A route to aircraft-like reusability for rocket engines.

[Exoscientist]

A route to aircraft-like reusability for rocket engines.
https://exoscientist.blogspot.com/2024/04/a-route-to-aircraft-like-reusability.html

 An interesting discussion on longevity of jet engines:

 The question I raise is whether this could also increase the reuse capability of rocket engines.

 Near the end I suggest, SpaceX is using this principle of running the engines at lowered power to increase engine life for the purpose of increasing the reliability of the Raptors. If they are, then they should explore the potential of this principle to also extend rocket engine reuse capability.

  Bob Clark
 

Edited April 19 by Exoscientist

[SunlitZelkova]

Color me skeptical that jet engine maintenance can be translated to rocket engines. The stresses out on them in flight and the temperatures they operate at are completely different.

3 hours ago, Exoscientist said:

Near the end I suggest, SpaceX is using this principle or running the engines at lowered power to increase engine life for the purpose of increasing the reliability of the Raptors. If they are, then they should explore the potential of this principle to also extend rocket engine reuse capability.

What is your basis for suggesting this?

I am also skeptical of this. Current Raptors that fly are not intended for reuse right now, so there would be no reason to run them at lower power.

[magnemoe]

4 hours ago, SunlitZelkova said:

Color me skeptical that jet engine maintenance can be translated to rocket engines. The stresses out on them in flight and the temperatures they operate at are completely different.

What is your basis for suggesting this?

I am also skeptical of this. Current Raptors that fly are not intended for reuse right now, so there would be no reason to run them at lower power.

Now its an argument of lowering trust rater than shutting down engines later on. 
But rocket engines burn for short time compared to jet engines who might burn over half the time daily for years.
Who is true for passenger planes but not for fighter jets, they are much more highly stressed and need more maintenance. 

[Terwin]

Wikipedia says the burn-time of f-9 first-stage using FT engines is 162 seconds.  Even just a quick visual inspection of the engines of a Falcon 9 would mean maintenance time > run time. 

So if each first stage took less than 1 hour of inspections/repairs per launch, it would still be getting maintenance > 2x the total burn time of the engines, something that would likely put most airlines out of business.

(9*162 / 60 = 24.3 minutes of total burn time per launch.  May not include <1m for landing burn))

 

If SH takes 2 hours from landing to launch, it could still average > 10x maintenance time per flight-hour. (F9 takes just less than 10 minutes from launch to landing for RTLS)

I do not see much value in taking SH maintenance down to even a 1:1 ratio of flight time to maintenance time, as the falcon 9 needs >30 minutes to load fuel, and I would not expect SH to be faster than that.

(As refueling is often counted as maintenance, taking SH down below 10:1 maintenance time:flight time might well be impossible even if the only maintenance is refueling)

 

 

Edited April 19 by Terwin

[Codraroll]

On 4/18/2024 at 11:36 PM, SunlitZelkova said:

What is your basis for suggesting this?

Raptor is assumed to have reliability issues.

It is assumed that there exists evidence to support this.

It is also assumed that the previously debunked evidence/arguments still hold true despite aforesaid debunking.

The conclusion is assumed to stand firm regardless of the amount of evidence presented against it, the weakness of evidence/arguments presented in favour, and the lack of any solid evidence.

The basic assumption may be questioned, but no answer will be provided.

Such is the way.

Edited April 19 by Codraroll

[Exoscientist]

On 4/18/2024 at 5:36 PM, SunlitZelkova said:

Color me skeptical that jet engine maintenance can be translated to rocket engines. The stresses out on them in flight and the temperatures they operate at are completely different.

What is your basis for suggesting this?

I am also skeptical of this. Current Raptors that fly are not intended for reuse right now, so there would be no reason to run them at lower power.

 

 The sea level Raptor 2 sea level thrust is 230 tons. For its vacuum thrust estimate it as proportionally higher by how much higher the vacuum Isp is over the sea level Isp: 230*(353/327) = 248 tons.

 The Raptor Vacuum thrust is given as 258 tons. Then the total thrust for the 3 sea level Raptors and 3 vacuum Raptors should be 1,520 tons.

