New Mars Forums

Void,

Absent exceptionally powerful new propulsion technologies, likely involving fusion, TSTOs like Starship are always going to have their place in the expanding menu of orbital launch vehicle options.

The trends are clear:
1. full reusability with minimal repair and refurbishment
2. streamlining of operational costs by reducing repair costs and time
3. substantial improvements to engine performance as we push towards the theoretically achievable Isp values and combustion efficiency limits for conventional chemical engines
4. exploration of non-rocket launch methods for lower value cargo such as propellants and metals
5. new materials for thermal protection

That said, no amount of materials science is going to overcome basic orbital flight physics.  A TSTO will have a higher payload mass fraction than a SSTO, period.  Thus far, no miraculous new strong but light material nor more efficient engine technology can compete with the ability to discard 2/3rds of your vehicle's inert / dry mass at booster burnout, because pretty much any tech you can apply to a SSTO can also be applied to a TSTO, to even greater effect, especially if full reusability is important.  The only area where TSTO cannot compete with SSTO is the simple fact that a SSTO nominally stays in one pretty piece, from launch to landing, and doesn't require multiple separate recovery, transport, and stacking operations.  Starship Super Heavy partially reduces that problem by returning the booster to the pad it launched from.  That's a great first step, but the upper stage stilll has to land somewhere and get stacked on top of a booster again.  Whatever savings there are in terms of propellant cost reduction, it probably cannot offset all that added operational complexity.

SpaceX wants to launch Starship several times per day.  With a HTHL SSTO, after the vehicle lands, it's towed back to the gate, inspected, gassed up, towed back to the runway, and then it's ready for its next flight.  That might feasibly be done in the span of a few hours.  Even a VTHL SSTO can be pretty straightforward to relaunch.  A VTVL TSTO requires multiple pads with supporting infrastructure, if only because the upper stage cannot readily land on top of its own booster, multiple cranes are needed for stacking, inspecting two vehicles vs one must be done (only a major time imposition because it probably needs to be done in two or more different places for a large fleet operation), etc.  All that adds up to more time, more touch labor, and more cost.

I want SpaceX to succeed, or any body else to succeed for that matter, but launching this way 1,000+ times per year (at least 3,500 per year to send 1,000 Starships to Mars every 26 months, or about 10 per day) is going to be challenging without a standing army of support personnel and sprawling infrastructure investment.  Doable?  Yes.  Very cheap and practical with limited head count?  Almost certainly not, at the scale required- too many simultaneously moving pieces.  Maybe AI and those Tesla automatons can coordinate the touch labor operations to the degree required, plus they can stay in a shelter on the pad, immediately ready to inspect the booster or upper stage the moment it's captured on the pad.  We will never run recovery and vehicle inspection operations that way with humans for obvious reasons.  Ground crew are not allowed anywhere near the vehicle during the launch and recovery.

All I know for sure about vertical launch and vertical landing is that it's very tricky to do correctly and routinely.  It's not impossible, obviously, but far from simple and easy.  Automated landing helps mightily, but what about 24/7/365 airline transport style operations?  We don't do that with rocket launches.  If a hair looks out of place, we stop the clock, reassess, and everyone involved weighs in before we proceed.  We run launch operations that way to stave off disaster.  The primary precious cargo we're transporting to Mars is lots and lots of people.  Losing even one vehicle per year is unacceptable.  Starship is clearly not an optimal solution for an airline transport style operation, even though it's acceptable for exploration missions.  Neither a SSTO nor the upper stage of a TSTO alone is the correct solution for colonization.  Refueling up to a half dozen times in LEO is impractical at the scale they want to implement.  We need some kind of fusion-based in-space propulsion system for practical large scale colonization operations.  A solid core NTR is woefully inadequate for that task.

Void,

33X 200:1 TWR Raptor-3 engines weigh 49,500kg.
33X 800:1 TWR RDEs would weigh about 12,375kg.

The propellant tanks weigh 80,000kg when made from 304L stainless, and the interstage weighs about 20,000kg.  If the propellant tanks were made from CFRP, then mass is around 20,000kg.  Those two changes can save around 97,125kg, cutting the dry mass of the Super Heavy Booster in half.  That is a massively consequential upgrade to both engine performance and propellant tank dry mass.

