Author Topic: On-Pad Explosions  (Read 24579 times)

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #100 on: 04/13/2017 05:28 PM »
A launch of a Zenit satellite from Plesetsk's LC-43/4 on December 3, 1971 failed when the core stage ingested loose debris and shut down almost immediately at liftoff. The booster continued to climb for about 20-30 seconds before breaking up, and according to a translated account of the launch, the weather that day was overcast with thick clouds, so it was soon out of visibility. One strap-on fell out of the clouds (literally) and impacted near the pad. The rest of the booster crashed a few miles away. The next launch from LC-43/4 took place three weeks later, so no serious damage occurred.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 30812
  • Cape Canaveral Spaceport
Re: On-Pad Explosions
« Reply #101 on: 04/13/2017 07:29 PM »

An R-7 ICBM test on July 10, 1958 failed when one strap-on shut down at liftoff and broke away from the stack. The strap-on fell onto the pad while the rest of the vehicle crashed nearby. According to some sources, it carried some of the avionics intended for the Luna 8K72 booster, and the failure was caused by high frequency combustion chamber vibration which would end up being a nagging problem over the next two years.


Are you going to cite your sources or just blatantly cut and paste from wikipedia and bypassing its sources?

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #102 on: 04/13/2017 08:42 PM »
Are you going to cite your sources or just blatantly cut and paste from wikipedia and bypassing its sources?

If you wanted that list of Plesetsk R-7 failures, it's right here.

http://forum.nasaspaceflight.com/index.php?topic=17286.235

Offline JAFO

  • Full Member
  • ****
  • Posts: 462
Re: On-Pad Explosions
« Reply #103 on: 04/13/2017 09:54 PM »
When you saw this one video, did you see the strap-on boosters stripping off moments before impact?

If so, that would be the 1982 Zenit accident. An old thread here from a couple years ago had identified that video clip as the '82 failure. Otherwise the April 2, 1969 Proton accident is a possibility.

In the case of the '88 accident, the core stage engines malfunctioned almost immediately at liftoff, but the booster computer system was blocked from sending a shutoff command until T+20 seconds so as to prevent a pad fallback. When the 20 second mark was reached, the command was unblocked and the core and strap-ons shut down. This turned out to not be enough time to get the booster away from the pad, which was severely damaged. According to a translated document I read, it seems like the core stage suffered a control rather than a propulsion system failure.

The same document also noted that the '87 accident would not have been survivable on a manned launch, and apparently necessitated improvements to the Soyuz SAS abort system (at least that's what I could discern from the translated text).
Video clip from the launch failure on 15.05.1982
http://www.kosmonavtika.com/lancements/1982/15051982/15051982photos.html

That's the one.

-Wonder what the uphill velocity was when the engines cut out was vs the downhill velocity at impact?

-Must of been scary as heck to be near the pad and see the rocket suspended in mid-air, juuuuuust before it started down.
Anyone can do the job when things are going right. In this business we play for keeps.
Ernest K. Gann

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #104 on: 04/14/2017 12:14 AM »
http://kik-sssr.ru/IP_4_Turatam_old_Razdel_1.htm

This site (use Google Translate) also has some nice bits of info about various R-7 failures. I had intended to post the link earlier but forgot.

http://cosmopark.ru/r7/prig8.htm

This one describes the 1962-63 Zenit accidents and the cause of them.
« Last Edit: 04/14/2017 12:21 AM by WallE »

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #105 on: 04/14/2017 03:54 AM »
A real on-pad explosion (Nicolas' website) http://www.kosmonavtika.com/lancements/1990/04101990/04101990photos.html
4.10.1990 pad 45/2 destroyed and never rebuilt

The Zenit booster got off to a bit of a rough start since the RD-171 engine was new technology and understandably had teething problems, however, it was the only survivor out of an ambitious series of next-generation Soviet LVs and manned spacecraft planned for the 1980s and beyond. Under original plans, it would have launched Zarya, the intended successor to Soyuz.

That 1990 disaster is easily one of the top ten rocket crashes of all time and the total devastation of the pad was largely due to a poor design that resulted in the blast wave having nowhere to disperse. If I remember correctly (and I could be wrong about this), there was only a small flame trench under the launch stand as opposed to the giant flame pit on R-7 pads.

