Author Topic: Neutron Design speculation  (Read 114796 times)

Offline Lars-J

  • Senior Member
  • *****
  • Posts: 6809
  • California
  • Liked: 8485
  • Likes Given: 5384
Neutron Design speculation
« Reply #40 on: 03/06/2021 04:42 am »
1 engine with multiple chambers/nozzles doesn’t solve their pump issue (turbo/electric/hybrid), is just makes it worse! Because now you don’t need to design a Merlin class pump, you need an F-1 class pump. You aren’t solving anything that way, just making it more complex.
« Last Edit: 03/06/2021 04:42 am by Lars-J »

Re: Neutron Design speculation
« Reply #41 on: 03/06/2021 04:51 am »
1 engine with multiple chambers/nozzles doesn’t solve their pump issue (turbo/electric/hybrid), is just makes it worse! Because now you don’t need to design a Merlin class pump, you need an F-1 class pump. You aren’t solving anything that way, just making it more complex.

You're circumventing all of the combustion instability problems the F-1 had, and making it much easier to throttle down. That's problems solved.
« Last Edit: 03/06/2021 05:07 am by JEF_300 »
Wait, ∆V? This site will accept the ∆ symbol? How many times have I written out the word "delta" for no reason?

Offline ArbitraryConstant

  • Senior Member
  • *****
  • Posts: 2014
  • Liked: 628
  • Likes Given: 311
Re: Neutron Design speculation
« Reply #42 on: 03/06/2021 05:47 am »
1 engine with multiple chambers/nozzles doesn’t solve their pump issue (turbo/electric/hybrid), is just makes it worse! Because now you don’t need to design a Merlin class pump, you need an F-1 class pump. You aren’t solving anything that way, just making it more complex.

You're circumventing all of the combustion instability problems the F-1 had, and making it much easier to throttle down. That's problems solved.
I don't think it's true that dual chamber engines are easier to throttle down.

Re: Neutron Design speculation
« Reply #43 on: 03/06/2021 06:22 am »
1 engine with multiple chambers/nozzles doesn’t solve their pump issue (turbo/electric/hybrid), is just makes it worse! Because now you don’t need to design a Merlin class pump, you need an F-1 class pump. You aren’t solving anything that way, just making it more complex.

You're circumventing all of the combustion instability problems the F-1 had, and making it much easier to throttle down. That's problems solved.
I don't think it's true that dual chamber engines are easier to throttle down.

I don't think so either. I'd refer you to my original post on this:
I can imagine a 7 chamber engine, with a "landing mode" where the pumps are throttled down and the flow to all but the center chamber is cut off.
Wait, ∆V? This site will accept the ∆ symbol? How many times have I written out the word "delta" for no reason?

Offline Lars-J

  • Senior Member
  • *****
  • Posts: 6809
  • California
  • Liked: 8485
  • Likes Given: 5384
Re: Neutron Design speculation
« Reply #44 on: 03/06/2021 07:07 am »
This idea of yours seems like solution in search of a problem.

It all ignores the established fact/wisdom that electric pumps don’t scale. So having one pump for all nozzles just makes that problem even worse.

Now you may say that such a statement (not scaling) is just conventional thinking... well, we are about to find out. Whether the masters of electric pumps (Rocketlab) abandon it for this vehicle, that should tell us.

Re: Neutron Design speculation
« Reply #45 on: 03/06/2021 03:21 pm »
This idea of yours seems like solution in search of a problem.

It all ignores the established fact/wisdom that electric pumps don’t scale. So having one pump for all nozzles just makes that problem even worse.

Now you may say that such a statement (not scaling) is just conventional thinking... well, we are about to find out. Whether the masters of electric pumps (Rocketlab) abandon it for this vehicle, that should tell us.

Problem: Peter Beck wants to have as few engines on the first stage of Neutron as possible, but it would be difficult to throttle deeply enough to land propulsively on one big engine.

Solution: Make it a multi-chamber engine with a chamber in the center, and cut off the other chambers and throttle back the pumps when you want to land, effectively turning the big engine into a smaller one.

Yes, I ignore the electric pumps problem, because I ignored the pumps entirely because any would work. This concept could use electric pumps or a gas generator or be staged combustion or even expander-bleed. That wasn't my point.

And I'm also not saying that this is the best idea or that this is what they'll do. It was just an idea that I thought was interesting. I'm kind of confused by the outright rejection of it.
« Last Edit: 03/06/2021 03:25 pm by JEF_300 »
Wait, ∆V? This site will accept the ∆ symbol? How many times have I written out the word "delta" for no reason?

