Massive Reveal From SpaceX’s Starship Flight 13!

Two months ago, SpaceX’s version three Starship left us with more questions than answers.

A booster that couldn’t stick its return.

A ship down an engine.

And a whole community asking the same thing: Is this new design actually the one?

On Thursday, SpaceX flew again.

Here’s what happened.

My name is Felix, and welcome to “What About It.”

Let’s dive right in.

Starship updates.

What a time to be alive.

Starship flight 13 has occurred, and there is much we can analyze and take away from this second version three Starship flight.

Is version three the solution SpaceX has been hoping for?

Can SpaceX fulfill all its goals moving forward, or are we looking at yet another year of hard work and slow progress?

Let’s take a look together.

It all started with a full 33-engine static fire for Booster 20 on July 10th.

It lasted for a full 24 ground-shaking seconds, the longest test for a Super Heavy booster ever.

The following night, the crew got to work on the chopsticks, addressing an actuator that hadn’t performed as intended.

That work wrapped up the next day with the actuator replaced, and the path to a full stack was clear.

With every preparation made, it was finally time to fly once again.

On Thursday, July 16th, SpaceX and the entirety of Starbase were ready for Starship’s 13th test flight, the second outing for the new and massively different V3 generation.

Flight 12 hadn’t managed to tick off every item on its mission plan, and now it was time to try to prove once more that V3 was fully operational.

As we started our live stream, the tank farm had already come to life, prepping for the fuel transfer into both Starship stages.

The Countdown Begins

“Uh, we have flight 13 today. It is absolutely incredible that we’re talking about this, what, five weeks after the last launch?”

Those preparations got the pipes and valves ready for the volumes of supercooled pressurized liquid that rush through the system as fueling begins.

The fuel transfer speed on these vehicles is still worth mentioning again.

Both stages were fully fueled in under one hour.

In fact, just 37 minutes and 30 seconds, which is even quicker than the flight 12 timeline.

SpaceX is improving with every test, and that speed isn’t a luxury.

Truly rapid reusability depends on it.

As mentioned in the last episode, pad two, according to SpaceX, is designed to be used continuously every 60 minutes.

It’s all already in the design and in active testing, even if launches aren’t rapid yet.

Here’s another detail that aligns with the goal of making launches as routine as airliners take off.

The flight director runs the first major poll, the one that clears the vehicle for propellant load, inside the final hour before launch—roughly 50 minutes before T-0.

In contrast, SLS does the fueling poll 10 hours and 50 minutes before flight.

That’s a full 10 hours earlier.

There is a different philosophy in plain sight.

And if you’re saying that’s because it’s an older design or because it’s not reusable, think again.

New Glenn starts fueling around 4 hours before launch, even though a Starship has a three to four times larger fuel capacity.

Here is a quick mind-blower: Starship has a combined fuel capacity of 5,250 tons, according to SpaceX’s official numbers.

Fuel flow isn’t uniform during the tanking process, but for simplicity, let’s just assume it is.

Considering the 37 minutes and 30 seconds, we’d end up with an average fuel flow of 2,333 kg per second, or a staggering 140 tons per minute of propellant flowing into the full stack.

140 tons per minute.

The reason, again, is rapid reuse.

It’s already in the design, even if flights are not rapid yet.

In spaceflight, you almost never trust a single instrument.

Critical systems get backed up, and the smart move is always to compare your sources before you commit to anything.

A New Player in AI

Today’s sponsor brings that same habit to AI.

This segment is brought to you by Mammoth AI.

Mammoth puts the best AI models in one place: Claude, ChatGPT, Gemini, Mistral, Grok, DeepSeek, and more.

Instead of looking into a single provider, you can run a prompt through several and compare the answers or just always reach for whichever model is sharpest that week.

It also generates videos, reads your PDFs and spreadsheets, runs deep research, and has a lighter lower energy mode for when you don’t need the heavy hardware.

And because it’s a European company based in Paris with servers in Germany, your prompts aren’t stored or used to train anything.

All of that starts at just $10 a month.

You’ll find Mammoth at mammoth.ai, and you’ll find the link in the description.

Try it out now and support “What About It?”

Back to Starship and flight 13.

But not everything was ready for a flight.

SpaceX proceeded through fueling with blistering speed.

Everything looked excellent.

Then came the ignition, and roughly one or two seconds in, the booster pulled the brake.

Hard stop.

Abort after ignition.

During the actual launch on July 24th, we got the real reason: ice had built up inside the engines, likely caused by the static fire just a day before.

The Ice Issue

“So, four of Super Heavy’s 33 Raptor engines aborted at startup due to issues with their ox turbo pumps.

Those rapidly pressurize and feed liquid oxygen into the engine for combustion.

For those of you that like to pay real close attention to the on-screen telemetry, you saw this at T0 during that first attempt.

We dug through the data, and what we saw was some off-nominal spin response in the ox turbo pumps on six of the engines.

So, the most likely root cause was determined: we were collecting some moisture inside of the pumps from previous operations, which then froze when we introduced that cryogenic propellant, and that either slowed those pumps down or stopped them completely from spinning.”

