Why are rocket engines so complex?

[king of nowhere]

well, I know rocket engines are the most complex and expensive part of a rocket. recovering them is the main cost saver. all that pile of criogenic propellant is actually small chip besides it. And I know that they are extremely complex. Just look at all the piping on this thing

 

However, I can't just figure out why rockets are so complex.

I mean, you take a fuel and an oxidizer, you mix them together, put in a spark, and they burn. You orient the exhaust gases backward with the nozzle, and it will provide propulsion. Basically, you should need a pipe for the fuel, a pipe for the oxidizer, a combustion chamber and a nozzle.

Obviously, you need so much more. Can someone explain me why?

[JoeSchmuckatelli]

There is a good video I saw recently about turbopumps and fuel mixing /premixing - did not specifically speak to the complexity of the whole system - but it's a start. 

 

Don't remember where I saw it - but if I find it I will link it 

[mikegarrison]

2 hours ago, king of nowhere said:

well, I know rocket engines are the most complex and expensive part of a rocket. recovering them is the main cost saver. all that pile of criogenic propellant is actually small chip besides it. And I know that they are extremely complex. Just look at all the piping on this thing

 

However, I can't just figure out why rockets are so complex.

I mean, you take a fuel and an oxidizer, you mix them together, put in a spark, and they burn. You orient the exhaust gases backward with the nozzle, and it will provide propulsion. Basically, you should need a pipe for the fuel, a pipe for the oxidizer, a combustion chamber and a nozzle.

Obviously, you need so much more. Can someone explain me why?

You need to control how much fuel you pump in. You need to control how much oxidizer you pump in. You need to mix them. You need to burn them VERY QUICKLY. You also need to extract some power from the combustion in order to run the pumps. And you need to keep all the parts cool enough so that they will survive long enough for you to burn all that fuel.

AND you need to do this as efficiently as possible in order to make the best use of your propellant. And you need it to be as light as possible. And you need it to be as cheap as possible. But also as reliable as possible.

Whether you use hydraulic power or electrical power to control all your valves, you also need plumbing and wiring for that. You probably also need a hydraulic pump and an electrical generator.

Edited October 17, 2021 by mikegarrison

[Spica]

2 hours ago, king of nowhere said:

I mean, you take a fuel and an oxidizer, you mix them together, put in a spark, and they burn. You orient the exhaust gases backward with the nozzle, and it will provide propulsion. Basically, you should need a pipe for the fuel, a pipe for the oxidizer, a combustion chamber and a nozzle.

Well this is basically the textbook case of "the devil is in the details".  In order to build a working rocket engine you need to figure out exactly how to get the propellants into the engine, mix them, ignite them, and send them down the nozzle. 

Here are just a few of the problems you need to solve to make a working rocket engine:
The pressure in the tanks is lower than the pressure in the engine, how do you make sure the propellants only flow "uphill" from the tank to the combustion chamber?
The gases in the combustion chamber are burning at temperatures of over 3000K, how do you stop the engine (this includes any fuel injectors or ignition system too) from melting or otherwise falling apart?
You need to control it, how? With what? 
 

Remember, the precise shape of every part much be chosen to make it as light at possible, what material is strictly necessary? Where isn't it necessary?

Here's an unofficial cutaway diagram of the raptor engine's systems, each one of these solves some problem mentioned above, along with countless others that aren't mentioned.

Edited October 17, 2021 by Spica
additions

[kerbiloid]

As KSP descriptions clearly explain, they are made of parts found by the side of the road.
So, full of random junk left from bikes, old fridges, microwave ovens, truck mufflers....

To hide this fact, they draw various clever schemes.

[Scotius]

Well, at least solid fuel engines are a bit simpler. A bit

[king of nowhere]

let me get this straight, though: if you had a simple system with just the two pipes for fuel and oxidizer, both with a valve and a turbopump (electrically powered, with batteries aside), and of course you still used the cold fuel to cool the nozzle; would the system still work, albeit at a lower efficiency?

[kerbiloid]

No.

