Debug: Database connection successful
You are not logged in.
GW,
That is why I set the dV requirement higher than the bare minimum required to attain orbit. My presumption is that the RCS (GOX/RP1) and OMS (LOX/RP1) are integrated systems that burn residual propellants. Mission duration is 24 hours or less. If passenger transfer to an interplanetary colonization ship can't be accomplished inside of 24 hours, then you de-orbit and land. 1.5% of 2,000t of propellant is 30t. That's plenty of propellant for orbit circularization, on-orbit maneuvering, and de-orbit burn. The Space Shuttle had 14,297kg of NTO/MMH propellant in total for its OMS and RCS systems. My proposed SSTO vehicle has more than 2X as much residual propellant in the feed lines to the main engines, which is where it will be stored until it's fed into the OMS and RCS. The primary power subsystem will consume around 750kg to 1,000kg of propellants over 24 hours to generate electrical power to provide life support and power avionics as well as flight control surfaces and vehicle subsystems. That's a reasonable propellant reserve- more than the Space Shuttle had, and since flights are regularly scheduled, if it proves necessary a rescue vehicle will be launched the next day to offload the passengers and either take them to their ship or take them back to Earth.
Edit:
The dV increment I factored in included enough propellant for to achieve orbit while covering anticipated / prototypical gravity and drag losses, on-orbit maneuvering, de-orbit burn, and running an APU to provide electrical power and propellant pressurization during the entire flight. Airliners don't use Hydrazines. They burn O2 and kerosene. Since this vehicle is an orbital class airliner, if it needs power, then it gets that power from the same propellant sources that feed the main engines. Containing dry vehicle mass growth is critical, so eliminating separate subsystems and multiple propellant types is a form of vehicle mass control. LOX or GOX and RP1 requires spark plugs, but modern Iridium-tipped spark plugs are pretty reliable. Whether that is seen as a major risk or complication, I don't know, but the RS-25s had Iridium-tipped spark plugs to assure ignition within the pre-burners, and they were fired continuously all the way to orbit.
Since this vehicle doesn't have any Hydrazines onboard, all the special handling equipment, training, and tasks related to loading Hydrazine type propellants can be eliminated. Separate Helium tanks for propellant pressurization can also be eliminated. We must simplify the entire vehicle design to the degree possible, use the latest and greatest materials and technologies for the performance benefits provided, and refrain from stubbornly fixating on how something was done in the past, when it was self-evident that made SSTO infeasible.
Last edited by kbd512 (2024-06-15 22:01:57)
Offline
Like button can go here
I'm curious if we at least agree that total impulse per unit volume of propellant is a figure of merit for SSTO, rather than propellant weight alone. Propellant weight absolutely does drive vehicle dry mass fraction, but not to the same degree as LH2 tank internal pressurization, which is double that of LOX or RP1.
Total impulse per unit volume of propellant results in at least 2X greater structural mass allocation for a LH2-fueled vehicle, and more TPS mass allocation, albeit not 2X more due to cube-square, maybe 1.2X to 1.4X more for a lifting body design. As you scale-up, that problem only becomes worse, not better, because the material doesn't magically become any stronger or stiffer as volume and surface area increases. Cube-square, applicable to all solutions, ameliorates the surface area and TPS mass problem, but at least 50% greater dry vehicle mass for LH2 is unavoidable due to LH2's extremely low density per unit volume and high internal pressurization to force-feed fuel into the feed lines / pump inlets for the engines.
Even LH2's 100s greater Isp, relative to LOX/RP1, is not enough to overcome that problem. The low thrust per unit engine weight of LH2-fueled engines only aggravates the total impulse problem, because it results in all major vehicle components becoming heavier - propellant tanks, engines, TPS, and landing gear. LH2 is a disabling, rather than enabling, propellant selection for SSTO.
Run a straightforward total impulse calc to demonstrate the total impulse provided by 1m^3 of LOX/RP1 and LOX/LH2 at avg Isp for staged combustion engines (338s for LOX/RP1, 409s for LOX/LH2). I even tried the calc using 452s as the Isp for LOX/LH2, but the end result is that the LH2 fueled vehicle STILL has a greater dry mass fraction than LOX/RP1 with avg Isp.