Losing two engines during the initial ignition means there is no safety margin anymore.

It’s just too risky.

Even if the rocket could make it, that’s why the onboard computers decided to cancel the launch.

SpaceX rolled the booster back to Megabay 2 and the ship to Megabay 1.

Engines were replaced.

It looks like more than just those two; a fast abort can cause damage to engines.

Pressure spikes and cavitation are the biggest problems here.

After having the engines replaced, SpaceX dodged bad weather on the 23rd.

The reason for the scrap on Thursday was that SpaceX was planning to image the Starship from the ground while it was going through a faster than usual Max-Q to test the heat shield even further.

If you have solid cloud coverage, imaging the rocket from the ground is simply impossible during Max-Q.

So, they moved the flight to the 24th and Friday.

The Big Day Arrives

Friday was that day.

All was ready for the big show.

Roughly 50 minutes before launch, SpaceX went through the fueling pool.

All looked good.

When the frost lines appeared, we had visual confirmation that cryogenic oxygen was rushing in.

Thirty seconds later, the booster’s oxygen tank followed.

One of the biggest upgrades on the V3 booster is the key to this faster tanking.

There are now two quick disconnects at the bottom, one for liquid oxygen, one for methane.

That lets oxygen and methane load as separate operations, and methane loading started less than a minute later.

At T-40 minutes, Ship 40’s methane tank was the last to join in.

Now, every tank was filling at once.

140 tons per minute.

Before you start a Raptor engine, you have to make sure it’s ready, and that means pre-chilling it so the engine’s material can survive the flood of supercooled high-pressure liquid to come.

All 33 Raptors on the booster and six on the ship began that treatment at T-minus 21 minutes.

When the booster loading procedure wrapped up, there were 2 minutes and 50 seconds left on the countdown.

Forty seconds later, within the same minute, the same was true for the ship.

With the vehicle fully fueled, the countdown ran down toward the 30-second mark.

That’s the point where the countdown stops for one final check of every system involved in the launch and the flight.

“Our primary test objectives for today’s test flight include a successful liftoff, stage separation.

Uh, we also plan to deploy 20 Starlink V3 satellites.

We have real satellites on board.

We expect to have about 20 minutes of available time to speed run some tests and really allow those Starlink engineers to put them to the test in that time.”

A series of polls runs in the background, and only once everything comes back go, does the flight director speak the words we’d all been waiting for.

“Prop load now closing out on both vehicles.

We have no holds on the board.”

“All right, we are through the gate.

Still no holds on the board.”

“Fifty seconds to go.

Starship getting ready to fly.”

“Go for launch.”

At T-minus 17 seconds, the flame diverter deluge system came to life, flooding the trench with water to soften the shock waves of the engines.

650,000 gallons or 2.4 million liters of water per minute.

“Three.”

“Three seconds before liftoff, the booster’s engines received their startup command.”

“Two. One.”

And that was it.

No way back.

The only way from here is up.

“The prop load is good.”

Booster 20’s 33 brand new Raptor 3 engines lit, hitting the flame diverter with a combined thrust of 8,240 tons.

We’ve seen this trench work before on static fires and on flight 12, but I still can’t get enough of this.

I’m glad that SpaceX had the pad one launch tool design ready fast, and I’m equally glad that it’s gone now.

Booster 20 with Ship 40 on top jumped off the launch table and started its climb into the Texan sky.

The speed it builds is hard to overstate.

Starship, even though it’s the largest rocket ever, clears the tower surprisingly fast now that SpaceX has reached version 3.

Performance Metrics

Here’s a small comparison to bring this into perspective.

Starship has a thrust-to-weight ratio of around 1.35.

New Glenn sits at roughly 1.3.

SLS is slightly higher at around 1.45, and Falcon 9 is the road runner at 1.7.

The higher the number, the faster it accelerates.

Throughout development, SpaceX was able to considerably ramp this number up from 1.25 on version 1, which took noticeably longer to clear the tower.

Version 2 hovered around 1.3.

Just like on flight 12, clearing the tower meant pad 2 was done for the day.

Once again, there were no plans to bring either stage home.

Yes, we still have to wait for a ship catch, but this flight was the missing piece of the puzzle that had to be set in place first.

Since the first sign of trouble on flight 12 was the loss of a booster engine, our eyes were on the engine diagram as the mission climbed.

At the 50-second mark, you can see something ripping off the right aft flap.

I don’t think that’s ice.

It might be the heat tiles on the underside of the flaps.

That’s why they went through this tougher Max Q to test the tiles under a higher aerodynamic load.

Less than a minute in, Starship reached Max Q, the moment of peak aerodynamic stress on the whole journey.

And this rocket is quick, just 58 seconds to Max Q.

So, this is usually the spot where I tell you to like, subscribe, and become a channel member, but today I want to thank those who gave super chats during the launch stream.

You people made Jordan flip.

You made it happen.

We’ll get a $5,000 gimbal stabilizer for him to make stable videos in the helicopter.

