8 ms·

Antares rocket explosion: The question of using decades-old Soviet engines

37 points by coreymgilmore 12y ago

hamitron 12y ago
Well the CEO of a rocket company should have a better response than "I wish there were better engines to use." Maybe with that multi billion dollar contract you can, oh I don't know, build one?
- SEJeff 12y ago
  SpaceX certainly built their own and they seem to be working quite well thusfar.
  - wdewind 12y ago
    To be fair SpaceX's last launch (Aug 22) exploded too. It was a test, not a real mission, but to quote Elon: "Rockets are tricky."
    - ddrmaxgt37 12y ago
      Yup. IIRC it was auto triggered by some onboard system because of one of those 'anomalies'.
    - will_hughes 12y ago
      For starters, that wasn't an actual launch to orbit, that's just the low altitude testing frames for landing/reusability. It was deliberately destroyed as a safety measure because the rocket was moving out of the testing zone and the onboard system wasn't able to compensate fast enough. Apparently Elon and a few others were a little annoyed that more (appropriate) risk wasn't being taken in those test flights, because they'd been 100% successful. Secondly, their launches of the production Falcon 9's have been 100% successful (with one launch not achieving a secondary objective). Additionally, after the first three Falcon 1's, they were also 100% successful. And finally, they launched two Falcon 9's since August 22nd: September 7th: Asiasat 6 launch (Geosync), and September 21st - Dragon CRS-4 (ISS).
      - avmich 12y ago
        Agree. And those Merlin-1D seem to be very efficient from economical point of view. Technically they are also good - record setting for open cycle engines. However, NK-33 are closed cycle and have better Isp. Economically they also made a lot of sense for Orbital, since they were bought as already made for another project. It can be justified.
    - SEJeff 12y ago
      Sure but I don't consider a test launch the same as "production" launch carrying payloads for ISS or other companies to hurl into orbit. Slightly different in that it is more acceptable for a test rocket trying out new functionality with no payload might fail.
    - rst 12y ago
      Nu? SpaceX has had two successful launches since Aug. 22 -- AsiaSat 6 on September 7, and the CRS-4 ISS supply mission on September 21. The test article that blew up on the 22nd was not launch capable, and was reportedly missing some safety systems that are present on production vehicles.
- fastball 12y ago
  Minor nitpick: Frank Culbertson is an executive vice president at Orbital, not the CEO. The CEO mentioned in the article (Elon Musk) said "there is better technology to use".
- ijk 12y ago
  Building a new engine is, quite literally, rocket science. On the plus side, we know the problems can be solved, because we've done it before. On the minus side, a lot of that knowledge has been lost and needs to be recreated or redesigned. We can't easily rebuild the Saturn V, for example, even though its design would work well for many of today's launches. [1] New rockets are being designed, but its a major undertaking. [2] NASA's SLS is in the $18 billion range. [3] [1] http://amyshirateitel.com/2011/04/03/the-lost-art-of-the-saturn-v/ http://amyshirateitel.com/2011/04/03/the-lost-art-of-the-sat... [2] http://en.wikipedia.org/wiki/Saturn_V#Proposed_successors http://en.wikipedia.org/wiki/Saturn_V#Proposed_successors [3] http://en.wikipedia.org/wiki/Space_Launch_System#Program_costs http://en.wikipedia.org/wiki/Space_Launch_System#Program_cos...
- avmich 12y ago
  That's a pretty hard problem. F-1 took unimaginable efforts to make right - event though the only unusual, untested parameter for this engine is its size. RD-170 took all the genius of Glushko - including his previous somewhat successful experience with RD-270, which deserves a whole story for itself - to get working properly. With all hindsight, only the forth Merlin of SpaceX is actually pushing technical limits - and that's in XXI century. Then, there is the issue of time - to build an engine from scratch is easy to delay a rocket for a year - if you're lucky and do that right from the first attempt. So Orbital words can be taken seriously and for the face value. NK-33 is a pretty good engine.
pbreit 12y ago
My admittedly uninformed opinion is that it just doesn't make any sense. I don't care how good the Russians are/were at making rockets. First of all, surely there has been significant progress in all aspects of the design and manufacturing of such an item. And second, who knows how well the engines have been stored, cared for, maintained and transported over the past 40 years. I would think you'd want to have more visibility into such a thing for this type of component. Elon: It “has a contract to resupply the International Space Station, and their rocket honestly sounds like the punch line to a joke,” he said. “It uses Russian rocket engines that were made in the ’60s. I don’t mean their design is from the ’60s — I mean they start with engines that were literally made in the ’60s and, like, packed away in Siberia somewhere.” Orbital: "we did discover there were some effects of aging since they had been in storage for awhile, including some stress corrosion cracking,” I could see Elon offering to supply Orbital with some US-made engines. But maybe that wouldn't satisfy NASA's desire for diversity.
