Why SpaceX Catches Rockets Instead of Landing Them?

Añadido: 2026-05-11

[00:00:04]What you're seeing on screen right now is Superheavy, the most powerful rocket booster ever flown.

[00:00:12]200 tons of steel and fire falling back from the upper atmosphere toward the ground. And right before impact, two steel arms reach out from a 146 m tower and catch it midair.

[00:00:30]October 13th, 2024. A maneuver that no one in the history of spaceflight had ever attempted, let alone pulled off.

[00:00:39]Agencies doubted it. Engineers debated it. The internet called it flatout impossible. But here's what makes this weird. SpaceX had already landed rockets per their own mission logs nearly 600 times using landing legs, a proven system, a system that works. So why would they throw all of that away? Why abandon what works to gamble everything on two steel arms and a tower? The answer isn't just engineering. It's a billiondoll bet that rewrites the economics of leaving Earth.

[00:01:24]To understand why SpaceX abandoned landing legs, you first need to understand what landing legs actually cost. Not in dollars, but in weight.

[00:01:34]Falcon 9 has four landing legs. Each one weighs roughly 500 kg.

[00:01:40]actually closer to 480 depending on the variant, but let's call it 500. That's about 2,000 kg total just for the legs.

[00:01:49]The entire Falcon 9 booster when empty weighs around 22,000 kg. So, the legs alone account for nearly 10% of the booster's dry mass. 10% might sound manageable on Falcon 9, but Superheavy is a completely different animal. At 71 meters tall and 9 meters wide, it dwarfs Falcon 9. Based on publicly available mass estimates, its dry mass is approximately 275 tons, more than 12 times heavier. Run the numbers. If landing legs take up roughly 10% of dry mass, same ratio as Falcon 9, Superheavy need legs weighing somewhere between 25 and 30 tons. That's the weight of a fully loaded semi-truck bolted to the bottom of a rocket. And that truck has to fly every single mission. And in rocketry, weight doesn't add up, it multiplies. Every kilogram of dead weight needs extra fuel to lift it. That extra fuel needs a bigger tank. A bigger tank needs more structural support. More structure means more weight, which needs more fuel. Engineers call it the tyranny of the rocket equation. But the weight's only part of the problem.

[00:03:10]Look inside a Falcon 9 landing leg. At the core, there's an aluminum honeycomb structure called a crush core. When the booster touches down, this core deliberately crumples, absorbing the impact energy so the rocket doesn't shatter on contact. The problem, it's single use. Every landing destroys the crush core. Every turnaround requires a replacement. At Falcon 9 scale, that's manageable. At Superheavy scale, you're replacing a shock absorber the size of a car after every flight. And this is where SpaceX's core design philosophy kicks in. A principle Elon Musk repeats constantly. The best part is no part. If you can eliminate a component entirely, you eliminate its weight, its failure modes, its maintenance, and its cost all at once. So, the legs had to go. But removing 25 tons of dead weight, that only solves the first problem. The second one, something no engineer can solve by cutting parts. It's time.

[00:04:20]Okay, so a Falcon 9 booster lands on a drone ship. Successful touchdown.

[00:04:25]Mission accomplished, right?

[00:04:28]Not quite. Because what happens next is where the real bottleneck begins.

[00:04:35]First, you wait for weather. Then, a tugboat towes the drone ship back to port. That alone can take 2 to 3 days.

[00:04:43]Once docked, a crane lifts the booster off the ship and onto a transport stand.

[00:04:49]Then the booster is transported over land to a processing facility. Engineers inspect all nine Merlin engines. They check the tanks for leaks. They examine the heat shield and the landing legs.

[00:05:02]Those can't retract on their own. SpaceX needs a custom crane jig just to fold them back in, which honestly that's a lot of hardware just to undo one landing.

[00:05:15]Per SpaceX's own turnaround records, the fastest they've ever turned a Falcon 9 booster around is 9 days, and that's the record. Normally, it takes 2 to 3 weeks from landing to the next launch.

[00:05:30]For Falcon 9, that's fine. They've got a fleet of boosters. While one's being refurbished, another one flies. But for Starship and Super Heavy, SpaceX isn't building a fleet. They're building a launch system. And a system only works if the turnarounds measured in hours, not weeks.

[00:05:51]So, picture Mechazilla. The booster comes back to the exact same spot it launched from. The tower catches it.

[00:05:59]Automated systems run diagnostics in minutes instead of days. If everything checks out, the booster is lowered onto the launch mount. A new Starship stacked on top. Fuel flows in and it's ready to fly again.

