(Orbital Sciences Corporation)

TSX-5: Another Step Forward for Space-Based Research

Michael L. La Grassa and James R. Farmin

Following on the heels of the failed STEP-4 satellite launch, the TSX-5 research satellite was successfully placed into orbit on June 7, 2000. Aerospace provided timely contributions to the overall mission success, including verification of solar-array deployment, validation of critical components, mitigation of potential failure modes, thermal modeling, contamination analysis, anomaly resolution, and more.

The Tri-Service Experiment 5 (TSX-5) is the sixth in a series of small-satellite missions for space research and experiments commissioned by the U.S. Air Force Space and Missile Systems Center as part of its Space Test Program (see sidebar). The mission generated significant interest for two important reasons. First, the vehicle was based on the Space Test Experiment Platform (STEP) spacecraft bus, whose final mission, STEP-4, failed in October of 1997. Second, the satellite was commissioned under acquisition reform—a streamlined cost-conscious procurement approach. Another failure would deal a serious setback to this program. Therefore, to help maximize the probability of a successful launch, the Air Force asked The Aerospace Corporation to validate many of the components, materials, and design concepts used for the space vehicle and its experiments. The investment clearly paid off, and TSX-5 was successfully launched into orbit on June 7, 2000.

Research Mission

TSX-5 hosts two Department of Defense (DOD) payloads—STRV-2 (the Space Test Research Vehicle-2), sponsored by the Ballistic Missile Defense Organization, and CEASE (the Compact Environmental Anomaly Sensor), sponsored by the Air Force Research Laboratory (see sidebar, TSX-5: Spacecraft and Payloads).

Engineers prepare the TSX-5 satellite for launch. (Orbital Sciences Corporation)

STRV-2 is a bundle of seven subexperiments, each with distinct scientific objectives: The first, LaserCom, attempted to achieve high-data-rate laser communications between a satellite and other platforms and establish the practicality of laser communications in free space (it has not achieved this goal because of hardware problems); MWIR (the Medium-Waveband Infrared imager) is designed to detect and classify military aircraft flying beneath the satellite;

RadMon, a radiation monitor, is designed to characterize the radiation environment (part of the MWIR experiment); SAMMES (the Space Active Modular Materials Experiment System) is providing long-term time-variant performance data on materials exposed on the spacecraft shell while quantifying the effects of contamination deposition; ETB (the Electronics Test Bed), which includes the meteoroid and debris impact monitor and the meteoroid impact sensor, was deployed to characterize the space environment at altitudes relevant to the proposed Brilliant Eyes surveillance satellite; ACESS (the All-Composite Experiment Spacecraft Structure) is an experiment to assess the performance of composite spacecraft relative to conventional designs; VISS (the Vibration Isolation, Suppression, and Steering system) is designed to provide an ultraquiet environment for sensitive optical sensors and transmitters (including the MWIR experiment).

CEASE is a miniaturized suite of sensors and particle detectors for monitoring the environment surrounding a spacecraft. The instrument is designed to give an indication of the probability and severity of faults caused by surface charging, deep dielectric charging, single-event upsets, and radiation-dose effects. TSX-5 marks the first mission for CEASE and primarily serves as a proof-of-concept flight.

Acquisition

TSX-5 was developed under acquisition reform with limited documentation and minimal contract oversight. The contract specified a firm fixed-price-plus-incentive fee of $25 million for the spacecraft bus and the commercial launch vehicle. Engineers were therefore challenged to ensure that the space-vehicle design, integration, and testing would meet requirements and be able to fulfill the mission objectives at a minimal cost. Aerospace played a crucial role in bringing the system to fruition, providing critical development and operational expertise. Total Aerospace involvement, including program office and engineering support, was approximately 12 staff-years.

