View Full Version : Deep Impact: The Mission

Nowhere Man
2005-Feb-16, 11:53 PM
From the navigator (or something like that) of Deep Impact, with his permission:

Good morning all,

Today we're doing our first Trajectory Correction Maneuver (TCM). We've been working hard on this activity for the last couple of weeks.

That sentence doesn't really do it justice. We've been working 14 hour days 7 days a week since before launch and we've been working 18 hour days for the last couple of weeks on this.

Flight Director, Ed Swenka has just polled us for our GO to start the TCM sequence. One by one, we give him our go's.

Charlie Schira, Steve Waydo and I have been here for about an hour looking over our telemetry pages, setting up realtime plots and making talking through what we expect to happen.

The burn should take about 280 sec and will put 28.6 m/sec of velocity on the spacecraft. We'll be thrusting mostly away from the earth, speeding us up in our orbit around the sun, making up for the small shortfall in velocity that the Delta-2 launch vehicle left us with on launch day. Without this burn, we would arrive at the encounter 14 hours late and miss the comet by 380000 km.

To thrust in the desired direction we have to turn the spacecraft almost 180 deg to an attitude where we can only get 2000 bits per second of telemetry over a low-gain antenna.

While we wait, we fret about the burn time. The spacecraft is designed to measure the accumulated delta-V with onboard accelerometers and shut off the burn when we get to the desired 28.5684 m/sec.

We also have a parameter to limit the maximum thruster ontime. It's our safety net. If the burn continues too long, it shuts off the thrusters before we do too much damage to the trajectory.

We're worried that we've got the parameter set a little too tight. That it might shut of the burn early even though things are really going fine. Earlier this week, I computed that the thrusters should be on for a total of 985.03 sec and put 1103.53 sec in the command sequence for our max burn time, but since then, we found a few subtle effects that could cause the burn to run a bit long. Since the command sequences had already been reviewed and tested, we decided not to change the value and take our chances.

We'll find out shortly, the thruster catbed heaters will come on in about 10 minutes and the burn will start about an hour after that.

Steve Collins Deep Impact Attitude Control
Steve Collins was also navigator on Deep Space One, as well as (IIRC) at least one of the MERs.


2005-Feb-16, 11:56 PM
Here's hoping...

Nowhere Man
2005-Feb-17, 12:00 AM
From the navigator (or something like that) of Deep Impact, with his permission:

It's 18:25 UTC. The command sequence for the TCM has started to clock out. The catbeds are on and we can see their temperatures coming up.

A few minutes later, we switch to the low-gain antenna.

Uh oh... Turns out a commands we built in a rush yesterday has a problem. It was supposed to disable some fault protection on the temperature of the thruster brackets. We expect these to get hot after the burn as the heat from the thrusters soaks back through the plumbing without the cooling effect of the flowing propellant. If they get above 50 C we will trip fault protection and go into safing.

It's 3 min until we start the 30 min turn to the burn attitude.

I throw the relevant temp channels into a Tball plot so we can keep an eye on them.

The Tball display of the spacecraft shows that we have started our flipover turn to the burn attitude.

Folks are running around building a replacement for the errant fault protection command.

Telecom is talking about uplink margin. The Earth goes between two low gain antennas during the turn and there is a period of time when we can't command.

Fault protection and the sequence team are in high gear getting the new command built.

Telemetry has dropped out during the turn as expected.

A few minutes later, Jim Taylor, the telecom guy, says carrier is back in lock now.

"Flight, ACS, I can report we're on track to the burn target."

Folks huddle around Ben Toyoshima's command console double checking things as he uplinks the rush fault protection command.

3 minutes till the burn starts...

I report that we have reached the burn attitude.

Tick Tock Tick Tock...

Then things start to happen fast. The burn begins. As usual, the first indication is the doppler display. It reports the change in the frequency of the radio signal coming down from the spacecraft. Half a minute later, we start to see indications in telemetry.

The Charlie and I watch the control errors like a hawk. They are staying under .15 deg. Much better that we usually see in the testbed. Our plot of accumulated delta-V slowly ramps upward. I report the delta-V every few seconds, thinking how I sound like some kind of rocket scientist or something.

The plot of the pressure drop seems to be matching my predicts pretty well, not falling through the floor the way they would if we were going to burn for too long.

