Ariane rocket suffers from rare launch anomaly | Hacker News Reader

Ariane rocket suffers from rare launch anomaly

bbc.com

66 pointsgeertj46 comments

While not a huge deal in that both satellites seem to have reached an orbit (even if not exactly the desired orbit), it's worth noting that NASA picked the Ariane 5 as the launch platform for the JWST at least partially because of its extreme reliability. It would be a damn shame to lose an $8.8B instrument in development for over 23 years due to a launcher malfunction.

This is "kind of good news" in that regard: It's extremely rare for the same failure to occur twice. You can't prove there are no bugs (and Ariane 5 may have been just too lucky), but you can prove to a certain extent past problems are avoided. This is why planes are so safe nowadays.

But there were two problems, not just one. 1. The communications link broke; 2. The satellite was delivered in the wrong orbit.

To clarify, the parent poster isn't affected by the gambler's fallacy; rather, they're pointing out that it's common practice for any accidents involving spacecraft to be thoroughly scrutinized to ensure that the failure mode is completely understood and can be avoided in the future.

Yes, that's exactly what I meant, thanks.

From what I can see on the NASA website, JWST is "planning to launch between March and June 2019". I wonder if that will be delayed - I imagine they will want to track this issue down quite thoroughly and reach high confidence it won't happen again.

I've always felt that when building these incredibly valuable satellites one should consider building 2. I know JWST is probably not a good example with the work required to make the mirror but but in a lot of cases one could probably replicate a rare satellite during production for much less than 2x the cost. Then you have a flight backup, and alternatively you can later double your science with another launch if everything goes smoothly on launch one.

As said probably not a reality for JWST but these high risk launches always concern me with no backup option. What a tragedy it would be.

Satellite insurance is probably cheaper (though rates may rise?)

https://en.wikipedia.org/wiki/Satellite_insurance

http://spacenews.com/space-insurers-warn-that-current-low-ra...

You can’t really insure a scientific career.

Let’s say you work on the design and implementation of a novel scientific instrument for a satellite for over ten years. That’s a job that, relatively, isn’t paid well, but if launched, the instrument will get you papers and recognition.

If the rocket launching your toy blows up, all you can get from an insurance is money, and that’s not something you value that much in life.

Some satellites also have fairly tight launch windows, for example if you want to intercept a comet.

> Let’s say you work on the design and implementation of a novel scientific instrument for a satellite for over ten years. That’s a job that, relatively, isn’t paid well, but if launched, the instrument will get you papers and recognition.

That's true, but building two of everything causes other problems. For a principal investigator (PI), they spend years drumming up support for a concept then hope that it continues to get funded, built, launched, and becomes operational. If we build 2 of everything, let's say it adds 2/3 to the cost. That means less money for other missions overall. That would make many scientists unhappy.

Government payloads are not insured. The government is "self-insured."

This happens sometimes. For example, GPS and DMSP (Defense Meteorological Satellite Program) had many blocks of identical satellites. One of the DMSP spacecraft will never launch and has been ordered to be scrapped, since it is now so out-of-date after being in storage for so long.

https://en.wikipedia.org/wiki/Defense_Meteorological_Satelli...

http://space.skyrocket.de/doc_sdat/dmsp-5d3.htm

The out-of-date aspect probably explains why this doesn't happen more. I worked on JPSS-1, a polar orbiting weather satellite. It was a follow on to NPP, but it was not an exact copy. After having experience using NPP in production (in space), naturally improvements and better technology could be integrated. These two satellite launches were separated by 6 years and a new satellite takes 4-5 years to design, build, and integrate. Think how much technology changes in 6 years.

Another instance of this:

https://en.wikipedia.org/wiki/2012_National_Reconnaissance_O...

"Hey, we've got a pair of better-than-Hubble telescopes just sitting around. Want 'em? You have to promise not to look at Earth with them, though."

Did you watch the movie Contact?

"Why build one when you can build two for twice the cost?"

So you'll often build two or three test versions of the sat that you subject to serious, and sometimes destructive, testing to simuluate launch stresses, etc. So if a given satellite blows up on the pad, they can use these test articles as a backup. So they don't really start from scratch..