Online edkyle99

  • Expert
  • Senior Member
  • *****
  • Posts: 15391
    • Space Launch Report
  • Liked: 8566
  • Likes Given: 1356
Re: Neutron Design speculation
« Reply #46 on: 03/06/2021 03:58 pm »
That's fascinating. In that case, it's fair to call RocketLab a US launcher, IMHO, Ed Kyle.
The rocket itself, Electron, was developed and tested and is assembled and launched in New Zealand by New Zealand crews as I understand things.  Engine "sub-assemblies", if not entire engines, are currently produced in California, but this is still a New Zealand engine design.  I believe that Electron structures are made in New Zealand.

Neutron might be different.  I tend to categorize things based on where the primary structures are made.  Antares confuses using this method, with a Ukrainian first stage, Russian engines, and US second stage.  Electron is much easier to categorize, but Neutron might be more like Antares.  We'll see!

 - Ed Kyle

Online edkyle99

  • Expert
  • Senior Member
  • *****
  • Posts: 15391
    • Space Launch Report
  • Liked: 8566
  • Likes Given: 1356
Re: Neutron Design speculation
« Reply #47 on: 03/06/2021 04:20 pm »
Antares 230+: Current version of Antares. Has a payload to LEO of ~8t with a liftoff thrust of 3,844 kN. Keep in mind that Antares has a heavy solid upper stage weighing it down.
Castor 30XL only weighs 25 tonnes or so at launch and probably a couple tonnes-ish at burnout.  Less than 9% of total liftoff weight.  Much lighter than Falcon 9's second stage.  Comparable to a Soyuz Blok I stage.

 - Ed Kyle
« Last Edit: 03/06/2021 04:22 pm by edkyle99 »

Online ZachF

  • Full Member
  • ****
  • Posts: 1610
  • Immensely complex & high risk
  • NH, USA, Earth
  • Liked: 2604
  • Likes Given: 531
Re: Neutron Design speculation
« Reply #48 on: 03/06/2021 05:54 pm »
Antares 230+: Current version of Antares. Has a payload to LEO of ~8t with a liftoff thrust of 3,844 kN. Keep in mind that Antares has a heavy solid upper stage weighing it down.
Castor 30XL only weighs 25 tonnes or so at launch and probably a couple tonnes-ish at burnout.  Less than 9% of total liftoff weight.  Much lighter than Falcon 9's second stage.  Comparable to a Soyuz Blok I stage.

 - Ed Kyle

I don't think he means "weighing it down" in the literal sense, just that a single solid upper stage is going to give poorer performance.
artist, so take opinions expressed above with a well-rendered grain of salt...
https://www.instagram.com/artzf/

Offline ArbitraryConstant

  • Senior Member
  • *****
  • Posts: 2014
  • Liked: 628
  • Likes Given: 311
Re: Neutron Design speculation
« Reply #49 on: 03/06/2021 06:37 pm »
This idea of yours seems like solution in search of a problem.

It all ignores the established fact/wisdom that electric pumps don’t scale. So having one pump for all nozzles just makes that problem even worse.

Now you may say that such a statement (not scaling) is just conventional thinking... well, we are about to find out. Whether the masters of electric pumps (Rocketlab) abandon it for this vehicle, that should tell us.

Problem: Peter Beck wants to have as few engines on the first stage of Neutron as possible, but it would be difficult to throttle deeply enough to land propulsively on one big engine.

Solution: Make it a multi-chamber engine with a chamber in the center, and cut off the other chambers and throttle back the pumps when you want to land, effectively turning the big engine into a smaller one.

Yes, I ignore the electric pumps problem, because I ignored the pumps entirely because any would work. This concept could use electric pumps or a gas generator or be staged combustion or even expander-bleed. That wasn't my point.

And I'm also not saying that this is the best idea or that this is what they'll do. It was just an idea that I thought was interesting. I'm kind of confused by the outright rejection of it.
I would agree Lars's objection is not well founded, but there's good reasons to think the massively multi-chamber approach will not happen.

The trends we've seen more recently is that a multi-chamber engine doesn't actually trade that well, and people that are perfectly capable of doing the trades like the Russians, SpaceX, Blue Origin, etc don't seem to prefer them these days. They exist today primarily in legacy vehicles that would be problematic to re-engine. My suspicion is this has to do with improved computers being able to handle the control problem plus the mass needed to connect all the chambers with connections that need to handle high pressure. The problem of isolating a single chamber would make this even worse.

I think RL would just use a bunch of separate engines even though they prefer not to rather than do the multi-chamber thing. There are far simpler ways to achieve low throttle points at sea level, like a dual bell nozzle.