TSX-5 was launched into orbit from a Pegasus XL rocket. It is shown here after integration with the launch vehicle. The folded solar array forms a six-sided shell around the satellite. (Orbital Sciences Corporation)

One of the primary Aerospace tasks was to ensure that recommendations from the failure investigation of the earlier STEP-4 mission were implemented on TSX-5 wherever possible. STEP-4 was successfully placed into orbit, but the ground team was unable to establish contact. The cause of this malfunction is still a subject of debate. Failed solar-array deployment, failed space-vehicle initialization sequence, vibration-induced failure of nonredundant critical components—these are all possible culprits. Because the space vehicle did not respond to commands and no telemetry was ever received, it's impossible to know for sure. The STEP Mission 4 Safety Investigation Board generated a list of recommendations, most of which were implemented on TSX-5 (even though much of the development work had already been completed).

Solar Arrays

Aerospace began by examining the solar arrays. TSX-5—like STEP-4—originally had solar cells on only one side of the solar panels (because only one side would face the sun when fully deployed). This meant, of course, that the solar arrays had to be deployed to get electrical power; if they failed to open properly—a possible failure mode with STEP-4—the space vehicle would have no electrical power other than what the batteries could supply. Aerospace therefore advised the TSX-5 design team to include outward-facing solar cells on the undeployed arrays. Thus, if the mechanism did not deploy as planned, the space vehicle could still generate at least a small amount of power, allowing critical operations to proceed while efforts to deploy the solar arrays could be performed.

While the additional solar cells would prevent a dead-on-arrival space vehicle, they would not save the mission if the arrays never deployed. The TSX-5 team was therefore understandably keen to test the solar-array deployment sequence. Unfortunately, the solar panels were connected by a new type of hinge designed to function only in the near-weightlessness of space. That meant the arrays could not be reliably tested in a laboratory on Earth because the hinge could not support the weight of the panels. Aerospace devised a way to "test" the hinge using DADS (Dynamic Analysis and Design System), a computer simulation tool for predicting the behavior of complex mechanical systems.

Electrical integration of the STRV-2 experiment module and the TSX-5 spacecraft. (Orbital Sciences Corporation)

The results indicated that overtravel in the hinge could allow the two solar-array wings to collide, potentially damaging the delicate solar cells. In order to prevent this, Aerospace recommended that the two wings be deployed in a staggered sequence, rather than simultaneously, with a sufficient delay to allow each component to reach stasis. The analysis further revealed that the articulation joints would not adequately secure the panels, which were not deployed in a planar configuration. Models showed that gaps between panels could potentially exceed hinge length. Ultimately, simply moving the joints closer to the spacecraft solved the problem. Although these issues were relatively easy to resolve, they would not have been discovered prior to launch if Aerospace had not detected them using the DADS tool.

Primaries and Backups

Aerospace also performed dynamic X-ray inspection of critical components within the power subsystem after one of these components experienced a relay failure. This technique uses real-time X-ray to generate an image of a part as it is rotated, allowing the inspector to see, for example, contaminant particles in a relay case that could lodge between the contacts.

This testing is not routine and is usually done when a suspect set of parts is identified. Testing identified a suspect lot of relays. Sixteen relays in the two mains power control boxes and four in the launch-vehicle interface unit were replaced because of particles (weld splatter from a deficient weld process) noted in some of the X-rays.

Redundancy in critical subsystems was reviewed and implemented consistent with a Class C (medium-risk) space vehicle. Greater redundancy was built into the CADACS processors (command and data- handling/attitude determination and control) and communications processors as well as the mains power control, the core power control, and the batteries. Aerospace further demonstrated the adequacy of a marginal circuit design used for the backup deployment device. Engineers were concerned that the circuit would not generate the desired output under all conceivable conditions; however, a SPICE analysis (Simulation Program with Integrated Circuit Emphasis) and breadboard test conducted by Aerospace revealed that the circuit could be expected to perform as designed, given the conditions it would face.

