SpaceX's Falcon 1 Destroyed During Maiden Voyage

legolas writes "SpaceX's Falcon 1 is the world's first privately funded satellite launch vehicle. After a successful static engine test on Wednesday, it was launched today. Unfortunately, the rocket was destroyed shortly after launch."

Comments

Early days

[SteelFist]

Sounds a bit like the early days of our space program.

Impact or RSO destroyed?

[sconeu]

Anyone know if it crashed or the RSO destroyed it?

Of course, it's never a good thing when your downward-pointing cam shows sky and clouds - spinning...

Re:Impact or RSO destroyed?

[nacnud75]

I think the RSO terminated the thrust due to excessive rolling (there is no 'self destruct' on the faclon 1, then engin just awitches off)

I had wondered...

[Shadow Wrought]

I thought it a bit odd that the static test was for only three seconds and took place the day before the launch. I would not be surpised if the accident was a by-product of them pushing their schedule.

Re:I had wondered...

[Intron]

Just as bizarre was that they had a payload on their first launch attempt. That's like trying to ship the Beta! That would never happen in software.

Re:I had wondered...

[roystgnr]

Just as bizarre was that they had a payload on their first launch attempt.

The payload cost hundreds of thousands of dollars, true, but lifting it into orbit costs millions. As long as they had a better than 10% chance of success, it was a good risk to take.

Re:I had wondered...

[david.given]

Just as bizarre was that they had a payload on their first launch attempt.

The payload massed 20kg (the Falcon could have lifted about 700kg) and was built by Air Force Academy cadets. I suspect it was being launched, er, would have been launched for free. After all, you have to test rockets with something, and you may as well launch something useful rather than a dumb telemetry package.

Re:I had wondered...

[interiot]

Well, the payload was a microsatellite, so the actual hardware didn't cost a huge amount really (~$100,000?). Anybody have any idea how much the USAF Academy paid for the launch?

Also, a large part of satellite cost is in the R&D, so if there are further funds, then building a duplicate would cost a fair bit less than the first one, right?

Re:I had wondered...

[shawb]

TFA put the satellite at $750,000. A lot of money, but not an insane amount as compared to the cost of actually putting the rocket up.

Re:I had wondered...

[iamlucky13]

Since the payload was a student project, it has no doubt already accomplished it's primary mission: to give engineering and science students experience in a large-scale, real world project designing and building a satellite and it's experiments. It's sad that they won't get to see the final fruits of their labors and the product of their effort was destroyed, but this doesn't really affect their overall education. The science loss is pretty small, as I'm pretty certain other satellites have studied similar phenomena in the past.

I think the Air Force giving SpaceX a launch contract was partially throwing them a bone to help get another launch provider off the ground (no pun intended), and partially saving money. No doubt had SpaceX not happened to be up-and-coming as they are, this would have gone up on a Pegasus or piggybacked with another satellite on a bigger rocket, like I believe the first Falcon-Sat was.

NASA's first failed attempts at orbit also had payloads on board.

Re:I had wondered...

[iamlucky13]

The on pad static fire was just a chance to confirm that all the systems worked together. It was not a proving for the engines or the flight software (which has to be exposed to a myriad of conditions you can't replicate during a static hold-down fire). In fact, firing the tanks dry with the rocket still on the ground receiving back blast probably wouldn't be a good idea. Over the past two years, however, SpaceX has tested the engines thoroughly on stands in the desert, logging several times what it takes to reach orbit on single engines. As Elon has said, they are pretty confident in the capability and reliability of the engines. I think faulting the schedule, especially when they've already been willing to delay a couple times for relatively small issues, is premature and a little unfair.

They have also thoroughly simulated the flight software, I believe with the hardware hooked up under simulated loads, as well. Of course, it's impossible to truly predict every contingency that the software will have to deal with, and given that the rocket began to exhibit uncontrolled roll rather than loss of power or anything like that, I suspect the problem does ultimately lie in the software rather than the power plant. We will have to wait for them to discuss their analysis to find out. I understand they have a relatively small code base, so hopefully they will be able to track it down quickly.