There’s nothing more to say but thank you so much for all your help.

You rock.

To everyone else, like, subscribe, become a channel member or a patron.

Thank you so much.

Back to the launch.

Flight 13 had one important addition for SpaceX.

The heat shield carried load sensing tiles that record the force the ship experiences.

Why now after 12 previous flights?

Because this time the vehicle was deliberately flown through higher dynamic pressure on ascent than any Starship before it, putting extra stress on the tiles and their attachments.

If that approach holds up, it opens the door to carrying more payload to orbit.

The less you slow down during Max Q, the less energy you waste.

There’s one thing to highlight during the ascent.

One of the center engines on Booster 20 had a pretty spectacular hiccup.

It’s not visible in the live stream, but NSF’s cameras captured it, and we took pictures of it.

Thank you, Amy.

This might just be a temporary thing, but it is worth noting.

Starship kept accelerating on its way up, heading for the next milestone a bit over a minute later: MECO.

It arrived right on cue.

At 2 minutes and 18 seconds into the mission, most of the Super Heavy booster’s engines shut down.

That is a Starship specialty.

Normally, MECO stands for main engine cutoff.

On this rocket, it is most engines cut off.

The difference comes down to hot staging.

Most rockets separate like this: shut down the first stage, split the stages with mechanical or pyrotechnic systems, and only then light the second stage.

Starship hot staging holds onto more of the thrust already invested in reaching separation altitude.

The first stage keeps some engines running while the upper stage lights its own and leaps off the booster.

Think of a cheerleader pushing off the hands lifting her at the same moment they lift her.

Two motions in one direction combine into a single push.

Just 3 seconds after MECO at 2 minutes and 21 seconds, hot staging began.

Ship 40’s Raptor lit and pushed the upper stage off the top of the booster.

Exhaust streamed through the fixed strut structure of the V3 hot staging hardware we got to know last flight.

Now came the part we’d all been waiting for because if you watched flight 12, you’ll remember exactly where Booster 19 got into trouble.

Last time, slight differences in how the ship’s engines started up threw the booster’s direction off by roughly 90 degrees.

Then, during the boost back burn, five of its 33 engines refused to relight cleanly, cutting the burn short and ending the booster’s day with a hard splashdown.

SpaceX went to work on both.

The startup sequence was reworked to be far more robust to timing variability, so the booster flips the way it’s supposed to.

And the engines got hardware modifications to improve relight reliability with the engine alarms and aborts retuned to match what the vehicle actually sees in a multi-engine environment.

So, SpaceX did its homework.

The Booster’s Return

It was now time to put it to a test.

Jordan was in the heli again during launch, something you can do too.

These aerial images are essential for my reporting.

If you’re down at Starbase, make sure to pay our official partner SPI Helicopters a visit and book a flight.

The same views our photographers get.

It is unforgettable and worth every penny.

Click the card or the link in the description.

Help Y in return.

Check it out today.

At 2 minutes and 25 seconds, the booster executed its flip and lit its engines for the boost back burn.

This time, the flip was clean, and the engines came alive exactly as intended.

The burn ran its full course and shut down at roughly 3 minutes.

Just what we wanted to see, and the best was still to come.

Booster 20 settled into its long coast back toward the Texas coast.

That’s the point at which things went south for Booster 20 this time.

Everything up to this point was perfect.

You can see the booster gliding through the atmosphere guided by its three grid fins.

The landing burn started at 6 minutes and 25 seconds.

The ignition pattern was erratic.

Usually, the pattern starts with 13 engines, then five in a symmetrical pattern, then three.

But as you can see, the booster had trouble igniting all 13.

Five of 13 didn’t light.

Then it wanted to go down to five and already started compensating by leaving other engines on to replace the faulty ones.

It did try hard, but it wasn’t enough.

We got a spectacular onboard view of the booster making it into the exclusion zone, but too fast for a soft splashdown.

It hit the water hard.

Something in there didn’t work the right way.

Maybe SpaceX will give us an official statement later.

We’ll have to wait and see.

This, unfortunately, was the end of Booster 20’s life.

It was never meant to be caught by Mechazilla’s chopsticks, and it went into the water offshore.

Not because it couldn’t have been caught, but because it’s a different design from the one they already caught, and it still needs to show that it’s as good and reliable.

But the booster was only half the story.

Now, it was Ship 40’s turn, and this second V3 upper stage still had a lot to prove.

At 8 minutes, the ship shut down its engines on schedule.

And that timing matters because last flight, the ship lost one of its three vacuum-optimized Raptors about 40 seconds after separation.

It still reached its trajectory on engine cutout capability, but now the latest upgrades were on trial.

This time, all engines carried the ship through to a clean cutoff on the planned suborbital trajectory.

No drama in the engine bay.

As with every Starship flight so far, the ship wasn’t aiming for a stable orbit.

It rode a ballistic trajectory that would bring it back into the atmosphere on its own.

Disclaimer: This content may be created by Al for entertainment purposes. Any resemblance to real persons, events, or places is coincidental.