Between those two night-and-day performance enhancements to our chemical rocket technology, SSTOs become not merely technically feasible but reasonably practical.  Flight physics dictates that a TSTO will always soundly beat a SSTO on payload performance unless dry mass approaches zero.  We really can't do much about that, and any SSTO mass reduction will likely carry over to a TSTO design with better payload performance.  However, TSTO does not beat SSTO in overall simplicity of operation.  By its very nature, any vehicle that deliberately breaks apart into multiple pieces during its flight, with separate recovery locations for the different pieces of the vehicle, is going to be more complex, time consuming, and expensive to operate.  When a mixed propulsion SSTO can take off and land on a conventional runway like other jet aircraft, VTVL TSTOs will be relegated to specialist heavy cargo delivery roles because we won't be able to justify the operating expenses or dangers associated with operating them for routine commercial orbital passenger service.

So long as it's not too extreme, justification for the additional propellant consumption of a SSTO vs TSTO for passenger service is little different than justifying the additional fuel consumption of much faster jet aircraft capable of carrying more passengers over shorter duration flights vs slower piston engine aircraft that carry fewer passengers.  Reduced propellant consumption is always desirable, but by mass most of the propellant is oxidizer rather than fuel.  That is why a mixed propulsion scheme that significantly offsets the onboard oxidizer mass by consuming atmospheric Oxygen is the most likely path to a commercially successful SSTO that is at least GLOW-competitive with TSTOs.

I view RDEs, mixed propulsion consuming atmospheric Oxygen, advanced high strength composites, and advanced ceramic composites for engine components and thermal protection as the technological underpinnings required for the success of the National AeroSpace Plane (NASP) Program.  NASP began in earnest in the 1980s, although conceptual studies began in the early 1970s.  Here in 2025, it's starting to look like we finally have the technology to make a practical NASP vehicle, which was intended to be a SSTO with mixed propulsion.

It took about 50 years to go from Wright Flyer to Boeing 747, 50 years from Saturn V to a rapidly reusable super heavy lift launch vehicle (Starship Super Heavy), and about 50 years yet again, to go from a 1.5 stage-to-orbit semi-reusable space planes (Space Shuttles) to rapidly reusable SSTO space planes (NASP), so time-wise I would say we're on-track.

RAND Report on NASP:
The National Aerospace Plane (NASP): Development Issues for the Follow-on Vehicle - Executive Summary

Void,

Brief impulsive burns are the preferred way to operate all rocket engines to minimize gravity losses, but some engines don't generate enough thrust to do that.  The insufficient thrust problem primarily applies to the various electric propulsion schemes, but sometimes applies to low power chemical propulsion for interplanetary spacecraft and satellites as well.

Gravity loss reduction is critical when attempting to achieve orbital velocity from a launch pad.  Some user here, maybe PhotonBytes, has proposed using an air-breathing mixed propulsion scheme whereby a spaceplane repeatedly dips into the upper reaches of the atmosphere to obtain Oxygen for combustion, following a sinusoidal trajectory around the Earth until sufficient velocity is achieved that very little pure rocket propulsion (onboard oxidizer plus fuel), at greatly reduced Isp, is required to attain orbital velocity.  Gravity losses will be higher in this scheme simply due to total thrusting time required to attain orbit, but the Isp is so much higher from not having to carry most of the oxidizer onboard the vehicle that so long as said spaceplane can tolerate the extreme aerodynamic heating it will be subjected to, the end result is a significant propellant consumption reduction.  How well that works in real life remains to be seen.  I place this scheme in the "easier said than done" category.  I think it's probably attainable, but at what cost?

Gravity loss is highest (near or at 100%) when thrusting "straight up" to clear the dense lower atmosphere, as is typical of a TSTO design that uses most or all of the propellant in the booster stage to rapidly ascend through the lower atmosphere before the gravity turn and firing of the second stage, which provides the majority of the delta-V necessary to achieve orbit.  If your rocket was powerful enough and you only intended to attain escape velocity, you could theoretically thrust "straight up" or at least "straight to your intended interplanetary target", and leave Earth's gravity well that way.

