Jump to content

One engine vs clustered engines


totalitor

Recommended Posts

Clusters are cheaper. Smaller engine = lower price. Larger production run = lower price.

If one engine fails, others can compensate and the mission goes on.

However, more engines = more stuff to go wrong. In case of an explosion of a motor, pieces of motor may damage other motors.

In the end with the clusters, if you need a bigger rocket, you can just add more engines. No need for costly and time consuming R&D. Just tweak the control software.

Link to comment
Share on other sites

3 hours ago, Shpaget said:

In the end with the clusters, if you need a bigger rocket, you can just add more engines. No need for costly and time consuming R&D. Just tweak the control software.

Accurately controlling multiple engines, with precooling and spin-up-times is no small feat. Expect a lot of time consuming R&D too.

Link to comment
Share on other sites

3 hours ago, Shpaget said:

Clusters are cheaper. Smaller engine = lower price. Larger production run = lower price.

It's not really that simple. The initial price of the engine (most of it R&D and man-hours) does go up the bigger it is, but the actual individual price increases at a lower rate.

Also, while there is the economies of scale, there's also the dis-economies of scale, where per unit costs go up instead of down, when you make more. Larger production run = higher price, in those cases.

Quote

If one engine fails, others can compensate and the mission goes on.

Certainly. This was actually the plan that the N-1 was going to rely on, but in the end it was too complex.

Quote

However, more engines = more stuff to go wrong. In case of an explosion of a motor, pieces of motor may damage other motors.

That's a good point. I would like to add the increased complexity. You have to have more pipes and manifolds with clustered engines.

Quote

In the end with the clusters, if you need a bigger rocket, you can just add more engines. No need for costly and time consuming R&D. Just tweak the control software.

You can't really "just" add engines, you have to do a lot of stuff first. Redoing the plumbing, redesigning the thrust structure, and making sure that you can produce enough engines.

 

Link to comment
Share on other sites

Just now, Frozen_Heart said:

Also if you want to land a rocket a single engine can't throttle low enough so a cluster is required. Even on a Falcon 9 a single Merlin if too powerful to hover.

Or they would just use 1 big engine to ascent (700 t) and a quad of small to land (25 t).

Link to comment
Share on other sites

1 hour ago, kerbiloid said:

Or they would just use 1 big engine to ascent (700 t) and a quad of small to land (25 t).

And then having to carry what is essentially dead weight during launch? Much more efficient to use all engines during ascent, and only use the small ones for landing.

Link to comment
Share on other sites

Single engine is more reliable than a cluster. For a given failure rate of an engine, the likelihood of at least one failing goes up with each engine added.

Single engines are longer for a given nozzle expansion design.

Single engines are able to use more area of a given rocket diameter for nozzle area.

Multiple engines permit roll control through gimballing. 

Multiple engines provide survivable failure modes in some cases.

Multiple engines allow the use of an engine design in different stages with little modification, just use more of them on larger stages. Classic example is the Saturn V, which used five J-2 engines on the second stage and one J-2 in the third stage.

Multiple engines means smaller engines, so sometimes this avoids some square-cube law issues.

Link to comment
Share on other sites

Big engines mean you need a big test stand and more development money, which if you're a commercial outfit can be a big problem. Beal Aerospace, in many ways SpaceX before SpaceX, failed because they couldn't raise the funds for a test stand for their giant single first-stage engine. They'd already done a bunch of tests on their second stage engine, which was still much larger than Merlin; if they'd stuck with clustering that they might still be around.

Link to comment
Share on other sites

4 minutes ago, shynung said:

And then having to carry what is essentially dead weight during launch?

As F9 launch mass is ~25-30 times greater that its first stage landing mass, such quad in total should give ~9/25 = 1/3 of one Merlin engine thrust.

Edited by kerbiloid
Link to comment
Share on other sites

Just now, kerbiloid said:

As F9 launch mass is ~25-30 times greater that its first stage landing mass, such quad should give ~9/25 = 1/3 of one Merlin engine.

What I was trying to say was: why not use these small engines to aid the larger main engine, maybe as verniers? They'll at least accelerate their own mass to avoid burdening the main engine.

Link to comment
Share on other sites

Just now, shynung said:

What I was trying to say was: why not use these small engines to aid the larger main engine, maybe as verniers? They'll at least accelerate their own mass to avoid burdening the main engine.

That might lower the effective ISP of the launch sequence.

Link to comment
Share on other sites

36 minutes ago, Bill Phil said:

That might lower the effective ISP of the launch sequence.

That would. But it can also reduce complexity, by allowing the main engines to remain fixed/gimbal-less.

