Why Starting A Rocket Engine Is So Hard!

Hi, it's me, Tim Dodd, the Everyday Astronaut. Liquid fueled rocket engines operate by flowing fuel and oxidizer into a combustion chamber at ridiculously high pressures in order to throw as much mass out the flaming end of the rocket as quickly and thereby as efficiently as possible. But how does it get that way? How on earth, and even more difficultly, how off earth do you get an engine up to operating pressures and temperatures? Engines often utilize pumps which themselves rely on the engine to be running in order to get power and having a rocket engine run

and getting a rocket engine running are two wildly different things. There is often an extremely delicate dance to get the palms, valves, temperatures, and pressures all up to operating conditions perfectly. Get it wrong by just a few milliseconds even, and you can cause a detonation that can completely destroy the engine. So today we're going to do a deep dive on how exactly you start a rocket engine. We'll cover pretty much everything from a simple solid rocket motor to all of the intricacies of liquid fueled rocket engines, including thermally conditioning the engine, the spin up process,

the scary transient regions, and the actual ignition of the propellants. Then we'll talk about some of the extra challenges like starting a rocket engine in space, and then we're going to actually go step by step inside a rocket engine during its startup. Now this is a very long video, so here's the time stamps for each section, and we've got the YouTube timeline broken up into those sections too. And if you're more of a reader, we've got you covered with an article version up at everydayastronaut.com with links and sources. That being said, let's get started. 3,

2, 1, liftoff. Right up front, I need to suggest that you watch my videos "Why don't rocket engines melt?" and "How to power a rocket engine" as those videos will help with a lot of concepts that we talk about in this video. Trust me, don't even try to watch this video without having watched those because they lay a lot of vital groundwork. And lastly, before we get started, I've got some great news. Our highly detailed one 100 scale metal Falcon 9 rocket models are finally back in stock. Get them while you can everydayastronaut.com/shop. Okay, assuming

you have all the basic knowledge of how a liquid fueled rocket engine works, let's actually start off with the most simple rocket engine to start a solid rocket motor. By far, the easiest rocket to start has to be a solid rocket motor. In fact, there's actually a good chance that you've done this yourself. If you've ever launched a model rocket or even fireworks, you've already pretty much experienced the ignition sequence of a solid rocket motor. Solid rocket propellant is exactly what it sounds like. It's a solid that is rocket propellant. It's usually like a sludge

of pre-mixed fuel and oxidizer, and all it takes is energy to get the combustion process started. In the case of a small model rocket motor, that energy usually comes from the heat from a set of small wires. Bigger solid rocket motors will usually require more energy to get started, so sometimes they'll have some black powder, which is basically gun powder that will initially ignite from the wires, and then a chain reaction will occur that will then light the solid propellant where it will continue to self sustain. But once you start talking about large orbital class,

solid rocket boosters like those on the Space Shuttle or maybe those on the Space Launch System, otherwise known as the SLS rocket, which is at the heart of the Artemis program, there's basically a bomb sitting on top of the booster that starts the ignition process. There's a device called a NASA Standard Detonator or an NSD, and this is a little device used since the Gemini program for things like separation events on booster separation and fairings and frangible bolts, or in this case, they're the device that they actually use to start the chain of reactions that'll

get these massive solid rocket boosters going. When it's go time, A signal is sent to not just one NSD, but a completely redundant NSD to ensure the boosters light. After all, if one doesn't light, it would be very, very bad. And don't forget, once you light a solid rocket booster, there is no shutting it down. The NSDs burst through a thin barrier seal, which then lights a pyrotechnic booster charge. Then that booster charge is what ignites propellant inside and igniter initiator, which then lights the main booster igniter, and that lights the entire surface of the

core of the booster virtually simultaneously. So for a little overview here, it's basically a waterfall of explosives that cascade down from the top of the booster until eventually the entire booster is lit pretty much all at once. Basically stepping up from a small charge, to a medium charge, to a large charge, to booster ignition. Like I said, it's basically more or less a bomb that's lit at the top of the booster or otherwise known as the forward segment that starts the ignition process. It's really just about that easy. That's kind of their whole point. They're

super simple and because of their easy ignition process, they're very reliable, but they definitely have limitations such as not being able to be shut down, and they aren't nearly as high performing as a liquid fueled rocket engine. But starting a liquid fueled rocket engine is just not anywhere near that easy. There's a ton of steps you need to take, but perhaps the first and most important thing to do is to purge the engine and thermally condition it for the extreme environments it's about to encounter. Before you get a liquid fueled rocket engine running, it has

to be prepared for the ludicrous temperatures it's about to face. And believe it or not, I'm not talking about the extreme heat. I'm actually talking about the extreme cold. Not only do most orbital liquid fueled rocket engines run propellant through the walls of the engine to keep the combustion chamber and nozzle from melting, but the pumps themselves will flow upwards of thousands of liters per second of cryogenic propellants, which can make the metals, valves and bearings brittle and failure prone. This of course, is specifically true of any engine that uses cryogenic propellants, so any liquid

oxygen, also known as LOx powered rocket engine, and especially those that run on hydrogen or hydrolox and methane or methylox for their fuels, which are also cryogenic propellants. These propellants all make up the majority of rocket engines that we see on orbital rockets. But before you even begin to chill down the engine with the propellants, you need to purge the engine from moisture and air pockets, and this is usually done with nitrogen. If there's any water vapor in the lines before cryogens are introduced, it'll freeze and cause damage to the engine with the potential for

clogged orifices, damaged seals, clogged injectors, et cetera, et cetera. But if it's a hydrolox engine; liquid hydrogen is so cold, it can even turn any residual nitrogen into a solid block of ice and do all of the bad things we just mentioned. And an ice blockage can be really hard to investigate because the evidence of the root cause of the failure melts away by the time a human can even get close enough to investigate. Because of this hydrogen powered rocket engines are usually purged with another inert fluid, helium. Helium is the only element that has