One other possibility I think fairly likely is vulnerability of their communications inside the rocket. Supposedly this is the first rocket to rely principly on ethernet, which reduces cost significantly over propriety methods. This is untested in flight, and interference or vibration may have caused problems.

I'm pretty bummed out by this, but their progress in the last couple of years is still impressive, and I'm looking forward to their eventual announcement of a second launch date. I wonder if it was a non-issue the recovery ship was out of position...or a good thing they moved it.

Re:I had wondered...

[DerekLyons]

I thought it a bit odd that the static test was for only three seconds and took place the day before the launch. I would not be surpised if the accident was a by-product of them pushing their schedule.

Pushing the schedule? This launch is eighteen months behind schedule.

"partially reusable rocket"

[ds_job]

That will be reusable as in "We are all reused parts of supernovae" or "We all have a billion atoms of Julius Caesar's body incorporated into our own" and not "Just pick it up , dust it off and we're ready to go again!"

Re:"partially reusable rocket"

[roystgnr]

Actually it is the "partially" in the sense that the first stage should parachute and be reused, but the second stage is space junk.

This is the right way to do it, too. Lower stages are larger, so you can save more money making them reusable. They're less important for overall rocket performance, so when making them reusable reduces their performance it's not so bad. They don't reach orbital speeds, so you can recover them without reentry shields or even without flyback capability. If we're going to move toward reusable rockets (which could be a very good idea) at a gradual pace (which the Shuttle program has proved is a good idea), the way to start is from the bottom up.

Guidance?

[TopSpin]

Watching the webcast it looked to me like the vehicle had a guidance problem; the on-board view seemed to be spinning. The feed didn't really provide enough to tell, however.

It definitely cleared the pad and I think it got to a few thousand feet.

Re:Guidance?

[twostar]

It also looked like it went unstable with a wobble. Spaceflight Now is reporting that the flame looked funny right before video was cut but this could be due to a weird orientation of the vehicle relative to the flight path. (ie sideways)

Hopefully we get more info soon and Elon flies the next one as soon as they figure it out. Take a page out of NASA's early history and just keep putting them up until you get it right. Luckily at $6 million a pop they're pretty reasonably priced compaired to other vehicles out there.

On the bright side...

[moochfish]

Now it's the world's first privately funded satellite crater.

Woohoo!

[MyLongNickName]

I've got two tickets for the maiden flight of Falcon 2! I guess this means I should get my ride soon, huh?

Re:Woohoo!

[MyLongNickName]

Hey! My wife bought the tickets for me. For some reason, though, she isn't able to come with me that day. Said something about this being less expensive than a lawyer. Not sure what she meannt by that.... but I digress.

How could I possibly return the tickets?

This isn't...

[Tackhead]

C'mon, Elon! This isn't rocket sci... oh, right. Well then, there we are.

And to add insult to injury, we'll link your web server from Slashdot.

Seriously, Elon. Good on you. SpaceX is doing something risky and interesting. Make as many mistakes as it takes to get the job done. Unlike NASA, the bulk of your funding comes from a free market, and you're therefore motivated to learn from your mistakes. The day you build something your investors are willing to let you slap a "man-rated" label on, I'll be in line with tickets to fly on it.

Re:This isn't...

[Anonymous Coward]

Wait..you're saying private investors are more likely to have safety concerns?

*boggle*

I mean, we're talking about the same sort of folks who, in other industries, constantly push companies to release products as early as they can in order to start realizing a profit. Go look at the drug industry. As private companies have increasingly gained influence over the FDA through lobbyists, the number of things slipping through has increased. Private companies cannot be relied upon to have the best interests of anything but their own pocketbook.

More details of the flight

[nacnud75]

Live video was shown of the vehicle's ascent from an onboard downward pointing camera. Within a few seconds the feed started to become intermittent. The small amount of imagery available showed a bright yellow glow protruding away from the normal exhaust pattern, as the rocket began to roll violently. The ascent profile also appeared to be more horizontal than what would be expected for that stage of the ascent. The video then cut out completely - with SpaceX confirming the rocket had been lost just moments later. - http://www.nasaspaceflight.com/content/?id=4394

Re:First flight with a paying customer?!