Thermal vacuum testing revealed several potential problems with TSX-5—notably, anomalies in the pyrotechnic fire circuit, noncommanded processor switchovers to the redundant sides, and an anomaly in the downlink filter caused by arc-over in the vacuum environment. (Orbital Sciences Corporation)

Thermal Modeling

Meanwhile, the Aerospace Heat Transfer Laboratory was developing thermal models for testing various spacecraft materials and components. For example, Aerospace engineers measured the thermal conductivity across the hinges that connect the solar arrays to the core plate. All avionics on TSX-5—including batteries—are mounted on the core plate, so it was important to understand the conduction coupling between the core plate and solar arrays. Another set of experiments modeled the normal and lateral conductivity through the 2.54-centimeter-thick core-plate panel itself. Aboard TSX-5, conduction through the core plate is used for passive temperature control of the boxes and avionics; therefore, knowledge of the core plate's thermal behavior was essential in determining the effectiveness of this technique. Similarly, Aerospace measured the normal thermal conductivity through the 1-centimeter-thick solar-array aluminum honeycomb (sandwiched between the panel face sheets) to ensure that thermal activity would have no effect on the boxes and components mounted on the core plate.

Temperature did, however, play a role in a problem affecting the DRAM (dynamic random-access memory). The spacecraft contractor had discovered that when the satellite's experiment interface processor was powered off and on again, the software was not starting from a known reset condition because mass memory in the processor had somehow been "retained." To find the cause of the problem, Aerospace performed a series of temperature tests on the DRAM devices and did indeed confirm that in the narrow range of 0–5 degrees centigrade, memory would be retained, possibly for several minutes. This confirmed the hypothesis and eliminated the possibility of any other hidden failure mode. Fortunately, a practical fix was easy: if the processor needed a reboot while in the critical temperature range, operators were advised to wait long enough to allow the memory to "dissipate."

Environmental Testing

Aerospace engineers were also concerned about the drop transient—the mechanical stress experienced during the initial drop of the Pegasus XL from the L1011 aircraft. Vibration-induced failure of critical components might have been a factor in the failure of STEP-4. Therefore, the environmental test regimen began with a base shake test, simulating the critical vibration frequencies and duration of the drop.

Validation of the space vehicle continued with a 48-hour operational "day-in-the-life" test conducted during the thermal vacuum test—a step that was performed less thoroughly for STEP-4. Thermal vacuum testing checks the integrated space vehicle operations and, to a limited extent, the experiment payloads, in a simulated space environment. The initial test revealed a number of problems that needed to be resolved—most notably, anomalies in the pyrotechnic fire circuit, noncommanded processor switchovers to the redundant sides, and an anomaly in the downlink filter caused by arc-over in the vacuum environment. Once these deficiencies were corrected, a second thermal vacuum test was performed to validate the rework.

Consequently, the integrated TSX-5 was subjected to two random-vibration tests. The first was performed prior to the thermal vacuum test at relatively low vibration levels. Aerospace successfully argued that the second vibration test should be performed at higher levels to validate the rework performed to correct the problems identified during the first thermal vacuum test.

The anomaly in the telemetry downlink system is typical of the problems that can only be discovered through thorough thermal vacuum testing. While the chamber was being depressurized for the first test, the downlink telemetry dropped out for several minutes, but then recovered. The radio-frequency filter, which was not vented, was high on the suspect list. Upon examination, the filter showed evidence of internal corona discharge—apparently triggered when the transmitter was turned on while the filter still held a partial pressure. A new filter was procured, and two vent holes were drilled into it. The contractor's estimate was roughly ten minutes to get past the critical pressure for corona effects. Aerospace performed calculations for the critical pressure, based on the filter's electrical and mechanical properties. Using a critical pressure of about two torr, the estimated time to vent was calculated to be approximately six minutes, which would satisfy the on-orbit requirements for when the downlink transmitter would normally be turned on. Aerospace also recommended a stand-alone test of this filter configuration in a small thermal vacuum chamber to verify its operation. The venting time became a constraint for test conduct during the second thermal vacuum test, which was successful.