[00:06:14]That's the long-term target, under 1 hour from catch to launch. Think of it this way. You're an airline. Every time a plane lands, you tow it by boat from the middle of the ocean back to the airport. Then you disassemble the landing gear, inspect every bolt, replace the shock absorbers, and put it all back together. That takes 2 weeks.

[00:06:39]That's Falcon 9.

[00:06:42]Now, the plane lands directly at the gate. Passengers step off. New passengers step on. It takes off again.

[00:06:50]That's Mechazilla. Same rocket, same pad, no ocean, no truck, not even a crane. SpaceX calls this zero touch reusability. No human hands on the booster between flights. Just catch, inspect, stack, fuel, and fly. Operate a rocket the way airlines operate a 737.

[00:07:15]And if that sounds overly ambitious, well, so did landing a rocket on a boat until it wasn't. If you're the kind of viewer who wants to understand the engineering behind the spectacle, not just watch the spectacle, you're in the right place. Hit subscribe. The Cosmic Rush goes deep, where other channels stop at the surface. Lighter, faster.

[00:07:45]But there's one more reason SpaceX ditched the legs, and this one's hiding inside the rocket itself.

[00:07:53]Here's something most people don't realize. Landing legs aren't just legs.

[00:07:59]They're an entire ecosystem of components that comes with them.

[00:08:04]To deploy a landing leg, you need a pneumatic system, high pressure helium, to push the leg out. You need actuators to control the deployment, sensors to confirm it's locked in position, wiring to connect all of that to the flight computer, and structural reinforcement at every attachment point on the hole because that's where the entire weight of the rocket concentrates on impact.

[00:08:31]Each attachment points a stress concentration zone. The hole has to be thicker there, stronger, which means heavier and more complex to manufacture.

[00:08:43]Four legs means four reinforced zones.

[00:08:46]Four potential weak points that engineers have to design around, inspect after every flight, and worry about for the lifetime of the vehicle.

[00:08:56]Remove the legs and all of that disappears.

[00:09:00]No actuators, no helium system, no crush cores, no sensors, no wiring, not even the reinforced zones. The hole becomes a clean, uniform cylinder. Simpler to build, faster to inspect, and cheaper to produce.

[00:09:19]This isn't just about saving weight anymore. It's a manufacturing philosophy. Every part you remove is one less thing that can break, one less thing to inspect, one less supplier in the chain, one less reason a launch gets delayed.

[00:09:38]Now, the honest caveat, SpaceX didn't eliminate complexity, they moved it. All the precision, all the engineering, all the risk of catching a 200 ton booster, that's now on the tower.

[00:09:54]Mechazilla isn't a simple machine. But here's the key difference. The tower never leaves the ground. The booster does hundreds of times. So, if you have to choose where to put the complexity, you put it on the thing that doesn't have to survive re-entry.

[00:10:13]Three reasons to kill the landing legs, but all three only matter if you can actually catch a 71 m 200 ton booster falling from the sky.

[00:10:25]And that is where things get really interesting.

[00:10:38]So, how do you actually catch a rocket?

[00:10:42]Let's start with the tower itself.

[00:10:44]Mechazilla stands roughly 146 m tall, taller than the Statue of Liberty, pedestal included. Mounted on its frame are two steel arms, each approximately 36 m long, give or take, depending on which iteration you're looking at.

[00:11:03]SpaceX calls them chopsticks.

[00:11:05]These arms ride on a carriage system.

[00:11:08]Electric motors and steel cables move them up and down the tower. They can open, close, and adjust position with precision measured in centime.

[00:11:20]But here's the part most people get wrong. The booster doesn't aim straight at the tower on its way down. That'd be insane. If anything goes wrong, you destroy your entire launch infrastructure.

[00:11:33]Instead, the booster targets a point near the tower. Only when every single catch commit criterion's met, engine health, position, velocity, tower readiness, does the booster shift its trajectory toward the arms. To thread that needle, the booster uses a layered guidance system. GPS gives it global position. An inertial measurement unit tracks acceleration and rotation.

[00:12:02]Onboard cameras and LAR provide close-range spatial awareness. All of that data is fused in real time, and the flight computer makes corrections dozens of times per second. Steering through the atmosphere, the booster relies on grid fins, but not the titanium ones you've seen on Falcon 9. Superheavy uses welded stainless steel. Heavier, yes, but far cheaper, and they handle the heat just fine.

[00:12:32]Based on SpaceX's latest configuration, the current design uses three fins instead of four, each about 50% larger than the previous generation.

