Post mortem for CRS-3 Antares notes turbopump failure
Engineers are methodically working through stacks of data related to last week’s failure of the CRS-3/OrB-3 Antares launch vehicle. Orbital managers have noted that early data points to the failure of a turbopump on one of Antares’ AJ-26 engines as the leading reason the rocket failed just seconds after launch. The company are now working towards “upgrading” the propulsion system with a new engine, ready for 2016.
Antares – along with her Cygnus spacecraft bound for the International Space Station (ISS) – were lost seconds after launch during Thursday’s mission from Pad 0A at the Mid-Atlantic Regional Spaceport on Wallops Island, Virginia.
The rocket made a nominal ascent off the pad, as Antares conducted her designed “Baumgartner Maneuver” – a subtle sideways step to ease the vehicle away from the close proximity of the Transporter/Erector/Launcher (TEL) – and ascended into the night sky.
Once clear of the pad, the first sign of trouble was a distinct change to the appearance of the plume generated by her AJ-26 engines at T+14 seconds – suggesting oxidizer starvation in at least one engine – prior to the aft of the vehicle exploding a second later.
Additional videos of varying quality showed a “ghost image” or visual artifact on the footage, which may provide extra insight into what transpired.
An enhanced and cropped view, as seen in this .gif, was created on L2. This imagery was subsequently requested – and handed over – to Orbital managers to be used in their investigation.
The next event was sadly obvious, as propulsion was lost and the vehicle fell in flames, exploding near the pad from where she had just departed.
The Flight Termination System (FTS) was activated just moments prior to the vehicle impacting the ground.
It is understood the FTS command was successfully sent via the Range Safety system that employs fully steerable antennas, which track the rocket, as well as omni antennas which radiate in all directions. This ensures the rocket receives the command.
This command results in the puncturing of the tank domes on the first stage and the motor dome on the second stage, causing the depressurization on each stage and rendering the vehicle non-propulsive.
The root cause is being worked by the Antares launch failure Accident Investigation Board (AIB), which is being led by Orbital under the oversight of the Federal Aviation Administration (FAA). The board includes former Space Shuttle Program manager Wayne Hale and other industry experts.
“The Investigation Board (AIB) is making good progress in determining the primary cause of last week’s failure. A preliminary review of telemetry and video data has been conducted and substantial debris from the Antares rocket and its Cygnus payload has been collected and examined,” noted Orbital on Wednesday.
“While the work of the AIB continues, preliminary evidence and analysis conducted to date points to a probable turbopump-related failure in one of the two Aerojet Rocketdyne AJ26 stage one main engines. As a result, the use of these engines for the Antares vehicle likely will be discontinued.”
It was later noted that the AJ-26 could still be used if they were proven to be completely flight-worthy.
Clearly, a failure of this type was always going to place focus on the AJ-26 engines, although the investigation has also been looking at all first stage propulsion items, such as the feedlines that supply the AJ-26’s with their lifeblood.
The AJ-26-22 main engine is a rebuilt version of Soviet NK-33, originally intended for the massive Soviet N-1 launch vehicle. The engine is fed with a LOX/RP combination, producing 3,265kN of thrust at Sea Level.
Around 40 of these engines were purchased by American company Aerojet in the mid-1990s, before refurbishing and modifying the engines under contract with Orbital. In total, 200 NK-33 engines were built and 575 engine tests conducted, totalling more than 100,000 seconds of test time.
Additional media focus on the engines is likely related to a couple of undesirable events on the test stand.
A previous failure of an AJ-26 occurred in June, 2011 – when the fourth Antares engine caught fire on the E-1 Test Stand. The fire was believed to have been caused by a kerosene fuel leak in an engine manifold, with the root cause was subsequently determined to be stress corrosion cracking of the 40-year old metal.
Additional information acquired via L2 adds that the first Stennis failure occurred in the highest pressure part of the fuel system just downstream of the turbopump. The pump over-sped and the chamber was starved of coolant in the ~170ms between the event and shutdown.
The most recent failure occured in May of this year, when an AJ-26 – set to fly on a future Antares mission – failed on the Stennis test stand mid-way through its burn.
Orbital conducted a complete review of the failure and deemed the current engine stock as safe to use.
The company never released any specific information into the test stand failure, although a turbopump was also believed to be the root cause.