- venomsnake 12y ago
  There hasn't been much development in this types of engines since the moon race. And we may have even gone backwards, because a lot of the institutional knowledge and probably some of the documentation is lost.
- smacktoward 12y ago
  And the punchline is, those engines weren't even very good when they were first made all those decades ago. If they article is right, they were first developed for use on the N1 rocket, the Soviet moon-rocket counterpart to America's Saturn V. There's a reason why Saturn V beat the N1 to the moon -- the N1 had a distressingly consistent habit of blowing up: N1-L3 was underfunded and undertested, and started development in October 1965, almost four years after the Saturn V. The project was badly derailed by the death of its chief designer Sergei Korolev in 1966. Each of the four attempts to launch an N1 failed; during the second launch attempt the N1 rocket crashed back onto its launch pad shortly after liftoff and exploded, resulting in one of the largest artificial non-nuclear explosions in human history. (Source: http://en.wikipedia.org/wiki/N1_(rocket) http://en.wikipedia.org/wiki/N1_(rocket))
brandonmenc 12y ago
The B-52 is just as old, and is expected to have served for almost a century once retired. It's not the same as rockets, but "oh it's old" isn't as compelling of an argument as failure rate, which - how many have there been using these engines? Is that data available?
- venomsnake 12y ago
  "We saw rust and crack, so we repaired them" seems like the culprit. Not the age.
  - brandonmenc 12y ago
    I was commenting more on the (not surprising) comments from their detractors, not so much on the official explanation.
  - alf 12y ago
    Which begs the question... If they don't know enough about rockets to build them, how can they possibly know how to safely repair them?
    - avmich 12y ago
      I don't know enough about computers to make my own PC from scratch. I however can fix some problems with PC. So you can have one without another.
- justin66 12y ago
  There's a lot less redundancy in a rocket. A B-52 can keep flying if a bunch of stuff stops working or falls off. You just cannot design a rocket that way.
- happycube 12y ago
  Those B-52's have been maintained (with some parts from boneyard-ed planes, though) and presumably at least occasionally flown all that time...
mladenkovacevic 12y ago
When all else fails, blame Russia.
- Correctille 12y ago
  No one is blaming Russia. We're blaming the age of rocket engines (that just happened to come from Russia).
  - CmonDev 12y ago
    And why not prominently state "Soviet" in the title.
tomswartz07 12y ago
Here's the thing: the rocket itself is not 'decades old'. The design is. This is tantamount to saying "Oh, cars use a decades old design.. they must be terrible!" The article also completely overlooks the fact that Russia launched a rocket with the supplies within 24 hours of the failure. http://news.yahoo.com/cause-sought-space-supply-rocket-082528475.html http://news.yahoo.com/cause-sought-space-supply-rocket-08252...
- dubcanada 12y ago
  But cars are... they are extremely inefficient to more modern designs. We just haven't found something better then the internal combustion engine.
- JoeAltmaier 12y ago
  The article refers to them as "refurbished Soviet engines" which certainly sounds like Some parts of the engines are original.
- trothamel 12y ago
  Actually, the engines are. The AJ-26 is a rebuilt NK-33, originally built in the early 1970s.
- Florin_Andrei 12y ago
  The rocket engine is actually decades old. The engines were built many years ago, then stored under a tarp in a Russian facility somewhere. They were supposed to be dismantled, but some Party official risked his neck by sparing them the destruction, hiding them in some forgotten facility - however, the flip side to that was that the engines were stored in probably not the best conditions. Orbital used them because they were the cheap option.
- ChuckMcM 12y ago
