Mars Pathfinder Navigation
Welcome to the WWW home page for the Mars Pathfinder navigation system. This is the gateway for accessing various types of information on MPF navigation for the general public.
The main responsibility of the Navigation Team is to maintain the spacecraft on the planned trajectory for the duration of the mission. In the development phase, the team helped to design the interplanetary trajectory that achieves the mission goals within various constraints such as propellant usage and planetary protection requirements. The team provides the project with predictions of this trajectory for the spacecraft and with orbit data for the planets and Martian satellites. In flight, the team is providing best estimates for the actual past trajectory of the spacecraft along with the predictions for the future trajectory. Based on these solutions, they plan and generate the trajectory correction maneuvers (TCMs) required to maintain the spacecraft on the desired trajectory. For Mars Pathfinder, the navigation team will also provide information used to execute a successful descent through the Martian atmosphere and landing on the surface.
The primary responsibilities of the Mars Pathfinder Navigation Team are:
To reduce costs, Pathfinder NAV has kept new software development to a minimum. The majority of software is inherited from previous missions such as Galileo, Mars Observer and TOPEX-Poseidon.
Previous navigation teams were sub-divided into specialists in orbit determination, trajectory analysis and maneuver design, requiring a NAV team of 6 people or more. Pathfinder has a NAV team of 3 people, each one cross-trained in these three disciplines. The current members of the NAV team are Pieter Kallemeyn, David Spencer, and Robin Vaughan. This team will be aided by Bobby Braun from NASA's Langley Research Center during the critical entry, descent, and landing preparations.
Pathfinder is using traditional radiometric data types (Doppler and range) with an enhanced data filtering technique. This is done with the MIRAGE-ODP, a program set developed by the Navigation Systems Section of JPL.
The flight path during Entry, Descent and Landing is being modeled using the Atmosphere Entry Program (AEP), which simulates the atmosphere and gravity of Mars with a model of the heatshield, parachute, and other components of the Pathfinder flight system.
The maneuver process for past missions was typically 7-15 days from the receipt of DSN tracking data to execution of the maneuver. By integrating the navigation software with maneuver command software, Pathfinder has reduced this turnaround time to 5 days.
Automated procedures are being used, where possible, to reduce and edit the tracking data, relieving NAV analysts of these time-consuming tasks.
The navigation team uses some clever techniques from estimation and filtering theory to determine the spacecraft's trajectory. A model of the forces acting on the spacecraft, such as the gravitational attraction of the planets and the Sun, is constructed from basic physical principles. Tracking data are obtained from the spacecraft in flight and compared to predicted data from the model. Orbit determination is the process of "tuning" the filter and model parameters to obtain the "best fit" of the tracking data. The trajectory that best fits the data is the best estimate of the actual trajectory of the spacecraft.
Two basic data types are obtained to locate the spacecraft precisely: doppler and ranging. Each data type provides a different kind of information; when used in concert, the accuracy of spacecraft position and velocity relative to celestial bodies can be very high.
Doppler is a way to measure the speed at which an object is approaching or receding from the Earth. In simple terms, a Deep Space Network antenna sends a radio signal up to the spacecraft which is then directly returned. If the spacecraft is approaching or receding from the tracking station, the signal is returned a tiny bit faster or slower, respectively. If you've noticed how a car's beep, or the sound of an airplane engine sounds lower after it passes you by, you understand how doppler works. Measuring this difference in frequency can help pin down the spacecraft's speed in the solar system, and therefore give navigators clues as to precisely where it is headed.
Ranging uses the fact that light has a finite speed to determine the distance from the Earth to the spacecraft. Signals sent to the spacecraft are received and quickly returned, and the delay between when the signal is sent from the Earth and when the same signal is received back on Earth is proportional to the distance from the spacecraft to the Earth. Ranging is similar to (but much more precise than) mailing a letter to yourself to see how long the postal service takes for delivery. When used together with Doppler, the spacecraft's position and speed can be determined very accurately.