I believe the Nova booster concept, which was to be equipped with 8 F-1 engines in its first stage, was designed for a "direct ascent trajectory".  For various reasons, mostly related to energy and monetary economics, direct ascent was ultimately rejected as the most practical means to achieve the lunar landings during the Apollo Program.  Direct ascent requires a much more powerful rocket.  The Saturn V, with only 5 F-1 engines in its first stage, was the "economy model" moon rocket.  We had to master orbital rendezvous and docking to use a comparatively less powerful Saturn V, but those on-orbit maneuvers were subsequently proven to work quite well, and are used today every time a spacecraft visits ISS.  Saturn V could deliver 140t to LEO.  Saturn C8 (Nova) could deliver 210t to LEO, and the SpaceX Starship Super Heavy V3 aims for about 200t with full reusability or up to 300t when fully expended.  If we were to expend the upper stage of Starship and include a Hydrogen powered third stage, then we'd have a very powerful direct ascent moon rocket.

After you're already in orbit, you can perform multiple burns, firing the engine at certain points along a highly elliptical orbit that are most favorable to increasing velocity (gravity assist) along your intended trajectory.  This typically applies to escape maneuvers, but is also used by satellites going to higher orbits to reduce propellant consumption.  GW will correct me if I'm wrong, but I believe his space tug concept also capitalizes on gravity assist thrusting to take the payload to just shy of escape velocity so the upper stage can be refueled and reused.  Any engine capable of multiple restarts makes this possible.  RL-10 upper stage engines could restart 15 and later up to 25 times.  Sometimes restart limitations are imposed due to ignition, especially when using TEA/TEB or other pyrophoric compounds to achieve ignition, and sometimes stress on various engine components or cumulative damage.  This will vary from engine to engine.  RS-25 engines use the equivalent of a spark plug, so if there is enough electrical power then the number of possible restarts are not limited by the count of onboard pyro cartridges.

To the best of my knowledge, RDEs can achieve thrust-to-weight ratios (TWRs) better than conventional chemical rocket engines, so engine thrust would not be a limiting factor.  Gravity and atmospheric drag, however, are always going to be limiting factors.  After you're in orbit, as you noted, a RL-10 thrust level engine, provided that all components can handle the stress of firing, could perform long duration burns or multiple burns.  Examples of small engines performing extended duration burns include engines powered by storable chemical propellants.  Russian interplanetary space probes have upper stages with small engines that fire for quite some time, for example.  Some engines can fire for cumulative times measured in terms of multiple hours, although most have single burn duration limits because their nozzles are made from uncooled refractory metals that can only withstand burns of a given duration.  If the engine combustion chamber and nozzle were regeneratively cooled or made from Carbon-Carbon that can readily withstand the temperatures generated by even the hottest burning propellants, so long as the turbopumps / valves / seals / injectors / combustion chamber / nozzle and any other attached machinery can withstand the forces and temperatures involved, such an engine could theoretically fire for hours on end to achieve the desired Delta-V, if need be.

At least in theory, since RDE fuel injection systems operate in a manner more similar to a piston-driven internal combustion engine with electronic fuel injectors, they don't use turbopumps to feed the propellants.  RDEs have very few components and fewer moving components than typical turbopump-fed rocket engines.  They do require electronic propellant feed control with exceptionally precise injection timing.  When sequenced correctly, it's as if the detonation of the propellant forms a sort of "standing detonation wave" that endlessly "swirls" around the combustion chamber.  After initial ignition, in steady state operation no igniter is required to maintain combustion.  Just as the first propellant injection and detonation sequence completes the second detonation sequence begins, and in so doing generates continuous thrust using a very rapidly pulsing rather than continuous fuel injection.  Heat from the combustion chamber and nozzle can also be used to convert cryogenic liquids into gasses.  You have more heated surface area to work with, as compared to a conventional rocket engine design with a bell nozzle, so the engine can generate more thrust than a typical pump-fed expander cycle engine like the RL-10, before a conventional expander cycle engine runs into heat transfer rate limitations from the limited surface area available to provide input power.