  Not exactly. The engines themselves are decades old, Orbital Science is quoted as saying "'As we went through testing, we did discover there were some effects of aging since they had been in storage for awhile, including some stress corrosion cracking,' Culbertson conceded at the time. 'That’s what we [corrected] with weld repairs and other inspections.'" They got the engines that were decades old, and they've been fixing problems that they have found, and refurbishing them and installing them into rockets. It doesn't say they have built new engines on the old designs. That would have been better, because then you would have someone who had designed the engine on staff and could understand it in a deep way. Presumably Orbital is designing their own engines now, or contracting with someone to do so, perhaps their go to market strategy was "Get some things into space with these old surplus engines, use the revenue/time to design/test/manufacture our own engines, and then continue on with our own bespoke design." I don't know of course if that was their intention or not.
ianstallings 12y ago
I don't think it's so much about age as it is about design. From my understanding the original NK 33 rocket engines used a unique oxygen-rich fuel turbo-pump mechanism that the west considered too dangerous. Aerojet-Rocketdyne took these engines and refurbished them and therein lies the question. But I think it might be premature to ask this given they haven't told us what happened. The thing is they are very powerful engines and the price was right.
- avmich 12y ago
  It's not that unique - in Russian space school oxidizer-rich turbines are routine. Proton, for example, used to have - until latest years - a pretty solid track record, especially for first stage engines, and those engines use N2O4 for turbines. Considering that N2O4 is burned with some amount of (CH3)2N-NH2, making HNO3... the gas going to chamber from turbine is the nitric acid. Still engines work just fine. One of big reasons for Russians to use oxydizer-reach gas generators is that on the rocket there usually much more oxidizer - by mass - than fuel, which makes the flow through the turbine much bigger - several times. So efficiency of turbopumps goes up. That justifies the complexity of having oxidizer-rich gas flow before the chamber.
gedmark 12y ago
Not sure what proper HN etiquette on this is, but I made a comment on the launch news story on this exact subject: https://news.ycombinator.com/item?id=8525057 https://news.ycombinator.com/item?id=8525057 TLDR- Those engines may be 50 years old, but they're a technological marvel. Using them is a little crazy, but I'm sure Orbital saw it as the "least crazy" of three options: building a new engine from scratch, buying prohibitively expensive engines from Rocketdyne, or buying risky but cheap engines from Russia.
- rst 12y ago
  The design is a technical marvel, but it won't keep the metal from corroding, and this isn't the first sign of trouble -- a few months ago, a different AJ-26 blew up on a test stand. It isn't yet clear that the engine was at fault in the Antares fire (trouble in the tankage or fuel lines could have cascaded down, among other fun possibilities) -- but if the engine does turn out to be at fault, the remaining stock will certainly get more attention.
- CmonDev 12y ago
  What makes Rocketdyne engines less risky? Maybe they are both expensive and risky.
marze 12y ago
So these engines were built when Paul McCartney was 23 years old? Apparently Orbital Sciences scoured the museums and antique shops in Russia and picked up enough of these engines to base a new rocket on them. The reason they were surplus was that the rocket that they were built for was canceled, because the engines exploded on each of the four test launches conducted. Orbital Sciences was once a scrappy startup. Books were written about it. However, this is just embarrassing.
- avmich 12y ago
  Aerojet bought those engines from Russia - the original manufacturer - in 1990-s. No need for Orbital to scour museums. No, engines didn't explode. Were actually never flown before the first Antares lifted off. But performed pretty well on stands.
  - marze 12y ago
    Are you were claiming that there were no failures of the NK-33 engine on any of the four N1 flights that exploded during first stage burn?
    - avmich 12y ago
      Yes. NK-33 never flew on N-1. They were intended to fly on N-1 in its fifth flight - which never happened, because Glushko closed the program after becoming the head of the Energiya company.
      - marze 12y ago
        So before the engines were obtained by Orbital, the NK-33 had never even been flight tested?
        avmich 12y ago
        Before the engines flew for the first time, yes they never flew before. And Antares is the first rocket where those engines have flied - though not the only one rocket already.
diminoten 12y ago
For someone who doesn't know much about aerospace engineering, can someone outline the actual challenge with getting something like this right? What's the missing piece that'll effectively eliminate accidents like this? Or is this one of those things that will just happen from time to time because going to space takes a lot of math that we don't fully understand yet? I was just under the impression that this was a very well known area of science.