Link to comment
Share on other sites

Big engine = combustion stability issues = lots of development money to resolve.

Clustered engine = synchronization&control issues = lots of development money to resolve.

Make your pick :-)

AFAIK main reason why clusters are favored is because they can be developed in steps. One  can take small, but well tested and reliable engine and multiply it in cluster, but you can't just inflate it into bigger engine. You can go halfway and create a multichamber engine (single turbopump assembly feeding multiple combustion chambers) but I bet even those have their share of troubles.
 

Link to comment
Share on other sites

I understand this topic is in the spaceflight forum, thus talking about real life launches (with good info!). I wonder how does the logic apply to KSP, however. I've always used single big engines, and apart from the gimbal roll authority and low techtree positioning, I never found clusters to be better than a large one. They're prettier, tho.

Link to comment
Share on other sites

It's like airplanes.

Single engine - If it quits you quit too. Easy to do though, I mean, you can make one in your backyard.

Double engines - If one quits in general you can still fly somewhere. But it's harder to operate (most of the propeller version are variable-pitched).

More engines - One failure means nothing. Increasingly complicated operations anyway, as well as servicing.

The problems, for single engine, are non-scalability. With multiple engines, insane harmonics.

In KSP ?? MUAHAHAHA what are you talking !? Strap the biggest as much as you can !! (I mean, of all the jet engines, I always use the most powerful no matter how awful it is ! Not the massive turbofan though.)

Link to comment
Share on other sites

9 hours ago, Bill Phil said:

Certainly. This was actually the plan that the N-1 was going to rely on, but in the end it was too complex.

I think the wiki for N-1 (or similarly unreliable source) claimed an absolutely impossible requirement for the N-1 control logic.  Note that "impossible" was for the 1960s, I suspect you could get something off the shelf around 2000 from Cisco, but it really pushed the control network beyond comprehension.  Even today, having to get such a complex network to work right from the first launch would be an immense feat.

Has anyone noted the "engine out" feature of falcon 9?  If your engines are light enough, that is a huge advantage.  If you can pull it off, suddenly your cluster went from 1/10th the reliability of a big engine to much higher.

Link to comment
Share on other sites

  • 2 weeks later...
On ‎18‎.‎07‎.‎2016 at 8:32 PM, wumpus said:

I think the wiki for N-1 (or similarly unreliable source) claimed an absolutely impossible requirement for the N-1 control logic.

The Soviets built the first ever digital controller for it. Processing power wasn't the issue. The issue was that they did not and could not test-fire the entire assembled first stage.

On ‎18‎.‎07‎.‎2016 at 3:49 PM, shynung said:

That would. But it can also reduce complexity, by allowing the main engines to remain fixed/gimbal-less.

Seems like history disagrees with you. There has been a move from verniers to gimbals.

Link to comment
Share on other sites

5 hours ago, DDE said:

Seems like history disagrees with you. There has been a move from verniers to gimbals.

I think that is to reduce the number of engines needed to be designed, to save on development costs.

Link to comment
Share on other sites

4 hours ago, shynung said:

I think that is to reduce the number of engines needed to be designed, to save on development costs.

Nope; even if verniers are part of the primary engine (e.g. RD-107/108) they apparently still are heavier than a gimbal.

Link to comment
Share on other sites

1 hour ago, DDE said:

Nope; even if verniers are part of the primary engine (e.g. RD-107/108) they apparently still are heavier than a gimbal.

Which is exactly my point. I think main engine gimbals are simpler to develop than vernier engines. Not only there are less chambers and nozzles to worry about, it also ends up lighter.

Link to comment
Share on other sites

  • 6 months later...
On 18/07/2016 at 9:27 AM, Bill Phil said:

It's not really that simple. The initial price of the engine (most of it R&D and man-hours) does go up the bigger it is, but the actual individual price increases at a lower rate.

Also, while there is the economies of scale, there's also the dis-economies of scale, where per unit costs go up instead of down, when you make more. Larger production run = higher price, in those cases.

Certainly. This was actually the plan that the N-1 was going to rely on, but in the end it was too complex.

That's a good point. I would like to add the increased complexity. You have to have more pipes and manifolds with clustered engines.

You can't really "just" add engines, you have to do a lot of stuff first. Redoing the plumbing, redesigning the thrust structure, and making sure that you can produce enough engines.

 

Regarding failure one example was falcon 9 3rd or 4th launch, one of the engines failed and the flight carried on nominal, you can see a dark streak in the trail!

also the quantity of production in the merlin engines reduced their price a lot, that's the reason they swapped from the krestel engine to merlin vacuum for the upper stage

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...