Offline Alter Sachse

  • Full Member
  • **
  • Posts: 253
  • Germany (near Heidelberg)
Re: On-Pad Explosions
« Reply #106 on: 04/14/2017 04:09 PM »
http://kik-sssr.ru/IP_4_Turatam_old_Razdel_1.htm

3.12.1971:Block A failed T+0.5
rocket continued the flight and crashed at the village Karasye (7...8 km north of launch complex 43)

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #107 on: 04/24/2017 01:42 AM »
I think that a few years ago Joel Powell did an article on Atlas pad crumples. Not all of them went boom. In fact, I think he was more interested in the ones that crumpled but did not explode, because they were not recorded regularly.

Another such failure mode was Atlas bulkhead reversals. If the pressure levels in the propellant tanks got too high or too low, the bulkhead separating the propellant tanks would reverse/rupture/collapse, causing all of the LOX and RP-1 to mix, turn to gel, and explode instantaneously like a giant stick of dynamite.

This happened on Atlas 5C, 7D, 5D, and 81D. The first three resulted from loss of RP-1 tank pressure, the second two from overpressurization of the LOX tank. If you'd seen the video of 7D, the explosion happens really fast, almost instantaneously while on, say, Atlas Centaur AC-1, there's a much slower moving fireball with very little propellant mixing.

Kerolox has this unfortunate property not experienced with other propellant combinations. Hypergols burn on contact, reducing their explosive potential, while hydrolox, in the absence of an ignition source, will merely turn into water vapor if it mixes.
« Last Edit: 06/21/2017 12:11 PM by WallE »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 30812
  • Cape Canaveral Spaceport
Re: On-Pad Explosions
« Reply #108 on: 04/24/2017 02:50 PM »
Kerolox has this unfortunate property not experienced with other propellant combinations. Hypergols burn on contact, reducing their explosive potential, while hydrolox, in the absence of an ignition source, will merely turn into water vapor if it mixes.

Wrong, Kerolox and hydrolox have the same issue. 

if there is no ignition, then hydrolox remain H2 and O2.  Kerolox also requires an ignition source  and would remain O2 and RP-1 without one.  hydrolox has more flammability and lower ignition energy.

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #109 on: 04/25/2017 02:23 AM »
Kerolox also requires an ignition source  and would remain O2 and RP-1

Alright, that part is correct. LOX and LH2 should turn into gel when mixed because LOX will do that when it's mixed with any flammable liquid. Most importantly, it turns into a shock-sensitive gel. An ignition source isn't required to set it off, you just have to bump/strike/hit it.

In the absence of an ignition source, then no, the propellants shouldn't go off if they're not mixed. The Atlas 190D/GAMBIT pad crumple didn't result in a fire or explosion because they'd removed the LOX when the mishap occurred, so the RP-1 just spilled onto the launch stand. If the LOX had still been present... kaboom. It was mentioned earlier in this thread that Thor on-pad explosions were less destructive on average than Atlas ones because of the separate tanks that reduced the chance for the propellants to inertially mix.

Hypergols aren't as explosive simply because they burn on contact and can't mix first. This was a factor in the decision to not use an escape tower on Gemini--they reasoned that a Titan II couldn't explode with as much force as an Atlas or a Saturn.
« Last Edit: 04/25/2017 08:53 AM by WallE »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 30812
  • Cape Canaveral Spaceport
Re: On-Pad Explosions
« Reply #110 on: 04/25/2017 04:15 PM »

Alright, that part is correct. LOX and LH2 should turn into gel when mixed because LOX will do that when it's mixed with any flammable liquid. Most importantly, it turns into a shock-sensitive gel. An ignition source isn't required to set it off, you just have to bump/strike/hit it.


Wrong again.  The warmer LOX would make the LH2 turn into a gas.  There would be no gel.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 30812
  • Cape Canaveral Spaceport
Re: On-Pad Explosions
« Reply #111 on: 04/25/2017 04:16 PM »
If the LOX had still been present... kaboom.

Not true either.

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #112 on: 04/26/2017 03:27 AM »
Not true either.

It doesn't take much imagination to assume that an Atlas splitting open on the launch stand and releasing thousands of gallons of RP-1 and LOX won't result in at least some mixing of the propellants.