I watch the thruster time creep upward toward that spooky cutoff limit. I sure hope we put in enough pad...

The doppler plot begins to curve. That's a sign that our planned "taper" has begun. Reducing the thrust so that we can cut off more accurately. As it does, I can predict ahead and am relieved to see that we're going to stop the burn on the accelerometers before we reach the burntime cutoff.

Steve Sodja, our prop guy, says he can see the RCS thrusters have started to fire. The burn is over and we're cleaning up any residual pointing error.

I check the delta-V. 28.56 m/sec. The spacecraft thinks we hit it on the money. The accumulated thruster time is 992.5 sec. Close to my predict of 985 sec and well below our 1103.5 sec cutoff time.

There is much rejoicing. Somewhere in the middle of it, we start the turn back to the cruise attitude. NAV comes on the net and reports a "quick look" of a 1.5% underburn.

Hmmmmm... That seems a little big, but I'm not too worried. Our telemetry is saying we did better than that and I know NAV has to collect data for a while before they can make a high accuracy estimate.

Now the game becomes watching the thruster bracket temperatures. As predicted, they are shooting upward now that the burn is finished. We think we got the fault protection turned off, but if we didn't, we'll go into safing when the brackets reach 50 C.

Fault protection sees the bracket temp pass 50 C. Good call. If they hadn't been able to get that rush command in, we would have gone into safing.

The turn back to the cruise attitude is done and the command sequence takes us back to the High Gain antenna. Jim Taylor, our telecom guy, quips that the only way he can tell that we've done a TCM is that the frequency has dropped by 1515 Hz.

I shake a lot of hands. Keyur Patel, our deputy project manager taunts me once again about my outfit. I've costumed myself "Apollo era" today: Black slacks, white short sleeved shirt, narrow black tie. In place of a slide rule is my Kyocera 6035 with RPN calculator software. It's starting to seem pretty retro itself. As I hand it to Charlie to do some calculations, I reflect that it has more onboard computing power than the entire Saturn V stack.

NAV comes back on the net and reports their newest estimate is a 6.6 mm/sec underburn. That's only 0.02% error. Sweeeeeet!!

Ed, Charlie and I can't stop laughing.

Steve Collins Deep Impact Attitude Control
TBall is a spacecraft navigation simulator that Steve developed for DS1, and adapted for Deep Impact.


Nowhere Man
2005-Jul-02, 01:21 PM
From the navigator (or something like that) of Deep Impact, with his permission:

I'm still alive after what has turned out to be a very long year. The Deep Impact project has been, well, let's say "challenging". I'm currently keeping myself going by repeating the phrase "Vaporize Impactor Star Tracker" in a Zoolanderesque mantra.

Current mission status: It could work.

The encounter sequence is onboard and clocking out. Navigation looks good and we've seen 2 sizable outbursts from the comet. It's huffing and puffing as it rotates in the sun. The IR spectrometer is giving excellent data and indicates that the signature of water went way up during the outbursts. The science folks are seeing some other intriguing features in the excellent spectra being proved by instruments.

The light curve of the comet is consistent with the ground based estimates of the comet rotation period. Structure in the light curve would seem to imply that there is substantial shading in concave areas.

At today's Navigation Team "Daily Show" they report that the estimates of our trajectory relative to the comet are hardly moving at all. Just a few tenths of a km. It looks like I may not have to spend Saturday updating TCM-5 with a new burn target. They show a movie made from the optical navigation pictures taken over the last several days. You can see the comet growing brighter as we approach and the outbursts we happened to catch. Somebody viewing the black and white sequence (ok it was me...) observed that it was like watching a Chaplin movie.

Tomorrow at 5:00 PM Pacific TCM-5 will clock out, turn us briefly away from the comet and perform a burn that gives us about 0.3 m/sec perpendicular to our in-bound velocity as a final correction to the impactor's trajectory. Then at about 11:00 pm we'll fire the pyros to launch the impactor and "hit the jets" to perform the divert maneuver.

Divert is a much bigger burn, over 100 m/sec that will re-target our trajectory (we miss the comet by 500 km) and slow the flyby spacecraft down ( so we arrive after the impact in a position to get good pictures). If it all comes together, we'll end up with a cool time sequence of the impactor blowing a RoseBowl sized crater in the comet.