Like all other JPL interplanetary missions, Pathfinder uses antennas at the 3 Deep Space Network complexes at Golstone, CA, Canberra, Australia, and Madrid, Spain to obtain the Doppler and ranging data, as well as spacecraft telemetry. Most of Pathfinder's tracking will be done with the 34-m antennas at each site (34 m is the diameter of the antenna dish). Occasionally, these will be supplemented by the 70-m antennas, the largest of the DSN antennas. For most of the mission, the spacecraft will only be tracked 3 times each week, typically once by each complex. Additional coverage occurs during periods of critical activities. A summary of Pathfinder's tracking schedule is shown below:
Launch to Launch + 30 days 3 passes per day (continuous coverage) Launch + 30 days to Mars - 45 days 3 passes per week 3 days before to 3 days after each TCM 1 pass per day Mars - 45 days to Mars arrival 3 passes per day (continuous coverage)
Despite our best efforts, the spacecraft will not follow its planned course exactly. Small deviations in its flight path from the desired one can grow into large errors at Mars arrival. Also, constraints imposed on the mission prevent us from following the desired path directly from the beginning of the mission. For these reasons, the spacecraft will occasionally be commanded to fire its thrusters to change its velocity at certain points during the cruise to Mars. These thruster firings, or burns, are called trajectory correction maneuvers or TCMs. The velocity changes caused by the thruster firings will alter the spacecraft's future trajectory so that it returns to the desired path and arrives at Mars with the proper geometry for atmospheric entry.
A total of 4 TCMs are planned for Mars Pathfinder. The first 2 of these are scheduled in the first 2 months of the mission while the spacecraft is still relatively close to Earth. The final 2 TCMs are scheduled near the end of the cruise phase when the spacecraft is close to Mars. Contingency plans allow for a fifth maneuver to be executed just a few hours before atmospheric entry, if necessary. The table below gives a summary of the MPF maneuver schedule. For a more technical discussion see the text under TCMs 3 and 4, and TCM 5.
| Maneuver | Time | Calendar Date | Mean Velocity Magnitude | Comments |
| TCM 1 | Launch + 37 days | January 10, 1997 (delayed from January 4, 1997) |
33.3 m/sec | Remove injection bias, correct injection errors |
| TCM 2 | Launch + 60 days | February 4, 1997 | 2.08 m/sec | Correct TCM 1 errors |
| TCM 3 | Mars - 60 days | May 7, 1997 | 0.432 m/sec | Target to final Mars atmospheric entry point |
| TCM 4 | Mars - 10 days | June 24, 1997 | 0.138 m/sec | Correct TCM 3 errors |
| TCM 5 | Mars - 12 or 6 hours | July 4, 1997 | 0.2 -> 2.0 m/sec | Correct any remaining errors |
Mars Pathfinder must satisfy 2 NASA planetary protection requirements designed to minimize potential contamination of the Martian environment. The first requirement is that the probability of the unsterilized launch vehicle upper stage impacting the surface of Mars be less than 0.0001 (or 1/10,000). The launch vehicle stage could be carried to Mars with Pathfinder if there were some failure in the separation procedure after the final burn that places the spacecraft on target to Mars. For this reason, the targetted state to be achieved by the launch vehicle, called the injection target, is biased away from Mars - just enough so that the impact requirement is satisfied. This means that the launch vehicle will not place Pathfinder on course for its final desired arrival state at Mars. It's up to the spacecraft itself to move toward this final aimpoint during its 7-month cruise. TCM 1 is the first step towards achieving the desired entry conditions at Mars.
The second planetary protection requirement imposed on Mars Pathfinder is that the probability of Pathfinder itself impacting Mars at a speed greater than 1,000 ft/sec be less than 0.001 (or 1/1,000). This requirement is met by designing TCMs 1 and 2 to a target state that is also biased away from the final desired arrival conditions - again, just enough to meet the requirement. The target state is chosen so that if control of the spacecraft is lost following either TCM 1 or 2, the spacecraft will enter the Martian atmosphere with a shallow entry angle, allowing the atmosphere to slow it to below 1,000 ft/sec at impact (assuming the parachute does not deploy).