a lower melting point than hydrogen. So in the presence of liquid hydrogen, it won't freeze solid. Purging and chill down of the engines can be anywhere from a few hours before launch to a few minutes, but this varies very much from engine to engine. So listen for something like engine cool down or engine chill down, called out on comms... Stage one engine chill has started. All right, there's that call that stage one engine chill has started. Okay, so once the engine has been purged of moisture, we can start to actually thermally condition the engine with

the cryogenic propellants. This process involves opening what's known as a pre-valve, which is the big valve that connects the engines to the propellant tanks. The propellant will fill the pumps and fill all the way down to the main valves, which is usually one or a number of valves for fuel and oxidizer. Then the pumps will just soak in a bath of cryogenic propellant until the pumps eventually get down to those frigid temperatures. In comparison to cryogenic propellants, the engine is actually scorching hot under normal, everyday human comfortable ambient temperatures. Even if an engine was hanging

out in Antarctica during a deadly blizzard, it would still pretty much immediately boil off any cryogenic propellant that touches it. So there needs to be a vent line to remove the boiled off propellant. Oxygen can usually just be dumped overboard, which poses little to no risk when being vented into the atmosphere. But in general, a cryogenic fuel that now has boiled off into a gas like methane or hydrogen needs to be recaptured and either chilled and pumped back into the rocket, or it can be vented away from the rocket in a controlled manner where a

flare can prevent a buildup from accidentally detonating because if you just vent gaseous methane or hydrogen out into the atmosphere in an uncontrolled manner, don't expect very good things to happen. This is why you see those giant sparkler lit before Space Shuttle. SLS or Delta IV Heavy launches. It's not to light the actual engines and start combustion, it's to burn off the gassiest hydrogen that will exit the engine during the startup process before main combustion chamber ignition. These are called radial outward firing initiators or roofies. They're just preventing a large cloud of hydrogen from gathering

and potentially detonating, but as you know, could be very, very bad. But the engine isn't chilled down just to protect the engine from the cold temperatures. It's also to protect the propellant from the relatively warm engine even when it's just resting at ambient temperatures before the engine is running. If the propellant is beginning to boil before it reaches the impellers and the pumps, it can cause cavitation or basically bubbles in the liquid, and those bubbles can actually chip away at the pumps and completely ruin them. Elon Musk talked about this a little bit when we

were talking about the Raptor engines pumps. So what you're trying to avoid is cavitation or bubble generation. If you start generating bubbles, the bubbles will actually eat away at your blades. Like it's weird like bubbles would chip away metal, but they will. And, and if you cavitate too much then you're gonna just be a bubble generator and you'll lose pressure and starve the engine. And not only can those bubbles chip away at the pumps, the bubbles can also cause the pumps to over speed and or starve the engine of the propellant and potentially create stoichiometric

conditions, which is the most energetic and highest temperature conditions and of course, those can be catastrophic. Getting an engine down to operating temperatures is absolutely vital and it's monitored very closely by the engine's computers that run the engine. In fact, it was a fear that one of the RS-25 wasn't cooling down prior to ignition that caused the first scrub of Artemis 1 on its very first launch attempt in August of 2022. It's a scrub. It is officially a scrub? Officially. All right. Launch director called it. Launch director called a scrub. Engine number three's temperature sensor

wasn't showing the engine had cooled down enough to the required temperatures, so the launch attempt was scrubbed for the day. It turns out the engine was likely indeed at the right temperatures, but the sensor itself was erroneous. But it just goes to show how important engine chill down is. If an engine's not cooling down as expected, no diligent launch conductor would proceed towards liftoff. Some liquid fueled rocket engines don't use any cryogenic propellants such as hypergolic propellants. Hypergolic propellants are those that will spontaneously combust when they come in contact with each other. But hypergolics are

happy as a clam at room temperature, and because of that, the engines don't really need to be chilled down. In fact, many intercontinental ballistic missiles can sit for years fueled up and ready to go, and can be fired within a moment's notice. There've been a handful of US-based hypergolic orbital rocket engines such as the LR-87 used on the Titan 2 or the AJ-10 on the Delta 2's upper stage. But the Soviet Union developed a ton of hypergolic rocket engines. For those of you who have watched my entire history of Soviet rocket engines, video may know

the Soviets well, specifically propulsion engineering legend, Valentine Glushko, loved hypergolic rockets. Just take a quick glance at this family tree that we put together. Any orange or yellow engine near the bottom of the chart is hypergolic. Yeah, that's a lot of hypergolic rocket engines. Some of the more noteworthy ones are the RD-275s on the first stage of the Proton rocket or the mighty RD-270. That was a full flow stage combustion cycle engine designed for a massive lunar rocket that sadly never left the drawing board. Okay, so we've chilled our engine down or we're using hypergolic.

So we're already at operating temperatures, but now we need to get the pumps spinning and get them up to operating pressures of potentially hundreds of Bar and tens of thousands of RPM. Perhaps the number one rule with rocket engines, and I guess the universe in general is pressure always wants to flow from high pressure to low pressure. So engines need to have extremely high pressure upstream in order to not send a flame backwards through the system, which would inevitably lead to a catastrophic failure. As you likely know, having a simple pressure fed engine is easy.

You just open some valves and ta da, you're running an engine. The rep propellant is already stored at high pressure, so it'll naturally flow into the combustion chamber at the necessary operating pressures. But what about an orbital class rocket engine that has turbo pumps that spin at absurd speeds? We're talking high power, high chamber pressure engines. Take a look at an open cycle turbo pump powered rocket engine while it's running. Notice the fuel pump and the oxidizer pump are connected by a shaft to a turbine that powers them. Turbines can be insanely powerful with some

engines making hundreds of thousands of horsepower from just a simple turbine. So what powers the turbine? Where does that power actually come from? Well, there's usually a preburner or a gas generator of some kind that's kind of like a rocket engine on its own right and it's sole purpose is to produce energy to power the turbine. But take a closer look at the gas generator. It's fed by the pumps. The pumps are powered by the gas generator, but the gas generator is fed by the pumps, but the pumps are powered by the gas- Okay, yeah,