NASA has experimented with Hydrogen, Methane, and RP1 fueled RDEs with regenerative cooling.  NASA has also experimented with Reinforced Carbon-Carbon (RCC) nozzles that don't absolutely require regenerative cooling.  RCC is 1/4th the mass of the commonly used Nickel-Copper alloys (2g/cm^3 vs 8g/cm^3) and won't melt if left uncooled.  The first goal behind using RCC was engine mass reduction / TWR improvement for upper stage engines.  The RDE scale engine models used thus far are about half as heavy as conventional rocket engines for the same thrust output and dramatically more compact.  That means a RDE fabricated from RCC could be about 1/8th the mass of a typical RL-10 for the same thrust output, so putting 8 RDEs on the back of an upper stage would provide about 880kN of thrust for the mass of a single RL-10.  Add a few more engines and you achieve J-2X / Merlin thrust levels from something that's not much more mass than the RL-10.  At that point, even with a fairly substantial payload, your burn is less than an hour in duration.

The current / active RL-10 versions are 170kg to 300kg, so a thrust-equivalent RDE would weigh about 85kg to 150kg.  If advanced materials such as RCC are used throughout the engine, then the engine might weigh as little as 21.25kg to 37.5kg.  That's a trivial amount of weight for a RL-10 thrust-class Hydrogen burning engine, mostly because it doesn't require an enormous Hydrogen turbopump.  At some point traditional turbopumps are likely required to feed enough propellant, but the 10kN to 44kN RDEs that NASA has experimented with thus far don't require them.  It's easy to see what the real advantages are.  There's a modest Isp bump from using detonation, but engine size, mass, complexity, and fabrication time / cost matters at least as much.  As we've already noted, as long as the engine can withstand the stress of firing, multiple hours of firing time are feasible and increased total burn time can be favorably traded-off against other engineering considerations such as the force that an upper stage engine will subject the entire vehicle to as the propellant mass is expended.  Maybe you don't want an excessive amount of acceleration applied to a large interplanetary vehicle equipped with fully deployed solar and radiator arrays.  A single large pump-fed engine can only throttle to a certain point.  Multiple smaller RDEs would give you more throttling range to work with.

Air-breathing RDE variants have been simulated to function up to ~5.8km/s, which is well beyond burnout velocity for all booster stages I'm aware of.  After that, you only need to add a couple more km/s using onboard oxidizer to attain orbital velocity.  You also get a badly-needed TWR improvement with Hydrogen burning RDEs for a SSTO vehicle, although there's still not much you can do to reduce the additional Hydrogen tank mass, nor TPS mass if you intend to make the vehicle fully reusable.  It's an interesting pathway towards a practical SSTO, but one that requires a lot more work.  Mixed propulsion with air-breathing Methane burning RDEs is probably the way forward for SSTOs.  You get to offload a lot of the oxidizer mass you'd need to carry in the vehicle, an Isp increase over RP1 without resorting to absurdly large LH2 tanks, and sky-high TWRs.  Raptor 3's TWR is 200:1.  Even if the best we can do is 400:1, the mass of the engine hardware becomes trivial, but 800:1 is at least possible.  We could implement almost arbitrarily large SSTOs (300t to 500t to LEO) with payload mass fractions comparable to today's TSTOs.

tahanson43206,

A much cheaper option would be using full ceramic bearings with minimal Silicone-based lubricant.  At 6rpm, the duty cycle is 3,153,600 cycles per year.

A drastically cheaper option would be using standard Babbitt metal bearings with lubricants.  Much hay is made over this, but the battery operated tools used by the astronauts to repair the Hubble Space Telescope, such as the bit driver they used to unscrew and re-torque bolts, did in fact use bearing lubricants.  The photovoltaic arrays on the ISS do in fact use sealed metal bearings with a vacuum-compatible grease.

That means there are 3 ways to solve this centering and bearing problem:
1. Contactless permanent magnets or electromagnets.
2. Dry or minimally lubricated ceramic bearings.
3. Lubricated metal bearings.

Option 1 is scientifically interesting because it will theoretically never fail from direct contact friction.
Option 2 can work, but is also expensive at the scale required and will still require periodic replacement.
Option 3 will work for many years before replacement and is exhaustively well proven on Earth and in orbit aboard ISS.

I think composite structures are going to have less difficulty resisting the forces applied to them than light alloys.  For starters, composites have sky-high stiffness (resistance to bending / deformation) and tensile strength, as compared to 2xxx series Aluminum alloys and 3xx series stainless steels.  Coefficients of Thermal Expansion are much lower for Carbon-based composites than they are for virtually any suitable metal alloys, so temperature deltas across areas of the structure exposed to full Sun vs shadow are far less.  Composites don't corrode from atomic Oxygen attack in LEO, so that's not much of a consideration, either.