TCM 1 is designed to move from the biased launch injection target state to the (still) biased Mars arrival state. The change required to move between these two targets is known, so that the nominal size of TCM 1 can be computed exactly. This type of maneuver is called "deterministic". However, there are several types of uncertainties associated with the design and implementation of maneuvers in flight. A second maneuver - TCM 2 - is planned to follow TCM 1 to correct any errors that occur in the design and/or execution of TCM 1. If we knew the spacecraft trajectory exactly and the propulsion system could execute the maneuver perfectly, there would be no need for TCM 2. This type of maneuver is called "statistical" since its characteristics can only be predicted by statistical analyses of the various error sources.
TCM 1 was originally scheduled for January 4, 1997, but was postponed so that some changes in the attitude control software could be implemented. These changes were made on the morning of Wednesday January 8, 1997. At that time, the spacecraft also performed a turn to the required attitude for TCM 1 execution. TCM 1 had been rescheduled for January 10, 1997 02:00 UTC, or 6 PM PST on Thursday January 9. Maneuver execution was delayed slightly due to some minor problems with hardware at the Deep Space Network complex in Madrid. TCM 1 was successfully executed at 03:40 UTC, or 07:40 PM PST. The thrusters were fired for about 1 and 1/2 hours, as predicted, to attain the necessary velocity change. Following the burn at 06:18 UTC (or 10:18 PM PST), another spacecraft turn was performed to point the spin axis at the Earth.
The navigation team has produced an orbit determination solution for the spacecraft's trajectory using data from launch through January 1, 1997. A total of 26,605 Doppler and 6952 ranging measurements were fit for this solution. Based on this trajecory, the team has calculated the velocity change required for this maneuver. The velocity change is expressed as a "delta-V vector", having both a magnitude and direction. This vector is shown below in the Earth Mean Equator and Equinox of J2000 coordinate frame:
delta-V vector for TCM 1: 17.300657 m/sec
-25.321327 m/sec
5.899257 m/sec
delta-V magnitude: 31.230 m/sec or 69.85952 mph
delta-V direction (unit vector): 0.55397
-0.81080
0.18890
The NAV team has been processing tracking data obtained since the execution of TCM 1. Our current best estimate for the magnitude of TCM 1 is 30.091 m/sec or 3.6% lower than the design value. The estimated direction is within 0.1 degrees of the design value.
For TCM 1, the spacecraft was turned so that its spin axis was pointing along the delta-V direction and the thrusters were fired to produce a force in that direction only. This is called an "axial" burn. The spacecraft can also fire its thrusters so that a force is produced normal to the spin axis. This is called a "lateral" burn. For TCM 2, the delta-V was implemented as a vector sum of an axial and a lateral burn. The directions and magnitudes of the two burns were chosen so that the vector sum of delta-V components for each burn equalled the overall desired delta-V vector. Both segments of TCM 2 were successfully performed on February 3, 1997. The spacecraft was turned to point in the direction of the axial delta-V on Friday January 31, 1997 at 8:25 AM PST. The axial burn was performed on Monday February 3, 1997 at 23:00 UTC (3:00 PM PST). The burn lasted 297 seconds, or just under 5 minutes. The lateral burn was then performed about two hours later at 00:52 UTC on February 4, 1997, which is 4:52 PM PST on February 3. Thrusters were fired in two pulses over a 30-seconds interval. Following these burns, the spacecraft was returned to an Earth-pointing attitude. The turn back to Earth was performed on 02:08 UTC on February 4, or around 6:08 PM PST on February 3.
The navigation team had previously produced an orbit determination solution for the spacecraft's trajectory using data from launch through January 25, 1997. A total of 8596 2-way Doppler points and 10006 range points from each of the 3 DSN complexes were used in this solution. Based on this trajecory, the team had calculated the velocity change required for this maneuver. The velocity change is expressed as a "delta-V vector", having both a magnitude and direction. This vector is shown below in the Earth Mean Equator and Equinox of J2000 coordinate frame:
delta-V vector for TCM 2: 1.119292 m/sec
-1.123117 m/sec
0.035196 m/sec
delta-V magnitude: 1.58602 m/sec or 3.548 mph
delta-V direction (unit vector): 0.70573
-0.70813
0.02219
The selected attitude for TCM 2 resulted in the following axial and lateral
delta-V components:
AXIAL DELTA-V
axial delta-V vector for TCM 2: 1.16063 m/sec
-1.08413 m/sec
-0.04731 m/sec
axial delta-V magnitude: 1.5889 m/sec or 3.554 mph
axial delta-V direction (unit vector): 0.73045
-0.68231
-0.02977
LATERAL DELTA-V
lateral delta-V vector for TCM 2: -0.04134 m/sec
-0.03899 m/sec
0.08251 m/sec
lateral delta-V magnitude: 0.01002 m/sec or 0.02241 mph
lateral delta-V direction (unit vector): -0.41263
-0.38918
0.82357
The NAV team has been processing tracking data obtained since the execution of TCM 2. Our current best estimate for the magnitude of the axial segment of TCM 2 is 1.598 m/sec, 0.6% higher than the design value. For the lateral segment, the magnitude estimate is 0.1020 m/sec, or about 1.8% higher than the design value.