As it's depicted above, it has two different turbines for the propellant components (liquid oxygen and liquid methane).

The turbines are spinned by the helium flow, so it needs two helium pipes and one more helium pipe for valve control. From the helium tank.

The fuel tanks should be pressurized  (several atmospheres?) by a gas to avoid the cavitation (gas bubbles in liquid above the sink whirl).

So, the liquid oxygen tank is pressurized by evaporated gaseous oxygen (the pipe on top from below), while the liquid methane tank with nitrogen (see the nitrogen pipe from the nutrogen tank).

Both nitrogen and helium flow should be adjusted, so needs control valves in turn, which are powered... probably by same fluids.

Both methane and oxygen have thin pipes to the ignition system, to first light a small torch which should ignite the main fire, fed from thick methanol and oxygen pipes.

The gymbal is powered by some dedicated hydraulic "work liquid", and as it can tilt in two planes, it has a set of actuators and hydraulic pipes.

The nozzle, the turbines, and other components are cooled by the flow of the fuel components, and require the cooling pipes (of different shape and diameter).

The fuel components get pre-vaporized to make the process controllable and feed the engine with gases. This needs the evaporation circuit.

And all of that require control pipes to put sensors and control valves to implement negative loopbacks to parry the sudden  pressure changes and to collect info about the engine system work.

The mechanical parts use also some lubricants, some of them may be liquid and need their pipes.

All these parts arre vibrating at their own frequencies, and the superposition of their vibration frequencies causes secondary frequencies of vibration.
Some of these frequencies can suddenly match the resonant frequency of some of these numerous parts and cause its destruction.
So, these frequencies must be researched, and supressed by the amortization system, which affects the shape of the parts and can include the springs and its own fluids and thus pipes.

***

Being simple, it would burn, or crash (and then still burn).

The simplicity is not a skeleton key. Try to weld the train carriages together. They will get simpler, but the train will overturn on the first turn.

[king of nowhere]

1 hour ago, kerbiloid said:

No.

As it's depicted above, it has two different turbines for the propellant components (liquid oxygen and liquid methane).

The turbines are spinned by the helium flow, so it needs two helium pipes and one more helium pipe for valve control. From the helium tank.

The fuel tanks should be pressurized  (several atmospheres?) by a gas to avoid the cavitation (gas bubbles in liquid above the sink whirl).

So, the liquid oxygen tank is pressurized by evaporated gaseous oxygen (the pipe on top from below), while the liquid methane tank with nitrogen (see the nitrogen pipe from the nutrogen tank).

Both nitrogen and helium flow should be adjusted, so needs control valves in turn, which are powered... probably by same fluids.

Both methane and oxygen have thin pipes to the ignition system, to first light a small torch which should ignite the main fire, fed from thick methanol and oxygen pipes.

The gymbal is powered by some dedicated hydraulic "work liquid", and as it can tilt in two planes, it has a set of actuators and hydraulic pipes.

The nozzle, the turbines, and other components are cooled by the flow of the fuel components, and require the cooling pipes (of different shape and diameter).

The fuel components get pre-vaporized to make the process controllable and feed the engine with gases. This needs the evaporation circuit.

And all of that require control pipes to put sensors and control valves to implement negative loopbacks to parry the sudden  pressure changes and to collect info about the engine system work.

The mechanical parts use also some lubricants, some of them may be liquid and need their pipes.

All these parts arre vibrating at their own frequencies, and the superposition of their vibration frequencies causes secondary frequencies of vibration.
Some of these frequencies can suddenly match the resonant frequency of some of these numerous parts and cause its destruction.
So, these frequencies must be researched, and supressed by the amortization system, which affects the shape of the parts and can include the springs and its own fluids and thus pipes.

***

Being simple, it would burn, or crash (and then still burn).

The simplicity is not a skeleton key. Try to weld the train carriages together. They will get simpler, but the train will overturn on the first turn.

this is exactly the kind of technical answer I was looking for.

