“We are in a little bit of a spin.” With those understated words from SpaceX’s Dan Huot, the world watched as Starship, the most powerful rocket ever built, lost control in the silent void above Earth its third consecutive flight to end in technical disappointment, and a vivid reminder of the daunting complexity behind Elon Musk’s vision for a multiplanetary future.
The stakes for Starship are huge. Not only for SpaceX, but for the entire space industry. This is the rocket that is supposed to reduce the cost of spaceflight by an order of magnitude, to transport the equipment for lunar bases, asteroid mining, and most boldly a Martian city. And so far, as of May 2025, Starship has not yet sent a single payload to orbit, and each mission has revealed new engineering weaknesses.
Tuesday’s test, Flight 9, was to be a milestone. The Super Heavy booster, fueled by 33 methane-burning Raptor engines, roared off the coast of Texas, sending the stainless-steel Starship upper stage sailing into the sky. The booster, a technical wonder in itself, had already become history as the first reused on a Starship launch, having flown and been refurbished since January a first that speaks volumes for SpaceX’s push toward rapid reusability. All 33 engines operated nominally, and the booster performed a hot-staging maneuver and a planned boostback burn. But while it was firing its final landing burn, the booster burst apart in the air, missing a controlled recovery during the descent phase.
The second stage, Starship S35, continued on. For the first time in three flights, all six Raptor engines burned for their full length, putting the vehicle on its intended suborbital course. However, when the moment arrived to jettison eight dummy Starlink satellites from the payload bay a key demonstration for future business missions the hatch mechanism failed. The door stayed firmly closed, and the release was aborted just after the midpoint of the coast phase.
Minutes later, the mission’s most significant failure was revealed. A leak in the primary fuel tank system made Starship lose tank pressure, depriving it of the power to maintain orientation. “Leaks caused loss of main tank pressure during the coast and re-entry phase. Lot of good data to review,” Elon Musk tweeted on X, emphasizing the iterative, data-driven mindset that underpins SpaceX engineering culture following the flight. But with attitude control gone, Starship started to slowly, uncontrollably spin. Plans to restart a Raptor engine in space crucial for future deorbiting maneuvers were abandoned. The vehicle spun back into the atmosphere and disintegrated over the Indian Ocean, away from inhabited areas.
For engineers and industry watchers, the root causes of these repeated failures are revealing. Previous January and March flights had culminated in disastrous explosions on account of various problems: resonance-caused fuel leaks, engine compartment fires, and a failure of hardware in one of the center Raptor engines of an upper stage to permit propellants to mix and burn while in ascent. Hardware improvements from SpaceX followed: upgraded venting in the “attic space,” upgraded engine plumbing, and new nitrogen purge systems. But as Tuesday’s breach exposed, every solution exposes new failure modes, a phenomenon that will sound familiar to anyone building complex aerospace systems.
The Raptor engine itself is a study in accelerated innovation and persistent iteration. It is designed to burn liquid oxygen and methane and boasts high efficiency and deep throttling capability, both needed for Mars missions and reusability. Musk has stated that “Version 3 of the Ship and Booster has quite a radical redesign,” with the next-generation Raptor engines (Raptor 3) to tackle joint preload and propellant management problems that have harassed recent flights in the latest test series.
Thermal protection is another key frontier. Starship’s reusability relies on its revolutionary heat shield, an arrangement of hexagonal ceramic tiles that can absorb the scorching temperatures of atmospheric reentry. Contrary to the brittle silica tiles of the Space Shuttle, Starship’s tiles are supported by a stainless-steel framework, capitalizing on steel’s high melting point (up to 870°C) and its enhanced strength at cryogenic temperatures due to its chromium-nickel alloy composition. The heat shield’s design includes micro-perforations for water to “bleed” out, enabling transpiration cooling a technique that could allow rapid turnaround between flights. On Flight 9, SpaceX intentionally removed 100 tiles to test the system’s limits, but the uncontrolled reentry prevented meaningful data collection on tile performance.
For SpaceX’s cost-reduction vision, these failures are more than technical asides. Starship’s potential for complete reusability and enormous payload capacity is at the heart of Musk’s ambition to bring the cost of material transportation to Mars from $1 billion a ton to $100,000 a ton based on recent market studies. Every failed launch postpones not only Mars plans, but the overall space economy impacting satellite deployment, lunar missions, and the growing market for orbital infrastructure.
But the engineering philosophy in SpaceX is one that involves “fail forward.” As Musk explained to Ars Technica, “The most important thing is data on how to improve the tile design, so it’s basically data during the high heating, reentry phase in order to improve the tiles for the next iteration.” Every flight, successful or not, funds a quick prototyping loop that has already seen breakthroughs in booster recoverability and engine reliability.
In the years to come, SpaceX will try to ramp up its launch rate, targeting a new Starship launch every three to four weeks. Subsequent versions Block 3 spacecraft and Raptor 3 engines will bring further improvements. But as Tuesday’s flight showed, the road to Mars is lined with bitter engineering lessons, and the margin for failure is still razor-thin.
For the thousands of engineers, investors, and fans following Starship’s development, the sight is more than a test program. It is a public, open experiment in space exploration’s future a one where every failure is a page in much larger book, and every data point gets the multiplanetary society dream that much closer to reality.