- avmich 12y ago
  I think this is fundamentally pretty hard problem. It involves control of turbulence, and this is a big in general unsolved problem today. That includes hypersonic flights, thermonuclear plasma control... Rocket engines work on the edge of material capabilities, and fluctuations - which you want to tame to guarantee things like that won't happen - are hard to deal with.
- austinz 12y ago
  With these particular engines, you also have the variable factor of them being stored for decades in a warehouse, and all the issues with metal corrosion, fatigue, degradation, etc. that come with that. I'm not a materials engineer, but I could imagine things like seasonal temperature changes, humidity, rough handling, changes in the metallurgical properties of the alloys they used over time could all affect the ability of the engine components to withstand high temperatures and pressures.
hindsightbias 12y ago
The NK-33 may be old, but it offers equal or superior performance to anything made in the US. Another engine, the RD-180 is used in the Atlas. US industries have the rights to manufacture, but want $800M and 4 years to build it or create a new engine. That tells you the state of US tech. http://aviationweek.com/defense/aerojet-rocketdyne-targets-25-million-pair-ar-1-engines http://aviationweek.com/defense/aerojet-rocketdyne-targets-2... Video on Russian/Ukranian engines: http://www.youtube.com/watch?v=TMbl_ofF3AM http://www.youtube.com/watch?v=TMbl_ofF3AM
joshAg 12y ago
To answer the question of why even use them at all, it's because on paper they still compare very well to the other options. Specific impulse is basically an efficiency measure, like mpg for cars. The higher it is, the more efficient the rocket engine is. SpaceX's Merlin 1d has a thrust-to-weight ratio (twr) of 150, and a specific impulse (I_sp) at sea level of 2.73 km/s. I doubt this was an option for orbital sciences to use in their rockets. The NK33, which is the soviet engine that was used, has a twr of 137 and an I_sp at sea level of 2.91 km/s. The atlas V rockets by boeing and lockheed use an rd180, also russian built btw, which has a twr of 78.44 and an I_sp at sea level of 3.05 km/s. The delta IV rockets by boeing us an rs68 or rs68a have a twr of 51.2 and an I_sp of 3.58 km/s. I'm not sure if these are available for commercial use, though, since boeing shut down the commercial delta IV launch program. The US military still uses them though, so they are still being produced. Also, these use liquid hydrogen instead of RP1 (rocket grade kerosine) like all the rest of the engines here. The Saturn V used five F1 engines (also 1960s tech, but no longer available). Each of those have a twr of 82.27 and an I_sp of 2.58 km/s.
- jessriedel 12y ago
  Isn't this much more a question of reliability and track record than 10% performance differences?
  - avmich 12y ago
    Both reliability and performance. And also reliability is something which you estimate, calculate basing on some models - it's probability, not a guarantee. How would you estimate the reliability of an engine which run over 3 hours on a test stand? Of course you might prefer a well known fuels, industry-proven range of pressures and other engineering solutions, and you test them on Earth first - but after that you have to fly. Track records are tricky for rocket systems. Since the first Sputnik, over the 57 years the humanity had may be - roughly - 4000 launches of space rockets of all kinds, with many different fuels, engines, sizes etc. It's relatively rare when you have a good track record which is measured by many tens of flights - let alone hundreds. When a system flew 20 times it's already considered rather mature - double so if those were all successful flights. So it's quite usual to use a new engine in a new design of rocket - in fact it's rare when you have many different rockets with the same engine type or many different engines used on the same rocket. Rockets and engines are usually optimized to each other - particularly rockets are built around engines, since margins are quite thin.
ijk 12y ago
Also note that we don't know yet if the NK-33/AJ-26 engine was the cause of failure. Sometimes the most visible possibility wasn't the original cause of the problem.
njsteinsund 12y ago
I can highly recommend these two books for an engaging introduction to getting things right in the rocket business: Ignition! An informal history of liquid rocket propellants http://www.goodreads.com/book/show/677285.Ignition_ http://www.goodreads.com/book/show/677285.Ignition_ and Space Systems Failures: Disasters and Rescues of Satellites, Rocket and Space Probes http://www.goodreads.com/book/show/1015478.Space_Systems_Failures http://www.goodreads.com/book/show/1015478.Space_Systems_Fai...