I hope I'm not bothering you if I ask some additional questions, since you are apparently very knowledgeable about this:

Quote

The turbines are spinned by the helium flow, so it needs two helium pipes and one more helium pipe for valve control. From the helium tank.

why use expensive helium instead of cheap nitrogen? or why not use an electric turbine?

Quote

So, the liquid oxygen tank is pressurized by evaporated gaseous oxygen (the pipe on top from below), while the liquid methane tank with nitrogen (see the nitrogen pipe from the nutrogen tank).

why not have the methane tank pressurized by evaporated gaseous methane? ok, I know the raptor is supposed to be fueled by supercooled methane, so not much will evaporate, but what's to be gained there? On second thought, the answer is probably above my ability to comprehend.

But why use nitrogen here and helium for the turbines? why not have a single nitrogen (or helium) tank feeding both the tanks and the turbines?

Quote

Both nitrogen and helium flow should be adjusted, so needs control valves in turn, which are powered... probably by same fluids.

again, why not use electric control valves?

I had a whole university exam on industrial chemical establishments, which included a lot of piping and valves, and I don't remember any mention of valves controlled by external fluids. unless we count no-return valves, of course

[kerbiloid]

2 hours ago, king of nowhere said:

why use expensive helium instead of cheap nitrogen? or why not use an electric turbine?

1. The helium atomic mass is the lowest between the inert gases (4 g/mol), so high pressure/energy ratio, less gas per same effort.

2. It's more cryogenic than oxygen, nitrogen, and methane. So, won't freeze in contact with them.

2 hours ago, king of nowhere said:

or why not use an electric turbine?

Too complicated, heavy and expensive for such effort, when they already have high-pressure gases in the system and just need to take a little.

The turbopump pressure is higher than pressure in the combustion chamber, so the electric one should be rather powerful.

2 hours ago, king of nowhere said:

why not have the methane tank pressurized by evaporated gaseous methane?

They prefer to spend the fuel asap. The oxygen can't decompose or turn into heavier molecules when passing through the hot parts of the engine.

If it was a kerosene engine, they could pressurize it with methane from a high-pressure tank (like they were going in Sea Dragon), but would highly likely just use the inert nitrogen, as it's anyway cheap and is required in small amounts.

2 hours ago, king of nowhere said:

But why use nitrogen here and helium for the turbines? why not have a single nitrogen (or helium) tank feeding both the tanks and the turbines?

Probably because they need much more gas to pressurize, and the efficiency (low atomic mass) is not critical, as the tank pressure is several atmospheres, while the turbopump pressure is hundreds.

Also, according to the scheme, the helium may be used only to start their spinning, while the further spinning is provided by some of the major components.

So, they save the expensive helium only for minor needs, where high mechanical efficiency is required.

(Though, say, in Nauka module, helium fills the fuel tanks from behind. But they are by orders of magnitude smaller.)

2 hours ago, king of nowhere said:

again, why not use electric control valves?

When it's possible, the rocketry uses pneumatic and hydraulic systems. They are cheaper, simpler, more powerful, more reliable, and utilize the exhaust gases which anyway appear.

2 hours ago, king of nowhere said:

I had a whole university exam on industrial chemical establishments, which included a lot of piping and valves, and I don't remember any mention of valves controlled by external fluids. unless we count no-return valves, of course

Amounts and pressures differ.

Also the rocket equipment works for a minute, then dies.

In case of reusable - for an hour in total, then dies.

Industrial equipment should work for years and usually in much softer conditions.

[mikegarrison]

5 hours ago, king of nowhere said:

let me get this straight, though: if you had a simple system with just the two pipes for fuel and oxidizer, both with a valve and a turbopump (electrically powered, with batteries aside), and of course you still used the cold fuel to cool the nozzle; would the system still work, albeit at a lower efficiency?

Maybe, but probably not.

Here's a very famous simple liquid fuel rocket:

This is basically what you are talking about.

[sevenperforce]

11 hours ago, JoeSchmuckatelli said:

There is a good video I saw recently about turbopumps and fuel mixing /premixing - did not specifically speak to the complexity of the whole system - but it's a start. 

