Search This Blog

Monday, September 23, 2013

Cygnus Delays ISS Berthing Following GPS Time Discrepancy

From Orbital’s Cygnus spacecraft was into the final leg of berthing with the International Space Station (ISS) on Sunday morning, prior to a discrepancy relating to the way the ISS and Cygnus determine GPS data. The fascinating issue can be fixed via an update to Cygnus’ software, allowing for a second rendezvous and berthing attempt, potentially as soon as Tuesday.

Cygnus was successfully lofted into space via the second flight of Orbital’s Antares rocket on September 18 from its Wallops facility.

Per L2 information, the post-flight review – known as the “24 hour quick look review” – noted no issues with the launch phase of ORB-D.

Bidding farewell to Antares – and completing what is known as the Integrated Launch Operations Phase (ILOPS), for the ascent and insertion element of the mission – Cygnus immediately began completing numerous first-time operations, stretching out its solar arrays before setting its sights on the orbital outpost.

The Cygnus vehicle – under the control of Orbital at MCC-Dulles – then prepared for a series of well-practiced demonstrations over its four day period of catching up to the ISS.

Known as the Phasing Operations Phase (POPS) – for the pursuit of the ISS – Cygnus has a task sheet of 10 specific demonstrations to complete during its four-day in-orbit journey to rendezvous with the ISS.

These demo elements have been intertwined with orbital operations such as Delta V burns to raise its orbit, the first of which was successfully completed early into the mission.

Cygnus’ flight profile – as seen in an expansive overview presentation acquired by L2 – involves chasing the ISS from behind and below.

Following another burn, the team carried out free drift and abort demonstrations – known as Demos 2a and 2b – marking the completion of the first demonstration milestones under the ORB-D requirements.

With Cygnus closing in on the ISS’ neighborhood, along with its altitude showing a large up curve per the flight profile, the spacecraft has completed a total of four Delta V burns to get within a four kilometers of the ISS’ altitude.

The trouble-free flight of the new spacecraft was marked by a NASA Mission Management Team (IMMT) providing approval for Cygnus to continue its pursuit of the Station, along with a provisional green light for berthing to take place early on Sunday.

However, Sunday’s rendezvous is arguably the most tricky part, as Cygnus uses its brain to track and communicate with the ISS.

Cygnus will eventually – from OrB-2 onwards – use the TriDAR vision system – designed by Canadian company Neptec, with the support of NASA and the Canadian Space Agency. This system provides real-time visual guidance for navigation, rendezvous and docking procedures.

The system has proved its worth by flying on three shuttle missions, previously with Discovery on STS-128 and STS-131, prior to its final shuttle trip with Atlantis on STS-135.

Its performance with Atlantis during rendezvous was impressive, with the TriDAR-created video (L2) of the acquisition of the International Space Station showing how the system tracked the orbital outpost from 34km out, all the way through to Atlantis’ docking.

For this mission, Cygnus is using sensors from Jena, scanning LIDARs that track retros – the same sensors that are used by Japan’s HTV resupply vehicle.

Overall, Cygnus will utilize three sensors on each flight. For the first couple of flights all of the sensors are supplied by Jena. However, during ORB-2 – the third flight of Cygnus – there will be two Jena sensors and one Neptec TriDAR sensor.  All subsequent missions will see the spacecraft use two Neptec TriDAR sensors and one from Jena.

The ISS and Cygnus also need to show they have a strong communication link, required not least for the ability to manually abort the approach – or at least retreat – in the event of problems. This phase of the mission is called the Joint Operations Phase (JOPS).

This critical approach period is called Proximity Ops, with Cygnus using the JEM PROX system for direct communications with ISS, effectively resulting in the use of the same system Japan’s HTV uses for arriving at the ISS, as much as there will be a number of different settings employed for Cygnus’ arrival.

As with SpaceX’s Dragon, Cygnus will stalk the ISS, slowly creeping up to its target via a large series of demos to test its systems. As the Orbital spacecraft sneaks up the R-bar, under the ISS, it will enter the KOS (Keep Out Sphere).

Once inside the KOS, Cygnus will demonstrate that he can hold and retreat, prior to receiving the go – via polling – to proceed.

With its sensors locked on – the tenth demo of the approach at this point – a final go will be given for Cygnus to approach to the capture point, mirroring Dragon’s own successful approach last month.

UPDATE: However, with Cygnus around 15km from the ISS, a problem was noted with the GPS readings between Cygnus and the ISS – one of the key demo requirements – resulting in an abort being called.

“This [Sunday] morning, at around 1:30 a.m. EDT, Cygnus established direct data contact with the ISS and found that some of the data received had values that it did not expect, causing Cygnus to reject the data,” noted Orbital. “This mandated an interruption of the approach sequence. Orbital has subsequently found the causes of this discrepancy and is developing a software fix.”

Expanding on the specifics, information acquired by L2 cite the two ways of specifying GPS time as key to the problem on Sunday.

GPS time is specified using week numbers, however there are two forms of GPS time – one based on the original 1980 ephemeris, and another based on an ephemeris designed around 20 years later in 1999, with the difference being exactly 1024 weeks.