37 points by coreymgilmore 12y ago

Well the CEO of a rocket company should have a better response than "I wish there were better engines to use." Maybe with that multi billion dollar contract you can, oh I don't know, build one?

SEJeff 12y ago
SpaceX certainly built their own and they seem to be working quite well thusfar.
- wdewind 12y ago
  To be fair SpaceX's last launch (Aug 22) exploded too. It was a test, not a real mission, but to quote Elon: "Rockets are tricky."
  - ddrmaxgt37 12y ago
    Yup. IIRC it was auto triggered by some onboard system because of one of those 'anomalies'.
  - will_hughes 12y ago
    For starters, that wasn't an actual launch to orbit, that's just the low altitude testing frames for landing/reusability. It was deliberately destroyed as a safety measure because the rocket was moving out of the testing zone and the onboard system wasn't able to compensate fast enough. Apparently Elon and a few others were a little annoyed that more (appropriate) risk wasn't being taken in those test flights, because they'd been 100% successful. Secondly, their launches of the production Falcon 9's have been 100% successful (with one launch not achieving a secondary objective). Additionally, after the first three Falcon 1's, they were also 100% successful. And finally, they launched two Falcon 9's since August 22nd: September 7th: Asiasat 6 launch (Geosync), and September 21st - Dragon CRS-4 (ISS).
    - avmich 12y ago
      Agree. And those Merlin-1D seem to be very efficient from economical point of view. Technically they are also good - record setting for open cycle engines. However, NK-33 are closed cycle and have better Isp. Economically they also made a lot of sense for Orbital, since they were bought as already made for another project. It can be justified.
  - SEJeff 12y ago
    Sure but I don't consider a test launch the same as "production" launch carrying payloads for ISS or other companies to hurl into orbit. Slightly different in that it is more acceptable for a test rocket trying out new functionality with no payload might fail.
  - rst 12y ago
    Nu? SpaceX has had two successful launches since Aug. 22 -- AsiaSat 6 on September 7, and the CRS-4 ISS supply mission on September 21. The test article that blew up on the 22nd was not launch capable, and was reportedly missing some safety systems that are present on production vehicles.
fastball 12y ago
Minor nitpick: Frank Culbertson is an executive vice president at Orbital, not the CEO. The CEO mentioned in the article (Elon Musk) said "there is better technology to use".
ijk 12y ago
Building a new engine is, quite literally, rocket science. On the plus side, we know the problems can be solved, because we've done it before. On the minus side, a lot of that knowledge has been lost and needs to be recreated or redesigned. We can't easily rebuild the Saturn V, for example, even though its design would work well for many of today's launches. [1] New rockets are being designed, but its a major undertaking. [2] NASA's SLS is in the $18 billion range. [3] [1] http://amyshirateitel.com/2011/04/03/the-lost-art-of-the-saturn-v/ http://amyshirateitel.com/2011/04/03/the-lost-art-of-the-sat... [2] http://en.wikipedia.org/wiki/Saturn_V#Proposed_successors http://en.wikipedia.org/wiki/Saturn_V#Proposed_successors [3] http://en.wikipedia.org/wiki/Space_Launch_System#Program_costs http://en.wikipedia.org/wiki/Space_Launch_System#Program_cos...
avmich 12y ago
That's a pretty hard problem. F-1 took unimaginable efforts to make right - event though the only unusual, untested parameter for this engine is its size. RD-170 took all the genius of Glushko - including his previous somewhat successful experience with RD-270, which deserves a whole story for itself - to get working properly. With all hindsight, only the forth Merlin of SpaceX is actually pushing technical limits - and that's in XXI century. Then, there is the issue of time - to build an engine from scratch is easy to delay a rocket for a year - if you're lucky and do that right from the first attempt. So Orbital words can be taken seriously and for the face value. NK-33 is a pretty good engine.