We can make fairly thick composite structures from very thin layers of Carbon Fiber tow / roving / tape material, such as IM7 fiber, because the composite's density falls somewhere between 1.5g/cm^3 and 2g/cm^3, with 1.7g/cm^3 to 1.8g/cm^3 being the most probable value for an IM7 fiber composite consisting of 60% fiber fill and 40% resin matrix.  Magnesium is 1.737g/cm^3.  The strongest Magnesium alloy I could find would yield around 50.75ksi.  IM7/5320-1 fiber/resin yields between 245ksi (compression) and 270ksi (flexure).  Maraging steel yields between 250ksi and 350ksi, dependent upon the steel's heat treatment, and it's density is around 8g/cm^3, or 4.7X less mass for the bulk composite for similar strength to C300 maraging steel.  The big and small of this matter is that we have an exhaustively proven aerospace composite combination available at affordable prices.  When used to fabricate cryogenic propellant tanks, the cost of the finished product was 25% to 30% below the cost of a comparable Al2195.  Most of the cost of composites is the energy consumed to create the materials.  The tape winding fabrication process takes place over a matter of weeks (typically 1-3 weeks, maybe 5 weeks for complex parts), and is no more time consuming than the months of milling and welding Al2195 "potato chips" into finished structures.  The curing process then takes place over the next month or two, but involves very little touch labor and no movement of the part or structure.  Additional painting or coating processes may be required, perhaps a fluoropolymer sealant to inhibit absorption of moisture, for example.

Hexcel HexTow IM7 Carbon Fiber Material Properties

Fiber Density: 1.78g/cm^3

Cytec CYCOM 5320-1 Epoxy Resin System Material Properties

CYCOM 5320-1 Cured Resin Density: 1.31g/cm^3

Wichita State University - National Center for Advanced Materials Performance (NCAMP) - Composites Lab Materials Processing Control Documentation

NCAMP 5320-1 Resin System Processing Documentation:
Cytec, a Syensqo Company 5320-1

NCAMP NMS 818 Carbon Fiber Specifications Documentation Repository for Qualified Materials:
NMS818 Carbon Fiber Specifications

NCAMP NMS 818/8 Fiber Specification for Hexcel HexTow IM7 Unidirectional Fiber Tape:
Document No.: NMS 818/8 Revision N/C, February 8, 2013 - for NMS 128/2 IM7 8552 Unitape

60% Fiber Fill: (1.068g/cm^3) + 40% Cycom 5320-1 Resin (0.524g/cm^3) = 1.592g/cm^3

I prefer to use 1.7g/cm^3 as my bulk density "fudge factor figure" in lieu of the manufacturer's exact stated material properties.  If I use 1.7g/cm^3, then I know I'm not going to end up with a primary structure component with greater mass than what I intended.  If the component ends up modestly lighter, that's fantastic.  If not, then I'm still covered.  Since the mechanical properties of this IM7 fiber used in conjunction with the 5320-1 resin system are so well defined, relative to many other composite materials, we can have high confidence in the mechanical properties of the components fabricated from this combination.  One of the major complaints engineers have about working with composite materials is not knowing exactly what they're getting.  I think that problem is mostly solved by using this combination.  We're not getting as much strength or as much stiffness as we possibly can, but we are getting a good mix of properties suitable for this application.

One underappreciated aspect of selecting these fabrication materials is just how thick our parts can be at reasonable mass.  If our parts range in thickness between 10mm and 25mm, we're going to have exceptional resistance to deformation relative to much thinner cross-section Aluminum alloys, with tensile strength very comparable to maraging steel.  The only mass-comparable structural metal alloys similar in density to composites are Magnesium alloys, none of which begin to approach the tensile strength of maraging steel.  The relationship between moment of inertia and resistance to bending / deflection under load means doubling the thickness of a part doesn't increase stiffness by only 2X, it's 8X as stiff.  If we were limited by the yield strengths and masses of structural metal alloys, this would quickly increase mass to impractical levels.  Thankfully, we're not.

This is from NPR, back in 2006:
Figures on Chinese Engineers Fail to Add Up - by Adam Davidson, June 12th, 2006

The China Academy -  How Did China’s Top Major Become the Worst?

