sevenperforce

That's what I said -- "you have to mix it with hydrocarbons to make it more like TNT", in terms of explosive yield. Just a small percentage of fuel by weight will double the yield. Detonation velocity in even the most perfectly-mixed grain dust explosions is under Mach 5 so I suspect all of the AN+whatever detonated well before the blast hit the silo. Pixel counting and a back-calculation of yield from the seismic data puts the fireball size at a radius of 80-90 meters. Since it appeared completely within a single frame, the minimum detonation velocity we are dealing with is Mach 7 (80 meters / 1/30 sec = 2400 m/s = Mach 6.99). Update -- according to this BBC article, people from the University of Sheffield estimate it at "about one tenth" of Little Boy, so definitely at least a kiloton. https://www.bbc.com/news/world-middle-east-53670839

He used both ANNM and AN+diesel; I was referring to the latter above. 70% of his bomb was ANNM; 30% was AN+diesel.

Those silos stored grain, not fertilizer. I believe they were also Lebanon's main grain storage facility which also sucks for the country right now. It is not clear whether the fireworks were stored in the same warehouse as the AN or an adjacent one. The AN was not in silos at all but was in one of the low warehouses. Don't know if it was in barrels or what. This was already touched on, but the AN is the fertilizer. You have to mix it with hydrocarbons to make it more like TNT or semtex. Mixing with fuel oil gets it up to around 70% TNT equivalent. The correct proportion of diesel (which is I believe what these types of ships use) can get it up to 78%; that's what McVeigh used in the OK bombing. Because of the way it went up, my money is on the entire warehouse catching on fire until the ambient temperature reached autoignition for the AN/ANFO mix. I doubt it was a primary-secondary at all. That is what happened with the Texas City disaster -- the AN in the hold of the Grandcamp was exothermically decomposing until heat and pressure reached a certain level, and then the whole thing went off at once.

A careless welder ignited material inside a fireworks warehouse, which rapidly spread out of control. Eventually, the fireworks began to explode, spreading the blaze to the warehouse containing the ammonium nitrate. As @Dragon01 pointed out, the AN had been there for a long time and gets touchier the longer it sits around. I suspect there was significant hydrocarbon contamination, either from manufacturing or from environmental effects in storage. The fireball size and radius of light blast damage -- out to 10 km according to some reports -- points to something close to a 2 kt blast. The only way to get that kind of blast out of 2750 tonnes of AN is if it has significant hydrocarbon content.

100% correct. Also see USGS statement: https://earthquake.usgs.gov/earthquakes/eventpage/us6000b9bx/executive As I mentioned above, a 3.3 on the Richter scale is comparable to a 1.3-kiloton nuclear warhead surface blast.

It's actually from the detonation velocity. Ammonium nitrate, while being less energetic per pound than ordinary black powder, has a detonation velocity of almost Mach 8, far far higher. It is at the lower end of what constitutes a high explosive (and if it is contaminated with hydrocarbons its energy can double and its detonation velocity can reach Mach 15). The spherical fireball is the result of the combustion occurring in a hypersonic mode which produces an extreme pressure wave, fully combusting the explosive. The shockwave slows as it moves through the air but it is still extremely fast. The original fire heated the air around and above the warehouse, but when the shockwave reached cooler, wetter air, the rarefaction of the pressure wave caused it to form a vapor cone like a jet breaking the sound barrier. Of course it was spherical, so it was a vapor sphere rather than a vapor cone. You can see the first pockets of cool, wet air being reached in a frame by frame: The red smoke contains high levels of nitric acid which has been confirmed as the cause of significant secondary injury in the city. The initial fire was started by a careless welder and spread into a fireworks warehouse, causing the initial explosion and all the small flashes. The initial explosion caught the AN storage warehouse on fire, and the rest was history. The impact to the grain silos is immense. You can see how much of the concrete collapsed and how much was simply vaporized:

However I should note that this is NOT smaller than a nuclear explosion, as many people are saying. It is most likely one of the five largest non-military artificial explosions in history, right up there with the 1969 N-1 failure. Tactical fission-based battlefield nukes were well under 100 tonnes TNT equivalent, the size of large conventional munitions. This explosion was comparable to medium-sized tactical nuclear weapons. For reference, this was around 900 times more powerful than the Oklahoma City bombing.

Correct. I did some pixel-tracing yesterday and came up with a yield of 0.8 to 1.5 kilotons, before there was any official discussion of how much AN had been in the warehouse. 2750 tonnes of AN at 42% TNT yield comes to 1.15 kilotons. However, hydrocarbon contamination can increase the yield of AN up to 78% (e.g. Timothy McVeigh) so that could account for higher yield. The seismic signature came to 3.3 on the Richter scale which would translate to about 1.3 kilotons. Definitely not nuclear for a lot of reasons. First, the explosion is too slow. Second, the smoke is red. Third, there was no blinding flash that caused thermal radiation burns. Fourth, no nuclear-capable country would drop a nuke into a warehouse fire in a residential area.