As explained in the preceding section, Pathfinder is not on target for its final Mars arrival conditions after execution of TCMs 1 and 2. Another maneuver is needed to remove the bias and target for the geometry necessary to successfully enter the atmosphere, descend, and land on the surface. TCM 3 is scheduled near the end of the cruise to Mars and is the first maneuver to target to the desired Mars entry conditions. The dynamics of entry impose strict constraints on the final target point. The spacecraft's position and velocity must be maintained within a narrow corridor; the bounds of this corridor represent regions where the spacecraft would either burn up in the atmosphere before landing or "skip out" of the atmosphere and return to interplanetary space. TCM 3, like TCM 1, is a deterministic maneuver since the offset between the biased target for TCMs 1 and 2 and the desired final target is known in advance. As for TCM 2, a second, statistical maneuver - TCM 4 - is planned to follow TCM 3. TCM 4 will remove any errors in design and execution of TCM 3, just as TCM 2 did for TCM 1.
We have executed our first maneuver to target the spacecraft for its descent through the atmosphere and landing on Mars. TCM 3 was performed on Wednesday May 7, 1997 00:00 - 04:00 UTC which was Tuesday May 6, 1997 at 5-9 PM PDT. All of the scheduled events executed successfully. Unfortunately, luck was not with us and the small execution errors accumulated while performing the maneuver have conspired to place the spacecraft on a trajectory that is just outside our desired entry conditions - if uncorrected. So, we will be doing TCM 4 in late June, ten days before arrival at Mars on July 4. See the following paragraphs for more details.
The navigation team produced the orbit determination solution used to design TCM 3 based on data through April 29, 1997. A total of 1806 2-way Doppler points and 3259 range points from each of the 3 DSN complexes were used in this solution. Based on this trajecory, the team calculated the velocity change required for the maneuver. The velocity change is expressed as a "delta-V vector", having both a magnitude and direction. This vector is shown below in the Earth Mean Equator and Equinox of J2000 coordinate frame:
delta-V vector for TCM 3: 0.076190 m/sec
0.063028 m/sec
0.036165 m/sec
delta-V magnitude: 0.10529 m/sec or 0.2355 mph
delta-V direction (unit vector): 0.72364
0.59863
0.34349
TCM 3 was executed somewhat differently than either TCM 1 or 2. For TCM 1, the spacecraft was turned so that its spin axis was pointing along the delta-V direction and the thrusters were fired to produce a force in that direction only. This was an axial burn. This could not be done for TCM 2 due to constraints on the spacecraft attitude. Instead, the spacecraft fired thrusters to produce forces both along and nearly normal to its spin axis. These axial and lateral delta-V components were sized so that their sum equaled the overall desired delta-V vector for TCM 2.
As was the case for TCM 2, thermal and power constraints precluded doing TCM 3 as an axial burn by turning the spacecraft directly to the delta-V direction. So, TCM 3 was implemented as a combination of axial and lateral burns. But in this case, we deliberately added an extra component to the delta-V breakdown. The reason for this is that our analysis for the late contingency maneuver - TCM 5 - has shown that, if executed, this maneuver would require a velocity change exclusively in the lateral direction. During the last 24 hours before entry when TCM 5 would be executed, the spacecraft has turned to the orientation required for safe descent through the atmosphere and cannot be turned from that orientation to do the TCM. And the geometry is such that the delta-Vs needed to target back to the desired landing site are nearly perpendicular to the entry attitude direction, so that TCM 5 will have to be performed as one or more lateral burns. In the worst case, 5 sets of lateral burns would be performed consecutively, each imparting a 0.4 m/sec velocity change. So far, the spacecraft has only performed a lateral burn once as part of TCM 2 and the velocity change for that burn was only 0.1 m/sec. The team decided to get some more data on lateral burn performance at larger delta-V magnitudes during TCM 3.