you, you get the point! Okay, here's the second step of starting a liquid pump fed rocket engine: spin up. Now, there's a handful of ways you can actually get the pumps up to speed, but perhaps the most common is by utilizing a separate high pressure gas to get the pumps spinning. This can either be supplied with an onboard system like helium stored in a high pressure COPV or Composite Overwrap Pressure Vessel, or it could be supplied by ground systems where weight doesn't matter since it doesn't contribute to the overall mass of the rocket. In either

case, high pressure helium or nitrogen is pumped into the gas generator to get the turbine spinning up to operating speeds. For a short period of time, the engine's pumps are being powered by basically a cold gas thruster, which isn't very efficient. A cold gas thruster has a low specific impulse or a low efficiency, and you would want to continue to run the engine off the system any longer than absolutely necessary. But it's simple and reliable since it's just opening up a valve and letting high pressure gas flow into the engine, into the preburner or gas

generator through the turbines until the pumps are spinning fast enough to begin combustion. Some engines will use a tiny little rocket motor called a starting cartridge to do pretty much the same thing. It's quite literally kind of a small rocket motor of either solid propellant or hypergolic propellants that act as a gas generator, almost like the cartridge in an airbag whose sole purpose is to generate a lot of gas to quickly inflate an airbag. So for engines that require multiple restarts, such as upper stage engines that might have multiple burns to get to their desired

destinations or even SpaceX's Merlin engine, which has to light up three to do that boost back re-entry and landing burn on a Falcon 9, you have to have enough nitrogen or helium or starting cartridges to get through every single one of those startups. We should probably take a quick little second here and mention the Rutherford on Rocket Lab's Electron rocket, that thing's electric pump fed. And of course we talked about that in my "How do you power rocket engine" video, but now in context of starting, just think about how much easier it would be to

start an electrically pumped edge engine. You literally just send electricity to pumps and work on the timing, and it's pretty much done from there on. Like it'd be so easy to tune and figure out the timing when it's all just direct correlated to the speed of a motor. Now, it might be tempting to think you could just use a small electric motor to get these spinning similar to how a car engine starts. Just get the pump spinning for a brief moment and then let the engine run itself. The problem is the power requirements for these

pumps is insane. We're not just talking about peak rpm, we're talking about horsepower, so RPM and torque. Or more specifically, RPM multiplied by torque. In fact, the RS-25s fuel preburner delivers about 200 horsepower per kilogram. Think about that. How big is a 200 horsepower engine or motor? I'm fairly certain that you won't find one that's only one kilogram, but there's another way to spin up an engine that doesn't require a separate source to get the pumps up to speed. It's called bootstrapping or tank head or sometimes called deadhead starting, and it's where you carefully allow

the engine to initially light up using only the tank pressure and the energy in the thermal difference between the propellant and the engine. The RS-25 on the Space Shuttle and SLS do this. At the beginning, the turbine will begin to spin up because the hydrogen flowing through the engine and the walls and the combustion chamber and the preburner boils off like we mentioned before. But instead of just having it bleed out and go to the flare stack, during the startup process, we can actually run that high pressure gas through the preburner into the turbine to

begin spinning it. Then there's a very delicate and precise dance to let in some oxygen and light the preburners under low pressures. That weak combustion will slowly gain pressure as the pumps begin to spin up over the next few seconds. As the pumps increase in speed, the pressure rises, which increases the power of the preburner, which then increases the speed of the turbine, which increases the speed of the pumps, which increases the pressures. Yeah, it just keeps going that way and keeps rising until the pumps are at adequate speed for main combustion. Fireflie's Tom Markusic

mentioned they're looking into trying to bootstrap their Alpha rocket's second stage engine, the Lightning, if at all possible to remove the nitrogen spin start system from the upper stage. We do a nitrogen spin start from the ground. Although we've recently did a series of tests where we're trying to figure out how low we can go on that, that spin start when before the bootstrap itself. So we've definitely been playing with the level of spin start that's required, but we do spin start them. And, yeah, what we found that we is, we can probably just tickle

this thing with a little nitrogen at altitude and get it to run much lower... very low spin start requirements. Yeah, the dream would be to deadhead start it, but we haven't gotten that brave yet. But bootstrapping is definitely a feedback loop that takes very precise control to get up to speed. In fact, bootstrapping is extra difficult because the engine spends a lot of time in what's called a transient, which is the time between being off and up to full power. And this is perhaps one of the most complicated and challenging pieces of this puzzle. Transients

are the in between moments. So in between the engine being stationary and at full power and vice versa, or even between throttle settings, the engine is actually in a transient state, but once the engine is running and at a steady state, it's relatively easy to keep running. But why is the transient so hard? Well, remember this whole chicken and egg scenario we had. We're just getting the pumps up to speed. How do the pumps run if they're powered by a gas generator that itself runs on the pumps? While the biggest challenge with the whole startup

process is the fact that everything is intertwined and a change anywhere in the system will immediately have an effect upstream and downstream from the change. So let's say you're trying to bootstrap an engine such as the RS-25, like we touched on earlier. The first thing you're going to do after you purge and thermally condition the engine is you're going to introduce some liquid hydrogen into the pumps and the preburner on its way in it will flash boil into a gas that gas expands and that expanding gas is what's going to start to blow through our

turbine and get our pumps spinning up. Stoke Space CEO, Andy Lapsa, explained this to me when I was standing with him on their test stand with their incredible engine that bootstraps. The start transient and to some extent also the shutdown transient are some of the highest risk moments in a rocket engine, right? And if you think about it, like our little engine, a thousand horsepower in the pumps bigger engines have tens of thousands of horsepower going through the pumps, and they go from zero to full power in maybe a couple seconds, right? So you've gotta

control all those horses and make sure they're running in the right direction during that time. So in early development is usually one of the higher risk activities for rocket engine. Quick little side note, wow, this is honestly one of my favorite conversations, and Andy Lapsa and his company, Stoke Aerospace, is working on an absolutely incredible rocket. Okay, but that boil off is spinning up the turbine, which spins the pumps, so now the propellant is starting to flow a little faster. As long as the pressure in the system continues flowing in this direction during this transient

region, you're golden. The problem is there's often slight delays between action and reaction. So it's almost like there's waves of pressure changes moving throughout the system, and if those get too extreme, the pressure gradient might actually start to flow backwards and you can stall out your startup sequence, or even worse, potentially blow up your engine. Take a valve opening, for example, as a valve opens and introduces say the oxygen into the preburners, that too will have an effect on the pressure and the flow, and then the preburner ignition will affect the pressure and the flow.