Take Atlas 71F. A little kerolox glob in the flame pit exploded and damaged the booster, resulting in eventual engine shutdown and loss of the mission. Now take a guess what an entire load of propellant can do.

Offline Stan-1967

  • Full Member
  • ****
  • Posts: 432
  • Denver, Colorado
Re: On-Pad Explosions
« Reply #113 on: 04/26/2017 04:07 AM »
Kerolox has this unfortunate property not experienced with other propellant combinations. Hypergols burn on contact, reducing their explosive potential, while hydrolox, in the absence of an ignition source, will merely turn into water vapor if it mixes.

Wrong, Kerolox and hydrolox have the same issue. 

if there is no ignition, then hydrolox remain H2 and O2.  Kerolox also requires an ignition source  and would remain O2 and RP-1 without one.  hydrolox has more flammability and lower ignition energy.

Are there any examples of pad failures where hydrolox or LOX + any hydrocarbon mixed together on the pad environment and failed to explode?   I'm recalling from some thread here on NSF the rocket scientist maxim that "ignition is free" when it comes to investigating explosion failures.

I can't think of any off the top of my head.  I recall various leaks of monoprops & hypergols making a mess on the pad, but not exploding.  I'm also not counting the buildup of gaseous hydrogen on the STS pad that was handled with the burnoff system prior to main engine ignition.

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #114 on: 04/26/2017 04:51 AM »
Are there any examples of pad failures where hydrolox or LOX + any hydrocarbon mixed together on the pad environment and failed to explode?   I'm recalling from some thread here on NSF the rocket scientist maxim that "ignition is free" when it comes to investigating explosion failures.

http://www.healthsafety.com/articles/man-lox-warning-video-contains-graphic-content

"LOX or liquid oxygen is not flammable on its own; however once you combine it with a flammable item, it can have explosive results. The boil off, or fog of oxygen that is produced can create oxygen enriched environments increasing the danger of fires and explosions. LOX has an expansion ratio of 1:861, meaning that one liter of liquid oxygen will expand to 861 liters of gaseous oxygen. LOX is considered a cryogenic compound, meaning that it instantly freezes biological material, like your skin, on contact. When LOX encounters petroleum based products, it can react violently causing explosions and fires. There have been reports of alcohol based hand sanitizers catching fire in oxygen-enriched environments at hospitals."

Jim is correct in his statement that LOX mixing with LH2 doesn't gel (still highly explosive). RP-1 sure is petroleum-based though, which is why it seems silly to think that an Atlas rupturing and dumping all of its propellants on the pad would not result in some very bad things happening.

Aside from Atlas 71F, the first Atlas D launch (Missile 3D) had a failure of the propellant fill and drain valves, causing the propellants to spill out and explode on the pad. It flew for 27 seconds until the booster section exploded and Range Safety blew it up, but as I understand, this was caused by unstable B-2 engine combustion due to the loss of LOX rather than the incident at liftoff.

I can't think of any off the top of my head.  I recall various leaks of monoprops & hypergols making a mess on the pad, but not exploding.  I'm also not counting the buildup of gaseous hydrogen on the STS pad that was handled with the burnoff system prior to main engine ignition.

When they were preparing Challenger for its first launch in early 1983, several leaks of liquid hydrogen ducting were discovered, which resulted in the SSMEs being removed and replaced. However, no fire or explosion resulted.

The attempted launch of Discovery in June 1984 failed because of a stuck SSME main fuel valve preventing the flow of LOX to the #3 engine. Liquid hydrogen pooled around the base of the thrust section and started a fire, but the water deluge system put it out. If the astronauts had tried to evacuate using the slidewire escape system, they would have plunged right into the invisible LH2 fire.
« Last Edit: 04/26/2017 08:06 AM by WallE »

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #115 on: 04/28/2017 04:51 PM »
Since we were on the topic of propellant mixing, Atlas 9C was particularly nasty because the propellant tanks lost pressure and the intermediate bulkhead collapsed, allowing all of the LOX and RP-1 to mix, gel, and explode with an estimated force of 60,000 pounds of TNT. Atlas AC-5, while it produced a huge blast, was actually much less destructive because the explosion started at the bottom of the booster. The RP-1 tank went off first and most of the propellant had ignited before the LOX tank blew, minimizing mixing. This would also apply to Samos 3 and several other pad fallbacks discussed in here.

