Launch Pad Explosion during Static Test Fire - Sept 1 2016

[Grendal]

Which seems to be what we are hearing from SpaceX: that the anomaly was a procedural issue. That changed their tanking procedure for this launch and very sadly paid the penalty for it. I'm much happier they discovered it now than at some future time during a crew launch. The crew may well have survived the destruction but the PR disaster from something like that would easily be twice as damaging as this has been.

Now that it has been somewhat pinned down, I look forward to them moving forward and seeing a new launch in the coming months.

Is the solution then to slightly reduce the LOX temperature, reduce the pressure, both, or construct the COPV tanks differently?

Probably a little of both. Basically, it will be a timing issue from what I've read from people that seem to know what happened.

 

[Mike1080i]

Does that mean that solid/frozen O2 formed inside the the carbon over-wrap material? In other words, under pressure LOX was forced into the tank wall material where it then froze solid and in solid form is capable of spontaneously igniting the carbon material?

Yes, it would have formed between the carbon fiber and the Helium tank's outer wall. As the Helium tank pressure increased the tank expanded and compressed the now frozen O2, eventually reaching the point of auto-igniting the carbon fiber, as I understand it.

 

[jkn]

After watching

and several other videos from TechX:

LOX at its boiling point -183⁰C would not cause condensation of O2 from air. SpaceX employs LOX at a temperature of approximately –207°C. Cooling the LOX to this temperature point yields a density increase of around 8%. (Falcon 9 FT – Rockets)

LOX at 24°C below boiling point will cause condensation of O2 from air.

It will also liquefy gas inside PU foam cells. So gap will open between insulation and LOX pipe or inside foam.

Perhaps: First bang was caused by O2 condensated from air reacting with PU. It broke the LOX pipe. LOX filled gab inside foam. Then another ignition created mostly vertical fireball seen in the first video frame with fire. This broke the rocket creating mostly horizontal fireball seen expanding in following frames.

 

[Ben W]

LOX at its boiling point -183⁰C would not cause condensation of O2 from air. SpaceX employs LOX at a temperature of approximately –207°C. Cooling the LOX to this temperature point yields a density increase of around 8%. (Falcon 9 FT – Rockets)

LOX at 24°C below boiling point will cause condensation of O2 from air.

Nitrogen has a boiling point of -196°C. Are you saying that LOX at -207°C would not also cause the nitrogen to condense? Liquid air per se (77% N, 23% O) should not be a problem, but clearly the ratio is dependent on temperature.

Perhaps manufacturing the PU foam with neon gas as a filler (boiling point -247°C) would be one workaround? No idea if this is feasible.

 

[jkn]

At surface of LOX pipe T would be -207. At outer surface of insulator perhaps T=0⁰C. So there is a region with T between -183 and -196 ⁰C collecting pure LOX.

Air= 78% N2, 21% O2, 1% Ar

Periodic Table of Elements and Chemistry
Neon: Cost, pure: $33 per 100g
Helium: Cost, pure: $5.2 per 100g

Neon might not escape as easily as He, so perhaps it would be a solution. I guess it could be used. PU mixed with LOX explodes. So getting rid of PU would be better.

 

[Ben W]

At surface of LOX pipe T would be -207. At outer surface of insulator perhaps T=0⁰C. So there is a region with T between -183 and -196 ⁰C collecting pure LOX.

Air= 78% N2, 21% O2, 1% Ar

Periodic Table of Elements and Chemistry
Neon: Cost, pure: $33 per 100g
Helium: Cost, pure: $5.2 per 100g

Neon might not escape as easily as He, so perhaps it would be a solution. I guess it could be used. PU mixed with LOX explodes. So getting rid of PU would be better.

$33 per 100g may sound steep, but the density of neon gas at STP is just 0.9g / L, so $33 goes a long way. A cubic meter of PU would require about $300 worth of neon. (For comparison, the bulk cost of ordinary PU looks like about $100-$400 per cubic meter, but no doubt SpaceX uses "special" PU.) Any idea what the total volume of PU in the second stage is? Also agreed that getting rid of the PU entirely would be ideal, if this is in fact the problem.

 

[jkn]

Helium is so much cheaper, that I guess they would use it. Even though it leaks faster than neon. They do have helium system anyway.

I don't know if they use any PU in Falcon. LOX pipe should have inner insulation from something better, only outer layer cheap PU. Perhaps it has. Perhaps LOX somehow got into outer PU.

TechX has published their recommendations:

 

[miimura]

After 2016 rocket explosion, Elon Musk’s SpaceX looked seriously at sabotage

My favorite Elon quote:
"We found things that looked like bullet holes, and we calculated that someone with a high-powered rifle, if they had shot the rocket in the right location, the exact same thing would have happened.”

Second favorite quote:
Early indications were that something caused an upper-stage helium bottle to explode, and at the SpaceX test site in McGregor, Tex., engineers were trying to replicate the explosion. But “we were having a hard time blowing these bottles up,” she said. So, instead, they got a rifle, “and we shot it,” Shotwell said. “And the signature on the bottle was just like the signature on the bottle that we recovered. That was an easy test to do. It’s Texas, right, everybody’s got a gun and you can blow stuff up.”