Offline Tomness

  • Full Member
  • ****
  • Posts: 660
  • Into the abyss will I run
  • Liked: 289
  • Likes Given: 737
Re: Neutron Design speculation
« Reply #50 on: 03/06/2021 07:35 pm »
Anybody body think Rocket Lab with Lockhead could buy the IP for PGA engine from ROC/Stratolaunch make really efficient upper stage engine?

Re: Neutron Design speculation
« Reply #51 on: 03/06/2021 07:56 pm »
Anybody body think Rocket Lab with Lockhead could buy the IP for PGA engine from ROC/Stratolaunch make really efficient upper stage engine?

Do I think so? No. But I do really really want that to happen. PGA engine would be really nice.

Although it's about Merlin sized as I recall, so you almost might as well just use it on the first stage at that point.
Wait, ∆V? This site will accept the ∆ symbol? How many times have I written out the word "delta" for no reason?

Offline edzieba

  • Virtual Realist
  • Senior Member
  • *****
  • Posts: 6104
  • United Kingdom
  • Liked: 9329
  • Likes Given: 39
Re: Neutron Design speculation
« Reply #52 on: 03/06/2021 08:35 pm »
I haven't done an real calculations, but from a gut level it seems like Neutron would need 5-6 Merlin 1D's - if using Merlin's  :) - to get the capability they announced.

So this is a good opportunity to have this conversation:

How much thrust does an 8 ton LV need on it's first stage? Let's look at some examples

Soyuz FG: I believe this is the Soyuz variant currently used for crew rotation. It's a ~7t to LEO rocket with a liftoff thrust of ~4,100 kN (I can't figure out if this number is vac or sea level). It's much heavier than it's needs to be, due to being massively overbuilt because it was designed by Soviets in the 1950s.

Atlas III: Atlas III was an 8-10 ton to LEO vehicle, depending on the upper stage. Had an RD-180, and therefore a liftoff thrust of about ~3,800 kN. It's design is about as mass efficient as it gets.

Delta IV Medium: The totally booster-less Delta IV (using post RS-68A specs) puts ~8.5t into LEO and has a lift-off thrust of "only" 3,140 kN. It get's away with this because the whole thing is hydrolox, and hydrogen is light.

Antares 230+: Current version of Antares. Has a payload to LEO of ~8t with a liftoff thrust of 3,844 kN. Keep in mind that Antares has a heavy solid upper stage weighing it down.

Falcon 9 v1.0: This was actually a 10t to LEO vehicle, with 9 Merlin 1Cs providing ~3,800 kN of thrust at liftoff.

So I generally estimate that we should be looking at 3,500 - 3,850 kN of liftoff thrust for a scratch-built modern 8t to LEO vehicle with a liquid upper stage. That's before any budgeting for reusability.


Edit for more: When it comes to thrust level, the Merlin is both the most convenient and most confusing existing option to use in a comparison. You have to remember that while the Falcon 9 started as a medium-lift launcher, it isn't anymore, and it's engine has scaled up accordingly.

Having run back through it like this, my numbers for max thrust for the engines (of a Neutron with 3-4 engines) were low to begin with, and then much lower still because I forgot about propulsive landing. it would be more like 1,000-1,750 kN engines.

Edit: made another edit to correct some numbers. There's actually a pretty strong consensus around 3,800 kN for expendable rocket's in this class.
Remember that Rocketlab are almost certainly going to be optimising for cost, not for efficiency. Of those launchers: Atlas III/IV, Delta IV & Antares were optimised for efficiency. Soyuz was designed around the R-7. Only Falcon 9 was designed to be low cost, and once re-use became a designed-for choice rather than a future option, thrust was uprated to 5885 kN (V1.1) and eventually to 7650 (Block 5). It is reasonable to Rocketlab to design and 'inefficient' stage around 'excessive' dry mass and thrust in order to pursue cost minimisation.

One engine layout worth considering is that of the initial Atlas, or the Zenit upper stage: combining fixed (non-gimballing) main engines with separate steering Vernier engines. Use Rutherfords for steering and landing, and any new engine now no longer needs to steer and can be hard-mounted, and can have simplified throttle requirements - or with enough Rutherfords to satisfy the desired stage throttle range could even be non-throttleable entirely.

Re: Neutron Design speculation
« Reply #53 on: 03/06/2021 11:31 pm »
Cross-posting this from the other thread.

https://www.nasaspaceflight.com/2021/03/rocket-lab-reveals-neutron/ - By Thomas Burghardt - talking to Peter Beck,

https://twitter.com/TGMetsFan98/status/1368354747770626048

Here's an interesting slice of that article, which relates to all the engine speculation we've been doing:
Quote
“Certainly additive manufacturing. But more traditional pumps are required at this scale.” Beck says that Neutron’s first stage will be powered by multiple engines, but not as many as Electron’s.

