Investigating the SpaceX Rocket Explosion of September 1, 2016

[liometopum]

As already posted on PF, and you have likely seen in the news, a SpaceX rocket exploded, September 1, 2016. Elon Musk is reaching out for help in finding out how it happened. http://www.csmonitor.com/Science/20...calls-on-public-government-in-explosion-probe

I took the video and placed it in this, to allow frame-by-frame examination:
http://rowvid.com/?v=_BgJEXQkjNQ
Just in case... you can set the speed, for example, at .25 by clicking that button. Use the < and > buttons to move one frame at a time.

You will see, that at 71.7 (seconds) all appears ok.
The next frame, at 71.74, only 4/100 of a second later, the explosion is well under way. That is part of the problem for them, as it happened so quickly.

I already have my idea of what might have happened, and sent it along to reports@spaceX.com, but I want to let everyone else look for themselves to see if they can deduce anything. Some of the smart people at PF might be able to see something that the SpaceX people have not noticed.

Have fun.

 

[Student100]

Did Musk also release detailed drawings of the Falcon 9 and minutes of the test? With just a video there's no way to come to any realistic conclusion that isn't pure conjecture.

 

[liometopum]

Musk is asking if anyone else has photos or videos, so conjecture away! They have all the technical info, and they are stuck and need help. The people here might make a difference for them, by seeing things differently.

Try it, for real. What can you deduce from the video? What do you see?

 

[jim hardy]

I don't know how to single frame Youtube, arrows only give me 1/4 second increments.
here's best i can get

 

[liometopum]

This is the 'raw' site:
You can't use YouTube to advance by frame. Use this link: http://rowvid.com/?v=_BgJEXQkjNQ
(You can use that site for other videos too. )
Or http://anilyzer.com/

With that said, you got the first frame of the explosion, or what I get, at least. But using rowvid.com is easy.

So, where do you think the explosion started?

 

[jim hardy]

Where'd it start ? well it's hard to say
My best guess follows.i find anilizer just typical recalcitrant software, a bundle of excuses to not do anything.

But using rowvid.com is easy.

well! What a pleasant surprise rowvid WORKS GREAT ! Sure beats trial and error on Youtube.

Okay that was first frame at 71.74
Second frame at 71.78

Seems to me a kerosene leak would have run down and set the whole rocket afire, slowly like charcoal lighter. Clearly this was a kerosene & LOX mix burning,

Whole 2nd stage got engulfed in a fireball and blew apart.
Would've taken failure of both tanks to let fuel and LOX mix
so
My guess is 2nd stage engine somehow got a start signal .
See my signature .....

what's your thoughts?
http://www.spacex.com/falcon9
http://www.spacex.com/sites/spacex/files/falcon_9_users_guide_rev_2.0.pdf

https://en.wikipedia.org/wiki/Falcon_9
Both stages use a pyrophoric mixture of triethylaluminum-triethylborane (TEA-TEB) as an engine ignitor.[51]
[see msds's linked below - old jim]

SpaceX uses multiple redundant flight computers in a fault-tolerant design. Each Merlin rocket engine is controlled by three voting computers, each of which has two physical processors that constantly check each other. The software runs on Linux and is written in C++.[52] For flexibility, commercial off-the-shelf parts and system-wide radiation-tolerant design are used instead of rad-hardened parts.

i've not yet found the tank that holds the ignitor fluid. See its MSDS 's here
http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=US&language=en&productNumber=257192&brand=ALDRICH&PageToGoToURL=http://www.sigmaaldrich.com/catalog/product/aldrich/257192?lang=en
http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=US&language=en&productNumber=257168&brand=ALDRICH&PageToGoToURL=http://www.sigmaaldrich.com/catalog/product/aldrich/257168?lang=en
Looks like it'd burn if it got loose. That's another possible single failure.

By the way - THANK YOU for the pointer about single-stepping at rowvid

old jim

 

[Nidum]

Just looking at the basic images may be of limited use .

More effective would be to find ways of processing the primary images to give a series of new images each covering a restricted range of light wavelengths .

Breaking down into primary colour images is certainly possible and that would be useful but ideally ways might be found of getting some IR images as well - even poor quality ones would be very useful indeed .

 

[Greg Bernhardt]

If a PF member could help that would be incredible! PF team assemble! :)

 

[RonL]

Using the four light rays in that first picture seems to show an X as in x marks the spot. pretty much where the steam or smoke was showing the frame before.

 

[TheQuietOne]

It looked like the fuel tank... I agree with Jim hardy

 

[Filip Larsen]

For what its worth, Scott Manley has posted a nice commented video analysis:

 

[.Scott]

Doing frame by frame, this is what I see:

50 to 71.73: Nothing of any significance.

71.77: In 40 msec or less, the "explosion" is already 17 feet high and wide. This is excluding illuminated sections of the booster that, on first glance, might appear to be part of the fire. Presuming that we started with a point source of ignition, this means that the illumination border has been moving at 8.5 feet in no more than 40msec or >210 feet per second or more.

