocity. Stardust was 6.1km/s, Deep Impact 11km/s vs Giotto 70?km/s That equates to an almost 50X kinetic energy content of similar sized dust particle that struck Giotto compared to DI.

[edstrick]

Giotto *LOST* the periscope during the flyby.
During post-encounter testing, the camera saw approximately no illumination, though the CCD and detector system were working, Also, the spacecraft <spin stabilized> balance behaved abnormally while they attempted to point the periscope. Inference was that most of the periscope and mirror were destroyed and debris blocked the light path between CCD and inboard optics and the mirror.

(this is from memory, I can't be more precise)

I'd ***LOVE*** some millennium to see what Giotto ended up looking like after the encounter.

[Bob Shaw]

There is also a very big difference in encounter velocity. Stardust was 6.1km/s, Deep Impact 11km/s vs Giotto 70?km/s That equates to an almost 50X kinetic energy content of similar sized dust particle that struck Giotto compared to DI.

Anyone have a handle on the distribution of dust sizes? As I recall, the Halley dust was made of very tiny - almost smoke - particles. A couple of lumps of sand or ice could really impart a kick!

[edstrick]

Comet dust seems to be made of stuff that's cigarette smoke sized. Look at the SUBSTRUCTURE of the dust-bunny clumps of cosmic dust we collect from the stratosphere. But this stuff does clump into clods, and even if a clod of comet dust has a density of 0.1 or maybe even 0.05 <like aerogel>... at 70 something km/sec, a 1 centimeter clod packs a whallop. Like the foam at 300 miles/hr and a shuttle wing, but with far far more bounce-per-ounce kinetic energy.

[djellison]

a 1 gramme clod at 11000 m/s is like a car hitting something at 30 mph or a Mach 1.5 golf ball

Doug

[helvick]

Just to put the problem of "cometary dust" in perspective.

0.05 grammes @ 70km/sec == kinetic energy of 1Kg at 1800kph.

Or about 60% of the KE of 1 round from the GAU-8, the gattling gun that the A-10 was basically built around (360 grammes @ 988m/sec). 6 of those is reckoned to be enough to take out a main battle tank, more or less.

50milligrams sounds small but it's about the mass of small piece of gravel, 5-6mm or so in diameter. Hopefully there's none of those.

The online material seems to indicate that nanometer to fractions of a micron in diameter particles are the order of the day at densities lower than the best vacuums on earth. I'd guess that it will be very nasty but much more like running into the top of a planetary atmosphere than hitting a shower of shotgun pellets.

[helvick]

Don't know why I was rabbitting on about 70km/sec. 10.2km/sec is what's up on the info page.

Yep ~30mph car it is, the idea of a Mach 1.5 golfball is too much to handle.

JoeM

[Bob Shaw]

Don't know why I was rabbitting on about 70km/sec. 10.2km/sec is what's up on the info page.

Yep ~30mph car it is, the idea of a Mach 1.5 golfball is too much to handle.

JoeM

So, there we are, on our bicycle...