The required delta-V for TCM 3 was broken down into 3 parts:
1. a lateral burn of 0.4 m/sec that tested TCM 5 execution
2. an axial burn
3. a 2nd lateral burn in a direction nearly opposite to that
of part 1. Its magnitude was 0.4 m/sec plus a little extra
that added together with the axial burn of part 2, gave the
total delta-V for TCM 3.
The selected attitude for TCM 3 resulted in the following axial and lateral
delta-V components:
1. LATERAL DELTA-V
lateral delta-V vector for TCM 3: -0.13925 m/sec
-0.33158 m/sec
-0.17511 m/sec
lateral delta-V magnitude: 0.4 m/sec or 0.89477 mph
lateral delta-V direction (unit vector): -0.348121
-0.828953
-0.437778
2. AXIAL DELTA-V
axial delta-V vector for TCM 3: 0.101451 m/sec
-0.02820 m/sec
-0.01259 m/sec
axial delta-V magnitude: 0.1060 m/sec or 0.23711 mph
axial delta-V direction (unit vector): 0.95666
-0.26592
-0.11872
3. LATERAL DELTA-V
lateral delta-V vector for TCM 3: 0.11424 m/sec
0.42342 m/sec
0.22419 m/sec
lateral delta-V magnitude: 0.4925 m/sec or 1.1017 mph
lateral delta-V direction (unit vector): 0.231940
0.859665
0.455170
Here's an outline of the events for TCM 3 as executed on the evening of May 6, 1997:
Event PDT on May 6, 1997 --------------------------------------- -------------------- Turn to TCM 3/Earth-pointing attitude 4:45 PM 1st lateral burn (to test TCM 5) 5:31 PM Axial burn 7:00 PM 2nd lateral burn 8:02 PMThe spacecraft executed all of its commands with no problems. Since then, the NAV team has received 2 weeks of tracking passes and has produced an assessment of the maneuver performance. The first lateral burn was approximately 0.71% below the commanded magnitude, while the axial burn was about 2% above the commanded magnitude. Both of these burns appear to be very close to the commanded direction. The second lateral burn was approximately 0.96% above the commanded magnitude and it appears to be have been pointed about 0.9 deg away from the desired direction. The predicted entry flight-path angle is currently -14.18 degrees, which is within the required range. However, the predicted landing site is outside of the desired area on the surface, so we will be doing TCM 4 in June to retarget the spacecraft to the desired entry conditions. The NAV team will refine its assessment of TCM 3 and the predicted entry and landing conditions in the next few weeks and publish updates as they become available.
Although TCM 3 was not quite optimal in total, the first lateral burn executed according to expectations and successfully validated our design and strategy for TCM 5 execution (should it prove necessary).
TCM 4 StatusWe successfully executed our second maneuver to target the spacecraft for its descent through the atmosphere and landing on Mars on Wednesday June 25, 1997 around 17:00 UTC or 10 AM PDT. The burn was performed as a sequence of lateral and axial burns followed by an attitude turn to maintain the spacecraft antenna pointing at Earth. All of these activities were performed with no problems.
The navigation team has now processed just under 24 hours of tracking data taken during and after TCM 4. Our latest solution done on June 26, 1997 show TCM 4 has moved us back to our desired landing site :-) As of today, it looks like we right on target! See the picture in our Trajectory Data (Technical) web page for our current estimate of where we're landing.
The navigation team has produced the orbit determination solution for TCM 4 based on data through June 23, 1997. A total of 5776 2-way Doppler points and 5259 range points from each of the 3 DSN complexes were used in this solution. Based on this trajecory, the team calculated the velocity change required for the maneuver. The velocity change is expressed as a "delta-V vector", having both a magnitude and direction. This vector is shown below in the Earth Mean Equator and Equinox of J2000 coordinate frame:
delta-V vector for TCM 3: 0.001047 m/sec
-0.001375 m/sec
-0.006827 m/sec
delta-V magnitude: 0.01858 m/sec or 0.04156 mph
delta-V direction (unit vector): 0.56349
-0.73994
-0.36736
As you can see this was an extremely small maneuver, representing just a
small "tweak" of the spacecraft's trajectory.