STRV-2, shown here during integration and testing, houses a suite of space experiments sponsored by the Ballistic Missile Defense Organization. Specific experiments include LaserCom, which unfortunately failed to demonstrate the feasibility of laser-based communications in free space, and the Medium Waveband Infrared imager, a new technology capable of identifying military aircraft flying beneath the satellite. (Orbital Sciences Corporation)

Systems Compatibility

On the command level, engineers from Aerospace developed the test plan and documentation for both the factory compatibility and launch-base compatibility tests. The factory compatibility test revealed that the Satellite Operations Complex would have difficulties processing experiment data, primarily because the data protocol was not fully compatible with the ground-system network. The data protocol was similar to that of a computer network that assumes continuous messaging from one node to another. Unfortunately, the ground-system network and satellite control equipment were not designed to accommodate this easily. Aerospace developed several options for resolving the problem and performed statistical studies to evaluate the total bit-error rate for the potential solutions. In the end, Aerospace suggested a mix of ground software redesign and new operational procedures. Though not elegant, the solution provided the required data quality at minimal cost. The recommendation was implemented, and to date, the system has provided excellent data quality.

Other command functions received a similarly rigorous treatment. For example, the fire, separation, solar-array deployment, and initiation sequences were tested live with the launch-vehicle interface attached. Again, this arose from the recommendations of the STEP Mission 4 Safety Investigation Board. One possible explanation for the failure of STEP-4 was that the space vehicle did not obtain any signal indicating that it was released from the launch vehicle. In that case, initialization never would have begun. For this reason, the TSX-5 team tested this critical interface and the space vehicle initialization process many times.

STRV-2

In addition to validating the design and integration of the TSX-5 vehicle, Aerospace also helped validate individual experiments. For example, the medium-wave infrared imaging experiment, which was sponsored in part by the United Kingdom Ministry of Defense, needed to reach perigee over the UK approximately 45 days after launch to achieve best results. Aerospace engineers used RAAN (Right Ascension of the Ascending Node) analysis to establish the optimal launch window for this component. This was used to set the launch day and time.

CEASE—shown here after integration with the TSX-5 spacecraft—will help satellite operators plan for and react to adverse space weather conditions. (Orbital Sciences Corporation)

One of the goals of the SAMMES experiment is to assess contamination deposition in space. Contamination is a critical enemy of many experiments, especially optical sensors. A contamination budget must be established to ensure normal operation amid known contaminants. Aerospace researchers performed the contamination budget analysis, developed the contamination monitoring software, and performed the contamination monitoring for this project. During thermal vacuum testing, where outgassing does occur, contamination monitoring is essential. Aerospace devised the allowable criteria to be used during the thermal vacuum testing and developed a software program to record and analyze the results. Fortunately, the testing indicated that contamination didn't exceed critical levels and hence was not an issue.

CEASE

For the CEASE payload, Aerospace engineers conducted the leakdown, or venting, analysis to ensure that the electronics box would properly vent atmospheric pressure before being turned on in space. This would eliminate the possibility of having a partial pressure, which could lead to arc-over and potential damage to the experiment.

The CEASE interface control document further stipulated that CEASE should be built to withstand a depressurization rate of one pound per square inch per second. Because this requirement was not met by the original design, Aerospace was asked to determine whether the requirement could be waived. An analysis was performed to establish the pressure capability of the CEASE unit and the expected gage pressure it would experience during the carrier aircraft ascent and subsequent launch via the Pegasus booster. The analysis showed that unit gage pressure was well below the pressure capability of the CEASE flight box, and Aerospace recommended that the program office waive the requirement, eliminating the need for a hardware redesign.

Against the Clock

Aerospace worked until liftoff to resolve the anomalies that invariably occur before a launch. When the space vehicle was at Vandenberg Air Force Base undergoing final testing for launch, an anomaly was detected in the spacecraft clock: approximately four to five times within a 24-hour period, the clock slipped by 868 milliseconds.