It's all just a constant back and forth feedback loop. But now we've touched upon another piece of the puzzle: ignition. We mentioned how the preburner will light under relatively low pressure, but how does it actually light what starts the ignition process? Just because you combine your fuel and your oxidizer doesn't mean you're going to get combustion. In fact, most repellants, to the dismay of rocket engineers, will kind of happily coexist in the same space and may even exit the system without unleashing their fiery goodness at all. Or even worse, they may accidentally ignite at the

wrong time or place, which is often catastrophic. Perhaps you've heard the term hard start. This is basically when the propellant combust at the wrong ratios or perhaps at the wrong time or place. The worst hard starts can severely over pressure the engine or maybe overspin the turbines or cause an energetic detonation that can completely destroy the engine and sometimes even the test stand and or the rocket that they're actually attached to. So the timing and precision of ignition is vital. Like most engines, rocket engines typically require an external source of ignition to begin the combustion

process. But what if we just use propellants that don't need an ignition source at all? Yes. You know, hypergolics, we've already touched on them earlier and, and almost every video I've made, but being able to spontaneously combust is one of their biggest advantages. And the fact that they don't need any additional considerations to keep liquid at cryogenic temperature is just really a pure bonus. I mean, don't get me wrong, they're a total pain in the butt in other ways like being super toxic and carcinogenic, but in the case of ignition and sheer reliability, they're hard

to beat. Which is why they're so common for maneuvering thrusters and on orbit engines where reliability and long duration missions matter since they won't boil off. But for LOx based propellants, if you put your fuel and oxidizer in contact with each other, it likely won't lead to combustion. If you took liquid methane and you poured it into a container of liquid oxygen, don't really expect anything to happen. Even if you swirled it around without a source of ignition, the two will coexist surprisingly peacefully. Even in gaseous form, this is true for the most part, but

as a gas, it's much more likely to ignite with just a small ignition source. So non-hypergolic fuels and oxidizers not only need to be properly mixed for stable combustion, they also need an ignition source to initiate the combustion process in the first place. But once stable combustion has been achieved, and assuming there's adequate and stable flow of your propellants, the initial ignition process is no longer necessary and it can actually be self-sustaining. The combustion in the chamber acts as a continual source of ignition. Okay, so we've actually got to light up our propellants. Perhaps the

simplest option outside of hypergolic propellants is what the Soviet Union decided to do with the R7 and what Russia still uses today on the Soyuz, which is basically just some giant matchsticks. Yep, literally just some large braces with a pair of pyrotechnics that are put inside each combustion chamber and are all lit prior to the propellants arriving in the combustion chamber. Now, this is also possible because the gas generator on the RD-107 and RD-108 is basically just a monoprop thruster, so it doesn't need a separate source of ignition. So you could, the main thing you

really have to worry about on those engines is actually igniting the main combustion chamber. ut it's not very elegant and it's probably not a great option once you're in space either. So for that, we'll have to look into some other options that are capable of space starts and potentially restarts. Perhaps one of the most common methods is to basically use a spark plug. Okay, maybe not your off the shelf spark plug from a Ford Fiesta, but the concept is the same, take a high electrical voltage and have it jump a gap, which provides the perfect

opportunity to ignite the propellant with ionized electrical energy. As you can probably imagine, it does require a good amount of electrical energy to power a rocket engine size spark plug. So rockets using this approach may need to carry heavy batteries or utilize ground support equipment to run current through the igniter at startup. Another method is similar to a spark igniter, actually, it's again similar to the automobile industry with diesel engines, and that's using a glow plug. Glow plugs also require a lot of electricity to convert to heat, similar to an incandescent light bulb, an electrical

heater coil, or a toaster. There needs to be enough heat energy on the surface of the glow plug to initiate combustion of the propellants. And again, once stable combustion has been achieved, the glow plug can just be turned off. Another method is to use a laser. Lasers can very precisely focus the beam of energy and tune it to excite the propellant molecules, so it has the potential to be more efficient than other electrically driven devices. Now, it's not very common yet, although Airbus does have some variance of optical ignition for some of their boosters. These

methods are all great and commonplace, but they all require massive amounts of electrical power, which is likely stored in big and not so energy dense batteries. So what if you actually could just use a chemical reaction to start the ignition process? This is actually pretty common and extremely effective. In fact, if you've ever seen a Falcon 9 or Heavy launch, you may have noticed a green flash right at the beginning of ignition. That, my friends, is the injection of a pyrophoric ignition fluid coming in contact with oxygen to begin stable combustion. Pyroforks are a type

of hypergolic. It's a fluid that will spontaneously and virtually instantaneously react with oxygen. So just inject some pyrophoric fluids into the combustion chamber along with oxygen and ta-da, you've got yourself a flame. SpaceX uses an ignition fluid called TEA-TEB or triethyl aluminum triethyl borine. It's stored in their own canisters onboard the rocket and allows for multiple restarts of their Merlin engines. Now, not every engine is connected to a TEA-TEB canister, only the engines that would do a relight, so three of the nine Merlin engines on the booster and the vacuum Merlin engine on the upper

stage. But on the ground, the Falcon 9 and Falcon Heavy are lit using ground supplied TEA-TEB because why not keep those systems and their weight on the ground if you don't need to carry that mass off the launch pad? Now, SpaceX has taken a little bit of a hybrid approach to ignition with their new Raptor engine. The Raptor engine uses what are called a torch igniters. Torch igniters are basically just a fancy version of the lighters that you might use to light a candle. You may or may not hear them called ASIs or Augmented Spark