71.81: In the next 40msec, the flames have continued to expand laterally, but the bottom of the flame has actually shortened! Moreover, a "cloud" created from cold temperatures along the bottom half of the booster is left undisturbed during this time - and for at least 10 frames that follow.

So at first, it would seem that what we are seeing in these first two frames is purely a flame propagation front. It would seem that an invisible and combustible mixture of gases was already there, hanging as a cloud in the air, when it became ignited.

But there's a problem. Flame propagation rates are measured in cm/sec, with our 210 feet per second being over 5300 cm/sec. Among the fastest propagation rates in air is a 38% mixture of hydrogen which tops out at roughly 480 cm/sec (http://www.comtherm.co.uk/CT-7a%20Fig.pdf). So, unsurprisingly, we are definitely working with something more potent than air.
Perhaps the chemists among us can come up with a mixture that will propagate at 6400 cm/sec. For comparison, the source above rates methane/air at about 70 cm/sec while this http://www.sciencedirect.com/science/article/pii/S0360544212005579 shows a oxy-methane mixture at about 295 cm/sec. Even assuming that hydrogen gets the same boost (haven't found a source on that), it would still fall short of our 6400 cm/sec. Besides, our gases should be on the cold side - slowing the burn rate down a bit.
So did something, perhaps turbulence or heat, enhanced the flame speed? Or perhaps the flame spread without incandescing for a few frames before becoming visible?

The flame propagation theory seems to have a problem.

Alternatively, both the the material that is burning and the oxidant could be ejecting laterally from the tank - carrying the flame with it. From what I can tell, this is exactly what is happening. The combustion started within the tank, created a lateral crack in the wall, and sprayed out.

 

[berkeman]

Musk is asking if anyone else has photos or videos, so conjecture away!

If a PF member could help that would be incredible! PF team assemble! :)

So the Mentors have been discussing this thread, and the thread should be fine as long as it sticks to the established facts and the science behind the rocket. The PF rules obviously apply, so please be like Jim in your posts, and show the science behind your speculations about what may have happened.

Jim -- could a false engine start signal have happened without it showing up on telemetry? There are a couple potential causes of a false engine start signal detection by the module, but hopefully SpaceX has done enough immunity testing to rule that out...

 

[liometopum]

As a side note on rowvid.com, I was puzzled by that domain name, so I looked into it.
From http://hackersome.com/p/CalumJEadie/rowvid-version-0
"RowVid was originally created by Calum Eadie, Andrew Ratomski and https://twitter.com/busterlj at an Entrepreneur First Hackathon, to help rowers analyse training and race videos.
Since then it has evolved into general purpose player for frame by frame and slow motion playback of any YouTube video."

It makes sense now. I assume they were rowers at Cambridge. See https://www.linkedin.com/in/calumjeadie

 

[Stephen Tashi]

It might be useful to have some links to videos of other similar sized explosions that had known causes.

 

[Tom.G]

Here is what it looks like with the intensity range restricted so the bottom half of the intensity range is compressed.
The contour lines are brightness/color contours.
After viewing the video forward, backward, and stop-frame, the ignition point appears to coincide with the crescent shaped area at the arrow head. This is partially supported by a subtle color change within the crescent area (less Blue).

Visually, from the video, the initial flame front seems to travel toward the 7 o'clock position but rapidly changes to about the 5 o'clock position. Based on the above image, the origin is 3 to 3.1 nose cone diameters from the the nose cone tip at the right edge of the rocket.

Viewing the video backwards, the fireball seems to converge to a point 3.3 diameters from the tip, at the center of the rocket body.

 

[Jonathan Scott]

Could this be a variant of the same failure as before? If a helium tank support strut broke because of buoyancy forces during loading of the oxygen, the excess pressure in the LOX tank could split the second stage, releasing both fuel and oxidiser and also generating enough heat to ignite the mixture.
However, in that case I'd have expected the tank pressure telemetry to have recorded enough information to make that clear.

 

[Jonathan Scott]

After thinking about it some more, it seems likely that something to do with the helium tank could be relevant, as a sudden failure would be able to rupture the second stage very abruptly in a way which would open up both the LOX and the RP-1 tanks. The fact that the LOX tank was only passing 70% full at the time of failure might be enough to rule out anything relating to buoyancy (but I don't know the details). If a strut broke as before, I would have expected a larger time interval between the initial problem and the rupture.

 

[Dotini]

Would a cosmic ray passing through the tanks be enough to cause concern?

 

[Nidum]

The oxygen tank on Apollo 13 exploded because a faulty electrical device in the tank sparked .

 

[mfb]

Would a cosmic ray passing through the tanks be enough to cause concern?

Every second, about 100 muons cross every square meter of the surface of Earth. No, those are completely irrelevant.

