INTUITIVE Explanation of Rocket Nozzles (Convergent Divergent)

[@JamesSeedorf]

Awesome explanation, about how much of the thrust for an engine like the Merlin or Raptor comes from each section?

[@kelsiehogan3400]

For Merlin only about 40% comes from the convergent section, 60% is from the divergent section

[@ConHathy]

Yep, the speed of sound in burnt RP1 at 3670 K is about 1150 m/s, the exhaust velocity is around 3000 m/s depending on altitude so it’s right in the 40:60 range. Thrust is directly proportional to velocity when perfectly expanded.

[@CollinPaul-tx7td]

Why does the fluid behave differently at Supersonic speed and why do we consider density to be constant at subsonic speed

[@Leo-h7c]

@@CollinPaul-tx7td Density is constant when M

[@Leo-h7c]

@@CollinPaul-tx7td It behaves differently because the fluid cannot send information backwards when it's supersonic

[@jamesrindley6215]

Thank you, this is brilliant. I'm an electronic engineer and I've always aimed to get an intuitive understanding first before getting into the maths. The equations make a lot more sense once there's a mental animation to go with it. Sometimes people claim you can only understand things via the maths and there is no intuitive model, but in most cases I reckon that's because they haven't thought long enough or haven't met the right teacher. You are that teacher, you did a great job here.

[@Andres-gr3us]

Best explanation of the thrust mechanism behind an engine I have ever seen on KZbin

[@tunatuna8877]

This is a fantastic explanation. I'm an ME and remember in gas dynamics going through the derivation and using the equations to find different properties, but no one ever explained what was physically happening. It's one thing to read the equations to understand how the flow is behaving, but this actually helps visualize what's happening.

[@gegetiti9373]

Lovely explanation, thanks to you I finally understand why the heck we need the diverging nozzle!

[@PramochanYaan]

This exact piece was missing from my mind, thank you!!

[@adamclark9588]

This video was super useful, thanks for the straight-foward explanation!

[@bensmith3304]

This explanation makes so much sense, thank you!

[@bcap]

Thank you for the video! Learned something new. Interesting enough, I actually find it _intuitive_ to have the nozzle. This video made me think about it and I believe my intuition came from the following: the nozzle forces the fluid to be closer to laminar flow instead of potentially going in several directions. That said, if that were true then a nozzle that is just a cylinder (no divergence) would work out as well, but now I understand that is not the case. I think there is an intersection of that with what you explained anyway. Nevertheless I quite enjoyed the information propagation explanation. Once you are at the speed of a wave, obviously the wave can’t propagate information back. Works in a very similar way to the causality principle I suppose A small correction/note on the video: at 7:38 you say weight of the cone but I think mass would be more appropriate? Thank you!

[@rkond]

An excellent video! This channel deserves much more views!

[@Illuminati2]

sick video dude. i think i get it now, the throat strategically contains sonic/supersonic flow which cuts off information, and each conic shape on each side of the throat takes advantage of the flow properties at that position

[@tazerface8659]

Thank you for making that make sense. Now that it does, I don't know how it didn't...

[@darkscarf]

Great attempt at the explanation, thank you, but there are some concepts still a bit in the air for me. Major one is the "information chain". Having trouble understanding the mechanics of it in the rocket engine example. In another example like "the supersonic plane" it is a bit more intuative to understand the break in the chain. The nose of the plane is moving faster than the disturbences it just crated so the upcoming air molecule in the way has no information until the nose hits it. In the plane example we are talking about the information exchange between things are coming closer (plane's nose and the upcoming air molecule). In the rocket engine example, information exchange is done between things that are moving apart (slow moving particles in combustion chamber and the faster moving particle in the throat) So the "information chain" and how it breaks should work differently? In the rocket engine example we say that the information chain is communicating "back pressure" upstream due to near sonic particles coming out of the throat being stopped by slower particles ahead in the nozzle. And we explain the behaviour chance by this chain being broken in supersonic state as there are no slower particles ahead to hit in the nozzle (they are all as fast or faster). This is all fine provided that the particles will continue to be supersonic forever. But what happens when they are finally subsonic again after the nozzle? Shouldn't that conversion into subsonic state somehow create "a" chain again? Does that not affect the rocket anymore? Watching similar videos, I found a comment that said the shockwave that forms when going supersonic to subsonic results in shocks and discontinuities that are transition layers where the plasma properties change from one equilibrium state to another. Still trying to understand what that means for the dynamics of this diagram :)

[@rabidlenny7221]