[hitchhikerjim]

$6 Million rocket. $800,000 payload. The cost of the payload is pretty small, all things considered. It's worth the risk to go ahead and fly the payload the first time. Saves you $5 million if it works, and cost you less than $1 million if it fails. ...and when you add in that everything's going to be insured, it makes finantial sense.

Space Aint Cheap

[Ash Vince]

Its interesting to compare this with the scram jet trials currently scheduled by Qinetic (British Defence Contractor thats just been privatised)

Qinetic are about to test fire a £1 million scramjet directly into the ground. If it works it will fire for 6 seconds before it hits earth at mach 7.

The problem with seeking venture capital is the the investors usually want a return of their investment within a specified (Probably too short) time frame.

Successful space exploration takes man decades not man hours.

Plume

[everphilski]

There was a plume coming out the side of the rocket in the last few frames of the SpaceX feed, normal to the body of the rocket - not the direction of flight. Most likely due to an engine/turbopump failure. This could possibly cause adverse roll/pitching. It looks like a physical problem; I doubt it was a guidance problem.

Re:Anyone know?

[iamlucky13]

"...before range safety got to it?"

That reminds me of an interesting talk I attended by an X-ray astro-physicist back in college. He had been involved in several launches. Not surprisingly, they are very personally invested in the payloads, since they spend quite a few years fighting for budgeting and designing and building, and plan to spend several more years analyzing data. He said there was one launch where the rocket went off course and the Range Safety Officer gave the order to blow it, but the lead scientist jumped on the guy in charge of the button in a rather desperate attempt to save his project (which was doomed anyways). Since then, the customers have been kept in a seperate room from the RSO's.

Smells of a tall tale, but probably based on fact.

Video URL

[antdude]

2.3 MB WMV.

Re:Crash and Burn Testing

[MurphyZero]

Shuttle NEVER flies flawlessly. It survives due to redundancy, the efforts of the people working on it, particular the foresight of some engineers, and in no small part, luck. When it fails, it fails due to lack of redundancy, a failure to be creative enough to foresee the failure mode, and an unforgiving environment.

So Elon was absolutely right but the true comparison is with software that may not be perfect but must at least handle problems gracefully (particularly with manned spaceflight) so that maybe the mission is degraded, but not finished. How do you get there? Shuttle still hasn't figured it out, so Elon can't really be faulted for a failure on the first try. He might even survive a second failure. Third time would be a death knell to commercial activity, even if he wanted to try further. It was mentioned elsewhere in the discussions that he'd stop if he got three failures and no successes--it'd be appropriate.

Our office (one of our jobs is to estimate rocket failure probability) pegged the likelihood of failure at 70%, so we weren't surprised. We were hoping he'd succeed, just realistic. Hopefully they'll learn from this one and succeed on the next one, but if you have an even money bet on his next launch, take failure.

Ariane 5

[JonBuck]

Consider for a moment the failure of Ariane 5's maiden flight in 1996.

Aboard Ariane 5 was Cluster -- a $500 million set of four identical scientific satellites that were designed to to establish precisely how the Earth's magnetic field interacts with solar winds.

The unmanned rocket was on its first voyage after years of intense development by some of Europe's leading scientists. The explosion was a setback for Arianespace, whose previous models of the Ariane rocket had been some of the most reliable vehicles for satellite launches.

The European Space Agency estimated that total development of Ariane 5 cost more than $8 billion.

Maiden flights are perilous things. They got a full minute of flight data that they didn't have before. I'm sure the next one will be a success.

A launch a day keeps the high costs away

[Baldrson]

One of the things that John Carmack does correctly is lots of small flights with the possibility to scale upon success. John Walker wrote a paper about this approach (restricted to expendables) called "A Rocket a Day Keeps the High Costs Away". It's good advice. It's too bad more people (to be fair, such as John Walker himself) don't take it to heart.