The implication from those articles is that some part of China's leadership has now recognizes that quality eventually has greater importance to their benefit to society than sheer quantity.  Beyond that, in the realm of true engineering there is no amount of fakery that will pass muster if your skills are put to the test.  Calling a guy who can do basic electrical work an "engineer" is a bit of a stretch.  By that metric, I would be considered an electrical engineer, except that I'm not and refuse to even pretend to be.

Dr Sabine Hossenfelder recently lamented that within the realm of physics, across the entire world, a lot of "scientific" publications pertain to "BS research".  There's nothing particularly useful about knowing the precise speed of light to 1,000,000 digits of precision vs 1,000.  If that is the sort of "scientific challenge" that your physicists are taking on these days, then whatever their true talents may or may not be, their time is being squandered on "trivial pursuits".  The results from such an experiment may be of academic interest, but for all practical purposes nothing fundamentally new will be learned.  We're still no closer to capitalizing on our new-found knowledge of precisely how fast the speed of light actually is.

We discovered this new miracle material named "Graphene", but we're not meaningfully closer to being able to use Graphene to construct a space elevator than when Geim and Novoselov first isolated that material 21 years ago at Manchester University.  Do we now possess vastly more economical methods for making Graphene today vs 21 years ago?  Absolutely.  What we presently know is still nowhere near "good enough" for a functional space elevator using Graphene support cables.  Quite a number of programs and research fellows from around the world, including China, have all contributed to humanity's Graphene knowledge base since then, but it remains a highly specialized material with few applications where the extreme cost of making it can be justified.

Is the next discovery, regardless of where it happens, poised to overturn the present paradigm and enable this miraculous new material to shine where previous efforts failed to produce the material in the quality and quantity required for Graphene macro structures?

Possibly, but I wouldn't hold my breath waiting for that to happen.

The point worth remembering would be that a slew of inter-disciplinary innovations are required to construct something like a space elevator.  You can throw as many people as you like at solving that problem, but it doesn't mean any particular scientific / engineering innovation problem will get solved any faster.  There is also an opportunity cost associated with pouring all your brightest young minds into Graphene innovation.  Other potential solutions get ignored out of necessity, to focus on a single challenging problem.

What if someone figured out how to economically generate enough thrust to accelerate upwards, without using reaction mass or megastructures, while all of your engineers were tinkering with Graphene space elevators?

A large part of that awesome research effort immediately became a waste of money, intellectual capability, and irreplaceable time.  The Chinese communist government is prone to repeatedly making such mistakes, because that's a "feature" of central planning.  When most of what you do is dictated by political power structure and you incessantly prioritize maintenance of that power structure over true innovation or useful engineering projects with coherent national priorities, as opposed to "just doing something to give your engineers something to do", you go from "Civil Engineering Is Everything" to, we're no longer attracting students because we have nothing for that many Civil Engineers to do that is minimally useful to society.  That's why they built multiple "fake cities" where nobody could actually live.

Moon to Mars (M2M) Habitation Considerations A Snap Shot As of January 2022

A high level slide deck to go along with the document from Post #12:
Transit Habitat Concept and Mars Analog in Cislunar Orbit

Imagine one of these ships had 2 sets of habitation modules rotating in opposite directions:

NASA has already designed the simplified form of it with 1 pair of rotating habitation modules rotating in 1 direction.  Imagine that we had a second pair rotating in the opposite direction.

Please see Page 8 of this document (look at the actual page numbers on the page, not what the PDF says the page number is):
Conventional and Bimodal Nuclear Thermal Rocket (NTR) Artificial Gravity Mars Transfer Vehicle Concepts

AN ARTIFICIAL GRAVITY CONCEPT FOR THE MARS TRANSIT VEHICLE

The "Lunar Gateway" has been designed as a purpose-built Mars Transit Vehicle that also happens to be a good miniature space station:
Gateway Lunar Habitat Modules as the Basis for a Modular Mars Transit Habitat

I've heard lots of people claim that the Lunar Gateway is a waste of time and money, yet NASA is quite literally designing and building it to support a crew of 4 people for up to 500 days in deep space without resupply, possibly more, use nearly closed-loop life support that doesn't hog too much electrical power or generate too much heat, have a solar storm shelter, incorporate artificial gravity, and use a combination of chemical and electric propulsion so as not to blow out the mass budget.  It's an Exploration class ITV by any other name, but wait...  They actually call it that, because that's what it's really designed to do.