My personal timeline calls for me to attend a 9:00 am meeting to see if we need to do the update to the TCM-5 target, then spend the afternoon doing last minute checks and nervous hand wringing. A relaxing afternoon nap to bank some sleep, then we'll gather to watch TCM-5.

Because of the burn direction we won't be able to point the High Gain Antenna at Earth and will have to watch the burn via agonizing 10 bits per second telemetry (Gaaah! I can type that fast!).

I'll spend an hour or two looking at the telemetry we recorded during the burn and then things will get really interesting as we gear up for separation of the impactor and the divert burn. If it goes OK, I'll try grab a few hours of shut eye in my office before attending a 4:00 am NAV meeting to report on the divert burn performance. If we need to, we can do a Divert Trim burn 10 hours from impact to make sure the flyby spacecraft is on the right trajectory to image the impact.

The real nail biting will be over the impactor. It has to stop its tumble, lock onto the comet, figure out if it's on target and then turn and perform 3 correction burns a few hours before impact to make sure that it hits the comet.

It needs to work, or we'll have to settle for the consolation prize, the highest resolution pictures and spectra ever taken of a comet nucleus. Plenty cool, but not what we came here to do.

More as it develops

Steve Collins Deep Impact ADCS

2005-Jul-02, 03:20 PM
How long after impact will it take before we know if the impactor impacted?

2005-Jul-02, 06:07 PM
How long after impact will it take before we know if the impactor impacted?
Boston Globe Story (http://www.boston.com/news/nation/articles/2005/06/30/nasa_set_for_a_crash_with_comet/)
a NASA probe traveling 6 miles a second will slam into a comet 83 million miles from earth.
83 million miles = 133.6 million kilometers.
At 300,000 kps we should know if it worked in 445 seconds, 7.4 minutes.

Nowhere Man
2005-Jul-03, 01:52 PM
From the navigator (or something like that) of Deep Impact, with his permission:

Spacecraft is behaving itself this morning, though we're having a minor problem at DSN Station 63 that's causing some data dropouts. They just turned "ranging modulation" off and it brought the signal up enough to keep the data flowing.

NAV has just given us a report on the orbit estimates they made overnight and nothing much has changed. The solutions are moving just a few hundred meters, so the trajectory correction target we have up on the spacecraft is good enough. That will earn Ed and I a little rest this afternoon. Thanks NAV.

The burn sequence is already onboard and running. It will clock out tonight at 5:00 pm PDT and move the aimpoint by 34.1 km and put us on the desired impact trajectory with the comet for Impactor release. I do some
last minute nervous checks that it will take us in the right direction. Looks ok.

Steve Collins Deep Impact ADCS

Nowhere Man
2005-Jul-03, 01:55 PM
From the navigator (or something like that) of Deep Impact, with his permission:

I run for home for a couple of hours, grabbing lunch and stopping off to buy an 8 inch Dobsonian reflector from a JPL retiree. I came across it yesterday in the JPL paper. I wanted something with enough aperture to see the comet if it ends up getting bright. I try to sleep, but only get a half hour before I need to head back to the lab to watch TCM-5.

Mike Hughes Charlie and I verify that the burn targets made it up to the spacecraft. Everything looks fine.

Kevin, our lead Fault Protection engineer comes by and hands Charlie a list of spacecraft "states" to verify. They have apparently found some AutoNav switches that aren't set quite right and we're going through everything one more time to make sure we're set up correctly for encounter.

The telemetry rate drops to 10 bits/second. We have to switch onto our Low Gain antenna because at the burn attitude, we don't have enough gimbal travel for the HGA to point at Earth. 10 bits is always frustrating, but being at this low rate for a critical burn this close to encounter is agonizing.

Eventually, someone puts a plot of the doppler up on one of the big projector screens. It's obvious from the plot that burn has completed. Someone calls the NAV on net and asks for an evaluation. They reply with a terse, "Dopper looks nominal" that makes everyone in the MSA laugh. Thanks for letting us in on your little secret guys! Usually, they beat us to the punch reporting every event because the doppler has none of the latency that exists in the processed telemetry stream.