Orbital has been looking to change the engines for the first stage of their Antares rocket for some time, with evaluations into a range of RD- engines ongoing, at least since August when the company was believed to be looking at the RD-181 (L2).
A switch to the RD-193 now appears to be favored – although Orbital officials do not wish to confirm the engine at this time. The new engine will debut on the Antares in 2016. In the meantime, one or two launches of Cygnus will take place on a different launch vehicle, possibly the Falcon 9 – among other options.
Orbital management noted they are talking to three launch providers, two in the US and one in Europe. There are 15 configurations of the vehicles they are in discussion with, for Cygnus launches next year.
“Orbital plans an early introduction of its previously selected Antares propulsion system upgrade in 2016,” added the company. “This will be preceded by one or two non-Antares launches of the company’s Cygnus cargo spacecraft to the ISS in 2015-2016, employing the spacecraft’s compatibility with various launch vehicles and its flexibility to accommodate heavier cargo loads as launcher capacity permits.”
Damage to the Pad:
Although the TV coverage of the failure showed a large explosion and multiple fires at the pad complex, the critical hardware appears to have endured the unscheduled return of Antares relatively well.
A photo taken of the impact site – after most of the critical Antares hardware had been removed – showed most of the infrastructure to be intact.
Although a lot of work will be required to bring the pad back up to launch operations, it is unlikely to be the main schedule driver for any hopeful return of Antares to the pad.
“The overall findings indicate the major elements of the launch complex infrastructure, such as the pad and fuel tanks, avoided serious damage, although some repairs will be necessary,” noted Orbital after an initial assessment.
“After up close visual inspections by the safety team, it still appears the launch site itself avoided major damage. There is some evidence of damage to piping that runs between the fuel and commodity storage vessels and the launch mount, but no evidence of significant damage to either the storage vessels or launch mount.”
Some rocket hardware remains at the site, which is being collected and sent for secure storage at Wallops for the investigation team to take a closer look.
“Based on initial sweeps conducted by an Orbital safety team, it appears a significant amount of debris remains on the site and it is likely substantial hardware evidence will be available to aid in determining root cause of the Antares launch failure,” added Orbital.
“Some of the Cygnus cargo has also been found and will be retrieved as soon as we have clearance to do so to see if any survived intact.”
With Cygnus now expected to ride on an different launch vehicle next year, her return to an upgraded Antares won’t be until 2016.
The immediate requirements of Orbital’s CRS obligations resulted in the development of a comprehensive plan to maintain the cargo supply line between Earth and the International Space Station, fulfilling Orbital’s commitment to NASA for the delivery of supplies.
The CRS-3 Cygnus was carrying 2,215 kilograms (4,883 lb) of cargo – with an additional 81 kilograms (179 lb) of packaging.
This cargo included 727 kilograms (1600 lb) of scientific equipment and 748 kilograms (1650 lb) of food and supplies for the crew. Around 637 kilograms (1400 lb) of the mass was spares and hardware for the station; mostly for the US segment but including around 30 kilograms (66 lb) for Japan’s modules.
A further 66 kilograms (146 lb) of hardware was being carried for the astronauts’ EVA equipment. The remaining 37 kilograms (82 lb) of cargo was computer equipment.
“Orbital is taking decisive action to fulfill our commitments to NASA in support of safe and productive operations of the Space Station. While last week’s Antares failure was very disappointing to all of us, the company is already implementing a contingency plan to overcome this setback,” noted Mr. David W. Thompson, Orbital’s Chairman and Chief Executive Officer.
“We intend to move forward safely but also expeditiously to put our CRS cargo program back on track and to accelerate the introduction of our upgraded Antares rocket.”
On the NASA side, work to change some of SpaceX’s SpX-5/CRS-5 Dragon cargo – which is deemed as time-sensitive for the Station – is also being evaluated.
The changes may impact on the December 9 launch date for Dragon’s ride on the Falcon 9 v1.1 – however, schedule information shows no change to NET (No Earlier Than) target at this time.
(Images: via L2’s Antares/Cygnus Section – Containing presentations, videos, a vast set of unreleased hi-res images, interactive high level updates and more, with additional images via Orbital and NASA).
(Click here: http://www.nasaspaceflight.com/l2/ – to view how you can support NSF and access the best space flight content on the entire internet).