Printable Version of Topic
Click here to view this topic in its original format
Unmanned Spaceflight.com _ Exploration Strategy _ Fuel Level Measurement On Spacecraft
Posted by: Doug M. Sep 7 2012, 07:44 AM
We got answers!
Here's what the good people at Mars Outreach have to say.
2) 1.9 kg is obligated for a follow-on OTM in January 2014. Odyssey utilizes approximately 1 kg / year and historically has budgeted 1 kg / year for unexpected safing events. So, propellant life should be between 5.2 to 9.2 years. These numbers are based on no onboard anomalies that would require propellant above the safing budget, no changes in Odyssey's basic operations (three operating reaction wheels, high gain antenna and solar array servos (motors) continue to function normally, etc.) and no further programmatic maneuvers. Consideration is being given to performing a procedure in the future that has the potential to refine the +3 / -5 uncertainty of fuel remaining.
When prop is exhausted, the mission ends because the spacecraft is no longer able to desaturate the reaction wheels and / or provide thruster control; the spacecraft can no longer be pointed to maintain communications and solar panel illumination.
3) The statement is basically correct: THEMIS, which is comprised of two imagers, visible and infrared, and two instruments of the Gamma Ray Suite, High Energy Neutron Detector and the Neutron Spectrometer, all continue to operate and gather science data. The Gamma Ray Spectrometer of the Gamma Ray Suite and Marie have both been decommissioned.
Thanks again for your questions and your interest in Mars exploration!
Wow! Okay, so what comes to mind from this?
Well, the first thing is that there's an awfully large uncertainty in the amount of fuel left. 17.3 kg +3 / -5 ? The range of uncertainty is almost half the total amount of fuel. I wonder why; and I also wonder what the "procedure" might be to narrow the range.
Second is, they have a fair amount of propellant left -- enough for at least another five years of operation. So, barring mechanical failure, Odyssey should see the arrival of MAVEN (2014) and quite possibly of Mars Insight (September 2016 if they make their launch window) and ExoMars Part One (ditto).
Doug M.
Posted by: mcaplinger Sep 7 2012, 01:24 PM
The amount of remaining propellant is only known by keeping track of the measured delta-V of all maneuvers and assuming a relationship between delta-V and used propellant mass, but that relationship is variable and imperfectly known.
Most ways of narrowing this down would involve using more fuel, which is clearly not a great idea.
A google search for "spacecraft fuel remaining estimate" turned up an interesting paper about various methods of determining remaining propellant, but most require special hardware that Odyssey doesn't have. Bookkeeping is usually good enough.
http://yspm.net/pdf/11_AnikE.pdf
Also, the usage rate is quite variable depending on torques, atmospheric density, solar activity, etc.
Posted by: djellison Sep 7 2012, 01:53 PM
The challenges on keeping track on the exact amount of fuel remaining was one of the reasons why the burned Stardust to depletion when they switched it off...to find out exactly how much was left compared to their various means of measuring it. It's not easy.
Posted by: mcaplinger Sep 7 2012, 02:10 PM
Re Stardust: good reference on that in http://www.spaceops2012.org/proceedings/documents/id1295605-Paper-001.pdf
Posted by: djellison Sep 7 2012, 02:23 PM
Oh - glad something got published on that - thanks for the ref Mike.
Posted by: Greenish Sep 7 2012, 02:56 PM
Also note that the initial propellant load was 348.7 kg (source: http://mars.jpl.nasa.gov/odyssey/mission/spacecraft/) so the uncertainty is only about 2% of that, even though it's a large fraction of what remains.
Posted by: Tom Tamlyn Sep 7 2012, 03:30 PM
What Mike said. The accuracy of the bookkeeping method decreases over time, because of the accumulation of errors due to the uncertainty of the calculations. According to the http://www.jpl.nasa.gov/news/press_kits/odysseyarrival.pdf, Odyssey launched with 353 kilograms of fuel. After using up 95% of its fuel over the course of a decade of maneuvers, a substantial percentage uncertainly as to the last drops is what you would expect.*
____________
* The Apollo mission struggled unsuccessfully to develop an accurate fuel gauge for the command module based on analysis of the attenuation of radiation sources transmitted through the fuel. Eventually the program manager, Joe Shea, killed the gauge project, saying that "'[w]e don't need anything that complicated.' Shea had a Karmann Ghia which didn't even need a fuel gauge, because it had a little reserve fuel tank that he could switch to. And that's what they ended up doing for the [thruster system] fuel tanks. They had a big tank that they used during the mission, and a little tank, not to be touched unless the big tank ran out, but that could get the crew home safely. No more nuclear gas gauge." Cox & Murray, Apollo: the Race to the Moon (1989) at 175-76. This is the finest book written for lay readers about spacecraft engineering and spaceflight operations, IMO of course.
Edit: Sorry Greenish, your reply wasn't there when I started working on mine.
Posted by: nprev Sep 7 2012, 03:54 PM
Pure speculation here...but I wonder if a good measurement of MODY's fuel could be made during this OTM by measuring the altitude gain from the burn (which will be of known duration). The delta between expected & actual might indicate the spacecraft's current total mass, subtracting that which was expended for the burn.