As was the case for TCMs 2 & 3, thermal and power constraints precluded doing TCM 4 as an axial burn by turning the spacecraft directly to the delta-V direction. So, TCM 4 was implemented as a combination of axial and lateral burns. The lateral burn was executed first. The selected attitude for TCM 4 resulted in the following lateral and axial delta-V components:
1. LATERAL DELTA-V
lateral delta-V vector for TCM 4: 0.0003454 m/sec
-0.014525 m/sec
-0.0071456 m/sec
lateral delta-V magnitude: 0.01619 m/sec or 0.0362 mph
lateral delta-V direction (unit vector): 0.021334
-0.897101
-0.441310
2. AXIAL DELTA-V
axial delta-V vector for TCM 3: 0.01013 m/sec
0.0007737 m/sec
0.0003181 m/sec
axial delta-V magnitude: 0.01064 m/sec or 0.0238 mph
axial delta-V direction (unit vector): 0.99661
0.07611
0.03129
From our June 26, 1997 navigation solution, there is less than
1% difference between the actual deltaV and the design value for the
axial burn and less than 4% difference between the actual lateral deltaV and
the design value. This still a preliminary assessmentl; the
solutions for the actual deltaVs may change a little
over the next few days as we get more tracking data following the
maneuver execution.
After execution of TCM 4, Pathfinder should be on course for a successful landing on Mars. If the trajectory following TCM 4 moves outside of the allowable corridor for successful atmospheric entry, an emergency maneuver - TCM 5 - can be executed up to 5 hours before entry. This maneuver can also be viewed as moving the predicted landing site back towards the target site on the surface of the planet. Current plans call for two "windows" for possible execution of a fifth TCM - one at 10.3 hours and one at 5.3 hours prior to entry.
TCMs 1 - 4 are "custom designed" in the sense that the navigation team computes the exact velocity change required to achieve the desired target and the rest of the flight team designs and executes new commands to implement each maneuver. Since this process takes about 5 working days, it's not possible to do this for TCM 5. Instead, a set of velocity changes will be chosen ahead of time and commands will be developed to implement each of these. The NAV team will choose whichever one is closest to the true velocity change indicated by the latest orbit determination solution.
Mars Pathfinder Trajectory Data gives a short summary of Pathfinder's current location. More detailed information, including orbital elements, is given at Mars Pathfinder Trajectory Data (Technical).
The last 48 hours before atmospheric entry will be a busy time for the navigation team. The team will regularly generate updated solutions for the spacecraft trajectory to determine the precise entry conditions. This will be done hourly during the last 24 hours. This information will be used to
The need to execute TCM 5 will be officially evaulated first at 13 hours before entry. If the trajectory solutions indicate it's necessary, it will be performed at 10.3 hours before entry. If TCM 5 is not performed at this time, another evaluation of the trajectory solution will be made. TCM 5 will be performed at 5.3 hours before entry if the new trajectory solutions computed between 14 and 7 hours out show that it is needed.
There are four designated times during these last 24 hours when parameters for the parachute deployment and RAD rocket firing and landing site location will be recomputed and sent to the spacecraft: (roughly) 36, 22, 10, and 4 hours before entry. In addition to the orbit determination software, this requires using two different sets of software to model the trajectory once in the Mars atmosphere. The Atmospheric Entry Program developed at JPL and the POST software from NASA Langley are used to verify these calculations. While similar, each of these programs has slightly different capabilities. They are used in a complementary fashion to insure that our calculations are as accurate as possible.
Robin Vaughan (rvaughan@mpfnav2.jpl.nasa.gov or robin.vaughan@jpl.nasa.gov)
Note: If you send e-mail during the week of June 30 to July 4, 1997, I may not answer you until after landing! The NAV team will be very busy during this last week and I must give priority to flying the spacecraft. Thanks in advance for your patience during this time.