It will be another 20 minutes before we start getting high rate telemetry, but all indications are that the impact targeting burn went just fine, maybe a 1% underburn.

A half hour later we get our recorded telemetry and Charlie and I spend some time using Tball to look through it for problems. We're briefly concerned about some wiggles in one of the plots and pull up a couple of earlier TCM's to compare. Looks like we've seen the behavior before. I report our findings on the voice net and the flight directo quickly moves on to the next major event, activating the encounter critical sequence.

I feel it's time that I return to Earth to feed...

Steve Collins DI ADCS

Nowhere Man
2005-Jul-03, 01:58 PM
From the navigator (or something like that) of Deep Impact, with his permission:

Yummy Japanese dinner.

Yikes! I just found a problem. Our attitude estimator is set up in OVERRIDE2 instead of....

Got to work now



We've finished the final poll before impactor sep and everyone is GO. ADCS was working a bunch of issues over the last hour so it's been kind of crazy. The OVERRIDE thing, temperature on the impactor gyros, etc... Nothing that would stop the SEP or divert burn that follows.

Steve Collins Deep Impact ADCS

Nowhere Man
2005-Jul-03, 02:03 PM
From the navigator (or something like that) of Deep Impact, with his permission:

Seven minutes to SEP.

It's out of our hands now. The laws of physics won't allow us to get a signal to the spacecraft in time to stop the event.

Rick Grammier, our project manager, walks the rows of consoles patting folks on the back and being encouraging. The room is almost silent. I'm poring over

Dave Spencer, our Mission Manager comes by and asks me about the angle from the Earth to the Sep direction. We conclude that the sep should be visible in the doppler signal.

The former project managers are gathered in the glass room next to the MSA hugging each other.

Time creeps by. Alarms pop up on our screen for each expected event. Inerting the hard line between the Flyby and Impactor, arming the pyro buss, electrical disconnect and finally separation. We won't know if
it worked for seven more minutes.

Finally, the traces on my realtime plots jump, NAV comes on the voice net and announces that they see the Sep in doppler. Then, the dreaded voice of Kevin, our Fault Protection Engineer:

"I'm sorry to report that we've run a fault response..."

We jump into a flurry of activity looking at telemetry.

It's "attitude error". That's *US*. The red lines showing me the error in the spacecraft pointing have exploded off the top and bottom of the plot. I zoom out till I can see what happened. Sitting beside me, Charlie reads the numbers off the plot and converts them from radians to degrees. "Four and a half," I repeat the number on the net for the rest of the room to hear.

FP Kevin comes back on the net. "The fault occured while responses were still disabled". If nothing else happens, we should be OK.

Looking at the Tball display of the spacecraft, I can see a turn starting. For a moment I fear we are turning to sun point. When I check, we've turned to the attitude for the impending divert burn. The attitude errors have jumped back down. I zoom the plot in to find that they're back to values that look normal.

I know that over in building 264, Mike Hughes, our ADCS lead must have his heart in his throat. He was nice enough to let me "drive" during Sep-Divert, but this is really his chair. Thanks Mike.

The situation on flyby spacecraft seems to be stable. Now our attention turns to the impactor. The time for signal acquisition comes and goes. Long moments pass with no contact from the impactor. The S-band telecom engineer reports "no lock".

Felicia Sanders, our brilliant Ground System engineer reports that she sees telemetery from the impactor. The room literally explodes with applause. After a few seconds, folks manage to stuff their excitement back inside and turn back to work.

Lew Kendall, the lead ADCS guy for the impactor works through his checklist and reports that the impactor is working. Folks practically sing their positive status reports on the net.

Keyur Patel, the Deputy Project Manager is beaming. "Very nice. Very nice," he says.

12 minutes after Sep, the divert burn starts. Charlie and I are glued to our screens. We have a 0.5 deg hang off in attitude error. Charlie does a quick computation and confirms that even if it persists, it won't affect the trajectory enough to worry about.

The burn takes 12 long minutes. We're putting 102.5 m/sec on the bird. I call out our progress every 33 m/sec with Steve Sodja, our Prop guy confirming that his hardware is right on his predicts. Then, the burn is done. The delta-V goes flat, and before I can zoom in to check the magnitude, Keyur is on the net asking about maneuver performance. I stall a few seconds to get him an accurate number. It's right on the money.