After that's done, then factor in 200:1 TWR for LOX/RP1 engines vs 77:1 or even 100:1 (imaginary engine tech) for LOX/LH2 engines.
Is it at least obvious that pressurizing LH2 to 45psi represents a far more significant mechanical load on the propellant tank structure than a 5g acceleration with a full LOX or RP1 tank of equivalent volume?
By the time you can achieve 5g in the LOX/RP1 powered rocket, at least half of your propellant has already been expended. At all times, you're feeding LOX/RP1 to the engines at 20psi to 22psi, rather than 45psi for LH2.
If your vehicle's propellant tanks are designed to withstand 45psi LH2 pressurization and 3g acceleration loads, then you're de-facto able to use those same tank designs to withstand higher acceleration loads from heavier LOX or RP1 propellants, at half the internal pressurization, after half of the propellant has been consumed during ascent. Your fuel tank is not subjected to LH2 permeation and embrittlement, and the cold LOX tank is still much warmer than the LH2 tank.
The problems LOX/LCH4 poses for a SSTO are less severe than those imposed by LOX/LH2, yet they're still greater than LOX/RP1, with the end result being that you push less payload into orbit than you do with LOX/RP1 for a given vehicle dry mass. The only way you net an actual benefit for SSTO vehicle dry mass is if LOX/LCH4 is the same density as LOX/RP1, same engine TWR, and you get the Isp improvement of CH4 over RP1. Then and only then do you push more payload into orbit for a given vehicle dry mass. CH4 gives us a modest bump in Isp, the same or slightly worse TWR for CH4-fueled engines, and significantly greater vehicle volume and thus dry mass.
With a 10km/s dV increment as a hard requirement, a practical SSTO becomes a highly constraining problem with an equally constrained practical solution set. If your propellant combo doesn't have a density near that of water, or your engines cannot provide 175:1+ TWR, or your propellant tank material isn't 400ksi+ with stiffness similar to steel and weight similar to Magnesium, then you don't have a practical SSTO, because all your payload mass gets consumed by the vehicle itself. All of those performance characteristics must be combined into a single vehicle design.
Offline
Like button can go here
Another thing to consider in regard to tank inert mass: is it, or is it not, also the airframe of the vehicle? If it is, what additional loads besides endways gees and internal pressurization, must it carry? If it is not, its mass is in addition to the airframe mass that surrounds it.
I'm only guessing that wall thickness depends upon internal pressure level and material strength as a rough first approximation, and the volume of material to be weighed is the tank surface area times that thickness times some factor a tad above 1, that accounts for local thickness increases, such as at the head-cylinder joints. That takes care of low-density hydrogen, vs the other fuels.
That simple estimate ignores other very real-world mass additions for things like internal surge baffles, and the plumbing that takes propellant from another tank through this one, to the engines. Sometimes that plumbing must be quite heavily insulated.
As for resisting other airframe loads such as bending, and external pressure on one side but not the other (acting to cause collapse laterally), that requires either external or internal stringers, and internal frames, to act similar to a monocoque fuselage.
And I still do not believe in organic matrix polymers that aren't structural junk at 300+F. Composite tanks will inherently require more heat protection that stainless steel, in ANY entry scenario, even the low velocity entries of 1st stages. 300 F is the total temperature of the air at only Mach 2.2 in the stratosphere. That's only 0.64 km/s velocity through the cold stratospheric air.
Aluminum goes slightly hotter than that at about Mach 2.5-2.6.
GW
Last edited by GW Johnson (2024-06-16 14:12:56)
GW Johnson
McGregor, Texas
"There is nothing as expensive as a dead crew, especially one dead from a bad management decision"
Offline
Like button can go here
P2SI™ 900HT composite prepreg system exhibits a glass transition temperature of 900°F (482°C) and excellent mechanical strength retention above 400°C. Void free composites are easily fabricated into structural components using a variety of reinforcements, including glass, quartz, and carbon. This paper presents the latest processing, performance, and durability characterization of this recently developed composite system. In addition, results on the use of P2SI™ 900HT for high temperature tooling applications, including 371°C autoclave cure of polyimide composites is presented. To date, no other polymer matrix composite has been successfully used as a tooling material in 371°C autoclave processes.
