Atmospheric Structure Experiment (ASE)
The PHX entry capsule contained two Inertial Measurement Units (IMUs) mounted on the underside of the lander on opposite sides of the spacecraft. Only one IMU was used during descent, entry and landing (EDL). The other IMU was a backup device that was not turned on. The Phoenix IMU is quasi-standard commercial product, sometimes called a Miniature IMU, or MIMU [GARAVELLIETAL1995]. This particular device was manufactured by Honeywell (Clearwater, FL) with model number YG9666BC. The IMU used three Allied Signal QFLEX accelerometers for linear acceleration measurement and three Honeywell GG1320 RLG ring-laser gyros for angular rotation measurement. However, the IMU was configured internally so that the linear accelerations were integrated and output as linear velocity changes while the gyro measurements were output as angle changes in a given sample time. No input from scientists was possible concerning the device selection and location within the lander, which was done by spacecraft engineers for the canceled NASA Mars 2001 Lander prior to the initiation of the Phoenix project. The Phoenix lander was based on the Mars 2001 Lander design. Consequently, the IMU used during EDL of the Phoenix spacecraft was neither located at the entry vehicle center of mass nor the spin axis, which otherwise would have been desirable to optimize noise minimization for scientific purposes.
The instrument attributes for the PHX IMU are discussed in [TAYLORETAL2008]. At a rate of 200 Hz, the IMU output the accumulated linear velocity change (the time integral of acceleration) in three Cartesian axes and accumulated angle change (the time integral of angular rate) about three Cartesian axes. Data was saved onboard PHX at the device output rate of 200 Hz and was not down-sampled.
The IMU had a mass of 3.52 kg and approximately 4051 cc volume. Each accelerometer within the IMU had a full-scale range of 36.8g (where g represents Earth's standard surface gravitational acceleration of 9.80665 meters per second per second). The accelerometer and gyro signals were electronically integrated inside the IMU to produce an output of accumulated delta-v (linear velocity change) and delta-q (angular velocity change) at 200 Hz. The accelerometers had two stages of digitization that resulted in pulses and then counts. The final raw digital output from the IMU accelerometers was in counts, in velocity units, where 1 output count = 0.0753 mm/sec. But the final output moved up and down in pulses, which are leaps of many counts, with each pulse equivalent to 2.7 mm/sec. Thus, the amplitude of one pulse was equivalent to 2.7/0.0753 = 35.86 counts, on average. Added to this was noise in the system, which was on the order of 1 count, so that jumps in the raw output were typically 36 or 37 counts. This determined the digital resolution of the raw 200 Hz data to be a delta-v of 2.7 mm/sec and typical noise level of 0.0753 mm/sec.
Individual gyros measure the angular rotation about one axis. Each raw IMU gyro output is a 16 bit signed integer (i.e., from -32767 to +32767) with the least significant bit (LSB) equivalent to 1 micro-radian. Because the data was generated at 200 Hz or every 5 millisecs, a full-scale value of 32768 would correspond to 375 deg/sec. (32,768 * 1 microradian /0.005 sec = 0.032768/0.005 radian/sec = 6.5536 rad/sec = 375 deg/sec). The smallest motion in a 200 Hz interval is one count, which corresponds to 0.01144 deg/sec (1 * 1 microradian /0.005 sec = 200 micro-rad/sec = 0.01144 deg/sec). Smaller angular motions can be measured over longer time periods by integration of the raw data. Noise on the raw gyro output is estimated as 45 micro-radians (3 sigma).