Good idea. Or even just make mutiples of the rare/expensive/long-lead-time components, optimize for a second build rather than actually build two.

Better still, use a swarm of cheap unreliable satellites for the same function (often not possible)

There is a docking ring on JWST, so it could be reboosted to a proper orbit if there's a problem with the launch.

A Dragon spacecraft (which will launch uncrewed with a docking capability sometime this year) or just some satellite bus with an NDS docking ring, should be capable of boosting it if there's a minor performance shortfall like this launch. $150 million, I would guess, would be the cost to address a minor performance shortfall like this. That's not too bad since JWST is a $10 billion spacecraft.

It should be pointed out that the electric-propulsion-equipped spacecraft actually launched will have no trouble reaching the correct orbit, it'll just take 4 weeks longer than planned.

The performance shortfall, based on guesses of the electric propulsion performance and the 4 weeks delay, is probably around 50 to 100m/s. JWST has a budget of around 150m/s, so JWST would, in principle, be capable of moving itself to its planned orbit, but this would signficantly impact its on-orbit lifetime (or require fancy mission planning... I'm always impressed by how clever mission planners can squeeze water from a rock when a problem like this occurs).

So JWST itself may actually have enough propellant in case this problem occurred again, and there is a docking ring on JWST that could, in principle, be used for reboosting it.

It should be pointed out that there was a similar problem with an AtlasV-Cygnus launch in 2016, but the payload was going to a low-energy orbit, so the rocket happened to have enough performance margin that it didn't affect the mission. But if the payload had used the full capacity of Atlas V (like this Ariane 5 launch), it would've been a similar performance shortfall.

Bottom line: no launch vehicle is perfect, not even Ariane 5 or Atlas V. But payload can make up minor performance shortfalls.

Note: based on the actually achieved apogee and perigee and inclination, it seems that the performance shortfall may be closer to 300 to 400m/s, which would make rescue much harder, except if you used electric propulsion. Luckily, the spacecraft have electric propulsion (but JWST doesn't...).

> There is a docking ring on JWST, so it could be reboosted to a proper orbit if there's a problem with the launch.

Is the JWST designed to accommodate thrust along the axis of the docking ring?

Apply less thrust. It survived assembly, transport and launch.

Everything costs money to put into space. You don't just attach a big heavy docking ring with scotch tape. You put is somewhere that material and strength can be useful. With the safety factors involved (wouldn't want to brick a billion dollar spacecraft because you came in a few mm/sec too hot) it makes the most sense to put the docking ring somewhere structurally significant.

It's not just a matter of how much thrust... It's also the direction of the thrust relative to the center of mass of the spacecraft... (it doesn't do you any good if the thrust you apply just induces a spin)

Interestingly, James Webb's sunshield is so ridiculously big, that it could actually be used as a solar sail.

JWST is 6200kg roughly. The sunshield is about 150 square meters. Sunlight is roughly 1360W/m^2, so given Einsteins relation between momentum and energy for light: P = E/c, and increase that by a factor of 2 due to reflection (as opposed to full absorption, which would be 1), then we have 9e-6 Newtons per square meter in sunshine near 1AU, for a total of 0.00135 Newtons over the whole thing. Over 5 years, that's 36 m/s of delta-v.

That could be enough to rescue the mission due to a small performance shortfall, if you can point this carefully in the right direction.

(And 5 years is a long time, but not unheard of if you're trying to recover from a major mission failure, for instance the Japanese Venus probe which failed its 2010 Venus orbit insertion and so had to go around the sun a bunch of times and successfully inserted on the next try in 2015 using the attitude control thrusters.)

The issue, IIRC, for JWST is that it's IR optics need to be cryogenically cooled, and even in the vacuum of space that coolant will gradually boil away. Hence why JWST will have a shorter lifetime than Hubble (L2 orbit, station keeping, and serviceability notwithstanding).

JWST doesn't rely on boiling coolant. It has an on-board cryocooler.

Space twitter suggests that there was no malfunction with the rocket, there was some process (read: human) failure that caused the rocket to be given the wrong instructions.

Further the satellite is going to have to burn years worth of fuel to get to where it's supposed to be.