 But the thrust for the ship of the current version is given as only 1,250 tons:

 So it was started at 82% of full  thrust. Commonly, you throttle down a rocket engine as most of the propellant burned off. The result is the overall average thrust was less than 75%.

  Bob Clark

[mikegarrison]

Turbine engines are usually limited by the combustor, first stage turbine nozzle, and first stage turbine blades. In other words, the hottest part of the engine.

All metals lose their durability as they approach their temperature limits. If you keep the temps low enough compared to the temperature resistance of the metal, the hot section can last for an almost indefinite amount of time.

But to get more thrust and fuel efficiency, you want that T4 to be as hot as you can stand for it to be.

So it's a tradeoff. Usually engines are designed for a certain number of hours on-wing, and then they push the temperatures and pressures up and the cooling airflow down until they hit that target. They could get them to last longer, but it would cost in performance and efficiency. They could get them to have higher performance and efficiency, but they wouldn't last as long.

[Meecrob]

On 4/18/2024 at 5:36 PM, SunlitZelkova said:

Color me skeptical that jet engine maintenance can be translated to rocket engines. The stresses out on them in flight and the temperatures they operate at are completely different.

I'm not sure the stresses/temps are all that different between Earth's atmosphere and space. The main source of temp/stress is the rapidly expanding heated gas escaping from the combustion chamber, not the external environment.
Also keep in mind that jet engines and rocket engines are actually close siblings. They both utilize the Brayton combustion cycle to spin turbine discs to insane RPMs. The correct term for jet engine is "gas turbine," and the turbopump in a rocket engine is a turbine. Granted, the rocket engine usually pumps liquid, not gas, but we are still firmly in the realm of fluid dynamics. This is an issue that is totally solvable, its just a question of the optimal way to solve for SpaceX's purposes. I'm not sure even they will know until they get more telemetry.

On 4/18/2024 at 5:36 PM, SunlitZelkova said:

Current Raptors that fly are not intended for reuse right now, so there would be no reason to run them at lower power.

Current Raptor 2's are only to get Starship into the sky. Engines and airframes do not necessarily develop at the same pace. There are countless examples of aircraft test programs where the production engine was not introduced until very late in the program. Hell, the F-14 had its "flight test" engine till they made the A+ and D models. 

They are developing Raptor 3 currently. It will be incorporating lessons learned from Raptor 2 as well as introducing features that required more engineering time to develop, and thus were not available for Raptor 1 or 2. Also, if there isn't a Raptor 4 in the works, I'm calling it right now, someone abducted Elon and the person in charge is an imposter.

 

On 4/19/2024 at 10:59 AM, Terwin said:

Snip

The example/trope you seem to be getting at is when someone adds in overhaul time and says "every hour of flight time requires 10 hours of maintenance" or something similar. A different way to manipulate the numbers would be to say that for a 4 hour cross-country flight in a Cessna, the only maintenance performed was a 15 minute pre-flight inspection by the pilot.

Both of those examples are technically true, but are meaningless for comparisons.

In your example you are saying that overall, the average per launch is 1 hour inspection time. Cool, lets use your round number. When designing the maintenance schedule, engineering would take note of all things that require inspection and schedule those inspections to coincide with each other. So lets say we have 10 items that require inspection to keep the round numbers going. If done individually, they would take 10 hours to inspect. However, since inspections require removal of fairlings/access panels, etc. if you already have them removed for one inspection out of 10, it means you can save 9 removal and re-installs of the fairings. So, if the fairings take 10 minutes each, we are now at 8 hours, 20 minutes of inspection. Now that number is for one person. If possible, maintenance will get multiple inspectors on the task. Lets assume they could only get two. That still cuts our 10 hours down to 4 hours 10 minutes.

So now we have 9 flights on schedule and one flight delayed 4:10. I don't have the numbers handy, but that's about how many flights get scrubbed due to weather, isn't it? 1 in 10-ish?
 

This is all before we make the distinction between line maintenance and depot/heavy maintenance. So far, we have only been talking about line maintenance. The big benefit comes when SpaceX optimizes Raptor removal/reinstallation and they can just swap engines needing inspection out for (heavy maintenance) overhauled engines. Once they have enough boosters of the same block for testing uniformity, I bet they will swap boosters, rather than engines, just like how airlines will just swap entire planes when possible if they go tech at the gate.