Don't remember where I saw it - but if I find it I will link it 

I have a series on this, actually.

And also….

 

[wumpus]

One big problem that any liquid rocket will have is that it has to avoid exploding while going from no flow of any fuel or oxidizer to full power.  It takes some time for the turbopumps to spool up, and you have to make sure that everything that enters the combustion chamber doesn't pool on the ground (and later explode).  I'd assume that electric turbopumps (see the electron rocket) have it easiest, but this hardly solves all the problems of getting from here to there.  And of course, once you finally get to full power, you are trying to ride a tiger of absolutely massive energy, any slight deviation from  nominal will try to make your rocket explode.

While normally we like to simply to say "mass is everything", don't forget that at the moment of launch, thrust is at least as important and if thrust isn't higher than weight, you won't be going to space today (see recent Astra failure, although that was more due to not having enough fuel once it finally burned enough fuel to get a TWR >1).  The other approach was "big dumb rockets", which was an idea that fuel was cheap and rocket engines were expensive.  True enough, but they  never seemed to get the engines cheap enough and the rest of the rocket light enough or big enough.

Edited October 18, 2021 by wumpus
left out word

[DDE]

6 hours ago, king of nowhere said:

let me get this straight, though: if you had a simple system with just the two pipes for fuel and oxidizer, both with a valve and a turbopump (electrically powered, with batteries aside), and of course you still used the cold fuel to cool the nozzle; would the system still work, albeit at a lower efficiency?

OK, so we're looking at Rocketlabs Rutherford engine. 

You're getting close enough. You're badly missing an injector and an ignition system. The blue thing is the hydraulic actuator for the gimbal, it's not strictly necessary and you need a source of hydraulic fluid (e.g. the fuel). Then you've got a myriad of sensors; I've actually chatted with those guys in an AMA and they told that full propellant utilization and mixture ratio control are a big reason why they use electric motors.

I'm also gonna guess that you need to cool the motors as well. Don't have a flow diagram for that engine handy, though.

What I've learnt from 700 pages of Sutton is that engines are getting simpper and simpler. The staged combustion cycle isn't without its flaws, and we've seen the abandonment of separate fuels for the gas generator (e.g. peroxide on R-7/Soyuz), of separate lubricant tanks and lubricant pumps, of gearcases between GGs and pumps, of solid-propellant start-up cartridges and separate engine start-up turbines.

[JoeSchmuckatelli]

1 hour ago, sevenperforce said:

have a series on this

Links don't appear to function 

[sevenperforce]

31 minutes ago, JoeSchmuckatelli said:

Links don't appear to function 

Crap, I hate how the forum handles those links.

Why Are Rockets Complicated?

https://vm.tiktok.com/ZM86TyBg5/

Power Cycles Make It Worse

https://vm.tiktok.com/ZM86TLYGV/

Those are each 60-second videos with a LOT of information. 

[totalitor]

Simple and working rocket engine can be made. I did it. Pressure-fed engine using ethanol/water and gaseous oxygen. Cooled and not very heavy.

I guess I was lucky?

You can watch how I did it on Youtube channel:

notmadrocketscientist

[Silavite]

5 hours ago, king of nowhere said:

why use expensive helium instead of cheap nitrogen? or why not use an electric turbine?

In addition to the points Kerbiloid mentioned, gaseous nitrogen also has the annoying tendency to dissolve in LOX. This dilutes your fuel and reduces performance.

Also, if I can speak to one particular aspect of rocket engines: Heat transfer in the thrust chamber. Particularly, heat flux at the throat (where it is maximized due to the low boundary layer thickness and high mass flux). A somewhat extreme example of this is the RS-25. As figure 6 demonstrates, the RS-25 has a peak wall heat flux of 163 MW/m² (or about 100 BTU/s/in² in freedom units).

To put that in perspective, let's consider the hot side wall temperature at the wall that would result from such a heat flux. Fourier's law of conduction for one dimension can be written as: q = κ * ΔT / L where q is the heat flux κ is the wall material's thermal conductivity, ΔT is the temperature rise, and L is the wall thickness. Let's approximate the cold side temperature as 20 K, the boiling point of LH2 at atmospheric pressure.