[00:12:43]Now, the moment of contact on the booster's inner stage, there are structural hard points called catch pins. These pins slot into catch rails on the chopstick arms. The arms close around them, and the booster is held.

[00:13:00]But it's not a hard grab. The catch rails are equipped with hydraulic pistons and pneumatic dampers that absorb the remaining kinetic energy.

[00:13:09]Think of it less like catching a baseball and more like a controlled cradle. The arms don't stop the booster, they decelerate it gently over a short distance.

[00:13:21]And what if something goes wrong? SpaceX designed for that, too. If any catch commit criterion fails at the last moment, the booster automatically diverts away from the tower and executes a controlled splashdown in the Gulf of Mexico.

[00:13:37]This isn't theory. On November 19th, 2024, flight 6, the tower's automated health check flagged an issue just seconds before the catch. The abort triggered instantly. The booster diverted and splashed down safely. No tower damage, no explosion. The system worked exactly as designed, just not the way anyone wanted. So SpaceX chose to move all the complexity to the ground, build a tower, build the arms, and trust the software. But across the industry, not everyone agrees. One company in particular is betting on the exact opposite approach.

[00:14:23]Blue Origins New Glenn. Same goal as Superheavy. Launch, separate, land the booster, fly it again, but their solutions fundamentally different.

[00:14:35]New Glenn keeps its landing legs, six of them, hydraulic. The booster carries them all the way up and all the way back down.

[00:14:45]Instead of returning to the launch site, the booster lands on a ship, a converted cargo vessel called Jaclyn, stationed hundreds of kilome downrange in the Atlantic. No tower, no arms, just legs and a landing pad on a boat.

[00:15:02]On January 16th, 2025, New Glenn flew for the first time. The upper stage reached orbit successfully, but the booster lost during descent. The landing attempt failed.

[00:15:17]And that's worth noting because it shows that landing legs at this scale aren't a solved problem either. SpaceX proved tower catching works. Blue Origin hasn't yet proven ship landing works at New Glenn's size. Both approaches carry risk.

[00:15:35]SpaceX moves complexity to the ground but depends on a single tower. Blue Origin keeps complexity on the rocket but can land anywhere there's a ship.

[00:15:46]Neither approach is wrong. They're different answers to the same engineering question. But only one of them allows a turnaround measured in hours. And as far as the economics go, that difference is everything.

[00:16:07]Now step back and look at the full picture.

[00:16:12]Remove the legs. You save 25 tons of dead weight. That weight becomes payload. More payload means lower cost per kilogram. Lower cost attracts more customers. More customers mean more flights. More flights generate more landing data. More data means higher catch precision. Higher precision means faster turnaround. Faster turnaround means even more flights. And yeah, it just keeps going.

[00:16:46]This isn't a feature list. It's a compounding loop. Each improvement feeds the next. And once a loop like this starts turning, it becomes very, very hard for anyone else to catch up.

[00:17:02]That more than any single technology is what makes Mechazilla dangerous. Not the arms, not the tower, the loop.

[00:17:21]But let's not pretend this is a perfect system. The critics raise real points.

[00:17:28]First single point of failure. If the tower is damaged, SpaceX can't launch or catch. With Falcon 9 and landing legs, you can land on any flat surface. With Mechazilla, you need that specific tower at that specific site.

[00:17:47]Second cost. Building a 146 meter launch and catch tower isn't cheap. That's a massive infrastructure investment before you've caught a single booster.

[00:18:00]SpaceX's answer, build more towers. A second tower is already under construction at Starbase.

[00:18:08]And the math works differently at scale.

[00:18:11]If each tower supports hundreds of flights per year, the cost per flight drops dramatically.

[00:18:18]And the abort system, flight 6 proved it works.

[00:18:23]This isn't all or nothing. It's high stakes engineering with a safety net.

[00:18:35]So, back to the question we started with.

[00:18:39]Why would SpaceX throw away a system that worked nearly 600 times?

[00:18:45]Now, you know, not because landing legs don't work, but because at this scale, they become the very thing holding you back. Too heavy, too slow, too complex.

[00:18:59]And SpaceX decided if you're going to build a system that flies to Mars and back every single day, you can't afford to carry dead weight.

[00:19:09]Literally, that tower standing in Bokeh Chica, Texas, isn't just a launch tower. It's a factory reset button for the entire space industry.

[00:19:23]SpaceX wants to catch a booster and relaunch it within an hour. When do you think they'll actually pull it off? Drop your prediction in the comments.

[00:19:33]And if this kind of deep engineering breakdowns what you're into, subscribe to the Cosmic Rush. We'll see you in the next one.