People like Dr Zubrin make a claim that goes like, "Hey, you know, with your capsule and my lunar lander and their giant rocket, we could do a lunar mission."  However, these people are talking to each other, via systems engineers, and there are systems integration engineers who take all the bits and pieces and transform them into coherent working vehicle designs with specific missions in mind.

This long duration interplanetary transport vehicle concept dates back more than 20 years.  That's how long NASA has wanted to build it, but didn't have the budget, or the giant rocket, or various other important bits like closed-loop life support that is durable and reliable.  We're almost there now.

GW,

Your complaints about Elon Musk, President Trump, and DOGE are duly noted, but this topic is about ITV technologies for exploration class Mars missions.  There is a political topic for politics.

I think space tugs and propellant depots are a splendid idea, but again, this topic is about ITV technologies.  I think we're already in agreement that all real exploration missions thus far have begun in LEO.  At this point, I would be pleased with any crewed mission to another planet, including a fly-by mission, irrespective of what orbit it starts in or what propulsion tech it uses.  A fly-by mission would at least demonstrate that our ITV is mission capable and that the crew is no worse for wear after the mission.

Propulsion is obviously a very important part of making all interplanetary missions possible, and I understand your affinity for those systems, which I share, but do you have any specific ideas on how we can create a more practical long-duration in-space vehicle to carry exploration crews to Mars?

I've already mentioned some of my ideas:

1. Counter-rotation sidesteps the precession and loss-of-control issues with artificial gravity.

2. Composites fabricated by robotic tape winding machines can create durable yet light vehicle hulls with excellent volume-to-mass ratios.

3. SEP, storable chemicals, and CMGs can provide adequate in-space propulsion and maneuvering after achieving escape velocity using already available cryogenic chemical propulsion upper stages.

Maybe we'll have ZBO tech perfected in another 5 to 10 years so that long duration on-orbit storage of cryogens becomes a more practical proposition.  ZBO tech development began over 25 years ago.  I think the required insulation tech is ready to apply, but not much else is.  My understanding is that insulation alone, and perhaps some modest cryocooler hardware to re-liquefy small amounts of boil off as compared to LH2, is sufficient for LOX and LCH4.  Unfortunately, NASA is dead set on using LH2.  If the daily loss rates weren't so high I understand the potential benefits from using LH2.  Thus far, long-term LH2 storage has proven to be a very tough nut to crack.

4. Rather than trying to store absolutely everything because we're afraid of equipment failures, capitalize on the excellent progress being made on low-temperature / low-power / high reliability CO2 scrubbers and water filtration systems.  I see these technologies as "de-risking" the mission, rather than adding risk.  Highly efficient air and water recycling is a mission enabler.  We already spent the money to develop this tech to a very high degree of reliability and readiness, so we may as well use it.  The mass reduction from water recycling alone is key.  IWPs provide a very consistent 98% water recycling rate.  CO2 recycling would provide another significant mass reduction.

There are near room temperature processes for CO2 splitting that leaves pure elemental Carbon dust floating on top of Gallium metal as the Gallium strips the O2 from CO2.  An electrostatic attraction process could remove the Carbon dust for storage, although even mechanical separation methods appear to work.  Heating the metal to 500C to 700C in a vacuum is sufficient to release 2 of the 3 Oxygen atoms.  Hydrogen gas will strip the remaining Oxygen, producing water and Gallium metal.  By the end of the mission we have 546kg of elemental Carbon from the respiration of 4 crew members.  Human feces is 40% to 55% Carbon, so an additional 200kg to 275kg of Carbon could be recovered.  This recovered Carbon is one of the important precursor materials for synthetic fibers, plastics, and rubbers.

5. Use the most modern avionics suite and sensors we have, because the reduction in power consumption and heat rejection are incredible.  Power-over-fiber combined with optical chips is a radiation hardened by design computing environment.  Some researchers here have already built a fully optical computer that performs all the necessary logic functions of legacy Silicon-based CPUs, to include RAM, but it does this with a 100GHz clock cycle.  Some simpler designs are operating at up to 800GHz.  That's 200 to 300 times faster than Silicon-based CPUs.  Even a very modestly capable CPU with a clock cycle that fast can still perform the same functions as more sophisticated Silicon-based chips.  The result is orders of magnitude power reduction.