“Nine engines is great if one shuts down, but otherwise is a bit of a pain,” citing manufacturing and testing required for every single engine. “A small amount of engines is optimal, but large engines have limited throttleability.” Beck says Neutron will have the “least amount of engines practical.

The big thing to note from that is that they will NOT, or at least aren't planning, to use electric pumps on the new engines.
« Last Edit: 03/06/2021 11:34 pm by JEF_300 »
Wait, ∆V? This site will accept the ∆ symbol? How many times have I written out the word "delta" for no reason?

Offline TrevorMonty

Re: Neutron Design speculation
« Reply #54 on: 03/07/2021 12:26 am »
Beck quote from local newspaper.

"For the US, engines are strong core of expertise. In NZ, structure, anaylsis and design is a key strength, in Canada (Sinclair) spacecraft elements."

Make sense for US develop engines, has large pool of expertise to hire from plus NASA test stands. Engine production is also there.

Going need Grasshopper for landing tests, could use Electron and fly out of Mahia. Have spare launch pad, just need to build landing pad.

Sent from my SM-G570Y using Tapatalk


Offline indaco1

  • Full Member
  • **
  • Posts: 283
  • Liked: 64
  • Likes Given: 37
Re: Neutron Design speculation
« Reply #55 on: 03/07/2021 12:34 am »
Forgive me in advance but why do you rule off mixed engines?

A couple of RD-191 or similar surrounded by some Rutherfords (say ~40, not many, just as 4 Electrons, giving about 1000 kN).

Rutherfords are external because used as attitude control thrusters. Most of them are cut of early after liftoff to save batteries.

Reentry burn and landing is performed restarting some Rutherfords.  Up to 1000 kN is well enough, electric-pump feed cycle provide the best control and you can throttle as you want.  Probably multiple engine-out capability at landing.

The big central engines are not restartable, possibly not throttleable and perhaps even gimbaling is not required.

No new engines to design, this way.  If it's really required not to purchase the big central engines they could just copy by a proven design .
 
Lego mode off, and sorry for that.
Non-native English speaker and non-expert, be patient.

Offline TrevorMonty

Re: Neutron Design speculation
« Reply #56 on: 03/07/2021 03:07 am »
Forgive me in advance but why do you rule off mixed engines?

A couple of RD-191 or similar surrounded by some Rutherfords (say ~40, not many, just as 4 Electrons, giving about 1000 kN).

Rutherfords are external because used as attitude control thrusters. Most of them are cut of early after liftoff to save batteries.

Reentry burn and landing is performed restarting some Rutherfords.  Up to 1000 kN is well enough, electric-pump feed cycle provide the best control and you can throttle as you want.  Probably multiple engine-out capability at landing.

The big central engines are not restartable, possibly not throttleable and perhaps even gimbaling is not required.

No new engines to design, this way.  If it's really required not to purchase the big central engines they could just copy by a proven design .
 
Lego mode off, and sorry for that.
Need one big engine for reentry and final braking burn, Rutherfords aren't powerful enough for high deceleration needed.

Sent from my SM-G570Y using Tapatalk


Offline trimeta

  • Full Member
  • ****
  • Posts: 1685
  • Kansas City, MO
  • Liked: 2147
  • Likes Given: 57
Re: Neutron Design speculation
« Reply #57 on: 03/07/2021 06:03 am »
Beck quote from local newspaper.

"For the US, engines are strong core of expertise. In NZ, structure, anaylsis and design is a key strength, in Canada (Sinclair) spacecraft elements."

Make sense for US develop engines, has large pool of expertise to hire from plus NASA test stands. Engine production is also there.

Going need Grasshopper for landing tests, could use Electron and fly out of Mahia. Have spare launch pad, just need to build landing pad.

Sent from my SM-G570Y using Tapatalk

From the article, Beck had this to say about hop tests and RTLS landing sites:

Quote
As far as achieving recovery for reuse, Beck does not anticipate a SpaceX-esque “hop” test program.

“Our priority is to get into service and deliver a customer satellite. Reusability can not work, and the mission can still be successful, so we can take some risk there.”

The first stage recovery is expected to take place on an ocean platform stationed downrange. While the option of performing Return To Launch Site (RTLS) landings is not off the table, it’s also not expected to be useful for Neutron.

“I’ve learned to never say never to anything. Otherwise, it results in some unpleasant dining experiences. But we’ve baselined a flight profile with a downrange landing.”