Holy Cow... A lot of spooky stuff just happened...

...And everything seems to be working.

Don Hampton, the Instrument engineer reports that a picture has come down showing the departing impactor. He makes it into a JPG and I throw it up on one of the screens.

The Project Manager declares that we can shelf the Late Release and Flight System Impact contingency plans. The room breaks into another round of applause.

Over the next hour the bigwigs who have gathered slowly diffuse away while the folks from Ball who built the spacecraft recall tales from assembly and test.

Ed Swenka and I make a late night run for food. In and Out is already closed. We consider raiding the semi parked out front and then move on. The food we eventually find works. After eating just a little, I am no longer hungry.

I return to my office to grab an hour of sleep before the NAV quick look meeting. Someone has thoughfully turned down my cot/bed, leaving me friendly note and a piece of chocolate on my pillow. The sleep is good. When Ed and I arrive at the meeting, it's clear that no Divert Trim maneuver will be needed. The Flyby spacecraft is within a few km of its intended aimpoint.

Now I'm on the way home for a few hours of suprise rest. Need to be ready in case tomorrow outer space is not as friendly as today.

Steve Collins Deep Impact ADCS

The Bad Astronomer
2005-Jul-03, 06:11 PM
Hey, are these online somewhere? This is a great running commentary!

I'm trying to find the image of the departing impactor. I can't find anything yet.

Nowhere Man
2005-Jul-03, 06:33 PM
Hey, are these online somewhere? This is a great running commentary!
Actually, I get them via a private mailing list, to which Steve Collins belongs. I'll let him (and the list) know what you think.

Steve gave us similar commentaries for Deep Sapce 1, back when.

Edit to add, one shot of the impactor is currently on the front page of http://www.jpl.nasa.gov/


The Bad Astronomer
2005-Jul-03, 06:40 PM
Yeah, I just found that image too. I also put up a blog entry about this.

And thanks!

2005-Jul-06, 09:05 PM
Good job, Steve.

2005-Jul-20, 07:00 PM
Deep Impact’s Flyby Craft Ready For Course Correction Today (http://www.space.com/missionlaunches/050720_flyby_update.html)

The Flyby spacecraft used in NASA’s Deep Impact mission to comet Tempel 1 is to be powered into a new trajectory today—a maneuver that could allow it to explore yet another comet in years to come.
Engineers here at Ball Aerospace & Technologies Corporation, designers and builders of the $333 million dual-element Deep Impact mission are delighted with the overall health of their Flyby vehicle. It watched the washing machine-sized Impactor as it collided with the target on July 4 at 1:52 a.m. EDT (0552 GMT).
Today’s trajectory correction maneuver—or TCM in spacecraft operations lingo—"will put the Flyby spacecraft in an orbit path that allows us to contact it at any time. This provides NASA time to investigate opportunities to utilize the Deep Impact spacecraft for future missions," Henderson told SPACE.com.
The trajectory change will place the spacecraft on an Earth-return heading, arriving in late 2007 or early 2008, said Donald Yeomans, Supervisor of the Solar System Dynamics Group at the Jet Propulsion Laboratory (JPL) in Pasadena, California. The Deep Impact project is managed by JPL.

"That will allow a gravity assist to comet Boethin in late 2008. Then we will have to submit—and win—a Discovery proposal for the extended mission funds," Yeomans added.

2005-Jul-21, 02:13 AM
Info on Comet Boethin.


2005-Jul-21, 06:11 PM
Follow-up to my previous post:

NASA Announces Deep Impact Future Mission Status (http://www.spacedaily.com/news/deepimpact-05o.html)

Dantzler announced today that all investigators interested in using the Deep Impact Flyby Spacecraft for further science investigations must submit proposals to the 2005 Discovery Program Announcement of Opportunity for a Mission of Opportunity.

"All proposals for use of the Deep Impact spacecraft will be evaluated for science merit and feasibility along with all submitted proposal for Missions of Opportunity," he said. "The spacecraft is being offered as is. Proposers must include mission management and spacecraft operations in the total proposed funding."

Manchurian Taikonaut
2005-Jul-24, 12:37 PM
What a great mission Deep Impact has been, I hope it can continue to bring back more great results