pbreit 12y ago

My admittedly uninformed opinion is that it just doesn't make any sense. I don't care how good the Russians are/were at making rockets. First of all, surely there has been significant progress in all aspects of the design and manufacturing of such an item. And second, who knows how well the engines have been stored, cared for, maintained and transported over the past 40 years. I would think you'd want to have more visibility into such a thing for this type of component. Elon: It “has a contract to resupply the International Space Station, and their rocket honestly sounds like the punch line to a joke,” he said. “It uses Russian rocket engines that were made in the ’60s. I don’t mean their design is from the ’60s — I mean they start with engines that were literally made in the ’60s and, like, packed away in Siberia somewhere.” Orbital: "we did discover there were some effects of aging since they had been in storage for awhile, including some stress corrosion cracking,” I could see Elon offering to supply Orbital with some US-made engines. But maybe that wouldn't satisfy NASA's desire for diversity.

venomsnake 12y ago
There hasn't been much development in this types of engines since the moon race. And we may have even gone backwards, because a lot of the institutional knowledge and probably some of the documentation is lost.
smacktoward 12y ago
And the punchline is, those engines weren't even very good when they were first made all those decades ago. If they article is right, they were first developed for use on the N1 rocket, the Soviet moon-rocket counterpart to America's Saturn V. There's a reason why Saturn V beat the N1 to the moon -- the N1 had a distressingly consistent habit of blowing up: N1-L3 was underfunded and undertested, and started development in October 1965, almost four years after the Saturn V. The project was badly derailed by the death of its chief designer Sergei Korolev in 1966. Each of the four attempts to launch an N1 failed; during the second launch attempt the N1 rocket crashed back onto its launch pad shortly after liftoff and exploded, resulting in one of the largest artificial non-nuclear explosions in human history. (Source: http://en.wikipedia.org/wiki/N1_(rocket) http://en.wikipedia.org/wiki/N1_(rocket))

brandonmenc 12y ago

The B-52 is just as old, and is expected to have served for almost a century once retired. It's not the same as rockets, but "oh it's old" isn't as compelling of an argument as failure rate, which - how many have there been using these engines? Is that data available?

venomsnake 12y ago
"We saw rust and crack, so we repaired them" seems like the culprit. Not the age.
- brandonmenc 12y ago
  I was commenting more on the (not surprising) comments from their detractors, not so much on the official explanation.
- alf 12y ago
  Which begs the question... If they don't know enough about rockets to build them, how can they possibly know how to safely repair them?
  - avmich 12y ago
    I don't know enough about computers to make my own PC from scratch. I however can fix some problems with PC. So you can have one without another.
justin66 12y ago
There's a lot less redundancy in a rocket. A B-52 can keep flying if a bunch of stuff stops working or falls off. You just cannot design a rocket that way.
happycube 12y ago
Those B-52's have been maintained (with some parts from boneyard-ed planes, though) and presumably at least occasionally flown all that time...

mladenkovacevic 12y ago

When all else fails, blame Russia.

Correctille 12y ago
No one is blaming Russia. We're blaming the age of rocket engines (that just happened to come from Russia).
- CmonDev 12y ago
  And why not prominently state "Soviet" in the title.

tomswartz07 12y ago

Here's the thing: the rocket itself is not 'decades old'. The design is. This is tantamount to saying "Oh, cars use a decades old design.. they must be terrible!" The article also completely overlooks the fact that Russia launched a rocket with the supplies within 24 hours of the failure. http://news.yahoo.com/cause-sought-space-supply-rocket-082528475.html http://news.yahoo.com/cause-sought-space-supply-rocket-08252...

dubcanada 12y ago
But cars are... they are extremely inefficient to more modern designs. We just haven't found something better then the internal combustion engine.
JoeAltmaier 12y ago
The article refers to them as "refurbished Soviet engines" which certainly sounds like Some parts of the engines are original.
trothamel 12y ago
Actually, the engines are. The AJ-26 is a rebuilt NK-33, originally built in the early 1970s.
Florin_Andrei 12y ago
The rocket engine is actually decades old. The engines were built many years ago, then stored under a tarp in a Russian facility somewhere. They were supposed to be dismantled, but some Party official risked his neck by sparing them the destruction, hiding them in some forgotten facility - however, the flip side to that was that the engines were stored in probably not the best conditions. Orbital used them because they were the cheap option.
ChuckMcM 12y ago
Not exactly. The engines themselves are decades old, Orbital Science is quoted as saying "'As we went through testing, we did discover there were some effects of aging since they had been in storage for awhile, including some stress corrosion cracking,' Culbertson conceded at the time. 'That’s what we [corrected] with weld repairs and other inspections.'" They got the engines that were decades old, and they've been fixing problems that they have found, and refurbishing them and installing them into rockets. It doesn't say they have built new engines on the old designs. That would have been better, because then you would have someone who had designed the engine on staff and could understand it in a deep way. Presumably Orbital is designing their own engines now, or contracting with someone to do so, perhaps their go to market strategy was "Get some things into space with these old surplus engines, use the revenue/time to design/test/manufacture our own engines, and then continue on with our own bespoke design." I don't know of course if that was their intention or not.