Overall, I think in say ~5 years, we will have clickbait-y SpaceX youtubers with videos titled "HOW SPACEX CAN LAUNCH WITH ONLY A 15  MINUTE INSPECTION!!!"

 

On 4/19/2024 at 6:56 PM, Codraroll said:

Raptor is assumed to have reliability issues.

Yeah, I gotta admit, one thing @Exoscientist is good at is throwing us red herrings. You know he's sitting there laughing at us for all thinking he was dumping on Raptor's reliability, when he was really saying something along the lines of "if this is the state of Raptor now, they need to improve reliability. One way to do that is to not run them full throttle."
 

I think this is where we all need to remember that what SpaceX is trying to do is fundamentally different than anything rocketry related we have ever seen. The age of the billion dollar delicate Rube Goldberg machine is finally dying off!

 

On 4/20/2024 at 9:51 AM, Exoscientist said:

The sea level Raptor 2 sea level thrust is 230 tons. For its vacuum thrust estimate it as proportionally higher by how much higher the vacuum Isp is over the sea level Isp: 230*(353/327) = 248 tons.

 The Raptor Vacuum thrust is given as 258 tons. Then the total thrust for the 3 sea level Raptors and 3 vacuum Raptors should be 1,520 tons.

 But the thrust for the ship of the current version is given as only 1,250 tons:

If I were Elon, I would purposefully put up graphics with numbers that don't add up just to see your reaction! Seriously though, given the digits, its possible its a typo?

 

On 4/20/2024 at 10:15 AM, mikegarrison said:

So it's a tradeoff.

Agreed. I could even see SpaceX going with different variants of Raptor once its operational. Reliability is different in LEO on an unmanned fuel tanker Starship vs. a crewed Starship on the Martian surface.

[Terwin]

52 minutes ago, Meecrob said:

The example/trope you seem to be getting at is when someone adds in overhaul time and says "every hour of flight time requires 10 hours of maintenance" or something similar. A different way to manipulate the numbers would be to say that for a 4 hour cross-country flight in a Cessna, the only maintenance performed was a 15 minute pre-flight inspection by the pilot.

Both of those examples are technically true, but are meaningless for comparisons.

I did not cite any examples or tropes, what are you talking about?

52 minutes ago, Meecrob said:

In your example you are saying that overall, the average per launch is 1 hour inspection time. Cool, lets use your round number. When designing the maintenance schedule, engineering would take note of all things that require inspection and schedule those inspections to coincide with each other. So lets say we have 10 items that require inspection to keep the round numbers going. If done individually, they would take 10 hours to inspect. However, since inspections require removal of fairlings/access panels, etc. if you already have them removed for one inspection out of 10, it means you can save 9 removal and re-installs of the fairings. So, if the fairings take 10 minutes each, we are now at 8 hours, 20 minutes of inspection. Now that number is for one person. If possible, maintenance will get multiple inspectors on the task. Lets assume they could only get two. That still cuts our 10 hours down to 4 hours 10 minutes.

So now we have 9 flights on schedule and one flight delayed 4:10. I don't have the numbers handy, but that's about how many flights get scrubbed due to weather, isn't it? 1 in 10-ish?

I was attempting to illustrate that rocket burn times per flight are so radically different from jet burn times per flight, that attempting to compare them is like comparing how you serve blueberries and watermelon.

Sure they are both firm berries that get soft when they go bad, but they are wholly different from each other in many practical ways.  I cannot imagine someone being satisfied with a slice or two of blueberry, nor would I expect to meet someone who frequently eats dozens of watermelons at a sitting.

Much like I would not expect rocket engines to have maintenance times < burn times like jet engines can manage.  Rocket burn times are just not on the same scale as turbofan engine run-times.

 

I also strongly suspect that chemical rocket engines tend to burn much hotter and higher pressure than turboprop engines, if only because ~80% of the turboprop working fluid is inert gasses.

Rocket engines clearly have a much higher fuel through-put than turboprop engines:

"A Boeing 747 or 777 typically consumes approximately 5,000 gallons of fuel per hour while in flight."