Igniters. The ASI has its own little small spark ignition and then its own supply of methylox that allows for the torch to burn. So it in itself is almost like a little flame thrower that's initially lit with a spark ignition. Then that torch remains lit through the rest of the startup sequence. So it's kind of like the cascading effect of the solid rocket booster ignition, this uses a smaller electrical ignition source to ignite a small baby flame thrower. But surprisingly, this is only done on the preburners because believe it or not, the Raptor engine has

no igniters in the main combustion chamber. So you got torch igniters for the main chamber but Raptor 2 has no torch igniters in the main chamber. So you can see it's much cleaner around the chamber area. How do you... how does it light then? Well, that's secret sauce! Although Elon was tight-lipped about how exactly they get away with ignition in the main combustion chamber, it's actually assumed that SpaceX is getting away with homogenous combustion or basically spontaneous combustion when the methane and oxygen come in contact with each other. But the main reason that they can

get away with that is because in the Raptors full flow stage combustion cycle, both the fuel and the oxidizer are preburned in their own preburners, and then they arrive in the main combustion chamber as hot gases already. So take the fact that it's a gas-gas interaction along with ridiculously high pressures seen in the combustion chamber, no ignition source is necessary, the propellant is already plenty hot and bothered and ready to burn. Well, I think that pretty much covers all the major parts of starting a rocket engine, but before we get into some examples, let's actually

cover one more challenge: starting a rocket engine in space. Starting a rocket engine on the ground is pretty challenging, but at least it has a few things going its way. First off, the rocket can utilize a lot of ground support equipment for startup. This means the rocket doesn't need to provide crazy amounts of high pressure starting propellant, electricity, ignition fluid, or whatever else you may need. And best of all, gravity is pulling down on the rocket and the rep propellant held inside. So the more dense liquid propellant will sit at the bottom of the tanks

while only at the very top of the tank. Opposite of the feed tubes will have gases. But the same thing isn't true when trying to light an engine in space. Well, not just specifically in space, but more when you are not accelerating or not on the ground because when not accelerating, everything is in the same inertial reference frame. In other words, when the engines are off and the spacecraft is just floating in space, there is no acceleration, meaning the propellant is more or less just a blob floating around in a tank. So in order to

start a rocket engine in space, we first need to make sure the liquid propellant is at the bottom of the tank, so the pumps engine and or the thrusters can suck up the liquid as intended and not get gas bubbles, which can be really bad. A common method for doing this is by using what's known as ullage thrusters. This is where you either use a small solid rocket motor or cold gas thruster to settle the propellants. You can almost think of it like a butterfly net trying to catch butterflies. By moving through space, it'll basically

catch and collect all of the liquid propellant. You'll see footage of ullage thrusters at stage separation for many rockets. This incredibly iconic imagery of the S-IVB ignition is actually from a Saturn 1B, but it's the same stage that was used on the upper stage of the Saturn V. Not only did the ledge motors help separate the stages, but it also did the propellant settling. This may also need to be done for any restart in space. Say you have a circularization burn or a geostationary transfer burn or a direct to geostationary orbital insertion burn, or whatever

you might have. If it's a liquid fueled rocket engine, you'll likely have to do an ullage burn. And similar to a rocket needing to do a spin start again. If you need to do multiple burns, you might need to either have multiple solid rocket cartridges or use a cold gas thruster so you can fire it up whenever you need to. Cold gas is pretty common for ullage thrusters since it might also be used as maneuvering propellant for stages. Even the Falcon 9 booster has to use cold gas thrusters to settle the propellant before boost back

and entry burns. You can use a cold gas thruster in space because it uses gaseous propellant stored at high pressure, so there is equal pressure throughout the entire tank, the gases won't blob about. But some spacecraft don't need ullage thrusters at all. In fact, many maneuvering thrusters run on liquid by propellants, usually hypergolics, and when they're in space in order to do precise maneuvering, they can't do an ullage burn every single time they need to do a little maneuver. Take SpaceX's Dragon capsule, for example. As it approaches the International Space Station, you'll see it firing

its Draco thrusters like crazy. These little tiny impulses help precisely control and maneuver the vehicle with extreme accuracy. The Dracos run on hypergolic liquid propellants. So how do they fire these engines in space when the tanks have liquids in them? There's a few ways to actually do this, but perhaps the most simple is with a propellant management device or a PMD. This can be a simple device like a vein screen or sponge to wick the propellant towards the tank outlets using surface tension and some other really cool tricks. Another way is with tank bladders. Yep,

believe it or not, many in-space tanks will have a bladder inside to separate the liquid from the pressure and gases. This way, the propellants can still be pressurized to necessary pressures and the bladder will help keep gas bubbles out of the liquid. These are most commonly used on spherical tanks, but there's also some piston tanks that do a similar thing, but these are more common on elongated cylinders, as you can imagine. Although it is a piston, it's usually just a moving seal. They're still actually pressurized by helium. Most in space thrusters and engines will utilize

one or a combination of a few of these devices to best handle propellants in space. But what if I told you that there's an even crazier way to light an engine in space without ullage thrusters or bladder tanks or anything? Have you ever noticed the graded fence looking section of the soy user proton rockets, or maybe you've seen these cutouts on the inner stage of the Titan 2? Well, these are necessary to perform the ultimate endearing space events, hot firing, otherwise known as hot staging. This is where you light the engines on the upper stage

while the stage below it is still attached and running. Yep, you just light that engine while there's still a rocket attached to the bottom of it. This is called "Fire in the Hole" for obvious reasons. But by doing this, propellant is already experiencing acceleration. So liquid will obviously be at the bottom of the tanks, but when you light an engine, the exhaust will have to go somewhere, and now you know why there's so much of an open gap between the stages. All right, that pretty much sums up how to start a rocket engine, but now

let's actually go over a real life example and see if we can follow along. Okay, my friends, I think in order to really put all this together and dive really deep into this, we need to get inside a rocket engine during startup and go step by step through the entire startup process of one of the most complicated rocket engines, the RS-25. If we can walk all the way through this one, I'm not gonna say that we've mastered it, but we should definitely have a very good understanding. Oh, what's best of all with the RS-25 is