The 1980 GPS time uses 10-bit week numbers, while the 1999 version, which uses 13-bit week numbers, was introduced to save old hardware-based receivers that were rolling over the GPS time “week number” – a problem similar to – but not exactly – the “Y2K” issue.

The relevance to the issue between the ISS and Cygnus is in the Japanese PROX system on the ISS, which transmit GPS data (including time) from the SIGIs (Space Integrated INS [Inertial Navigation System]\GPS) units located in the JEM (Japanese Experiment Module) to Cygnus.

The Cygnus then uses that data in a Kalman filter to produce a best estimate of the ISS-relative Cygnus position.

Before Cygnus uses the GPS time data from the PROX system SIGIs however, it first compares it with the GPS time derived from its own on-board SIGI units, which are the same make and model as those used on the ISS. The issue that led to the Cygnus-ISS GPS communication issue is that Orbital interpreted that that ISS PROX system was transmitting GPS times based on the 1980 ephemeris, and thus used the 1980 ephemeris in Cygnus’ on-board SIGI units too.

In reality however, the ISS PROX system uses the 1999 ephemeris for its GPS time data, which meant Cygnus could not understand the GPS time data being transmitted to it from the ISS, hence the reason for the aborted rendezvous attempt.

In all of the integrated Cygnus rendezvous tests that were performed on the ground, the incorrect 1980 ephemeris was used, and thus it was not until the JEM PROX receivers were put into the loop that the discrepancy was discovered.

While the JEM PROX system has been used on the four Japanese HTV missions flown to date, it appears that the HTV does not use the PROX-transmitted GPS week number at all, hence why the issue never surfaced before now, despite the fact that the PROX is a proven system.

SpaceX’s Dragon vehicle does not use the PROX system at all, and instead performs it’s Relative GPS (RGPS) navigation using the US SIGIs on the ISS, rather than the Japanese SIGIs of the PROX system.

As such, rather then being a “problem” with either Cygnus or the ISS, the issue is merely a simple case of the two vehicles trying to communicate with each other the wrong way, and thus the solution is fairly simple. The simple fix is to add “1024″ (the difference in week numbers between the 1980 and 1999 ephemeris) to the data received from the PROX system, which only requires modification of a single instruction in the Cygnus software.

As of Sunday morning Orbital began running regression tests to ensure nothing breaks as a result of this minor change, but the very limited nature of the modification can insure high confidence by means of inspection.

The only issue is that in order to execute the new code, the Cygnus avionics must be reset – and NASA would like Cygnus to be several hundred miles away from the ISS when it resets.

While the relatively simple issue may seem obvious now, it is a good example of the challenges of testing hardware for use on or with the ISS, where testing with actual flight hardware is not possible, leading to a situation where many issues can only be found or resolved in-flight (hence the need to perform demonstration flights in the first place).

The next attempt – per a call to the ISS crew – will not take place before Tuesday.

When Cygnus does manage to complete rendezvous with the Station, the ISS’ Space Station Remote Manipulator System (SSRMS) will then reach out and grapple Cygnus, prior to being berthed on the Station.

Once berthed, the ISS crew will begin vestibule ops and Cygnus activation via ISS power jumpers on rendezvous day, documented as a nine hour procedure.

Hatch opening and ingress will occur on the following mission day.

“Cygnus receives power via PVGF; overnight park position not required for capture delay. Current timeline gets through berthing, start of vestibule ops & Cygnus activation via ISS power jumpers on rendezvous day.  Hatch open and ingress is the following day,” noted an associated overview presentation (L2).

Cygnus’ hatch is very similar to a standard US segment hatch, albeit slightly smaller, making it a similar sight to the ISS crewmembers. A ventilation duct will be hooked up, and the spacecraft cleared of any dust prior to becoming safe to ingress without eye protection and masks.

One the hatch is open, crewmembers will begin cargo removal operations, during its month-long stay, a phase of the mission known as the Berthed Operations Phase (BOPS).

Cargo ops involves the crew removing the “top layers” on PORT and STBD pallets to make room in PCM. They will then remove components of the Secondary Structure as required, ahead of emptying the FWD and AFT pallets to gain access to the Standoff pallets, which they will empty and repack.

The reverse sequence will be employed until the vehicle has been repacked, although all the return cargo won’t be classed as downmass, because – unlike Dragon – Cygnus won’t be returning to the ground or water.

Instead, it’ll be sent on a path to a destructive re-entry.

The final phase of the mission – a reverse of the berthing procedures – is called the Descent & Reentry Operations Phase (DROPS), as Cygnus ends its life in a disposal corridor during entry, hopefully with a smile on its face, following a successful demonstration that paved the way for its siblings to each take a turn in providing full CRS operations.

The deal to carry out ISS resupply flights – under the $1.9 billion CRS contract – encompasses eight missions between 2012 and 2015 carrying approximately 20,000 kg of cargo to the ISS.

Preparations are already in full swing for the ORB-1 mission, with hardware already being processed at its Wallops base.

(Images: via L2′s Antares/Cygnus Section – Containing presentations, videos, images, interactive high level updates and more, with additional images via Orbital and Neptec).