I can't think of a direct technique to measure fuel levels in free fall/variable microgravity at all. You could do so during periods of acceleration in a variety of ways, though. Least intrusive instrumentation-based method would be an acoustic generator/receiver pair: basically, shake the tank at a predetermined high freq, measure the amplitude of the response; the more fuel mass remaining, the lower the amplitude.
Don't think you'd want to use aircraft-style capacitive fuel probes. They'd work, but they've never been the most reliable technology...plus, you gotta punch holes in the tank to mount them.
Posted by: Paolo Sep 7 2012, 04:07 PM
you could spin up the probe and measure its moment of inertia. that of course would require lots of fuel...
Posted by: djellison Sep 7 2012, 04:21 PM
The problem with the thrust/acceleration problem is that you're not exactly sure what the performance of the thrusters is any more. It's uncertainty is probably more than our fuel uncertainty. F=MA and all that...but while IMU's will give you A, F has error bars that then induce errors in M.
The thermal inertia idea is one that's been put out there as an option for Cassini as well.
Posted by: nprev Sep 7 2012, 04:30 PM
Ah. Did not realize that thruster performance predictability degraded that much over time; thanks, Doug.
That's it, then; the models are what we got. Can't make useful direct measurements with more than one variable uncontrained beyond the error limits of the models.
Posted by: RoverDriver Sep 7 2012, 05:12 PM
The thermal inertia idea is one that's been put out there as an option for Cassini as well.
I wonder if precise measurements of angular acceleration from reaction wheels could make the measurements more accurate.
Paolo
Posted by: nprev Sep 7 2012, 05:37 PM
Could be, Paolo.
I'm forced to question whether ANY onboard solutions would provide significant improvement in the margin of error, though. The more indirect the measurement, the more cumulative error builds up. Gotta look at off-platform means.
The best independent method would be precision real-time monitoring of position change during & after maneuvers with later data comparison with models as is done with Earth-orbiting spacecraft (GPS augmentation of same in particular).
Certainly not an option at Mars, nor for any solitary spacecraft in deep space, and there's still gonna be more residual error than you'd want to accept for your car's fuel gauge.
Very difficult problem.
Posted by: Doug M. Sep 7 2012, 06:05 PM
If you had multiple (>3) orbiters, could you use them to make precise measurements of each others' positions?
Doug M.
Posted by: RoverDriver Sep 7 2012, 06:13 PM
I'm forced to question whether ANY onboard solutions would provide significant improvement in the margin of error, though. The more indirect the measurement, the more cumulative error builds up. Gotta look at off-platform means.
The best independent method would be precision real-time monitoring of position change during & after maneuvers with later data comparison with models as is done with Earth-orbiting spacecraft (GPS augmentation of same in particular).
Certainly not an option at Mars, nor for any solitary spacecraft in deep space, and there's still gonna be more residual error than you'd want to accept for your car's fuel gauge.
Very difficult problem.
Hmm, what about monitoring Doppler shift at one of the DSN stations while using a reaction wheel? You don't get absolute position but the relative position and speed could be measured quite accurately.
Paolo
Posted by: RoverDriver Sep 7 2012, 06:17 PM
Doug M.
You would need to synchronize OD, MRO, MEX and the rovers to make such a measurement... like at an EDL event? I wonder if the telemetry gathered during MSL EDL could be used for that purpose. My knowledge on orbital mechanics is pretty limited, so I have no idea whether this is possible or not.
Paolo
Posted by: nprev Sep 7 2012, 06:30 PM
New topic for this old problem.
Posted by: stevesliva Sep 7 2012, 08:06 PM
Really interesting paper. Talks about more than the depletion burn and fuel estimates, but the relevant points:
The decommissioning burn showed that the PGS methodology had provided the most accurate estimate of remaining fuel for this particular situation. There had been many discussions within the operations and science teams about the conservatism of these results and the impact it was having on the mission planning. However, had
the project not adjusted its mission plan to account for this worst case prediction, the spacecraft would not have had enough fuel to complete its mission, and would not have captured the images the comet shown below.
the contribution of the thermal mass of the fuel to the transient thermal behavior. Such heating tests were implemented on the Stardust spacecraft in 2008 and 2009 the data was incorporated in the PGS model. This yielded a prediction of 10.8±0.4 kg remaining fuel, compared to the 12.9±1.0 kg predicted by the ‘book-keeping’ system at this stage of the mission
wow.
Posted by: chris Sep 10 2012, 11:44 AM
While Googling for information on the design of spacecraft fuel tanks (since I know little about them), I found this set of papers:
http://www.psi-pci.com/Technical_Papers.htm
Chris
Posted by: Tom Dahl Sep 10 2012, 04:41 PM
There is an entertaining http://www.psi-pci.com/Press_Releases/1812%20Ovature%20Burst.mpg on the psi-pci web site of tank burst tests. A couple of dozen exploding tanks are great fun to see.
Pressure Systems Inc. has been around for many decades and created thousands of aerospace pressure vessels. They produced most of the tanks on the Viking orbiter, aeroshell, and lander, for example.
Powered by Invision Power Board (http://www.invisionboard.com)
© Invision Power Services (http://www.invisionpower.com)