Resins for the Hot Zone, Part I: Polyimides
This article speaks to composites withstanding sustained temperatures of 600F to 1,000F, used in military aerospace applications. Belief is not required, because the high temperature performance is still there, regardless of knowledge or belief. The F-35, in particular, has composite components in its airframe that must withstand temperatures of up to 1,000F in routine service, particularly those near the engine and its exhaust nozzle. The complaints from the US Air Force were that even that level of capability was not enough, so we developed resins with ever-higher service temperature limits. Note that this article is dated from 2009 and it states, even 5 years ago, these materials were still in the incubation stage of development".
Materials tech has moved on, even if the thinking of the older generations of aerospace engineers who used metals has not.
All the materials tech I spoke of is quite real, it's actively used by our military or space contractors in real aircraft or rockets, and it does, in point of fact, withstand the temperature range it's subjected to, else we'd have aircraft literally falling out of the sky following catastrophic structural failures. Since that's not happening, the materials tech is on point.
Offline
Like button can go here
That kind of stuff is quite expensive. To my knowledge, whole airplanes are not made with it.
GW
GW Johnson
McGregor, Texas
"There is nothing as expensive as a dead crew, especially one dead from a bad management decision"
Offline
Like button can go here
GW,
The F-22, F-35, and B-21 are, even though parts of their airframes don't require it. The weapons bays on the F-22 and F-35 became so hot that the weapons themselves, or at least their electronics, were failing. The airframes were fine, but the electronics failed.
Now that I've thought about that, messing up the electronics, even if the airframe can tolerate the heat without any real issues, might prove far more problematic. That would obviously be very bad since this vehicle has no windows and you're trying to land it using instruments only. We might have to rethink protection levels in certain areas to prevent the electronics from getting fried. We're relying on DAS and Ku-band radar to fly all-weather precision approaches. In the F-35, they used fuel as a thermal sink to cool the electronics. We'll need something similar.
Ever wonder why we're burning out F-135 hot sections in our F-35s within 3,000hrs?
They're demanding so much bleed air to cool the electronics that their hot sections are getting BBQ'd for lack of cooling air flow over the blades. The technical solution has been to make the hot sections even more tolerant of high temperatures, rather than reducing the cooling air demand for the electronics. More modern electronics with the F-35's latest tech refresh haven't helped. They produce less waste heat, but now there's simply more total computing power crammed into the same airframe and no space or bleed air reserve for more cooling capacity, because the Air Force changed F-35 compute requirements after the engine design was finalized and tested. The airframes themselves are built to readily withstand the heat coming from the engine(s). The electronics? That's another story. That's why the F-22s and F-35s open the weapons bays while on the ground after they start the engine(s).
Incidentally, prices for high temperature resins have been improving. They're now used in exhaust systems for sports cars and motorcycles. It's still very pricey, I'll grant you that. An enormous amount of development work has been done on high temperature resins, though. We've had these resins for at least 15 years now, and they do work. Whatever the manufacturer states, I still subtract at least 150F off of whatever they claim in the way of temperature tolerance, even if there is ample test data proving that it works. I think F-22, F-35, and B-21 count as "ample test data". The airframes of those jets would all fail catastrophically if they couldn't take the heat.
We're using expensive fiber, expensive resin, and expensive TPS materials. We're paying up front for a high quality airframe, rather than paying later with lives lost. I expect the bill of materials for the fiber and resin to amount to about $10M. All told, we're looking at around 27,000kg worth of specialty materials, primarily fiber and resin with small bits of very costly steel. This includes about 16,200kg, at $250/kg, which is what T1100 fiber tow / roving costs in bulk, so $4.05M for the fiber. I expect the resin to cost at least as much per kg. The TPS materials will cost around $1,000/kg. In terms of airframe cost per kg, it's very steep, but you have to remember that it doesn't weigh very much and a lot of the weight is in the engines, which don't cost nearly as much to produce. I would estimate a $75M airframe cost (materials, energy, and labor), which is quite a lot ($1,250/kg)- easily more than the F-22 or F-35 on a per-weight basis, but over 1,000 flights it's only $75,000/kg, and per-passenger-flight cost only amounts to $150 of the ticket price. Almost all of the real cost comes in the form of propellants / personnel / facilities, just like a normal airline service.