Say that we want to make our thrust chamber out of Inconel, an alloy known for its resilience to high temperatures. κ = 15 W/m/K for Inconel. Supposing that the chamber wall is 3 mm thick (about 1/8"), that results in a wall side temperature of 32,600 K. Obviously, this won't work. Let's try copper, which has an excellent thermal conductivity of κ = 390 W/m/K. This results in a wall side temperature of 1,250 K, which is still too high. Suppose that we want to limit the hot side temperature to 600 K (about 327 deg C or 620 deg F) to prevent the copper wall from losing too much of its yield strength (the relevant line on this chart is for Cu-DHP, which is very nearly pure copper).

Spoiler

With a hot side temperature of 600 K, we can say that the chamber wall must be no thicker than 1.4 mm (slightly thinner than 1/16"). This is an approximate upper limit on wall thickness from thermal considerations. There will also be a lower limit on wall thickness from structural and manufacturing considerations. For the RS-25, I'd imagine that the margin between those two limits is very slim indeed.

Addendum: The problem is both worse and better than this analysis predicts. Worse because this neglects the fact that the cooling channels have a finite thickness, so there will be some fraction of the wall which doesn't have coolant behind it. Better because the structural load between the thrust chamber and cooling channels is relatively small; the LH2 in the channels will be at a slightly higher pressure than the LH2 in the thrust chamber, thus pressure differential between the thrust chamber and cooling channels is much smaller than the pressure differential between the thrust chamber and outside.

Edited October 17, 2021 by Silavite
"losing" not "loosing"!

[Lewie]

1 hour ago, totalitor said:

Simple and working rocket engine can be made. I did it. Pressure-fed engine using ethanol/water and gaseous oxygen. Cooled and not very heavy.

I guess I was lucky?

You can watch how I did it on Youtube channel:

notmadrocketscientist

Wait, that’s you? I’m a huge fan!!!!!! 

[cubinator]

I'm building a 'simple' rocket engine with my classmates. It's fueled by liquid kerosene and gaseous oxygen. We have two valves in sequence on both lines, for controlling the startup of the engine. They are actuated by pressurized nitrogen gas, and the fuel tank is also pressurized by nitrogen. The pressure pushes fuel and oxygen into the combustion chamber through an injector plate. We have two more small valves at the fuel tank so that we can safely vent its pressure when needed.

At various points along the plumbing are pressure sensors, and around the engine are temperature and force sensors, all connected to computers on the test stand and in the control room. It's a fair amount of work to get all that integrated and tested to get the equipment working properly and the proper procedure developed before we will be able to ignite the engine for the first time.

Using a more energetic oxidizer like liquid oxygen will require the ability to fill the rocket tanks remotely, different materials, and more valves to make sure we can purge the lines safely.

[magnemoe]

1 hour ago, Lewie said:

Wait, that’s you? I’m a huge fan!!!!!! 

Like this  

 

[Lisias]

There's an (excelent) thread on Forum about a dude that decided to build his own (pretty small) rocket engine.

I think that everybody that cames to this thread should have a look on it!

In special, this guy @totalitor, did some really impressive (home built!) attempts starting at this post. Note that he started on 2016, and his last post for this subject is from 2020. There're lots and lots of good information obtained by the good and faithful Trial and Error methodology!

Or you can go straight to his Youtube Channel:

https://www.youtube.com/channel/UCf2lLsBevPguEq7i2YKQ4Sg/videos

 

[totalitor]

Yes, that is me:) After making rocket engine I have designed and build actual rocket. Slowly, but I think launch is next year 2022.

I put videos about progress to Youtube.

Welcome to follow.

[JoeSchmuckatelli]

On 10/17/2021 at 3:48 PM, magnemoe said:

Like this  

 

That is a big ass fan

 

 

 

https://www.bigassfans.com/

 

Edited November 3, 2021 by JoeSchmuckatelli

[RCgothic]

Complex as opposed to what?