we have some incredible insights on how exactly this engine starts up because the startup process is actually really, really well publicly documented. So we can go really, really deep and actually explain what's going on inside the engine for the entire startup process. And for this, I actually bought a book called Space Shuttle Main Engine, the first 10 years by Robert E. Biggs, and it's an awesome book. It goes really, really in depth into this process. But the coolest thing about this book is it has these charts showing the pressure and valve openings during the startup

process. So we'll draw these out a little bit more cleanly along with a diagram of the engine so we can go just step by step really slowly and watch all the things change throughout the entire process. Now, I must warn you again, if you haven't watched my how to power a rocket engine video, you really will need to have a good understanding of the fuel rich closed cycle engine before you go through this portion of the video because we get really, really in depth with the exact components. But as a quick review, the RS-25 has

two preburners since it's fuel rich closed cycle, both pre burners are fuel rich, meaning all of the fuel will flow through the preburners, one preburner powers the oxygen pumps and the other powers the fuel pumps. Then there's a small boost pump on the oxygen side that feeds the preburners, the higher pressure liquid oxygen that they need to operate, and there's a handful of valves, the prevalve that shuts off the tanks to the engines. Then there's the main fuel valve, which feeds the pre burners and the regen cooling channels. But the regen system has a separate

valve called the chamber coolant valve, which can be throttled to redirect fuel between the main combustion chamber and the regen system. Then there's three oxygen valves, one that feeds each of the three combustion chambers, so both of the pre burners and the main combustion chamber. There's a few recirculation pipes that feed boiling off gaseous propellant either back into the tanks or vents them out into the atmosphere. There's also what's called ASIs or Augmented Spark Igniters. There are three sets of ASIs, one in each preburner and one in the main combustion chamber. These ASIs have their

own fuel and oxygen supply lines and will be the first items to receive each propellant in the system. Now there's obviously a million other small tubes and pipes and sensors littering the real engine, but to try to keep this understandable, I think we'll just leave it at this for now. As we know, the first thing a cryogenic liquid propellant engine needs to do is be purged and thermally conditioned for launch. So the RS-25 first goes into what they call the Start Preparation Phase. During this period, the oxygen side of the engine is purged of moisture

using nitrogen, and the fuel side is purged with helium. Then the engine can have the cryo propellants in its system, so they would open the main fuel pre valve that allows liquid hydrogen to flow through the fuel pumps down to the main fuel valve. There's a small recirculation flow that either dumps some of the hydrogen or might even pump some of it back into the fuel inlet. Liquid oxygen fills the oxygen side of the engine by opening the oxygen prevalve and then it flows through the oxygen pumps up to three valves, which will all need

to be very precisely controlled during the startup process. The propellants are inside the engine for an hour or more to fully condition the engine Throughout the stage, the main engine computer is constantly monitoring the pressures and temperatures 50 times per second to make sure everything is looking good. Then at about four minutes before engine ignition, there's a final engine purge with helium downstream of the main fuel valve. Now assuming everything is looking good to go and it's all within the predefined parameters, the computer will go into what's called the engine ready status At three seconds

before engines start, the bleed valves for both the oxygen and hydrogen lines are closed and the engine waits for the start command. The engine is now fully thermally conditioned, purged of all moisture from the atmosphere, and ready to take on the challenge of what's to come because here's where things get super complicated. The moment the start command is received, the first thing that will happen is the main fuel valve is fully opened. This valve actually takes about two thirds of a second to fully open even at its maximum rate. At the same time as the

main fuel valve is opening, the ASIs are powered up and ready to ignite any propellant that they come in contact with. So ignition of the ASIs needs to be before any mixed propellants are present in the system. As you can imagine, despite the engine being cold by human standards, the actual components downstream of the main fuel valve are still relatively hot, at least compared to the liquid fuel since they haven't been soaking in that cold cryo liquid propellant, like the parts of the engine that are upstream of the main fuel valve. This energy from the

latent heat of the engine is enough to begin spinning the turbine. The engine initially is starting up and beginning to spin up as an expander cycle engine, otherwise known as bootstrapping or deadhead starting like we mentioned before. But it's facing some pretty major thermodynamic instability. As the propellant flash boils, it'll create uncontrollable but predictable oscillations, and that's one of the hardest things here. Since everything has a little bit of a delayed reaction, you need to actually predict when the peaks and dips and the pressure will occur during this process. These oscillations will occur for about

1.5 seconds until the main combustion chamber reaches what's known as prime. Now prime in this example is when the pressure is stable on each side of the injector, well, specifically when the mass flow rate is stable. So prime occurs in all three combustion chambers when there's stable flow between the pumps and the chambers, each combustion chamber's targeted prime time is very important so as to keep things moving in the right direction. So we've introduced fuel into the system and mostly flowing through the pumps and the turbines and beginning to get the engine spinning. Fuel is

flowing through the three sparking igniters and they have electrical power. Now we've got to start introducing oxygen. The first thing that's going to receive oxygen is the igniter inside the fuel preburner. The system will start to flow liquid oxygen basically as soon as that fuel preburner LOx valve starts to even open. At just 5% opening of the fuel preburner oxygen valve, LOx is directed straight to the igniter. The timing of when the valve starts to open and when the oxygen will start flowing into the igniter coincides perfectly with the first dip in pressure during those

pressure oscillations. This helps make sure that you don't have backflow of hydrogen up through the oxygen system, but also ensures that you're setting up to have the right mixture ratio for the first bit of combustion. From here, the fuel preburner oxygen valve has to do a lot of work to kind of ride the waves of those oscillations to flow through the highs and the lows of the system, and again, they can't react to the pressure oscillations because of the delayed reaction between the valve opening and the events happening downstream. So these oscillations had to be