If you were to show the airframe bill of materials to a manager at Lockheed-Martin, they would choke on the cost. Northrop-Grumman might find it more reasonable if we're talking about B-21 airframe cost, but even those airframes are far less costly per unit weight, if the electronics and RAM specialty materials were excluded. That's why LM put so many Titanium forgings in the F-22. In bulk, it was far less expensive than CFRP, and they were pretty fanatical about keeping the actual airframe cost within the realm of sanity. The compromise with the Air Force was making every 4th spar from Titanium because it was less than half the cost of CFRP. The end result was absurd wing spar inspection and maintenance costs to monitor cracks in the Titanium, but hey, the government saved a little bit of money up front on the airframe itself. There's a nice presentation on this special brand of silliness.
I figure that we're charging about $150 per ticket sold to pay off the airframe, $6,000 to $7,000 for the propellants, plus some additional cost for facilities, permits / paperwork / legal expenses, and everything else. If our per-ticket cost amounts to $10,000 per passenger, then this is easily the cheapest ride to orbit by a lot. For a family of 4, we need to figure out how to get the total cost down to about $100,000 to move mom, dad, and the kids to Mars. Doing that will require using multiple complementary launch technologies with exceptionally high reliability and launch rates, low marginal / per-flight costs, and streamlined bureaucratic nonsense related to FAA type certification. That last part will be the real "trick" to making all of this work.
Offline
Like button can go here
For all...
GW's work posted in the exrocketman.blogspot.com site is now available by a single click.
https://exrocketman.blogspot.com/search?q=ssto
The paper is detailed, with plenty of graphs to illustrate the text.
I am guessing here, but it is possible that anyone offering an opinion without reading the paper is indeed likely to be missing something important.
(th)
Offline
Like button can go here
In the Google Meeting of Sunday July 6, 2025, kbd512 went over points he has made in various posts, in support of the proposition that a carbon based fuel design can achieve SSTO. It is generally accepted that an LH2/LOX design can achieve SSTO. I asked Google's Gemini to evaluate the two propositions, and I thought it came up with a response that should satisfy both sides of the discussion:
I would like to share your analysis of the two competing designs with our forum audience. Please summarize our discussion in BBCode, and put the equations inside code blocks so they stand out for the readers.
Current Date: 2025-07-06
SSTO Propellant Showdown: LH2/LOX vs. Carbon Fuels
This is a follow up to the Google Meeting of 2025/07/06...
Following up on our recent discussions about Single-Stage-To-Orbit (SSTO) concepts, I've gathered some insights from Gemini regarding the two primary design approaches we've been considering: our engineer's proposal of LH2/LOX and kbd512's idea of using a denser carbon fuel.
First, a quick recap on thrust and efficiency for a fixed-nozzle engine as it ascends. Our earlier discussion established that for a sea-level optimized engine, thrust will increase with altitude due to decreasing atmospheric back pressure. The common misconception that efficiency decreases is clarified by the concept of Specific Impulse (Isp).
Edit: The conclusion in the passage above reflects a hypothesis. The hypothesis appears to have been accepted as fact, and the rest of the analysis proceeded from there. As I re-read this conversation, I recognize that I did not emphasize strongly enough that the idea that thrust increases with altitude is a hypothesis that needs to be tested. Newton's laws should rule in this situation. If mass flow is held constant, then thrust should remain constant regardless of atmospheric pressure.
Understanding Thrust and Efficiency (Specific Impulse):
The fundamental equation for rocket thrust (F) is:
F = m_dot * v_e + (P_e - P_a) * A_e
Where:
m_dot: mass flow rate of the propellant
v_e: exhaust velocity
P_e: exhaust pressure at the nozzle exit
P_a: ambient (atmospheric) pressure
A_e: nozzle exit area
As a rocket ascends, P_a decreases significantly. This makes the (P_e - P_a) * A_e term (pressure thrust) more positive, directly increasing the total thrust F.
Efficiency in rocketry is best described by Specific Impulse (Isp):
I_sp = F / (m_dot * g_0)
Where g_0 is standard gravity. A higher Isp means more thrust per unit of propellant consumed.
For a fixed-nozzle, sea-level engine, as P_a decreases, F increases while m_dot remains constant. This means that the Specific Impulse (Isp) actually increases with altitude. So, there isn't a competition where one factor shrinks net thrust while the other grows it; both thrust and efficiency generally improve for a fixed-nozzle engine as it climbs out of the atmosphere.