Atlas 9C has to be one for the record books since there has never (to my knowledge) been another failure of a US launch vehicle that did this much damage to the pad. There have been Soviet failures that did more damage, but not American ones. The umbilical towers and the large service tower were completely flattened. The top portion of the service tower shown here weighed about a ton, and it was thrown a few hundred feet by the blast. As you can see, the concrete launch stand was also caved in. They didn't rebuild the service tower on LC-12; after restoration, the pad hosted the second and third Atlas-Able launches and ICBM tests before being converted for Atlas-Agena.

All because of one incorrectly installed piece of plumbing in the Atlas.
« Last Edit: 04/29/2017 11:01 AM by WallE »

Offline kevin-rf

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8408
  • Overlooking the path Mary's little Lamb took..
Re: On-Pad Explosions
« Reply #116 on: 04/28/2017 06:03 PM »
Atlas 9C has to be one for the record books since there has never (to my knowledge) been another failure of a US launch vehicle that did this much damage to the pad. .... The umbilical towers and the large service tower were completely flattened. The top portion of the service tower shown here weight about a ton, and it was thrown a few hundred feet by the blast. The concrete launch stand was also caved in.
I wonder how the SpaceX Amos 6 and Antares CRS-3 failures stack up. Both did significant damage.
Sharknado 5: Global Swarming
Make America Bait Again!

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #117 on: 04/28/2017 06:38 PM »
I wonder how the SpaceX Amos 6 and Antares CRS-3 failures stack up. Both did significant damage.

They didn't knock the umbilical tower over, and on CRS-3, the Range Safety Officer sent the destruct command about two seconds before pad impact.

The damage caused by AMOS 6 has been discussed at length in other threads, and earlier in this thread, it was pointed out that pad explosions in the 50s-60s were expected and planned for. It was pretty obvious that SpaceX didn't worry about such an event in this day and age, as exhibited by the fact that they had the satellite and upper stages on the booster, all of which were fully fueled, as a PFRF was being conducted. They didn't do this in the 50s-60s. Atlas 9C had an unfueled second stage, dummy third stage, and the probe was not on top of it.

Shown here is the PFRF for Wally Schirra's Atlas. No Mercury capsule on top. They didn't take any chances back then, especially since this firing had the added objective of making sure there wasn't a recurrence of a turbopump problem that had destroyed two Atlas Fs earlier in the year.

Offline WallE

  • Member
  • Posts: 94
Re: On-Pad Explosions
« Reply #118 on: 05/29/2017 05:13 PM »
The Thor was built with separate propellant tanks.  If it fell back onto the pad in a more or less vertical orientation, the vehicle descended below the pad deck through the center hole, hit the flame deflector and skidded along it, breaking neatly into two parts, one fuel and one oxidizer. One tank would break open below the deck, the other above. The propellants didn't inertially mix as thoroughly as was the case on other pads, such as Atlas.  While there was a still large conflagration, the blast effects were less than were seen on other stands.



Here's some high speed film of Atlas Centaur AC-5 from a couple different angles. The loss of booster engine thrust is clearly visible here and there's a LOX-rich explosion in the Quad III section at pad impact (probably ruptured low-pressure ducting) followed by a spreading thrust section fire and some LOX leakage out of the Quad I section. Also note how the vernier engine becomes intermittent post-impact. The final explosion begins just at the bottom of the RP-1 tank.

Also check out the photo here. The sustainer engine is badly burned, but the booster engines seem to have only impact damage, likely because they had shut off prior to impact/vehicle breakup.

Contrary to popular belief, damage to LC-36A was not that extensive, it was mostly fixed by June 1965.
« Last Edit: 05/29/2017 05:18 PM by WallE »

Offline rsnellenberger

  • Amateur wood butcher
  • Full Member
  • ****
  • Posts: 584
  • Houston, TX
Re: On-Pad Explosions
« Reply #119 on: 05/29/2017 05:29 PM »
Watching these early attempts to land an Atlas booster helps me appreciate the progress we've made...