precisely documented and predicted. In fact, every time you see the valve moving during this period, you can pretty much assume that there was an engine that blew up and changes had to be made from the lessons learned. By this point, the pumps are spinning pretty quick and the system is getting closer to reaching an equilibrium in all three chambers, or again, hitting prime. At 1.25 seconds, it'll do a speed check of the fuel pump turbine. It was found that the pump needed to be above 4,600 RPM in order to continue moving forward into fuel preburner

and main combustion chamber ignition. Otherwise, there wouldn't be enough hydrogen pressure to overcome the main combustion chamber pressure. At 1.4 seconds, the fuel preburner hits prime right when there's that large dip in pressure and then that causes a rapid rise in pressure. Now, this causes the fuel turbine to spin up very quickly. In fact, there's virtually no back pressure after the turbine from the main combustion chamber yet, since that hasn't hit prime at this point. So the turbine spins up ridiculously fast. If left unattended, it would actually overspeed the turbine and would likely cause a

catastrophic failure. So making sure the combustion chamber hits prime at exactly the right moment is extremely important to provide the necessary back pressure so the turbine doesn't spin up too fast. Now notice we are spinning up the fuel turbine and pumps first. This ensures that the whole system has higher fuel pressure and ratios that will ensure a cool fuel rich start. Obviously, it can't be so fuel rich that it floods a system and can't be lit, but it's better to stay too rich rather than to lean and be anywhere closer to stoichiometric. So let's actually

walk backwards now a bit to 0.2 seconds after engines start, that's when the main combustion chambers oxygen valve will start to open to flow oxygen into the main combustion chamber igniter. The main combustion chamber valve is slowly open to just under 60% open. The delay and the slow rate of opening makes it so the main combustion chamber igniter has oxygen at 0.85 seconds after engine start, and this will begin the main combustion chamber ignition at a nice safe fuel rich mixture ratio. Main combustion chamber hits prime at 1.5 seconds, which causes the pressure in the

main combustion chamber to rise rapidly and helps prevent over speeding the fuel turbine with an increase in back pressure and therefore resistance on the turbine. Okay, so let's actually go backwards again in time and go through the oxygen preburner system. The oxygen preburner valve initial opening is just 0.12 seconds after engines start, but it's designed in a way that its initial opening is all it takes to power the oxygen preburner igniter. The timing of this makes it so the oxygen preburner igniter is lit at 0.95 seconds, just 1/10 of a second after the main combustion

chamber igniter. Now, the oxygen preburner valve is designed to not really flow oxygen all the way through until about 46% open. Again, it's very important that the flow of oxygen is generally conservative. It's a careful balance between giving the system enough oxygen to begin combustion and help to provide the power necessary to run the engine, but not giving it too much oxygen where the engine can start to run lean and experience damaging temperatures. As you can imagine, the oxygen preburner valve helps control the power of the oxygen preburner, which controls the speed of the oxygen

pumps turbine, which is what controls the speed of the oxygen pumps, which is what controls the pressure in the oxygen system. So that one valve actually has a huge effect on the entire engine. The oxygen preburner is the last of the three chambers to hit prime at 1.6 seconds, and that's again done to ensure the oxygen pressure doesn't get too high in the system. At 1.7 seconds after engines start, the main engine computer verifies that all three combustion chambers had proper ignition and are operating normally. At the beginning of this phase, where the engine has

all three combustion chambers lit and primed, the main combustion chamber is at roughly 25% of its rated power level, but it's far from stable. And if you check out slowmo footage around this time, you'll actually notice the rocket engine's nozzle just oscillating like crazy. You might also notice there's some big spikes of flow separation going up into the nozzle, and that's because the pressure inside the nozzle is actually still lower than atmospheric pressure at sea level. So the air's atmosphere is actually kind of creeping back up into the nozzle and it has these big oscillations

as those shockwaves kind of form, and it has a bit of an instability at this point. So in order to increase stability and pressure and increase safety margin, the chamber coolant valve, which was fully open up until this point is throttled down to 70%. This forces more fuel into the main combustion chamber. It does this for 0.4 seconds to help absorb variations in the pressures and temperatures. Now, up until this point, the main engine computer's actually been operating in open loop control, meaning it's only receiving pre-programmed commands, it's not like reacting to anything, it's just

a set of orders basically. But at 2.4 seconds after engines start, the computer goes into a closed loop control, meaning that throughout the rest of the ramp up to rated power level, the main engine computer is actually reacting to the pressures and temperatures and making adjustments accordingly to try and follow the path to ramp up. Most of this is done by controlling the oxygen preburner oxygen valve because like we said, that really has a huge effect on everything else. But remember, again, this is all pretty tricky since there's a pretty big delay between reading the

combustion chamber and the temperatures and reacting to it. It's not actually like the reaction time, but just how long it physically takes to say open a valve and for those changes to make a difference downstream. So this all has to be done just incredibly carefully. At 3.8 seconds, the system goes into fully closed loop mixture control. So not just closed loop has it head been operating, but now it's fixed even with its mixture ratio, meaning only the fuel preburner oxygen valve is used to ramp up the correct mixture ratio in the main combustion chamber trying

to get to that 6:1 ratio, which will occur right around 5 seconds. And this also means the engine has fully reached operating power levels and during this ramp up period we're awarded with those gorgeous shock diamonds or mach diamonds, and they're just so perfect on the RS-25. It's honestly crazy to me that this all happens in such a short period of time and that they could ever achieve the reliability they did with a Space Shuttle and now with SLS. And now let's take a look back at the chart again one more time, now that we

hopefully understand a little bit of what's actually going on. We can see how each of those dips of pressure and the corresponding valve positions was obviously a learning lesson. There were 19 tests through 23 weeks with eight turbo pump replacements just to get through the first two seconds of startup. And it took another 18 tests, 12 weeks, and five more turbo pump replacements just to get up to full power. That's honestly actually fewer bits of hardware than I would've guessed, but it still helps appreciate how hard and how expensive engine development can be. Not to