"Raptor consumes 140kg of methane per second"

Conversion:

Jet A1 Fuel Conversion Chart

Lbs	Litres	U.S. Gallons
1000	568	150

5000 gal -> 33,333 lbs -> 15,119  kg

15,119 kg/hr / 3600 s/hr = 4.2kg/s

So 4.2kg/s(all 4 jet engines on 747) vs 140kg/s (per engine). 

This suggests each raptor burns ~ 133 times as much fuel per second as a 747 engine.  

So yes, that seems very different, even if the outside environment does not change much.

[magnemoe]

22 hours ago, Terwin said:

I did not cite any examples or tropes, what are you talking about?

I was attempting to illustrate that rocket burn times per flight are so radically different from jet burn times per flight, that attempting to compare them is like comparing how you serve blueberries and watermelon.

Sure they are both firm berries that get soft when they go bad, but they are wholly different from each other in many practical ways.  I cannot imagine someone being satisfied with a slice or two of blueberry, nor would I expect to meet someone who frequently eats dozens of watermelons at a sitting.

Much like I would not expect rocket engines to have maintenance times < burn times like jet engines can manage.  Rocket burn times are just not on the same scale as turbofan engine run-times.

 

I also strongly suspect that chemical rocket engines tend to burn much hotter and higher pressure than turboprop engines, if only because ~80% of the turboprop working fluid is inert gasses.

Rocket engines clearly have a much higher fuel through-put than turboprop engines:

"A Boeing 747 or 777 typically consumes approximately 5,000 gallons of fuel per hour while in flight."

"Raptor consumes 140kg of methane per second"

Conversion:

Jet A1 Fuel Conversion Chart

Lbs	Litres	U.S. Gallons
1000	568	150

5000 gal -> 33,333 lbs -> 15,119  kg

15,119 kg/hr / 3600 s/hr = 4.2kg/s

So 4.2kg/s(all 4 jet engines on 747) vs 140kg/s (per engine). 

This suggests each raptor burns ~ 133 times as much fuel per second as a 747 engine.  

So yes, that seems very different, even if the outside environment does not change much.

Yes, rocket engines are hungry. It burns even more LOX.  
This is the main problem with cross-feed and drop tanks, you have an flow usually only seen in hydro plants. Yes some large water handling systems might handle many ton's second but here pressure is low and bubbles and cavitation is not an issue. 

[Exoscientist]

On 4/19/2024 at 6:56 PM, Codraroll said:

Raptor is assumed to have reliability issues.

It is assumed that there exists evidence to support this.

It is also assumed that the previously debunked evidence/arguments still hold true despite aforesaid debunking.

The conclusion is assumed to stand firm regardless of the amount of evidence presented against it, the weakness of evidence/arguments presented in favour, and the lack of any solid evidence.

The basic assumption may be questioned, but no answer will be provided.

Such is the way.

 

 I am arguing that NASA was blindsided by the low 40 to 50 ton capability of the current version of the SuperHeavy/Starship. A couple of reasons why I say this. First, while Elon was extolling the “success” of the latest test flight in his recent update at Starbase on the Starship development, NASA soon after wards started making plans for use Starship in Artemis III that won’t use the Starship as a lander. (NASA did not openly reveal this; it had to be leaked.) Note also the proposed options NASA is considering also would not use refueling of the Starship. The low 40 to 50 tons to orbit would cause impractically large number of refueling missions. Then V2 or even V3 would be needed for this and I’m suggesting NASA believes neither of those would be ready by Artemis III.

 Note the SpaceX plan for a lander using multiple refuelings absolutely can not work if the Raptor can not operate reliably for both boostback and landing burns. Relighting, apparently, successfully at boostback is not sufficient if a Raptor explosion on landing causes vehicle RUD.

 So I’m also arguing NASA has no confidence SpaceX can solve the relighting reliability issue, that requires three Raptor firings per flight both for the booster and ship, by Artemis III.

 Beyond that, another key reason why I say NASA was blind-sided by the low payload capability of the current version is if you run the numbers SpaceX cited for the specifications on the current version, it should easily make 100+ tons to orbit even as a reusable. That it makes at best half that suggests the dry mass or Raptor values or more likely  both are significantly worse than the values cited by SpaceX.

  Bob Clark

 

Edited April 23 by Exoscientist