Photonic crystal displays can render painting-on-paper quality colors.  In the tech world, there's a recently established "fetish" for near-infinite color range and quality that looks as if the image was inked onto a piece of paper, even though it's been rendered onto a clear piece of glass.

Optical computers have been tested using specialized lasers and particle accelerators to withstand radiation doses in the range of 10s of megarads- a near-instant lethal dose for the crew.  There's even been some talk of putting optical computers inside the cores of fission reactors, or at least ones operated at steam generating temperatures, as integrated monitoring and control systems.  That makes them highly resistant to both damage and upset from normal radiation fluence ranges found in space.

I think we should pursue a modified minimum risk mission that uses existing air and water recycling technologies that we already paid through the nose to develop:
Minimum Risk Deep Space Habitat and Life Support

I think the technological risk associated with using regenerative life support is greatly over-played.  ISS wouldn't have been habitable for 20+ years if the tech didn't work.

The next generation life support tech, specifically the Amine Swingbed CO2 Scrubber and Ionomer Waste Water Processor have been quietly doing an even better job aboard ISS, using a lot less power and suffering fewer system casualties, for over 2 years now.  They're much closer to closed-loop systems, drastically lighter / more compact, and they've had very few teething problems.  If the ionic liquid CO2 scrubber experimentation pans out, the total amount of power both systems consume will be under 1kWe for a crew of 4.  The total power requirement should be 2.5kWe or less.  That means photovoltaic array deployment mechanisms are not required.  Waste heat management radiators can be fixed / hull-integrated designs as well.

600W - Counter-Rotation Motors (CRMs)

400W - Skylab Control Moment Gyros (for 8 CMGs at normal operating rpm)
Source:
[url=https://ntrs.nasa.gov/api/citations/19790007076/downloads/19790007076.pdf]NASA TM-78212 - 25 kW POWER MODULE UPDATED
BASELINE SYSTEM - December 1978 - Page Labeled 39 in lower right-hand corner[/url]

800W - Carbon Dioxide Removal by Ionic Liquid Sorbent (CDRILS) System
Source:
Scale-up of the Carbon Dioxide Removal by Ionic Liquid Sorbent (CDRILS) System
Novel Liquid Sorbent CO2 Removal System for Microgravity Applications

60W - Ionomer Waste Water Processor (Direct, Single Cycle, or Dual Cycle)
Development of Ionomer-membrane Water Processor (IWP) technology for water recovery from urine
Demonstration of a Full Scale Integrated Membrane Aerated Bioreactor- Ionomer-Membrane Water Purification System for Recycling Early Planetary Base Wastewater

140W - Avionics / flight computers, electronics, communications, sensors (redundant mobile chips using power-over-fiber sensors)

2kWe is a very modest amount of power for 4 crew members.  That can be supplied by hull-integrated thin film arrays.  Additional power can be stored in Lithium-ion batteries.

Starship Flight 9 delivered a smooth ride to orbit aboard a flight proven booster.  All indications were that the performance of the Raptor engines was nominal.  The Starship itself did not "blow up", so at this point the issues with the main propulsion system seems to be reasonably well sorted out.

SpaceX attempted an extreme AoA "broadsiding" maneuver with the booster, which failed.  Their simulations told them this was the most likely outcome, but they needed to confirm this through real world testing to obtain flight data in agreement with their simulations.

Starship itself still suffers from severe propellant leaks somewhere in its plumbing, quite likely due to excessive thermal load.  Both the Super Heavy Booster and Starship intentionally vented some propellant after the Raptor burn was completed, earlier in their respective flights, which did not appear to cause a problem.  However, Starship suffered another propellant leak later in the mission.

Starship needs an effective RCS system to prevent loss-of-control from propellant venting.  GOX / GCH4 would be pretty good candidate propellants since there's clearly an over-abundance of that stuff in the propellant tanks causing leaks.

There appears to be frozen condensation forming inside Starship's cargo bay, likely from the cryogenic propellant tanks below, or a propellant leak, which may have caused the payload bay door malfunction that precluded deployment of StarLink satellite simulators.

Edit:
I think Starship needs Spray-On Foam Insulation (SOFI) on the parts of the vehicle not covered by TUFI tiles, even though that will cost more and add more weight.