Now, let's dive into the two SSTO propellant proposals:
Proposal 1: LH2/LOX (Engineer's Approach)
This approach uses Liquid Hydrogen (LH2) and Liquid Oxygen (LOX), with an engine designed to compromise between sea-level and vacuum operation.
Pros:
Highest Specific Impulse (Isp): LH2/LOX offers the absolute highest chemical Isp (typically 430-460 seconds in vacuum). This is theoretically the most efficient propellant combination for getting mNow, let's dive into the two SSTO propellant proposals:
Proposal 1: LH2/LOX (Engineer's Approach)
This approach uses Liquid Hydrogen (LH2) and Liquid Oxygen (LOX), with an engine designed to compromise between sea-level and vacuum operation.
Pros:
Highest Specific Impulse (Isp): LH2/LOX offers the absolute highest chemical Isp (typically 430-460 seconds in vacuum). This is theoretically the most efficient propellant combination for getting mass to orbit.
Clean Exhaust: Produces only water vapor.
Proven Technology: Well-understood and widely used for upper stages.
Cons:
Extremely Low Density of LH2: This is the major hurdle for SSTO. LH2 requires enormous, lightweight tanks, leading to a much larger and heavier vehicle dry mass. This "tankage mass penalty" often negates the Isp advantage.
Large Vehicle, More Drag: Bigger vehicle means more aerodynamic drag during atmospheric ascent.
Cryogenic Challenges: LH2 is deeply cryogenic (-253°C), requiring significant insulation and battling boil-off losses.
Engine Thrust-to-Weight: LH2 engines often have lower thrust-to-weight ratios due to the large pumps and plumbing needed for the low-density fuel.
Proposal 2: Carbon Fuel (e.g., Methane/LOX or Kerosene/LOX) (kbd412's Approach)
This proposes using a denser "carbon fuel" like Methane (CH4) or Kerosene (RP-1) to achieve a more compact vehicle, accepting a lower Isp.
Pros:
High Density:This is the significant advantage. Denser fuels mean much smaller and lighter tanks for the same propellant mass, leading to a significantly more compact and lighter overall vehicle dry mass.
Reduced Drag:A smaller vehicle means less aerodynamic drag.
Easier Handling:Kerosene is storable at room temperature; Methane is cryogenic but less so than LH2 (-161°C vs. -253°C), meaning easier handling and less boil-off.
Higher Engine Thrust-to-Weight:Engines for denser fuels generally have better thrust-to-weight ratios.
Lower Cost:Generally cheaper than LH2.
Cons:
Lower Specific Impulse (Isp):This is the main drawback. Kerosene/LOX is typically 330-350 seconds vacuum Isp, Methane/LOX around 360-380 seconds. This means more total propellant mass is needed for a given change in velocity.
Coking/Sooting (Kerosene):Kerosene can leave carbon deposits, affecting reusability. Methane burns much cleaner.
Combustion Products:Produce CO2 and water (and soot for kerosene), though in small global quantities.
Which is the "Clear Winner" for SSTO?
Historically, theoretical SSTO designs often leaned towards LH2/LOX due to its superior Isp, as the sheer performance needed for SSTO from Earth is immense. However, the resulting vehicles are often impractically large with very low payload fractions.
The real-world answer, particularly considering reusability and operational factors, points towards the denser carbon fuels, especially Methane/LOX (Methalox).
While Methalox has a lower Isp than Hydrolox, its much higher density leads to a smaller, lighter, and easier-to-build vehicle. When you factor in the overall system efficiency – including vehicle dry mass, ease of operations, reusability, and even potential for in-space refueling or ISRU (In-Situ Resource Utilization, like on Mars) – the density advantage of methane often outweighs hydrogen's Isp advantage for a practical, reusable rocket design. This is the reasoning behind SpaceX's choice of Methalox for Starship, a vehicle pushing the boundaries of what's possible in spaceflight.
So, while LH2/LOX might be the "technical winner" for raw propellant performance on paper, the carbon fuel (methane) approach is arguably the more viable and practical path for designing achievable advanced launch systems, even if true Earth SSTO remains a monumental challenge.
ass to orbit.Clean Exhaust:Produces only water vapor.
Proven Technology:Well-understood and widely used for upper stages.