Re:Crash and Burn Testing

[O2H2]

As a designer of very successful rockts I can tell you why making a rocket fly is much harder than making darn near any other machine function properly...you are trying to harness enormous energies which do more than just push the rocket upward. The vibration and shock environments are beyond anything you probably have experienced. A simple connector can see upwards of 300 Grms without even trying. You can reduce polymeric materials to a puddle just with hysteretic heating from vibration. And you cannot simulate and predict everything. Weird system interactions are par for the course. You can only get first flight success with a lot of painful experience. SpaceX do not have this level of experience.

That is why demonstrated reliability cannot be replaced by calculation. Spacex bragged about their high reliability but it is all on paper. Successful rockets have tens of thousands of hours of debugging of problems built into them. You just never see it. Nothing can replace hours in the air. And they come slowly and at great expense.

Elon is now going to learn firsthand why spaceflight is so damn expensive. It is not the lack of innovation or intelligence at Lockheed Martin or Boeing- it is the brutal reality that nature imposes on lack of attention to detail and ignorance. It ain't the metal in the rocket - its the know-how in the people. We have to dig down to root cause on even the most innocuous anomaly - hence we know a lot more about flaws in parts than damn near anybody on the planet. But this knowledge is pricey.

Re:here's a hint

[penguin-collective]

Actually, "underwater ships" already existed by the time Jules Verne wrote his story.

As for traveling to the moon, that's just not comparable. The physics for going to the moon were well understood and within reach; that was just a question of technology and engineering.

For manned interplanetary and interstellar travel, it's not so much that we can make a reasoned argument against it, we don't even have a hint of the physics needed to make it work; current reactor, propulsion, and shielding technologies are many orders of magnitude away from what they would need to be for manned travel. And the technology being developed by SpaceX is completely irrelevant; it's a commercial launch vehicle, and an inefficient one at that--it has nothing to do with interplanetary or interstellar travel.

It's a different thing for unmanned interstellar travel: technologically, if we devote enough resources to it, we can probably send a small interstellar probe to a neighboring star system within the next century--it would be hugely expensive, but feasible.

Actually, I think the most likely path to manned space exploration is to reengineer people: radiation hardening, hibernation, vacuum resistance, and changes to the skeletal system, among others. If you do that well, you could send people in small pods and they might be able to work when they arrive. But I give it a century before people overcome their squeamishness to permit genetic engineering with people, and another century to do it. But you and I are never going to set foot on another planet.

Re:here's a hint

[Kjella]

For manned interplanetary and interstellar travel, it's not so much that we can make a reasoned argument against it, we don't even have a hint of the physics needed to make it work; current reactor, propulsion, and shielding technologies are many orders of magnitude away from what they would need to be for manned travel.

Say again? Interplanetary travel is quite well understood. It'd take some months but hardly out of reach. Now interstellar is a completely different ballgame. The solar system (diameter of Pluto's orbit) is about 80 AU wide, the nearest sun is 272000 AU away.

It's a different thing for unmanned interstellar travel: technologically, if we devote enough resources to it, we can probably send a small interstellar probe to a neighboring star system within the next century--it would be hugely expensive, but feasible.

As in arrive in the next century? Nope. With current tech we're talking about 75000 years or so. Even the most theoretical scenarios I've seen using ungodly amounts of antimatter as fuel takes about 20 years.

Actually, I think the most likely path to manned space exploration is to reengineer people: radiation hardening, hibernation, vacuum resistance, and changes to the skeletal system, among others. If you do that well, you could send people in small pods and they might be able to work when they arrive. But I give it a century before people overcome their squeamishness to permit genetic engineering with people, and another century to do it. But you and I are never going to set foot on another planet.

Interplanetary I don't see any reason why we couldn't do today. As for interstellar, I think it's far more likely we'll not actually send humans per se. Even with all the genetric modifications you suggest, sending humans is horribly inefficient. I think we'd send fertilized eggs and artificial wombs, or even just a DNA sequencer to do it on-site.