I personally think that’s a bad description. To me, the better description is: Take a section view of the exhaust flow as it travels down the divergent nozzle. Each slice obeys a conservation of energy and conservation of mass. As you expand, it is reducing in density. As it’s doing this, mass is conserved, but the area is increasing, meaning a drop in pressure as well. Drop the pressure of a gas with conservation of energy, that’s a temperature drop. So you have a pressure and temperature drop, essentially a loss of energy seemingly. However energy must be conserved. So where does it go? To kinetic energy, velocity. So when the goal is to throw this shit out of your rocket as fast as possible, that’s where & why you’ve got a divergent nozzle

[@ConHathy]

The issue I see with that is that it doesn’t explain why you still use convergent nozzles at compressible but subsonic speeds. Compressibility becomes a factor down around mach 0.3 but if you try to expand a flow at that speed it will slow down and not speed up

[@rabidlenny7221]

@@ConHathy well yeah, the convergent portion of the nozzle is generally intuitive to most folks, having practical experience with it in day to day life. And his description of it, and why you’re capped at Mach 1 was good enough for me

[@Illuminati2]

@@rabidlenny7221u suck at explaining

[@mecarloschavez6281]

Interesting but not sure to understand the information flowing thing - --@ what pressure you can start having supersonic flow out of a let’s say pressurized container ?

[@jeanknecht4098]

Great video! It makes more sense now!

[@unknown-ql1fk]

I know this is an oldevideo but THANK YOU. I was having a discussion with a real life "flat earth" person and i realized i unable to put int my words how this effect works

[@Elliot_97]

Just throwing this out there, but would it be right to say that the flow accelerates in the expanding section of the nozzle because all the velocity components of the gas in the radial directions are bouncing off the walls of the nozzle and being redirected out the back of the nozzle? So the net velocity of the gas is increasing in the direction we want it to, but the total 'velocity' of the particles isn't increasing, it's just aligning into a more productive direction. Like you could have a cloud of gas where all particles are moving 10km/s, but half are moving in one direction, the other half moving in the other direction, so your gas has a net 0 velocity.

[@ConHathy]

Yes that’s definitely a good way to think about it. The energy in the flow is fixed so you’re taking the random motion (temperature) and converting it to organized motion (net velocity). The more you convert, the more efficient your nozzle is

[@kingcosworth2643]

The explanation of the increase of velocity needs to incorporate the drop in temperature, a reabsorption of entropy is creating the acceleration.

[@amjadhani8929]

Good explanation

[@waysofchaos7076]

Very nice video, but could you tell what software you used for these animations in the video like at 7:21 sec? Thank you very much

[@ConHathy]

@@waysofchaos7076 Solidworks, back when I still had access through my university

[@waysofchaos7076]

@@ConHathy thank you

[@lucaswilkinson3662]

Hey mate, love the video. I've been looking for an explanation of thrust that doesn't simply say "the gas is faster at exit therefore there is more force". Way more intuitive to me that it's the increase in backpressure resulting from the restriction that builds more thrust force (despite as you said it actually adding a drag component). I did have one question though, at 5:30 you mention at the speed of sound the pressure wave cant get back to build the pressure, so far so good. However you then go on to say that if you keep decreasing the area you slow the gas upstream. This has me a bit confused as how can the upstream flow decrease if there is no way of information making its way upstream to tell it so? In my mind I would've thought a pressure spike/shock would build at the restriction causing the downstream flow to accelerate further (without increasing upstream pressure and increasing the force on the injector face). Either that or the downstream flow would speed up (conserving mass flow) and end up at a lower pressure than atmospheriic (the exit pressure being lower though couldn't slow it as again that information cant travel back upstream. I am thinking of a system where the mass flow in is being held constant and not sure if that assumption of mine is what is causing me issues. I'm a mechanical engineer with no background in this area, just wound up here whilst chasing a rabbit hole trying to get a more intuitive understanding of bernoulli so I am a bit off course haha but find all this super interesting.

[@ConHathy]

My explanation is looking at the steady state situation once everything is stabilized, but at the risk of over simplifying: when starting, the flow may hit supersonic speeds before reaching the throat, but then it will “struggle” to fit through the over restrictive section (run into the walls) and generate shock waves. Shock waves can travel faster than the speed of sound upstream but also are transient, so rather than giving you constant back pressure they will just slow down until it eventually stabilizes. It’s similar to what happens on supersonic airplanes, we usually say that they are traveling too fast for air to get out of the way, but in reality there is a shockwave a short distance a head of the plane that causes a rapid reduction in the speed of the flow right before it contacts the nose of the plane.

[@lucaswilkinson3662]

@The Con Hathy Channel ah OK, so the shock wave would be the information carrier in this scenario? So say you could decrease the throat area on the fly somehow, reducing past the choke point can still effect upstream conditions in this case? Appreciate the response and apologies if these are silly questions

[@ConHathy]

@@lucaswilkinson3662 oh not a silly question at all, supersonic flow is very weird stuff

[@gustverschelde9654]

at 6:22 you show a vid where can I find this please answer?

[@KennethCherisol]

Very well explained.