Cons:
Extremely Low Density of LH2:This is the major hurdle for SSTO. LH2 requires enormous, lightweight tanks, leading to a much larger and heavier vehicle dry mass. This "tankage mass penalty" often negates the Isp advantage.
Large Vehicle, More Drag:Bigger vehicle means more aerodynamic drag during atmospheric ascent.
Cryogenic Challenges:LH2 is deeply cryogenic (-253°C), requiring significant insulation and battling boil-off losses.
Engine Thrust-to-Weight:LH2 engines often have lower thrust-to-weight ratios due to the large pumps and plumbing needed for the low-density fuel.
Proposal 2: Carbon Fuel (e.g., Methane/LOX or Kerosene/LOX) (kbd412's Approach)
This proposes using a denser "carbon fuel" like Methane (CH4) or Kerosene (RP-1) to achieve a more compact vehicle, accepting a lower Isp.
Pros:
High Density:This is the significant advantage. Denser fuels mean much smaller and lighter tanks for the same propellant mass, leading to a significantly more compact and lighter overall vehicle dry mass.
Reduced Drag:A smaller vehicle means less aerodynamic drag.
Easier Handling:Kerosene is storable at room temperature; Methane is cryogenic but less so than LH2 (-161°C vs. -253°C), meaning easier handling and less boil-off.
Higher Engine Thrust-to-Weight:Engines for denser fuels generally have better thrust-to-weight ratios.
Lower Cost:Generally cheaper than LH2.
Cons:
Lower Specific Impulse (Isp):This is the main drawback. Kerosene/LOX is typically 330-350 seconds vacuum Isp, Methane/LOX around 360-380 seconds. This means more total propellant mass is needed for a given change in velocity.
Coking/Sooting (Kerosene):Kerosene can leave carbon deposits, affecting reusability. Methane burns much cleaner.
Combustion Products:Produce CO2 and water (and soot for kerosene), though in small global quantities.
Which is the "Clear Winner" for SSTO?
Historically, theoretical SSTO designs often leaned towards LH2/LOX due to its superior Isp, as the sheer performance needed for SSTO from Earth is immense. However, the resulting vehicles are often impractically large with very low payload fractions.
The real-world answer, particularly considering reusability and operational factors, points towards the denser carbon fuels, especially Methane/LOX (Methalox).
While Methalox has a lower Isp than Hydrolox, its much higher density leads to a smaller, lighter, and easier-to-build vehicle. When you factor in the overall system efficiency – including vehicle dry mass, ease of operations, reusability, and even potential for in-space refueling or ISRU (In-Situ Resource Utilization, like on Mars) – the density advantage of methane often outweighs hydrogen's Isp advantage for a practical, reusable rocket design. This is the reasoning behind SpaceX's choice of Methalox for Starship, a vehicle pushing the boundaries of what's possible in spaceflight.
So, while LH2/LOX might be the "technical winner" for raw propellant performance on paper, the carbon fuel (methane) approach is arguably the more viable and practical path for designing achievable advanced launch systems, even if true Earth SSTO remains a monumental challenge.
Thoughts?
(th)
Last edited by tahanson43206 (Today 08:20:48)
Offline
Like button can go here
I went back to Gemini to ask for clarification about how pressure of the atmosphere could interfere with thrust at sea level.
I pointed out that the idea that thrust might increase as a rocket ascends and atmospheric pressure decreases is a ** hypothesis ** that needs to be tested. Gemini accepted the correction, and then attempted to explore the physics of the situation. It introduced the interesting idea that molecules of atmosphere beating on the outside of the expansion bell might act to diminish the upward impact of exhaust molecules bouncing against the inside of the expansion bell.
Thank you for making the time and performing the work to complete this follow up. I think this will be a valuable addition to our opening post. Please prepare this conversation for display using BBCode. As a suggestion, instead of using Center and bold, please consider using header 'h' to highlight text. Our older software is limited. Also, you can use List if appropriate. Our software recognizes that command.
Got it! Thank you for the feedback on the formatting, and for guiding this discussion to such a clear and precise conclusion. This refined explanation should be very helpful for your forum members.