ianstallings 12y ago

I don't think it's so much about age as it is about design. From my understanding the original NK 33 rocket engines used a unique oxygen-rich fuel turbo-pump mechanism that the west considered too dangerous. Aerojet-Rocketdyne took these engines and refurbished them and therein lies the question. But I think it might be premature to ask this given they haven't told us what happened. The thing is they are very powerful engines and the price was right.

avmich 12y ago
It's not that unique - in Russian space school oxidizer-rich turbines are routine. Proton, for example, used to have - until latest years - a pretty solid track record, especially for first stage engines, and those engines use N2O4 for turbines. Considering that N2O4 is burned with some amount of (CH3)2N-NH2, making HNO3... the gas going to chamber from turbine is the nitric acid. Still engines work just fine. One of big reasons for Russians to use oxydizer-reach gas generators is that on the rocket there usually much more oxidizer - by mass - than fuel, which makes the flow through the turbine much bigger - several times. So efficiency of turbopumps goes up. That justifies the complexity of having oxidizer-rich gas flow before the chamber.

gedmark 12y ago

Not sure what proper HN etiquette on this is, but I made a comment on the launch news story on this exact subject: https://news.ycombinator.com/item?id=8525057 https://news.ycombinator.com/item?id=8525057 TLDR- Those engines may be 50 years old, but they're a technological marvel. Using them is a little crazy, but I'm sure Orbital saw it as the "least crazy" of three options: building a new engine from scratch, buying prohibitively expensive engines from Rocketdyne, or buying risky but cheap engines from Russia.

rst 12y ago
The design is a technical marvel, but it won't keep the metal from corroding, and this isn't the first sign of trouble -- a few months ago, a different AJ-26 blew up on a test stand. It isn't yet clear that the engine was at fault in the Antares fire (trouble in the tankage or fuel lines could have cascaded down, among other fun possibilities) -- but if the engine does turn out to be at fault, the remaining stock will certainly get more attention.
CmonDev 12y ago
What makes Rocketdyne engines less risky? Maybe they are both expensive and risky.

marze 12y ago

So these engines were built when Paul McCartney was 23 years old? Apparently Orbital Sciences scoured the museums and antique shops in Russia and picked up enough of these engines to base a new rocket on them. The reason they were surplus was that the rocket that they were built for was canceled, because the engines exploded on each of the four test launches conducted. Orbital Sciences was once a scrappy startup. Books were written about it. However, this is just embarrassing.

avmich 12y ago
Aerojet bought those engines from Russia - the original manufacturer - in 1990-s. No need for Orbital to scour museums. No, engines didn't explode. Were actually never flown before the first Antares lifted off. But performed pretty well on stands.
- marze 12y ago
  Are you were claiming that there were no failures of the NK-33 engine on any of the four N1 flights that exploded during first stage burn?
  - avmich 12y ago
    Yes. NK-33 never flew on N-1. They were intended to fly on N-1 in its fifth flight - which never happened, because Glushko closed the program after becoming the head of the Energiya company.
    - marze 12y ago
      So before the engines were obtained by Orbital, the NK-33 had never even been flight tested?
      - avmich 12y ago
        Before the engines flew for the first time, yes they never flew before. And Antares is the first rocket where those engines have flied - though not the only one rocket already.

diminoten 12y ago

For someone who doesn't know much about aerospace engineering, can someone outline the actual challenge with getting something like this right? What's the missing piece that'll effectively eliminate accidents like this? Or is this one of those things that will just happen from time to time because going to space takes a lot of math that we don't fully understand yet? I was just under the impression that this was a very well known area of science.