[@MikeJohnson-de3zf]

oh thank god for this video.

[@praveenb9048]

Is there any open source software that can do simulations like those?

[@ConHathy]

I know OpenFOAM is able to do supersonic flows, but I haven’t tried it myself

[@praveenb9048]

@@ConHathy Thanks! 👍

[@TheMrbubl3s]

Thanks, you made this make sense for me.

[@utilityzc2039]

If compressibility effects are responsible for this behavior of acceleration in diverging section for supersonic flow, why doesn’t the same happen in case of flow slightly below Mach 1, when the compressibility effects are significant as well?

[@ConHathy]

It’s not just compressibility, the converging nozzle is increasing the thrust by increasing the pressure in the combustion chamber, but that means information has to flow upstream from the convergent section into the chamber. The point at which information can no longer flow back up the flow is the speed of sound.

[@utilityzc2039]

@@ConHathy Thanks a lot...still trying to wrap my head around this! I am finding compressible flow topic a bit difficult to understand...Can you suggest some good resources/books to develop a good understanding?

[@drmikebass]

Question: you repeatedly said words to the effect, "a *little* bit of extra thrust", referring to the divergent section. However, it seems clear that it produces a *substantial* amount of additional thrust. Is the Merlin engine a special case? Is the additional thrust of the D (keep your dirty minds in check!) normally less than what is delivered by C section (don't you dare!)?

[@ConHathy]

Yeah, that was a poor choice of words. In another comment I figured it was probably 40% C and 60% D for kerosene (Merlin). You can estimate it for any fuel if you know the exhaust velocity and the speed of sound at the temperature of combustion. Thrust ends up being proportional to velocity, so the convergent section makes (speed of sound)/(exhaust velocity) and the divergent makes the rest.

[@ChemEDan]

This comment was posted 9 months ago...

[@matveyshishov]

Dude, you're awesome, thank you! I still have one question, if you could shed some light on it, please. Say we have two molecules, just created products of combustion. They have momentum component X and Y, perpendicular to axial, and component Z, the axial momentum. What happens to the X and Y components INSIDE of the combustion chamber, are they just lost to heat, or do they somehow manage to get converted into Z?

[@ConHathy]

Across the bulk of the flow the perpendicular movement is turned to be more parallel. The most efficient nozzle is the one that gets it the most parallel (obviously without going below the atmospheric pressure)

[@matveyshishov]

@@ConHathy Right, I understand the statistical representation. The problem is with individual molecules. The momentum is conserved, so when combustion generates motion, there is a non-zero momentum perpendicular to the exit direction. What happens to this momentum, is it just dissipated as heat?

[@ConHathy]

@@matveyshishov momentum is never dissipated, it is always conserved. For every molecule moving in one direction, usually there’s another moving in the opposite direction, meaning they cancel out (so the net momentum is all in one direction)

[@matveyshishov]

@@ConHathy Statistically - yes, they cancel out. But stats are based on individual molecules, and the question is about very specific real life physics of molecules. So that's the original question - what happens to the molecules which have kinetic energy in motion perpendicular to the axis of the nozzle? Even after having collided with other molecules, they'll still exchange this momentum, but never change it, right? So there exists a significant amount of kinetic energy "locked" in the perpendicular direction to the nozzle, inside the combustion chamber. Where does it go? How does it get converted into motion in the direction of the nozzle, given that in all images the combustion chamber has parallel walls?

[@ConHathy]

@@matveyshishov kinetic energy and momentum are very different things. Momentum is directional and always conserved, kinetic energy is non directional and can be dissipated as heat, or converted to other types of energy such as pressure If you only sent a few particles through a nozzle, then yes, you would probably see side force due to a net perpendicular momentum of the group, but the more atoms you add the less significant that error. (A Merlin burns something like 10^27 atoms per second which is why I keep going back to bulk properties) The kinetic energy can be turned (because it is non-directional) the mechanism of this is atoms bumping into each other causing random changes in direction. The geometry of a nozzle primarily lets out atoms moving axially, so the others will (probably) stay in the chamber until they bounce in such a way that they move mostly axially. Of course to get this to work reliably you need a lot of atoms, a single atom can obviously escape with even a tiny amount of axial velocity if there is no other atoms blocking it. As for where it all goes, the momentum goes into the walls of the engine, every bounce pushes the engine in the opposite direction a little bit. For kinetic energy, some is dissipated as heat but most is stored as pressure (a form of potential energy but again one that only emerges when you consider large numbers of molecules)

[@gisellesinclair6811]

Thank you! So frustrating when all ya get are the formulas.

[@ValidatingUsername]

Is the thumbnail asking why the bell of the nozzle get hotter and expands?