Aerospace engineers feared that the clock was degrading and could conceivably fail. The space-vehicle contractor, in contrast, attributed the problem to noise in a clock-compare circuit and recommended sending the satellite into orbit "as is," with a work-around to compensate in flight. Quickly working to ascertain the actual cause, Aerospace generated fishbone diagrams, evaluated noise characterizations on a card-cage model of the clock-compare circuitry, analyzed the timing circuit, reviewed the component noise-susceptibility data, and analyzed the archived test data. The researchers concluded that noise at the trailing edge of the timing pulse did indeed trigger the anomaly. Moreover, a survey of the archived data revealed that the problem had occurred before during thermal vacuum testing and was possibly linked to temperature. A timing-circuit review did not turn up a plausible hardware degradation or failure that would match the anomaly signature. The final recommendation was to proceed with the launch using the operational work-around to correct the clock timing as needed.

TSX-5 was placed into orbit from a Pegasus launch vehicle, shown here in a clean environment being prepared for launch. (Orbital Sciences Corporation)

On-Orbit Operations

Aerospace's involvement in the TSX-5 program did not end with the satellite launch. On the contrary, Aerospace continues to support on-orbit operations and system optimization.

For example, a number of electronic boxes on the space vehicle have experienced anomalous behavior. In particular, the GPS (Global Positioning System) receiver and the fiber-optic gyro seem prone to upsets in their operation. Analyses by Aerospace and the CEASE research team have shown that these upsets are strongly correlated to inclement space weather such as magnetic storms and other high-particle-flux events. The GPS receiver is essentially a commercial, off-the-shelf product, and thus has no radiation hardening. The fiber-optic gyro is the same as the one flown on the National Aeronautics and Space Administration's Deep Space 1—which also exhibited a propensity for single-event upsets. Such glitches cannot currently be prevented, but Aerospace worked with the ground operations and the spacecraft contractor team to develop space-vehicle contingency plans to minimize their operational impact.

Interestingly, these upset-inducing phenomena are precisely what CEASE was sent to investigate. Unfortunately, data from CEASE is evaluated after the fact, and as such could not be used to predict conditions aboard TSX-5; however, the information has been enormously useful in correlating certain space-vehicle anomalies to high-particle-flux events, thereby providing plausible cause and effect.

Ephemeris

VISS (Vibration Isolation, Suppression, and Steering system) is a self-contained vibration-control device for use with sensitive optical sensors and other detection devices. VISS achieves vibration isolation and suppression greater than 20 decibels, with plus or minus 0.3 degrees steering of the payload at a rate of 2 Hertz. (Orbital Sciences Corporation)

One of the responsibilities of the Satellite Operations Complex is to provide several days of predicted ephemeris to help researchers plan their experiments. Unfortunately, the research team found that the predictions were not as accurate as needed. Aerospace performed extensive analyses of the Satellite Operations Complex's ephemeris products and did comparisons with products generated by Aerospace tools. The ephemeris propagation improved after Aerospace provided updated range biases and suggestions for better estimating atmospheric drag. Still, this improvement didn't always provide the desired precision. The problem was that the accuracy of the predictions depended on the stability of the atmosphere during the forecast period. For a quiet atmosphere, the ephemeris predictions should be sufficiently accurate, but during periods of high atmospheric activity, the accuracy requirement could be exceeded well within a day. After more analysis, Aerospace determined that there were simply no tools that could achieve the desired level of accuracy. As a result, engineers were forced to change their operations concept to accommodate the achievable accuracy of the ephemeris products.

Conclusion

Acquisition reform places a greater burden on the insightful testing and validation of any space mission. As STEP-4 clearly showed, a failed mission is neither faster nor cheaper—and certainly not better. These checks cannot be omitted, but must be performed more efficiently and with greater confidence. To achieve these goals, testing methodologies must often be as innovative as the technologies they seek to prove. As for TSX-5, the experiments have met their initial six-month mission requirements (with the exception of LaserCom, which experienced hardware problems from the outset), and the Air Force approved an additional six months of support toward the one-year mission goal.

To Summer 20001 Table of Contents