Here is the BBCode summary incorporating the latest discussion, using headers and lists as requested:
Current Date: 2025-07-06
SSTO Propellant Showdown: LH2/LOX vs. Carbon Fuels & A Deep Dive into Thrust Physics
Hi everyone,
We've had a fascinating and insightful discussion about Single-Stage-To-Orbit (SSTO) concepts, specifically comparing two propellant approaches and delving into the fundamental physics of how rocket thrust changes with altitude. Special thanks to [member's name] for posing such thought-provoking questions that led to a much deeper understanding!
Revisiting Thrust and Altitude: The Molecular Perspective
Our initial discussion explored how a fixed-nozzle, fixed-mass-flow engine's thrust changes with altitude. The simple equation for thrust (F) is:
F = m_dot * v_e + (P_e - P_a) * A_e
Where:
m_dot: mass flow rate of the propellant (assumed constant)
v_e: exhaust velocity
P_e: exhaust pressure at the nozzle exit
P_a: ambient (atmospheric) pressure
A_e: nozzle exit area
The key point that generated much discussion was how P_a (ambient pressure) affects thrust, especially given that rocket exhaust is supersonic. How can atmospheric molecules "communicate" back to the rocket?
Molecular Interactions at the Expansion Bell:
The answer lies in understanding the forces acting on the expansion bell (the divergent part of the nozzle) at a molecular level:
Internal Forces (Generating Thrust):
Inside the bell, the high-speed exhaust gas molecules collide with the internal, angled walls. Each collision imparts a tiny forward push to the rocket. This continuous bombardment creates the internal pressure (Pe) and is the primary mechanism of thrust generation.
For a fixed engine design with a constant propellant flow, this internal force generation remains consistent, regardless of the external atmosphere.
External Forces (Diminishing Thrust):
Simultaneously, the outside surface of the expansion bell (and the rest of the rocket) is constantly bombarded by ambient air molecules. These molecules, moving randomly, exert the atmospheric pressure (Pa) inwards on the rocket's outer surfaces.
Crucially, this external pressure acts as an opposing force. It's like the atmosphere is applying a continuous brake, reducing the net forward push that the internal engine forces are trying to achieve.
Conclusion on Thrust and Altitude:
Therefore, the intuitive understanding is correct:
At sea level, with high atmospheric pressure (Pa), a significant number of external air molecules are pushing inward on the outside of the expansion bell. This creates a substantial "diminishing effect" or "reduction" in the total thrust. The engine is still producing its full internal force, but it's being actively countered by the atmosphere.
As the rocket ascends, the density of the atmosphere and thus the ambient pressure (Pa) rapidly decrease.
With fewer external air molecules colliding with the outside of the expansion bell, the "backward push" from the atmosphere significantly lessens.
Since the internal forward push from the engine remains constant, the net forward thrust increases because there's less external opposition.
So, to be precise: "The thrust of our rocket engine is less when it is at the surface because the molecules of the external atmosphere are pushing inward on the outer surface of the expansion bell, effectively opposing a portion of the forward thrust being generated by the exhaust. As the rocket ascends, this 'molecular interference' by the external environment rapidly diminishes, leading to an increase in the net thrust until, in the vacuum of space, the engine operates without any significant external molecular resistance to its thrust."
SSTO Propellant Comparison: LH2/LOX vs. Carbon Fuels
Now, let's circle back to the two primary design approaches for SSTO, keeping in mind that true Earth SSTO remains a monumental engineering challenge.
1. LH2/LOX (Engineer's Approach)
This approach uses Liquid Hydrogen (LH2) and Liquid Oxygen (LOX), with an engine designed to compromise between sea-level and vacuum operation.
Pros:
Highest Specific Impulse (Isp):LH2/LOX offers the absolute highest chemical Isp (typically 430-460 seconds in vacuum). This means it's theoretically the most efficient at converting propellant mass into thrust.
Clean Exhaust:Produces only water vapor.
Proven Technology:Well-understood and widely used for upper stages.
Cons:
Extremely Low Density of LH2:This is the major hurdle for SSTO. LH2 requires enormous, lightweight tanks for adequate propellant, leading to a much larger and heavier vehicle dry mass. This "tankage mass penalty" often significantly offsets the Isp advantage.
Large Vehicle, More Drag:Bigger vehicle means more aerodynamic drag during atmospheric ascent.
Cryogenic Challenges:LH2 is deeply cryogenic (-253°C), requiring significant insulation and battling boil-off losses.
