OFJ Field Journal from Steven Tyler - 10/7/95
PRACTICING DIFFICULT TASKS TO DETERMINE POSSIBLE PROBLEMSHere's what I did for the last couple of days...my journal topic for now:
December 7 of this year will be a very busy day for Galileo. We'll have our encounter with Io, our closest approach to Jupiter, our passage through the most dangerous part of Jupiter's radiation environment, the relay of data from the Galileo Probe (which separated from our Orbiter in July and will fall into Jupiter on December 7), and the biggest burn of our "big" 400 Newton engine (which supplies about 100 pounds of thrust). This burn will put Galileo's Orbiter into an orbit around Jupiter; if the burn fails, we will fly on past the planet, and our opportunity for getting into orbit will be lost.
You might wonder about a couple of things. First, what if our burn is off by just a little bit? Can we correct it afterwards? Second, do we practice for problems like this with some sort of "dry run" of the actual work?
It turns out that we do. One of the most interesting dry runs happened in the last few days. We pretended that we'd had a little bad luck and that our trajectory near Io was a little bit off. In addition, we pretended that the accelerometer (which is supposed to measure how much our velocity has changed) was also a little bit off, and that these two errors added. In this case, we would have to figure out what commands to send to the Orbiter, and then send them. The whole process would have to be done in less than 18 hours? Could our team do it?
Why do we have only 18 hours to send up a maneuver to the spacecraft? You may think that we'd need to do it before Galileo gets too far past Jupiter, but that isn't the real problem. The real problem is that every day, the Sun gets closer to getting between us and the spacecraft. When the Sun is right between us and Galileo, we can't get commands to the Orbiter. When the angle between us, the spacecraft, and the Sun gets too small, the commands we send start to get too much noise in them. This is scary: you tell the Orbiter to do one thing and it does something else...something unexpected. We've decided that once the Earth-spacecraft-Sun angle gets to be less than 7 degrees, we won't send any more commands until Galileo gets out from behind the Sun (back to an angle of at least 7 degrees). We'll reach this 7 degree angle on December 9, just 2 days after our big Orbit Insertion burn, and we don't get back out until December 28. So we have only a short time to figure out how bad our burn was, design the commands to clean up our mistakes, and send them. If we can't get the work done in time, we'll need to wait another three weeks, and by then it may be too late to get back to the Tour we've planned.
In the first part of our exercise, the Navigation Team takes practically all of its allotted 5 hours to figure out what maneuver we'll need to perform (on the basis of rather limited information). This is a very tough test for them, and they do a great job. I'm sure they must be thinking that the rest of the exercise will be easier than their part was. We merely have to do what they tell us to.
The Navigation Team is aware of the kinds of problems the Orbit Engineering Team may have in creating the appropriate sequence. They know that we have to decide how to divide up the pulses of the little 10 Newton thrusters, make sure that the spacecraft is pointing the right way, doesn't overheat or get too cold, and so forth. They have taken a little time to make sure that we can fit all these engine pulses into the roughly 17 hours we have to perform the maneuver. They also know that the commands will take up space in the spacecraft computer, and they want to be sure that we'll have enough space to fit everything in.
I'm the coordinator for the Orbiter Engineering Team. We get the proposed maneuver from the Navigation Team and have 5 hours to design our burn sequence. Unlike the Navigation Team, we're not completely prepared. We haven't decided on a strategy for fitting a big maneuver like this into a small box. If we put too many pulses into each segment of the maneuver, before we readjust the spacecraft spin and pointing, it might endanger the whole maneuver. Of course, this is just a test ... we won't send it to the spacecraft ...so we could try something a little risky now, and change our minds for the real event.
We decide to try a very conservative maneuver for this test, leaving out none of what would be our normally scheduled spin and pointing corrections. I like this: if we had to send such a maneuver to the spacecraft, it would be less risky. The downside is that rather than give the Navigation team all (or maybe 99%) of what it wants, we're giving them about 60%. But in December, our maneuver would save the Tour, while doing nothing would lose about a third of it.
Now, can we get the work done in 5 hours?
No, we can't. Someone else is using one of our computer files, and it takes 2 hours to access it. There's some confusion about what spin state we belong in, and our first try at a sequence overruns the computer data space we've been given. When the dust clears, we're more than 3 hours late! But the test isn't over. There's over 4 hours of pad in the schedule before the maneuver has to be sent to the spacecraft. If we miss our first "uplink window", there's a backup window which ends just minutes before the start of our maneuver.
The Sequence Team has its turn next. They aren't ready either. They have to merge our sequence with the proper background sequence. But right now, it's October, not December, and the final touches have not been applied to their background sequence. They don't have the right one. This costs them two hours. But they aren't late: they were allotted 5 hours and get done 30 minutes early!
Now, the Orbiter Engineering Team gets 4 hours to review the Sequence product. All of us have previously prepared checklists of questions to ask about the sequence, and we inspect the product to make sure it's all right. This time we make our deadline easily. Most of us are ready in 2 hours. Our maneuver is presented to the Project (which already saw it after Navigation proposed it and after we designed it). Since we're not really sending this to the spacecraft (not now, in October, at any rate), Project does not approve it. However, had this been December, with the Sequence Team having the right background sequence, this maneuver might well have been on time and approved.
Tomorrow, we get to tell the Galileo Project management what we learned from our test. We've learned plenty about what we'll need to do to get our job done in time. Much of the work will need to be done well in advance, or we won't have a chance. I'm going to argue that we put a couple of sample maneuvers on our shelves, and make only the minimum changes between one of them and the real one when December comes around.