Entry State and Timing
The frictional drag of the atmosphere upon the entry vehicle resulted in a reduction in the speed of the entry vehicle. This deceleration was measured by the accelerometers within the IMU, while the orientation of the entry vehicle was provided by the gyroscope measurements. Some ancillary parameters are necessary to use the IMU data scientifically. These include the wet mass of the Phoenix probe at atmospheric entry, which was estimated as 572.743 kg. (Further mass properties during descent to landing will be documented in the DOCUMENT directory of this archive). Another parameter is the entry probe cross-sectional area of the entry probe, which was 5.515 m**2, based on an aeroshell diameter of 2.65 m. To integrate the IMU data to derive atmospheric structure, requires an entry state boundary condition. Entry state information was derived by the Lockheed spacecraft team from navigation communications using two-way Doppler, ranging (the light time of signal transmit and receive to determine distance), and delta differential one-way ranging. The latter technique uses a fixed natural radio source such as a quasar and two stations on the Earth to allow difference ranging, which gives an accurate angular separation of the spacecraft from the fixed celestial source. From the angular separation the position of the spacecraft can be determined. All this information was then propagated to the time and point of entry defined as 3522.2 km radius from the center of Mars. The best estimate of the Phoenix probe entry state was as follows: An entry position of 69.3660 deg N (1 sigma 0.0040 deg), 197.7160 deg E (1 sigma 0.00031 deg) at a universal time on 2008-05-25 of 23:30:57.7330 (1 sigma 0.002 sec) with respect to a planetocentric Cartesian frame. The equivalent spacecraft clock (SCLK) time was 896225523.703. Also the best estimate of the entry velocity was a speed of 5600.34 m/s (1 sigma 0.02 m/s) with a flight path angle below horizontal of -13.010 deg (1 sigma 0.00015 deg) and flight path azimuth, clockwise from north, of 77.6720 deg (1 sigma 0.00055 deg) in the same planetocentric frame. The best estimate of the time of the first IMU data acquisition (with an uncertainty of 0.02 seconds) was a universal time on 2008-05-25 of 23:30:47.913, equivalent to a spacecraft clock (SCLK) time of 896225513.881. However, the first EDR data point for delta-velocity and delta-angle was derived from a difference between the first two IMU readings, which gave a time stamp 0.005 seconds later on 2008-05-25 at 23:30:47.918. Thus, the EDR data points began 9.82 seconds before the nominal atmospheric entry state position as defined above.
Platform Mounting Description
There were two IMUs mounted on the underside of the Phoenix Lander deck, on opposite sides, denoted by Side A (or 1) and Side B (or 2). The location of the IMUs under the lander deck is described in the EDR SIS. However, only the IMU on side A was used in collecting data. This IMU is referred to as IMU-1 or IMU-A in PHX project documentation.
The IMU for the PHX mission was not selected for characteristics needed for optimal atmospheric structure reconstruction but was inherited from NASA's canceled Mars 2001 lander mission. The IMU system was part of the engineering instrumentation and was selected, configured, installed, and operated by members of the Phoenix lander engineering team from Lockheed Martin. The IMU provided entry vehicle deceleration and orientation information that was used for triggering EDL events on the entry vehicle such as parachute deployment. However, the Atmospheric Science Theme Group (ASTG) of the Phoenix Science Team decided that the IMU data should be used for atmospheric structure reconstruction. David C. Catling from the ASTG became the cognizant scientist that led the effort to retrieve and archive the PHX IMU data to enable atmospheric reconstruction.
The IMU is an engineering device from which science can be derived. The measured delta-velocity values provided by the IMU, as a function of time (which needs to be used to determine the height above the surface) can be employed to deduce the vertical structure of density and pressure, and ultimately temperature along the atmospheric trajectory traversed by the Phoenix probe as it descended through the atmosphere.
EDL is a mission critical phase (if EDL fails, the mission is lost) and since the IMU was an active sensor of the EDL process, its operational characteristics as described above were wholly driven by EDL engineering considerations. These choices constrain the scientific return of the instrument.
The IMUs were delivered with a factory calibration based upon individual unit testing after assembly. Outputs of velocity change from accelerometers and angle change from gyros are internally compensated for biases, scale factors and alignments based on coefficients determined by factory calibration tests. From cruise data, the Lockheed Martin spacecraft team estimated that there was a zero offset on the IMU output expressed in the [X, Y, Z] Cartesian axes of the cruise body frame of reference of the spacecraft of [-3.47845e-4, 2.42877e-4, 4.38682e-4] in units of m/s/s.
From the time of IMU turn on through the EDL process of atmospheric entry to landing, IMU data were provided at 200 Hz and saved at this rate by the on-board software. There were no gain changes or offset changes throughout this time period.