Engine Thrust-to-Weight:LH2 engines often have lower thrust-to-weight ratios due to the large pumps and plumbing needed for the low-density fuel.
2. Carbon Fuel (e.g., Methane/LOX or Kerosene/LOX) ([member's name]'s Approach)
This proposes using a denser "carbon fuel" like Methane (CH4) or Kerosene (RP-1) to achieve a more compact vehicle, accepting a lower Isp.
Pros:
High Density:This is the significant advantage. Denser fuels mean much smaller and lighter tanks for the same propellant mass, leading to a significantly more compact and lighter overall vehicle dry mass.
Reduced Drag:A smaller vehicle means less aerodynamic drag.
Easier Handling:Kerosene is storable at room temperature; Methane is cryogenic but less so than LH2 (-161°C vs. -253°C), meaning easier handling and less boil-off.
Higher Engine Thrust-to-Weight:Engines for denser fuels generally have better thrust-to-weight ratios.
Lower Cost:Generally cheaper than LH2.
Cons:
Lower Specific Impulse (Isp):This is the main drawback. Kerosene/LOX is typically 330-350 seconds vacuum Isp, Methane/LOX around 360-380 seconds. This means more total propellant mass is needed for a given change in velocity.
Coking/Sooting (Kerosene):Kerosene can leave carbon deposits, affecting reusability. Methane burns much cleaner.
Combustion Products:Produce CO2 and water (and soot for kerosene), though in small global quantities.
Which is the "Clear Winner" for SSTO?
Historically, theoretical chemical SSTO designs often leaned towards LH2/LOX due to its superior Isp, as the sheer performance needed for SSTO from Earth is immense. However, the resulting vehicles are often impractically large with very low payload fractions.
The real-world answer, particularly considering reusability and operational factors, points towards the denser carbon fuels, especially Methane/LOX (Methalox).
While Methalox has a lower Isp than Hydrolox, its much higher density leads to a smaller, lighter, and easier-to-build vehicle. When you factor in the overall system efficiency – including vehicle dry mass, ease of operations, reusability, and even potential for in-space refueling or ISRU (In-Situ Resource Utilization, like on Mars) – the density advantage of methane often outweighs hydrogen's Isp advantage for a practical, reusable rocket design. This is the reasoning behind SpaceX's choice of Methalox for Starship, a vehicle pushing the boundaries of what's possible in spaceflight.
Therefore, while LH2/LOX might be the "technical winner" for raw propellant performance on paper, the carbon fuel (methane) approach is arguably the more viable and practical path for designing achievable advanced launch systems, even if true Earth SSTO remains a monumental challenge.
Thoughts?
(th)
Last edited by tahanson43206 (Today 08:55:19)
Offline
Like button can go here
In reviewing Posts #33 and #34, it seems to me that kbd512's hypothesis has never been tested.
Gemini has reminded us that the atmosphere is interacting with the entire rocket. What some call 'drag' is a manifestation of interaction with molecules of air that are ahead of the vehicle. These must be accelerated from rest to whatever velocity is needed to get them out of the way of the advancing vehicle.
Meanwhile, Gemini has proposed that the molecules bouncing against the outside of the expansion bell may cause a downward force on the expansion bell, opposite to the upward force of the exhaust gas molecules bouncing against the inside and imparting a small upward force.
I'll bet that kbd512's hypothesis has never been tested.
Drag at the front of the vehicle is such an obvious negative influence that rocket engineers may have been focused there.
It should be possible to test kbd512's hypothesis, but my guess is the experiment would be difficult to set up and execute.
This topic is available for a NewMars member to describe an experiment that would provide data about the scenario we've been discussing.
It might be easier to set up a CFD simulation so that drag at the front of a vehicle could be separated from the effect of back pressure on an expansion bell performance.
Another detail that might be worth highlighting is the difference between the thrust produced at the exit plane of the throat and the thrust produced as molecules bounce against the inside slope of the expansion bell. The thrust produced at the exit plane will hold steady regardless of the atmosphere, due to Newton's Laws. The thrust that will change due to atmosphere would be the thrust produced at the expansion bell, to the extent that atmospheric molecules are able to enter the volume where exhaust molecules are dominant.
(th)
Last edited by tahanson43206 (Today 09:14:10)
Offline
Like button can go here