Yep, I was about to say that. I suppose it could have gone up on Pegasus XL. Pegasus can take off from Kwajalein and get an equatorial trajectory without an inclination change. Initial IXPE renders seem to have assumed it would go up on Pegasus. XL can send 450 kg to a LEO reference orbit so I suppose it might be able to manage 320 kg to 540 km. A 540x540 km equatorial orbit from the Cape is going to be...exciting. I wonder if it will even circularize in its parking orbit. For that hefty of an inclination change it might want to do a 23-degree-inclined ascent to a 540 km apogee with the perigee still suborbital, then do the inclination change during circularization.

best wide-angle view I've seen:

I didn't think there was any threat to Starship; it just didn't seem nominal. But if it's a new system then that makes sense. Liquid nitrogen is much colder than liquid methane so it should be easy enough to run a condenser if you have somewhere for the warm(er) nitrogen to go.

Huge geyser of something coming out of the flare stack. Does not appear to be nominal.

Yep, we have a scrub/abort.

The LaPadre stream guys keep going on and on about the offset engine. It's not actually that crazy. They are going to put three engines on that thrust puck eventually, but the whole point of having three engines is to have engine-out capability. It needs to be able to maneuver, hover, and land on a single engine. That's why they have the mass sim up top; they are making sure they can handle a single-engine landing near tank depletion. It is nowhere near the full prop load (if it was, it would need all three engines). No reason to risk multiple engines when you only need to test one.

Siren.

Yep, pretty much! I would also have a lower bay door design with a small radiator acting as the door. So you have a decoupler with the radiator placed on it, and the drogues placed on that. Once the drogues stabilize the vehicle's descent, you pop the mains simultaneously with the decoupler, so that the drogues pull the door away to open up a path for the mains.

I just came to say that

LaPadre stream suggesting about 100 tonnes GLOW with a small propellant load.

Speculation on the 150m hop to take place today at 7 PM EST.

I find it so interesting that the drogues pop from above the capsule, but when they are cut they pull free the mains, which are stored in the service section and open up on lines around the cabin door. I suppose you could replicate that in KSP by placing the chutes on freely-moving hinges near the top of the capsule and then offsetting them down into the lower section. Seems like those lines-around-the-cabin-door would be a place where you would expect fouling but apparently not. Perhaps the mains are stored below just because there is no space up top?

That's a Smardan-class river monitor patrol boat, operated on the Danube by the Romanian Naval Forces. This particular one, the Opanez, was commissioned in 1990. It has a 3.9" main gun on the front taken from a T-55 Soviet battle tank, four auxiliary cannons, and a total of twelve 14.5 mm machine guns. The rocket launchers are 122 mm rocket-based artillery; each of the two independent launch platforms contains forty rockets.

On the topic of how big the solar system is, I've always found this to be helpful. https://joshworth.com/dev/pixelspace/pixelspace_solarsystem.html

But why? This image makes it clear that the two flaperons are, in fact, mated to opposite ends of this frame. I suppose it could be used to test actuation before mounting, under horizontal loading (which better matches what it will experience during free-fall). But it is odd, nonetheless. EDIT: So they just mounted the fins here for show? For fun?

Imma use this.....

I, for one, am very interested in this frame to which the flaperons appear to be mounted. I'm assuming from this that the frame itself is going to be mounted onto the belly (or perhaps back?) of Starship, which is definitely different from what we saw with Mk1, etc. It's also unclear from the images whether there are two frames like this, or whether the opposite flap is going to be mounted on this frame. Depending on which way it goes, we may see flap positions which are not actually 180 degrees opposite each other. If the opposite flap is mounted onto this frame and the whole thing is then mounted onto the back of Starship, then we would be seeing a shape like this: Having the flaperon center of pressure behind the center of mass would tend to make the whole thing much more stable on re-entry. In keeping with Elon's "the best process is no process; the best part is no part" mantra, I wonder if the aft flaperons will end up being "locked" into a small, selected number of positions, rather than being actively actuated during entry, not unlike the feathering system of SpaceShipOne. This way you would only need to actively adjust the positions of the smaller forward flaps/canards, which require less torque. Much more passive stability. You could have a position for launch (probably folded all the way windward), a position for Earth aerocapture, a position for Mars aerocapture, a position for Earth EDL, and a position for Mars EDL. A system which locks the flaps in place has fewer failure modes than one where the flaps are continuously actuated.