[@ConHathy]

Just why the geometry of the nozzle expands

[@evan._.7436]

Awesome video great explanation

[@nickmiller6915]

Hi trying to wrap my mind around this. So we have back force on the walls from the divergent nozzle which is increasing thrust, then you say that because the gas is producing thrust it MUST be accelerating, as such the sonic gas must accelerate through the divergent nozzle. I mean couldn't this explanation be used for subsonic flow as well? Wouldn't you get some (emphasis) thrust from the high pressure expanding gas leaving the nozzle? You could just put a divergent nozzle there, get some (emphasis) back force and use the same chain of logic?

[@ConHathy]

Maybe you could get some, but in a divergent section your subsonic flow will actually slow down and basically cause drag more than thrust

[@shaelynrounds4194]

Very cool

[@CollinPaul-tx7td]

Why does the density remain constant at subsonic speed and at supersonic speed the density isnt considered constant

[@ConHathy]

Well it’s never truly constant, it’s just that at slow speeds the changes are small. I think around Mach 0.3 is when it really starts to matter. At supersonic speeds it is compounded by the inability of information to flow upstream. For example if supersonic flow hits a blunt body it piles up in front of it, increasing density significantly and forming a bow shock.

[@CollinPaul-tx7td]

@@ConHathy Understood. Thank you so much

[@ricardoabh3242]

Mach 1 is in the neck, and should not move… How can you throttle your engine at supersonic without moving the position of the Mach 1/neck? I though of you change the pressure the position where Mach 1 is located changes and that a lost of efficiency?

[@cudedog]

As a really dumb explanation, can you say that the engine is riding an explosion (albeit a very weak one) that's pushing against the divergent section of the nozzle?

[@limejet2566]

thanks

[@alext8828]

4:45, time to throw some Bernoulli stuff in there. No?

[@paradiselost9946]

its far easier as just pressure... heres a tank. 30 psi (absolute), acting all directions. counteracted by tank in all directions. balanced. atmosphere, 15psi. punch 1 squin hole in tank. now the 30 psi acting outwards previously balanced by the tank wall is only balanced by 15 psi of atmosphere. 30-15/1 squin gives 15lbs of force. acting OPPOSITE to hole. the rate of acceleration is governed by the ratio of thrust to mass. how bigs the tank? how much does it hold? how long will it maintain the 30psi? the rate of flow is dictated by the size of the tank, the nozzle, and ability to maintain pressure. the bigger the hole, the faster it leaks out at the same speed for a given pressure. higher the pressure, higher the flow... until that limit of choke. and then you get strange things... as air expands it starts absorbing heat. and as it cools, it cant expand as much. the only energy available is the surrounds, and the pressure pushing it. and once you hit choke, you cant push any harder. 0.557 p1/p2 iirc... at the choke point, the pressure is LOWEST. gas doesnt like flowing from low to high. it needs to be pushed. the only thing that forces the gas through the nozzle is the pressure difference and its limited to a certain speed due to whats behind the nozzle. and that means you can only get so much thrust. regardless of the pressure. can only push so much gas through a hole before going supersonic. and the nozzle makes no difference with plain compressed ambient temperature air. whereas if the gas is hot... steam... combustion... it can release heat as it expands. if its just a plain orifice, flat plate with hole... it expands too much. you hit the speed limit due to pressure differential sooner. its as if the gas exiting "gets in its own way". whatever is leaving is slowing down, is gaining pressure. and gas doesnt flow from low to high... the delaval "tricks" it. the nozzle keeps getting larger, but not TOO large, at a rate so that pressure differential in any section is relatively constant. from choke point it never "sees" any higher pressure and happily just keeps flowing... gaining in velocity... unlimited by any sonic concerns of whats "in front of it" as its just flowing into "an extended choke point". yeah, maybe that is a bit un-intuitive? lol... but it RELIES on that excess of heat. here in the nozzle, the gas is this hot, has no pressure, and is supersonic. limited. as it expands, it cools, but rather than increasing in pressure, it increases in velocity above supersonic. you still flow so much through any given hole before it goes supersonic, but now you get to push that mass with extra force, accelerate it backwards... with more force, more velocity, than just pressure alone. the denser the gas, the more kick you can get. air again, is useless... steam and combustion products are relatively dense and hold far more heat... far more energy density.

[@style_of_experiment2009]

Habibi vala valaa

[@kelsiehogan3400]

nice bedroom

[@ricardoabh3242]

Nice static fire photo…

[@hytralium]

"It makes no sense" - Integeza

[@GamingKeenBeaner]

I feel like this would be better if you were dressed like a wizard. Otherwise I approve.

[@hafizuddinmohdlowhim8426]

Your explanation is not right. The thrust and pressure are two distinct physical properties. Convergence and convergence-divergence nozzles have the same thrust but vastly different speeds.

[@ConHathy]

Well just by the conversation of momentum we know that’s not true, same mass flow more speed guarantees more thrust