Improved star camera attitude data 5
6.2 Propellant consumption dependence on the selected primary star camerastar camera
The gaseous nitrogen (GN2) is consumed for thruster firings which are necessary to keep the satellites in the required formation and orientation. The twelve 10 mN attitude control thrusters are activated approximately 1000 times per day when control torques generated by the magnetic torquers do not suffice to maintain the desired attitude. More information about the GRACE attitude control can be found in Section 3.4. The torque needed to keep the satellite in the target attitude is computed based on the information about the instantaneous and the desired attitude. The instantaneous attitude is measured by the attitude determination sensors. In the science mode and back-up science mode, the attitude data are delivered to the AOCS by the primary star camera. Clearly, any inaccuracy in the measured attitude will directly propagate into the computed control torque which is subsequently generated by attitude actuators. When
"the bad" star camera is set as the primary camera, the instantaneous attitude is determined less accurately. This results in larger differences between the instantaneous and the desired attitude and hence bigger control torque is needed to keep the satellite in the desired attitude.
Consequently, the number of thruster activations as well as propellant consumption increase.
The dependence of the attitude control on the accuracy of the input attitude data is illustrated in Figure 6.3, where thruster events and their duration are shown for GRACE-A for the whole year 2008. The primary camera switches occurred on DOY 135 and 305. Between DOY 135 and 305, the primary camera was SCA head#2. In the other epoch, the primary camera was head#1. The dependence of the frequency of thruster firings and their duration on the selected primary camera is obvious at first sight, especially for pitch and yaw. The number of thruster firings per day increased more than 4 times for pitch and 2.3 times for yaw, when the SCA head#1 was set as the primary camera. The total duration of thruster firings per day increased 2 times for pitch and 2.8 times for yaw. As roll can be determined well by both star camera heads, the differences in number of thruster events and in the duration of thruster firings are much lower, characterized by a factor of 1.2.
In order to quantify the impact of the different performance of the two star cameras on the attitude control, the number of thruster activations, duration of thruster firings and propellant consumption, are shown for the period of 2006-2009 in Figure 6.4, according to the epochs defined by the selected primary camera.
As discussed above, the roll angle can be determined with high accuracy, hence the differences between the epochs are rather small. However, the number of thruster events is large, and more importantly, the duration of the firings is the biggest compared to pitch and yaw firings. This is because the efficiency of attitude control using magnetic torquers is extreme low along the geomagnetic equator. Although there are slightly more thruster activations in yaw compared to roll, their duration is significantly lower than for roll firings. The big differences between the epochs defined by the primary camera reflect well the different SCA measurement accuracy.
The yaw thrusters are most frequent and the yaw thrusters have already exceeded the nominal limit of 106activation cycles, cf. Table 3.2. Compared to roll and yaw, pitch thruster firings are rare (cf. also Figure 6.3) and hence they have the smallest effect on the propellant consumption.
The gas consumption depends on the absolute number of thruster firings in all directions as well as on their duration. In the epochs when the SCA head#1 was set as the primary camera, the gas consumption on GRACE-A is approximately 1.3 times bigger than when the SCA head#2 was the primary camera. Evidently, the impact on GRACE-B operation is even bigger than for GRACE-A. The propellant consumption in those periods, when SCA head#1 was the primary camera, is almost double. This is caused by the fact that the performance of GRACE-B SCA head#1 is the worst of all four GRACE star cameras (cf. Figure 3.13).
The selected period of 2006-2009 is special for two reasons. The first one is that the mission was operating during solar cycle minimum. The solar activity influence the overall satellite’s performance. Among others, it affects the air drag and the disturbing atmospheric torques, which are low during the solar cycle minimum. The other reason is that the onboard laboratory was nominally operating. This is no more the case in the recent years when the SCA measurement accuracy of some SCA heads has substantially deteriorated. Also the satellite was more thermally stable than in the later years, because the thermal heating was switched off in 2011. Therefore the results obtained from the chosen sample period reflect well the impact of the SCA performance on the factors limiting the mission lifetime.
The differences in propellant consumption vary during the mission operation period, cf.
Figure 6.5 which covers the epoch 2002-2014. For GRACE-A, the differences between the observed epochs decreased, but the overall GN2 consumption became almost double compared to early years. Further, since 2012, "the bad" camera is no longer SCA head#1, but SCA head#2. On GRACE-B, the propellant consumption keep growing in the periods when the SCA head#1 is set as the primary camera. This is caused by the degradation of the SCA measurement accuracy. When SCA head#2 is the primary camera, the gas consumption is kept nominal, i.e. between 2-4 g/day. The overall increasing trend in the gas consumption depends on multiple factors. One of them is the decreasing SCA measurement accuracy. The other factors are increased frequency of special attitude maneuvers, i.e. yaw maneuver for battery discharge, and higher disturbance torques due to increased solar activity and lower satellite’s orbit.
(a) roll (b) pitch
(c) yaw
Figure 6.3:Illustration of the dependence of the thruster activity on the selected primary star camera. Thruster firing events are plotted along the orbit for roll, pitch and yaw together their duration which is coded by color.
The positive and negative values indicate the sense of rotation. Based on THR1B data from GRACE-A for the whole year 2008. The primary camera switches occured on DOY 135 and 305. Between DOY 135 and 305 the
SCA head#2 was set as the primary camera
20060 2007 2008 2009 2010 500
roll
[nr/day]
20060 2007 2008 2009 2010
50 100
pitch
[nr/day]
20060 2007 2008 2009 2010
500
yaw
[nr/day]
(a) number of thruster activations
20060 2007 2008 2009 2010
500
roll
[nr/day]
20060 2007 2008 2009 2010
50 100
pitch
[nr/day]
20060 2007 2008 2009 2010
500
yaw
[nr/day]
(b) number of thruster activations
20060 2007 2008 2009 2010
50 100
roll
[s/day]
20060 2007 2008 2009 2010
5 10 15
pitch
[s/day]
20060 2007 2008 2009 2010
50
yaw
[s/day]
(c) duration of thruster firings
20060 2007 2008 2009 2010
50 100
roll
[s/day]
20060 2007 2008 2009 2010
5 10 15
pitch
[s/day]
20060 2007 2008 2009 2010
50
yaw
[s/day]
(d) duration of thruster firings
20060 2007 2008 2009 2010
1 2 3 4 5 6
[g/day]
SCA head#1 SCA head#2
(e) propellant consumption
20060 2007 2008 2009 2010
1 2 3 4 5 6
[g/day]
SCA head#1 SCA head#2
(f) propellant consumption
Figure 6.4: Dependence of thruster activity and propellant consumption on the primary star camera in 2006-2009 for GRACE-A (left column) and GRACE-B (right column). Based on THR1B and MAS1B data. All
figures show the given parameter averaged per day
20020 2004 2006 2008 2010 2012 2014 2
4 6 8 10 12
[g/day]
SCA head#1 SCA head#2
(a)
20020 2004 2006 2008 2010 2012 2014 2
4 6 8 10 12
[g/day]
SCA head#1 SCA head#2
(b)
Figure 6.5: Dependence of the propellant consumption on the selected primary camera in 2002-2014 for GRACE-A (a) and GRACE-B (b). The figures show the averaged GN2 consumption per day. The satellite
swap maneuver in December 2005 is taken into account