I guess SpaceX will monitor all future static fire tests and launches with a high-speed camera (maybe they do that already).

 

[liometopum]

Here is the failed Titan launch of 12 December 1959
https://en.wikipedia.org/wiki/List_of_Titan_launches
http://rowvid.com/?v=UBzigaTSPZY

 

[liometopum]

This compilation is nice. The first rocket, at 15.5 seconds, appears to explode from higher up than the Titan, in the previous post (which is included in this video compilation).
http://rowvid.com/?v=m6qJh9upqW8

You might notice that these explosions are very different from the SpaceX disaster.

Musk called it a "fast fire" http://www.cnbc.com/2016/09/02/elon...lode-but-instead-experienced-a-fast-fire.html

 

[jim hardy]

pieces flying away

Jim -- could a false engine start signal have happened without it showing up on telemetry? There are a couple potential causes of a false engine start signal detection by the module,

Anyone know how much delay there is in the telemetry ?



.

 

[liometopum]

Jim, that is not a piece flying away. Go to 71.94 and look at the point at the bottom left hand corner, and then click until about 72.54. You will see the point go to the top right corner. I think that is a jet taking off from a nearby airport.

 

[Jonathan Scott]

I think "fast fire" is misleading. I agree it's not detonation (that is, not as from high explosive), but rather subsonic (as for low explosive), but that's still an explosion, and yes it was clear that small bits were flying off.

I like the way the sound has been resynchronized on the Scott Manley commentary - that really helps convey what is happening.

I think it's interesting that although the initial flash of flame and smoke expands rapidly, the expansion is in a sideways direction and the rate of expansion seems to be rapidly decelerating. Also, the explosion is surprisingly symmetrical between left and right given that it seems to start on the right of the 2nd stage, although the left side of the expanding cloud appears to be darker (perhaps more fuel than oxidiser?). These factors suggest to me that the origin of the force which ejected the material was inside the second stage, and that the material then caught fire and exploded after it was ejected. That then suggests a failure of the helium pressure vessel, although presumably a more abrupt and total failure than in the previous case where a strut broke.

Edit: When I say "small bits" I'm referring for example to the bright point in the first frame just to the right of the strongback, not to the fast-moving dark specks which I assume are something small close to the camera.

 

[mfb]

There is a piece of the rocket going upwards, on the other hand, first visible at http://rowvid.com/?v=_BgJEXQkjNQ&t=72.1&s=1 and then in multiple images afterwards. It seems to be a rotating flat piece (it is not visible in some frames), with some flames visible beginning 72.58.

 

[Jonathan Scott]

I think that is a jet taking off from a nearby airport.

On that sort of time scale? I think it's a bug flying past the camera.

 

[liometopum]

Jonathon, yes, that probably is a bug, but it is not debris from the explosion.

Here is a blinking of the explosion in an attempt to locate where the first blast occurred. I used the last frame, before the fire, and the first frame of the fire.
I made a better video:


Original:


Use replay to repeat it.

 

[bsheikho]

Is it possible the fault occurred at another location, but the explosion occurs because that is the only area susceptible to exploding?

 

[CalcNerd]

I see a vent burst at 66.56 at the midway spot of the rocket, on the opposite side of the explosion. This would support the idea of a fuel leak. Also, it would mean the actual ignition may have happened earlier than the actual catastrophic explosion some 4-5 seconds later. It is a white puff jetting out from the lower midpoint seam between the first and second stage rockets.

 

[mfb]

SpaceX seems confident to have everything sorted out (and a working launch pad) until November:

SpaceX President Shotwell: We anticipate return to flight in November, meaning down for three months. Next flight from CCAFS, then to VAFB.

Source

CCAFS could be either the damaged launch pad or the new one they are currently preparing for Falcon Heavy. VAFB is an independent launch site in California.

Edit: More tweets. 1, 2

SpaceX's Shotwell: Nov return to flight is our best hope. We still haven't isolated the cause or whether its origin was rocket or ground.

SpaceX's Shotwell: We have been told that the Sept. 1 anomaly will not affect Falcon 9's insurance rates. So we expect no impact.

 

[liometopum]

SpaceX still does not know the cause. This article just appeared, from Nature World News, dated Sep 13, 2016 04:10 AM EDT

http://www.natureworldnews.com/arti...osion-most-difficult-complex-failure-ever.htm

 

[GeorgeDishman]

I wouldn't want to speculate on a cause but from the video, the centre of the fireball in the frame at 71.73 seems to coincide with vapour venting from a vertical tube alongside the craft, not internal. That venting is seen from time 49.86 but it already in progress, there is a fade of the video between 49.35 and 49.86 where the onset of the venting has been edited out. It looks as though there was a much lower rate prior to that. It's fairly obvious though so presumably must be part of the plan.

 

[Greg Bernhardt]

What I want to know is why there aren't super high res cameras covering the launch pad? All I've seen is grainy photos and video.