“It’s a propellant trade between RTLS and downrange landing. RTLS begins to grow the vehicle.”

So seems like the plan is "fly as soon as we can, all landing tests will be secondary objectives of actual paying missions." Which says something about their plans regarding pricing and manufacturing: the vehicle will likely need to be close to profitable without recovery, and with enough manufacturing cadence to launch regularly without expecting reuse. Although, later in the article it's suggested that they'll be making Neutrons at a lower rate than Electrons, because they expect a greater degree of change between each Neutron and don't want to have multiple vehicles in flow which are already obsolete. So I wouldn't expect too many launches per year until they solve reuse.
« Last Edit: 03/07/2021 06:04 am by trimeta »

Offline indaco1

  • Full Member
  • **
  • Posts: 283
  • Liked: 64
  • Likes Given: 37
Re: Neutron Design speculation
« Reply #58 on: 03/07/2021 07:49 am »
Forgive me in advance but why do you rule off mixed engines?
...
Reentry burn and landing is performed restarting some Rutherfords.  Up to 1000 kN is well enough, electric-pump feed cycle provide the best control and you can throttle as you want.  Probably multiple engine-out capability at landing.

The big central engines are not restartable, possibly not throttleable and perhaps even gimbaling is not required.
...
Need one big engine for reentry and final braking burn, Rutherfords aren't powerful enough for high deceleration needed.


I've not made accurate calculations, but assume a mass ratio of 15 (like F9), say 10% mass at landing including spared propellant. 

Rutherfords give about 20% of liftoff thrust: 4 MN by RD-191 (2 x 2000 kN) and 1 MN by Rutherfords (40 x 25 kN).  Suppose T/W at liftoff is 1.5.

This give a max T/W = 3 at landing (that is 20 m/s2 deceleration) throttleable to a single Rutherford so you can even hover and sucide burn is not required.

Most of deceleration came by aerodynamics. If you start landing burn when subsonic at 300 m/s it takes 15 sec and 2.25 km to stop.

To save propellant a small parachute could be used and jettisoned before burn or skydive or other aerobrakes, but it's for efficiency not because thrust is not enough.

As far as reentry burn is concerned we don't even know if they will do it.  Tests with Electron have been totally passive reenty and Neutron could be inox made and more heat resistant.   But even if done you could just start engines earlier during parabolic coasting, I have not access to a trajectory oprimizator but I don't think a longer reentry burn will be that inefficient.

Of course Peter Beck said they will use less engines as possible and it makes sense.  But I repeat, 40 reusable Rutherfords are the same number expended for just 4 Electron missions. 
« Last Edit: 03/07/2021 08:15 am by indaco1 »
Non-native English speaker and non-expert, be patient.

Offline trimeta

  • Full Member
  • ****
  • Posts: 1685
  • Kansas City, MO
  • Liked: 2147
  • Likes Given: 57
Re: Neutron Design speculation
« Reply #59 on: 03/07/2021 07:53 am »
Forgive me in advance but why do you rule off mixed engines?
...
Reentry burn and landing is performed restarting some Rutherfords.  Up to 1000 kN is well enough, electric-pump feed cycle provide the best control and you can throttle as you want.  Probably multiple engine-out capability at landing.

The big central engines are not restartable, possibly not throttleable and perhaps even gimbaling is not required.
...
Need one big engine for reentry and final braking burn, Rutherfords aren't powerful enough for high deceleration needed.


I've not made accurate calculations, but assume a mass ratio of 15 (like F9), say 10% mass at landing including spared propellant. 

Rutherfords give about 20% of liftoff thrust: 4 MN by RD-191 (2 x 2000 kN) and 1 MN by Rutherfords (40 x 25 kN).  Suppose T/W at liftoff is 1.5.

This give a max T/W = 3 at landing (that is 20 m/s2 deceleration) throttleable to a single Rutherford so you can even hover and sucide burn is not required.

Most of deceleration came by aerodynamics. If you start landing burn when subsonic at 300 m/s it takes 15 sec and 2.25 km to stop.

To save propellant a small parachute could used and jettisoned before burn or skydive or other aerobrakes, but it's for efficiency not because thrust is not enough.

As far reentry burn is concerned we don't even know if they will do it.  Tests with Electron have been totally passive reenty.   But even if done you could just start engines earlier during parabolic coasting, I have not access to a trajectory oprimizator but I don't thing a longer reentry burn will be that inefficient.

Of course Peter Beck said they will use less engine as possible and it makes sense.  But I repeat, 40 reusable Rutherfords are the same number expended for just 4 Electron missions.

If anything, "this could possibly work with 40 Rutherfords" makes me even more convinced that this isn't the plan.

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1