You gotta appreciate the guys who have to solve issues like why this rocket's second state lost communications. The actual hardware with the problem gets dropped into the sea and lost forever. All they have to go on is whatever telemetry they did receive from the second state, if they received any at all. Good luck figuring out the root cause of that one. Maybe they have some pre-flight test data or pictures taken of the hardware before launch they can look at for anything funny?

One advantage is that it's a relatively self contained system where you can understand what each component does and under want conditions it operates. Together with today's telemetry you have a lot of data to understand the issue.

Ironically SES-14 was originally slated for Falcon 9

https://www.reddit.com/r/spacex/comments/6wimhc/ses_switches...

After all, there doesn't appear to be much of an anomaly. Satellites are in nominal or near nominal orbit.

Not quite, they are a ways off from a nominal orbit, and there is still question if Yah-3 will be able to make it to it's correct orbit with fuel to spare. (SES-14 has electric propulsion so it's not much of a worry for them except for lost time).

At best, the Ariane 5 is causing these 2 sats a month of delays, and at worst it could render one of them useless. Just because there weren't any explosions or spectacular looking failures doesn't mean this isn't a big deal.

Edit: my apologies, it seems the article has been updated and Orbital ATK has stated that they are confident that the sat will be able to reach the right orbit. (it previously stated that Orbital ATK had not been able to say either way yet).

Still, loss of comms and an incorrect orbit insertion could prove to be disastrous for many payloads. With ArianeSpace's track record, I have no doubt in my mind that many steps will be taken to prevent this exact failure from ever happening again.

Try using recycled rockets like SpaceX. It seems to be working with a reduction in cost. https://latechnews.org/spacex-launches-bulgariasat-1-recycle...

Ariane currently has a better reliability record than SpaceX. Cost isn't everything - saving $100M on a $10B satellite like the JWST (which will use an Ariane) doesn't do you any good if the $10B evaporates in a launch failure.

As of now, Ariane has 82 successful flights and a partial failure. If the plans of SpaceX pan out, they will have about 60 successful flights by the end of 2018, with no partial failures, so it becomes harder to compare. They could even reach parity with Ariane by the time JWST launches.

JWST is a payload nobody wants to be responsible for blowing up, I don't think SpaceX necessarily wants it on their roster. Even a 1% chance of failure could be too much for what would be a public confidence disaster, likely to cut off significant govt funding and hamper long term prospects.

SpaceX has lost two payloads, CRS-7 and Amos-6. Ariane's record still looks pretty good, even for SpaceX fans like myself.

Ariane 5 had two failures and 3 partial failures, I'm only referring the the most recent streak of successes which Ariane just broke, and comparing with the same from SpaceX.

Amos-6 was an interesting one only because it blew up in the ground days before launch. Ultimately, it was SpaceX's responsibility, but does it count as a "launch failure"?

For those curious: CRS-7 was a resupply mission for the ISS, had a bunch of miscellaneous cargo on it (a bunch of food, experiments, etc.) Nothing too important as far as I can tell. 99% American/European/Japanese supplies. And a single thing for the Russians: a wrench :)

Amos-6 was an Israeli communications satellite that would've provided satellite internet over Europe and Africa.

So what you're saying is that the failed mission had a wrench thrown into it by the Russians? :)

I'm a EU citizen but I'm not surprised by this at all.

The "country equality" policy implemented for contract attribution on the Ariane 5 project by the EU is the culprit.

Once again prioritising feelings over efficiency or common sense.

Is it also the "culprit" for the 82 consecutive problem-free launches?

Sorry, I got confused with Ariane 6.

Ah, so the obvious "culprit" of this launch failure is... something you don't like that hasn't happened yet? Who's prioritizing feelings over common sense, again?

Yeah a rocket that is way, way over budget and late but hey looking at the facts is difficult.

You don't get to claim an Ariane 5 failed because of an Ariane 6 contracting/funding provision and lecture someone on "looking at the facts".

Side note: there haven't been many rockets that weren't late and over budget. It's hardly unique to the Ariane 6.

Dude Arianne 5 has one of the best success rates of any launch vehicle ever made, I think only soyuz is better. It's ridiculously expensive by today's standards, but that's thanks to Musk's moonshot project succeeding. I'm glad both exists, they're good at different jobs.

Reply on news.ycombinator.com