say there weren't additional lessons to be learned once they developed a reliable startup sequence, but operating at a steady state is orders of magnitude easier than the dynamic startup process. Well, I think that should do it for an example. I mean, that's about as deep as I can get, so I think it's time we wrap up by touching on throttling and the opposite of startup shutting the engine down, and then we'll do some final thoughts and a summary. Of course, startup isn't the only dynamic situation a rocket engine faces. Sure, once an engine running, it's

in a steady state and there shouldn't really be any real surprises. But what happens when an engine throttles down for maximum aerodynamic pressure or MaxQ maintaining desired peak G loads or even when it's landing? As you can likely imagine, this varies very much from engine to engine, and frankly, this is probably a topic for another video. But in short, engines generally throttle by reducing the flow to the preburner or the gas generator, so usually with one of the control valves and often by reducing the flow of oxygen to maintain a fuel rich state. The same

thing goes for shutting an engine down. It's another dynamic event, and the general rule of thumb is you never want to get close to stoichiometric conditions since engines usually run fuel rich, you do this by first reducing the oxygen flow and then the fuel flow. Generally, you want to shut an engine down pretty much as quickly as possible, but you often have limits on how quickly you can do that to avoid high G loads. For example, when the RS-25 was on the Space Shuttle, there was a limit to how quickly the thrust could decay so

as not to exceed the orbiter's structural limit. The initial oxygen preburner oxygen valve motion was limited to 45% per second. The main oxygen valve also had to be closed at a particular rate, but mostly to ensure there was sufficient back pressure on the turbines so they wouldn't accidentally over speed during the shutdown process. To me, it's just crazy how many considerations there are to absolutely every single input and condition. Like I feel like there's a million lessons to be learned that frankly can really only be learned the hard way by examining scraps of an engine

and trying to figure out what went wrong. I think it's kind of this mindset that drives SpaceX to test their Raptor engines so quickly. They truly believe that by just getting them out on the stand, not treating them as some one-off golden pony or something, they can learn more quickly. There's countless lessons that they have learned that have shaped safe operation during startup throttling and shut down. And I know that many of you are probably wanting to know how exactly the Raptor engine starts. Its startup process isn't as well publicly documented, of course, as say

the RS-25, but it does have some similarities. It of course, has two pre burners, like the RS-25, but the biggest difference is that one preburner is fuel rich and the other is oxidizer rich. So this means the interaction between the two preburners is even more intertwined. Changing the speed of one has a very direct correlation and impact on the other one. In the case of Raptor, you've got an oxygen power head and a fuel power head, and they're different shafts and you've got two turbines and two preburners and, and they're cross feeding one another. So

the start sequence for Raptor is insanely complicated compared to the slot sequence for Merlin. It has to be perfectly precise cause each one relies-. Basically you're doing this delicate dance between the fuel power head and the auction power head and if they get out of sync then you can go stoichiometric in the preburners and melt or explode the preburners. Once it's running, it's a much easier situation. But if you get anything wrong with that star sequence you're either gonna melt or explode the engine. Initial spin up of Raptor is done with either helium or nitrogen.

And as we mentioned before, there are torch igniters in the preburners and likely some kind of homogenous ignition in the main combustion chamber. And from there on, I honestly can't even begin to imagine what goes on internally to balance the startup process. It's no wonder they're firing up raptors about five times every day at this point. They've really gotta get all the kinks worked out because there's going to be so many engines going through startups simultaneously, it has to be perfect. So to summarize, starting a rocket engine is very hard. Some are easier than others,

but you can imagine why it's easy for a company to come up with a rocket engine or concept and very hard for them to get into production and operation. So when a company shares that they've successfully started up an engine and made it all the way up into operating power levels, it's very applause worthy as they've likely made it through the biggest hurdles and development. And it's really fun to see how different programs tackle starting an engine and what techniques they might employ to achieve a reliable engine. But of course, starting a rocket engine can

be extremely complicated, like having valves changing their position by just two degrees within a few milliseconds to avoid catastrophic conditions from pressure oscillations like in the RS-25. Mastering this sequence can take thousands of hours of development, which means millions of dollars in time and hardware, which can mean years and years of hard work and troubleshooting. It is pretty much a miracle to me that people have figured out how to make rocket engines that are so complicated and so reliable considering how close to catastrophe they are throughout the entire startup sequence. But that pretty much does

it. Hopefully this video helps give you an appreciation of what engineers and scientists have to solve before your favorite rocket has ever made it to the launchpad for the first time. It's just absolutely incredible how much work has to go into this. Let me know if you have any other questions or thoughts in the comments below. I owe a huge thank you to my Patreon supporters for helping make videos like this and everything we do here at Everyday Astronaut possible. I also have to give a little extra thank you to our Mission Directors and Commanders

who were there for some of the script read throughs. We did two of them for this script where I'm sitting there reading through the entire thing in Discord and people are giving me comments and feedback as we read. We kinda go like paragraph by paragraph. We developed that script a lot just from those read throughs. And then the rest of our Mission Specialists and Pilots ended up doing a script read through on their own and giving a lot of good feedback and this video has really evolved a lot and I learned a ton making this,

which to me is always kind of the bookmark on whether or not I think a video is going to be good. It's just based on how much did I learn and I learned so much. So I really hope you guys do too. So if you want to help give your feedback and be involved in the making of these videos or if you just want to help financially support, head on over to patreon.com/everydayastronaut. And while you're online, be sure and check out our awesome merch store for shirts like this, the Heliocentric Shirt, this is actually based

on the album art from an EP I released as Everyday Astronaut. You can find it on Apple Music and Spotify, et cetera, et cetera called Heliocentric. But also be sure and shop around and check out all of our new cool stuff, including our super highly detailed 1:100 scale metal Falcon 9 model rockets while we have them in stock. Or check out our new Full Flow Stage Combustion Cycle Shirt or awesome dress wear, our Future Martian collection, our Space Shuttle Ejection Suit Hoodie, cute clothes for your little space cadets, and all sorts of other fun stuff

everydayastronaut.com/shop. Thanks everybody. That's gonna do it for me. I'm Tim Dodd, the Everyday Astronaut bringing space down to earth for everyday people.