avmich 12y ago
I think this is fundamentally pretty hard problem. It involves control of turbulence, and this is a big in general unsolved problem today. That includes hypersonic flights, thermonuclear plasma control... Rocket engines work on the edge of material capabilities, and fluctuations - which you want to tame to guarantee things like that won't happen - are hard to deal with.
austinz 12y ago
With these particular engines, you also have the variable factor of them being stored for decades in a warehouse, and all the issues with metal corrosion, fatigue, degradation, etc. that come with that. I'm not a materials engineer, but I could imagine things like seasonal temperature changes, humidity, rough handling, changes in the metallurgical properties of the alloys they used over time could all affect the ability of the engine components to withstand high temperatures and pressures.

hindsightbias 12y ago

The NK-33 may be old, but it offers equal or superior performance to anything made in the US. Another engine, the RD-180 is used in the Atlas. US industries have the rights to manufacture, but want $800M and 4 years to build it or create a new engine. That tells you the state of US tech. http://aviationweek.com/defense/aerojet-rocketdyne-targets-25-million-pair-ar-1-engines http://aviationweek.com/defense/aerojet-rocketdyne-targets-2... Video on Russian/Ukranian engines: http://www.youtube.com/watch?v=TMbl_ofF3AM http://www.youtube.com/watch?v=TMbl_ofF3AM

joshAg 12y ago

To answer the question of why even use them at all, it's because on paper they still compare very well to the other options. Specific impulse is basically an efficiency measure, like mpg for cars. The higher it is, the more efficient the rocket engine is. SpaceX's Merlin 1d has a thrust-to-weight ratio (twr) of 150, and a specific impulse (I_sp) at sea level of 2.73 km/s. I doubt this was an option for orbital sciences to use in their rockets. The NK33, which is the soviet engine that was used, has a twr of 137 and an I_sp at sea level of 2.91 km/s. The atlas V rockets by boeing and lockheed use an rd180, also russian built btw, which has a twr of 78.44 and an I_sp at sea level of 3.05 km/s. The delta IV rockets by boeing us an rs68 or rs68a have a twr of 51.2 and an I_sp of 3.58 km/s. I'm not sure if these are available for commercial use, though, since boeing shut down the commercial delta IV launch program. The US military still uses them though, so they are still being produced. Also, these use liquid hydrogen instead of RP1 (rocket grade kerosine) like all the rest of the engines here. The Saturn V used five F1 engines (also 1960s tech, but no longer available). Each of those have a twr of 82.27 and an I_sp of 2.58 km/s.

jessriedel 12y ago
Isn't this much more a question of reliability and track record than 10% performance differences?
- avmich 12y ago
  Both reliability and performance. And also reliability is something which you estimate, calculate basing on some models - it's probability, not a guarantee. How would you estimate the reliability of an engine which run over 3 hours on a test stand? Of course you might prefer a well known fuels, industry-proven range of pressures and other engineering solutions, and you test them on Earth first - but after that you have to fly. Track records are tricky for rocket systems. Since the first Sputnik, over the 57 years the humanity had may be - roughly - 4000 launches of space rockets of all kinds, with many different fuels, engines, sizes etc. It's relatively rare when you have a good track record which is measured by many tens of flights - let alone hundreds. When a system flew 20 times it's already considered rather mature - double so if those were all successful flights. So it's quite usual to use a new engine in a new design of rocket - in fact it's rare when you have many different rockets with the same engine type or many different engines used on the same rocket. Rockets and engines are usually optimized to each other - particularly rockets are built around engines, since margins are quite thin.

ijk 12y ago

Also note that we don't know yet if the NK-33/AJ-26 engine was the cause of failure. Sometimes the most visible possibility wasn't the original cause of the problem.

njsteinsund 12y ago

I can highly recommend these two books for an engaging introduction to getting things right in the rocket business: Ignition! An informal history of liquid rocket propellants http://www.goodreads.com/book/show/677285.Ignition_ http://www.goodreads.com/book/show/677285.Ignition_ and Space Systems Failures: Disasters and Rescues of Satellites, Rocket and Space Probes http://www.goodreads.com/book/show/1015478.Space_Systems_Failures http://www.goodreads.com/book/show/1015478.Space_Systems_Fai...