Starship flight 4: the ship survived reentry

Starship IFT-4 did something no previous flight managed. The ship came back.

Let me say that again because the significance is easy to miss. A stainless steel spacecraft the size of a 16-story building reentered Earth’s atmosphere at 27,000 kilometers per hour, survived the heat and pressure, executed a controlled belly-flop maneuver, and splashed down in the Indian Ocean.

The heat shield lost some tiles. The flaps were visibly damaged. The livestream cut out multiple times. But the ship survived. It survived reentry. That sentence, simple as it is, represents something new.

What reentry means

Getting to space is hard. Coming back is harder.

When a spacecraft reenters the atmosphere, it hits air molecules at orbital velocity. The compression of that air generates temperatures around 1,500 degrees Celsius. For a spacecraft made of stainless steel, which melts around 1,400 degrees, the thermal math is terrifying.

SpaceX solved this with hexagonal heat shield tiles on the windward side of the ship. Thousands of them. Each one designed to absorb and radiate heat while protecting the steel underneath. On previous flights, missing or damaged tiles led to burn-through and loss of vehicle.

On IFT-4, enough tiles stayed on. Enough protection held. The steel didn’t melt. The ship made it through.

The footage from the onboard cameras during reentry was stunning. Orange plasma streaming past. Tiles glowing. The flaps fighting to maintain attitude through hypersonic buffeting. You could see individual tiles ripping away and the steel behind them starting to glow. And then, gradually, the colors shifted from orange to blue. The ship slowed. The atmosphere thickened from enemy to ally. And the ship was falling, not burning. Falling is fine. Falling is controllable.

The booster

The Super Heavy booster also performed well. It flew back toward the coast and attempted a landing burn over the Gulf of Mexico. It didn’t attempt the chopstick catch this time (that comes later), but the controlled descent and soft splashdown demonstrated that the booster can return to the launch site.

Two stages. Both survived. For the first time.

Why this matters for Mars

The Starship going to Mars needs to do exactly what IFT-4 just demonstrated, but at Mars. The Martian atmosphere is thin, about 1% of Earth’s, which means the vehicle will be traveling at high speed much closer to the surface. The heat shield needs to work not just once, but on repeated flights across interplanetary distances.

Every bit of data from IFT-4’s reentry teaches SpaceX something about how the tiles behave under real thermal loads. Simulations are good. Wind tunnels are good. But there’s no substitute for flying a heat shield through actual plasma at actual orbital velocity and seeing what happens.

Now they have that data. They know which tiles failed. They know where the heat was worst. They can fix it for IFT-5.

The iteration speed

IFT-1: April 2023. Exploded 4 minutes in. IFT-2: November 2023. Reached space, ship exploded. IFT-3: March 2024. Reached space, ship lost during reentry. IFT-4: June 2024. Both stages survived.

Fourteen months from “exploded at launch” to “both stages survived.” That’s not a normal pace for aerospace. That’s a pace that only makes sense if you’re building rockets the way software companies build products: ship, break, learn, ship again.

The next flight will attempt the booster catch. A 230-foot rocket, caught mid-air by the launch tower. I can barely type that sentence without my heart rate going up.

This is the year Starship becomes real. Not a concept. Not a test article. A vehicle.

Related thinking: