Trip to Mars on hypergolics

[fredinno]

33 minutes ago, AngelLestat said:

ok. in that case those tank should be well insulated.  But my point for big tanks is still valid. You can find it in literature.

All that insulation is pretty massive, and can easily push the higher ISP out compared to Hypergol. But we will never know for sure until we do the calculations.

[DBowman]

3 hours ago, AngelLestat said:

volume increase by cubic and surface square, so the volume/surface ratio increase a lot, each time you double the volume, the ratio increase like this: 8, 17, 33, 67, 133.    Why this is important?   because it tells you how much energy needs all that mass of fuel to rise few degrees its temperature, so if your tank volume is big, (5 to 10 times the shuttle tank) then it can last years, and hydrogen leak takes a lot of time too.

This is true only if all three dimensions can increase. The launch vehicle geometry usually limits two of the dimensions to the circular cross section of the launcher payload. So if you are assembling your craft by 'docking' lifted tanks then both surface area and volume will increase linearly with volume. You'd only get the square vs cube advantage if you could loft or assemble (and fill) in orbit huge spherical tanks.

1 hour ago, fredinno said:

All that insulation is pretty massive, and can easily push the higher ISP out compared to Hypergol. But we will never know for sure until we do the calculations.

NASA DRA 5 seemed to think cryo fuels would payoff though, even when constructing by 'docking' (the LH2 tankage was pretty huge though if I recall - maybe 30% of wet mass (for the return propellant)).

[DBowman]

7 hours ago, lobe said:

Yes I could use hydrolox but I wanted to see how large a hypergolic rocket would be to deliver the payload+hab. For delivering the lander separately, I figure a simulation mode might be able to be rigged up so the crew can practice enroute.

The constraints are part of what makes problems interesting & fun, on the other hand if the Apollo Applications program had moved ahead they would have made it as easy as possible on themselves and probably done something like what @sgt_flyer is talking about - Mars orbit/surface rendezvous of components sent as 'direct' as possible to get the best deltaV per Saturn. If you are going to assemble by docking you might as well do it in Mars orbit after using LOX+LH2 for the TMI, what tonnage could it throw to Mars? I could imagine them direct launch and aerocapturing separate Surface Hab+Descent Vehicle, Crew Descent/Ascent Vehicle, and TEI injection stages and being forced to LEO assemble CommandModule+TransitHab+TMIStages+MarsCaptureStages all hypergolic powered.

[kerbiloid]

On 15.03.2016 at 6:12 AM, AngelLestat said:

sabatier reaction is like a children play for chemist

Unless you put several human lives and a space expedition as a bet — rather than show to pupils "you see: this is mostly methane".

[AngelLestat]

On 16/3/2016 at 11:37 PM, fredinno said:

All that insulation is pretty massive, and can easily push the higher ISP out compared to Hypergol. But we will never know for sure until we do the calculations.

What you call massive?  because the best insulators are the most lightest materials in the world. But I agree that for small tanks does not worth it, but if the shuttle choose lh2 and lox for onboard fuel, is in part because it could be combined with the current engines that it had, and again.. cryo fuel is not a big issue as many want to believe, if you need to develop a sat or a probe, then of course you will not waste time and money with those issues, but anything bigger than that it would be silly to not go for cryo.

On 16/3/2016 at 0:59 AM, DBowman said:

This is true only if all three dimensions can increase. The launch vehicle geometry usually limits two of the dimensions to the circular cross section of the launcher payload. So if you are assembling your craft by 'docking' lifted tanks then both surface area and volume will increase linearly with volume. You'd only get the square vs cube advantage if you could loft or assemble (and fill) in orbit huge spherical tanks.

NASA DRA 5 seemed to think cryo fuels would payoff though, even when constructing by 'docking' (the LH2 tankage was pretty huge though if I recall - maybe 30% of wet mass (for the return propellant)).

I know what you said, but I did the math before, and even if one of those dimensions are 3 times longer than the other 2, the volume-surface relation change is negligible, only passing those values start to increase reaching its asymptote at the higher differences, so even for fuel tanks which long is 5 times the diameter would not be an issue.
And you need to prevent as you said add many tanks (for the same stage) because those benefits are lost, but even if you do, those tanks would be enough big to not represent any trouble in leak or boiling.
About Dra5, I guess it is a very good design to have as base for a mars case, one of the changes I will made would be the engines placement, I like the fact that is an horizontal aerodynamic shape with engines that point down, but engines would only work if the heat shield is release it, but you need those engines before the aerobrake  step ends to ensure you stay at certain altitude killing all your horizontal velocity before lose more altitude, staying at that layer only using aerodynamics is a very hard case in the thin mars atmosphere.
So the engines should be place it with a little tilt angle above the black heat shield (something similar to dragon v2 engines).

5 hours ago, kerbiloid said:

Unless you put several human lives and a space expedition as a bet — rather than show to pupils "you see: this is mostly methane".

Not sure if I understand you.. you mean you put humans in risk?  Again.. that is impossible, because you have sensors in the vehicle that are measuring everything, and you can even do a burn test, with hover and landing before send humans there, in case something fail, you send the same vehicle empty with the corrections needed, and the only thing you lost is a time window.
Sensors are light and very accurate, this is a common methane sensor:
http://www.bluesens.com/english/products/allgassensors/bcp-ch4-sensor.html
Then you can have: electrochemical sensors, optical sensors, chromatography and spectrometry,  mass sensors, etc.
Just with an optical sensors can tells you all the amount and mix of elements in a substance.

[kerbiloid]

2 hours ago, AngelLestat said:

Not sure if I understand you.. you mean you put humans in risk?  Again.. that is impossible, because you have sensors in the vehicle that are measuring everything

Then there is no need to control and certify the fuel for the Saturn/SLS/Falcon, so on.
Why to bother with the fuel department when it's possible to give to a security guard a sensor (like a metal detector) and let him decide:
"What's here? RP-1? Do you see that rocket? Come on, pour it there. And what's here? Petrol? No, AvGas? OK, let it be AvGas. Bring it to that hangar, those guys will pour it into the plane for astronauts."

If a MAV falls back to Mars because there was 2% of CO2 in the ISRU methane, who's guilty: a sensor or a space administration who "risk it on one turn of pitch-and-toss"?

[fredinno]

6 hours ago, AngelLestat said:

What you call massive?  because the best insulators are the most lightest materials in the world. But I agree that for small tanks does not worth it, but if the shuttle choose lh2 and lox for onboard fuel, is in part because it could be combined with the current engines that it had, and again.. cryo fuel is not a big issue as many want to believe, if you need to develop a sat or a probe, then of course you will not waste time and money with those issues, but anything bigger than that it would be silly to not go for cryo.

I know what you said, but I did the math before, and even if one of those dimensions are 3 times longer than the other 2, the volume-surface relation change is negligible, only passing those values start to increase reaching its asymptote at the higher differences, so even for fuel tanks which long is 5 times the diameter would not be an issue.
And you need to prevent as you said add many tanks (for the same stage) because those benefits are lost, but even if you do, those tanks would be enough big to not represent any trouble in leak or boiling.
About Dra5, I guess it is a very good design to have as base for a mars case, one of the changes I will made would be the engines placement, I like the fact that is an horizontal aerodynamic shape with engines that point down, but engines would only work if the heat shield is release it, but you need those engines before the aerobrake  step ends to ensure you stay at certain altitude killing all your horizontal velocity before lose more altitude, staying at that layer only using aerodynamics is a very hard case in the thin mars atmosphere.
So the engines should be place it with a little tilt angle above the black heat shield (something similar to dragon v2 engines).

Not sure if I understand you.. you mean you put humans in risk?  Again.. that is impossible, because you have sensors in the vehicle that are measuring everything, and you can even do a burn test, with hover and landing before send humans there, in case something fail, you send the same vehicle empty with the corrections needed, and the only thing you lost is a time window.
Sensors are light and very accurate, this is a common methane sensor:
http://www.bluesens.com/english/products/allgassensors/bcp-ch4-sensor.html
Then you can have: electrochemical sensors, optical sensors, chromatography and spectrometry,  mass sensors, etc.
Just with an optical sensors can tells you all the amount and mix of elements in a substance.

Look, the primary problem on a Mars mission is that going there will take at least 6 months, and so will. coming back. You're going to be at Mars for over a year, and unless you have 2 pads (FH does not, it only uses LC-39A, neither does SLS, which only uses LC-39B) you're going to be waiting for at least a month (probably more due to the size of Falcon Heavy, and larger rockets take longer to prepare and make pads ready for) for launch. Even if you DO have 2 pads, there can still be massive delays though- Constellation's Departure stage had to be rated for over 50 days in orbit despite using 2 pads to launch 2 rockets. 

Sure, it might still work for Earth departure stages, and Mars descent/ascent (if using ISRU), but unless you produce propellant at say, Phobos and fuel up there, you need to use hypergolic for Mars Departure and Mars Ascent. I know DRA5 used H2 propellants to depart Mars regardless, but that was more so it did not need to carry a second set of engines (and NTR is more efficient regardless). Generally by the time you arrive at Mars, your propellant will have boiled off to the point it becomes a significant problem, and a major addition to your mass. 

Insulation is not that heavy, but remember that EVERY GRAM COUNTS, especially when going as far as Mars.

 

And when it comes to something as essential (and without an abort option) as getting off Mars, you need to be 99.999% it will work, or you will have dead astronauts on your hands. Thus, you should incorporate a extensive program testing ISRU. You don't need a dedicated lander, you can add it onto other probes, like what is happening on Mars 2020 with Mars ISRU. Elon can also fund Mars Sample Return Missions to Mars and sell the samples and science. 

I would start with a basic Mars Orbiter>Mars Lander>Mars Rover>Mars Sample Return using Hypergol and no rover>Mars Sample Return with rover, extended payload capacity, and CH4/LOX>Mars Sample Return with ISRU and rover (+another 3 times)

This will both build up experience necessary to make a Mars Colony/Mars mission, help fund a Mars mission, and test ISRU extensively. Of course, it can be contracted to LockMart or Boeing to skip steps 1-3, but SpaceX's tradition of making everything in house goes against that.

[DBowman]

 @lobe you may not be aware of, and you might be interested in @GregroxMun's excellent The Apollo Applications Program challenge - stock recreate major Apollo milestone missions, and then the missions envisaged for an extended Apollo program. I know Stock is not what you are shooting for, and this is sci & spaceflight - but the mission profile is pretty on topic.

I had a lot of fun with it, culminating (so far...) in my Apollo 20 - Double Duna - two identical stacks of LM derived Duna lander + transit & surface habs + three man CM each launched with a Saturn V on steroids (modeled on Saturn MLV). The two stacks provided redundancy, a large six man crew, and company for the poor CM pilots - the two CM+transits habs docked for a single 'party time' return transit. The image below shows a cluster of habs, one of landers, and the docked CM+habs in orbit. 

[Rune]

On 14/3/2016 at 4:39 AM, lobe said:

I looked into this a little bit more, and came up with some still quite amazing numbers. To start off the mission profile is 670 days, now with 6 crew as per fredinno's recommendation and has a Skylab sized habitation module. Food, water, and air for this crew comes to 24.5 tons. The applicable delta-v budgets are 3.6 km/s for Trans-Mars Injection, 2.7 km/s for Martian orbital insertion, and 3.8 km/s for descent/ascent, as per the delta-v map I posted earlier in this thread. I will assume that the specific impulse for all engines are based on some variation of the Service Propulsion System engine, which is 314 seconds.

The lander is still two stage. The ascent stage I based off of the dry mass of the Apollo CSM, at about 12 tons, adding another 1.2 tons for supplies of a 3 crew for 30 days on the ground, coming to 13.2 tons. Fuelled mass is 45.4 tons. The descent stage takes that number and adds 10 tons, when this is fuelled there is a total lander weight of 190.4 tons. Yes, AngelLestat, this does and my previous calculations include the the total 7.6 km/s it takes to land and depart Mars, because aerobraking is for wimps. Also, this is tech pretty much from Apollo that I am using, and the end date for the tech I can use is up to 1980, the launch doesn't need to be in that time frame. By 1980 we already landed 2 probes on Mars (Viking 1 and 2, 1976), so I assume that the development program they used could be scaled up. 

Skylab weighed about 68 tons, adding 24.5 to that gives us 92.5 tons. This is added to the lander, and will assume that the tank Skylab is attached to is 25 tons. This brings the unfuelled mass to 307.9 tons, fuelled 740.4 tons. Since we are now moving something with the fuelled mass over the Proton rocket (693.8 tons) the tank and the amount of engines are going to be pretty massive, so I assumed another 50 tons for this departure stage. This now makes the rocket 2,546 tons fuelled. To put this in perspective, this is about 3 and 2/3 Proton rockets, or about 85% of a Saturn V. It would take 18(actually 18.1, but you aren't going to launch 0.1 of a rocket) Saturn V launches to complete this single spacecraft.

While refreshingly devoid of flights of fancy, I think that you went a bit too simplistic in that analysis. There are a great many ways of reducing that dV budget and minimizing IMLEO (Initial Mass in Low Earth Orbit, the most important figure for launch costs).

For example, you don't have to capture into a low martian orbit! A highly eccentric barely-bound trajectory will do just fine, you can lower the apoapsis over a long time with aerobraking (as opposed to aerocapture). Many unshielded martian probes have used it, and it doesn't even need a huge knowledge of the martian atmospheric conditions, just trial and error measuring the orbit after each pass. Takes a few months form the surface mission, but it gives time to do a proper orbital survey, and you are baselining a short surface stay anyway (I guess to eliminate the need for a surface hab).

You could even not lower your orbit at all, and shift all that the dV to the lander. You will get some aerodynamic help braking to the surface, whether you plan it or not, so there's that. But more importantly, you are carrying less payload, so you should save some total mass by doing that and keeping the massive transit hab really close to a Earth, energetically speaking, the whole time. It means a heavier lander, but it also means a much smaller Earth return stage for the Hab, and a smaller capture burn with the whole fueled lander counting as a payload, so the influence should be massive in IMLEO.

And those are just two tricks I pulled form Von Braun's book for a saner mission route, with a much lower IMLEO. I encourage you to peruse the mars mission proposals of the sixties and seventies, like Mars '69 by Von Braun or the old EMPIRE studies, lots of different Mars architectures to get ideas from. They are a gold mine, and it makes you realize that these things had already been studied to death a long time ago.

 

Rune. If we had stopped studying the idea, and started launching fuel tankers filled with hydrazine and N2O4, we would already be there, I concur. That was the point of the thread, right?

Edited March 20, 2016 by Rune

[lobe]

Hello again, back with the rest of the architecture. We were left with ship of 2546 tons for the mission, mind you that only delivered the lander and crew with no fuel for a return trip. This part now includes the fuel for the Mars hab and the return stages.

The hab I assumed would mass 22 and 2/3 tons, basically a Skylab chopped to a third. 10 tons for a landing apparatus and the mass fuelled is 112.3 tons. Add on the delivery stages (injection and insertion) and the mass comes to 1144 tons. About 8 Saturn V launches

The return stage is designed so the transit hab is swapped from the delivery stage. This saves some weight but not much as we will see later. I figure by the time its engines are lit the supplies are down to 11.6 tons, meaning about 80 tons to LEO needs to happen. The total return transfer stage mass would then be 674.4 tons. This now needs to be delivered to Mars. The mass of the entire delivery in LEO is 5826.4 tons, almost 6 million kilograms. 42 Saturn-V launches.

So, in total if everything was to be launched by Saturn-V to LEO and assembled there, it would require 68 Saturn V launches delivering almost 10,000 tons. Just in launches it would cost 12.6 billion dollars. Though I mean, this method does get around a lot of issues, like nuclear power in space, cryogenic storage, you can put the project down for a bit, proven technology, safe-ish assumptions of Mars's atmosphere (tenuous at best), time can be taken for the project with out worrying about stuff boiling or decaying, and it provides a decent thrust for its electical consumption.

[DerekL1963]

6 minutes ago, lobe said:

So, in total if everything was to be launched by Saturn-V to LEO and assembled there, it would require 68 Saturn V launches delivering almost 10,000 tons. Just in launches it would cost 12.6 billion dollars. Though I mean, this method does get around a lot of issues, like nuclear power in space, cryogenic storage, you can put the project down for a bit, proven technology, safe-ish assumptions of Mars's atmosphere (tenuous at best), time can be taken for the project with out worrying about stuff boiling or decaying, and it provides a decent thrust for its electical consumption.

It gets around a lot of issues...  but it significantly increases programmatic risk due to the large number of launches and rendezvouses required.  The length of time the first pieces will be on orbit prior to departure will be an issue as well.

[fredinno]

2 hours ago, lobe said:

Hello again, back with the rest of the architecture. We were left with ship of 2546 tons for the mission, mind you that only delivered the lander and crew with no fuel for a return trip. This part now includes the fuel for the Mars hab and the return stages.

The hab I assumed would mass 22 and 2/3 tons, basically a Skylab chopped to a third. 10 tons for a landing apparatus and the mass fuelled is 112.3 tons. Add on the delivery stages (injection and insertion) and the mass comes to 1144 tons. About 8 Saturn V launches

The return stage is designed so the transit hab is swapped from the delivery stage. This saves some weight but not much as we will see later. I figure by the time its engines are lit the supplies are down to 11.6 tons, meaning about 80 tons to LEO needs to happen. The total return transfer stage mass would then be 674.4 tons. This now needs to be delivered to Mars. The mass of the entire delivery in LEO is 5826.4 tons, almost 6 million kilograms. 42 Saturn-V launches.

So, in total if everything was to be launched by Saturn-V to LEO and assembled there, it would require 68 Saturn V launches delivering almost 10,000 tons. Just in launches it would cost 12.6 billion dollars. Though I mean, this method does get around a lot of issues, like nuclear power in space, cryogenic storage, you can put the project down for a bit, proven technology, safe-ish assumptions of Mars's atmosphere (tenuous at best), time can be taken for the project with out worrying about stuff boiling or decaying, and it provides a decent thrust for its electical consumption.

Less launches please. I think you should use h2 o2 cryogen for at least the mars infjection burn, and Rp-1 for all of the other burns (which is far less cryogenic)

Maybe even Rp-1 for the lander. However, you also need better insulation.

You should also use aerobraking for the in-space HAB to shave mass.

[DBowman]

3 hours ago, lobe said:

68 Saturn V launches

wow. 1969 was the highest cadence they managed; 4 launches - so about 8 launches per Mars window, you have 8-9 windows worth at that rate. I assume you'd be sending the return stages at least one window before the crew, I cannot imagine you'd want to stretch things out over more than two windows - so you'd have to factor quadrupling the 'launch manufacture and support pipeline' - whatever that would cost (and have stuff loiter in LEO for up to two years before it even sets out).

What kind of automation are you assuming? I imagine uploading some simple orient and burn schedule? so you could get stuff into Mars orbit ahead of time but not dock or land? Docking and landing to be handled by the pilots via remote control when they reach Mars orbit?

If you wanted to reduce mass and launches then doing the Mars injection with LH2 & LOX on any chunks that could be assembled in Mars orbit would be a technique that does not 'break immersion'.

[Rune]

21 hours ago, lobe said:

Hello again, back with the rest of the architecture. We were left with ship of 2546 tons for the mission, mind you that only delivered the lander and crew with no fuel for a return trip. This part now includes the fuel for the Mars hab and the return stages.

The hab I assumed would mass 22 and 2/3 tons, basically a Skylab chopped to a third. 10 tons for a landing apparatus and the mass fuelled is 112.3 tons. Add on the delivery stages (injection and insertion) and the mass comes to 1144 tons. About 8 Saturn V launches

The return stage is designed so the transit hab is swapped from the delivery stage. This saves some weight but not much as we will see later. I figure by the time its engines are lit the supplies are down to 11.6 tons, meaning about 80 tons to LEO needs to happen. The total return transfer stage mass would then be 674.4 tons. This now needs to be delivered to Mars. The mass of the entire delivery in LEO is 5826.4 tons, almost 6 million kilograms. 42 Saturn-V launches.

So, in total if everything was to be launched by Saturn-V to LEO and assembled there, it would require 68 Saturn V launches delivering almost 10,000 tons. Just in launches it would cost 12.6 billion dollars. Though I mean, this method does get around a lot of issues, like nuclear power in space, cryogenic storage, you can put the project down for a bit, proven technology, safe-ish assumptions of Mars's atmosphere (tenuous at best), time can be taken for the project with out worrying about stuff boiling or decaying, and it provides a decent thrust for its electical consumption.

Wait, what? You are using mass ratio 10 in one stage or something like that? Break that up! To deliver 112.3mT to Mars, you should need, at most, 6.6km/s going by the crappy map on the wiki (using no aerobraking, if not, you can make do with 5.2). That would be broken up in a sane word into two stages, to take advantage of the ~3km/s exhaust velocity, looking for the ideal mass ratio of around e. That would work out to be a mass ratio of ~2.86 on each stage, or an effective 8.18 for the whole craft. 920mT IMLEO, not 1144. Then again, now that I think about it, I think I fell into the classic mistake of ignoring tankage and engine mass, that probably accounts for those ~200mT of difference. Whoops .

Anyhow, if you use perfectly harmless aerobraking, you can get by with a single stage of MR~5.25, or ~590mT IMLEO. Now that is quite the difference.

Want to get another great weight-saving measure? Depart form EML1/2, or a barely-bound distant retrograde orbit. Boosters can put around half their payload to LEO into such trajectories, but doing that would drop the dV budget to a pitiful 2 km/s, way less than half. Magic! And considering the rocket equation... Mass ratio for the departure/insertion stages ends up being around 1.9, meaning ~100mT to Mars can be launched for ~190mT in EML1, which works out to a ~380mT IMLEO equivalent or ~5 SLS (~40mT to EML1/2, IIRC).

Why the magic? Well, such big boosters usually have very efficient upper stages, and you exploit the exponential nature of the rocket equation by breaking up the journey in legs that are lower than the effective exhaust velocity of your propulsion, or close to it. Plus, aerobraking. Seriously, solar-powered satellites have done it with the panels unfurled.

Oh, and coming back from Mars, you can do the same thing to capture around earth and come back down, meaning the return trip is more like 3.5km/s form low Mars orbit, or a mass ratio of almost exactly 3. Call it 120mT for a useful payload of under 40mT like Transhab, and a trip to Martian orbit and back would be 230mt in EML1/2, or ~460mT IMLEO equivalent, or 6 SLS. Just to send the crew there and back, mind you, I'd imagine the lander and surface equipment can make their own trips there, and these are enough numbers for a single post. In a lucky coincidence, BTW, both stages would have a similar mass, ~80mT for the return stage and ~90 for the departure... I'm sure these numbers can be refined so in the end you do both things with the same stage, I've played fast and loose with the roundings.

So, you know, not that daunting a task after all, as I said earlier. Note all captures are still done propulsively, no prior knowledge of Martian or Earth atmospheric conditions is required, or crazy entry precision, or in fact any special thermal protection or streamlining.

 

Rune. In fact, it sounds pretty darn straightforward.

Edited March 23, 2016 by Rune

[DBowman]

1 hour ago, Rune said:

Want to get another great weight-saving measure? Depart form EML1/2, or a barely-bound distant retrograde orbit. Boosters can put around half their payload to LEO into such trajectories

Is there anything special about EML1/2 or high retro? Is what you suggest more or less the same thing as using cryo upper stage to get the highest energy orbit possible? i.e. could be a GTO or something higher - whatever the booster is capable of? and then follow up with the storable stage? I see that a high energy Earth orbit would let you cheaply fix inclination and do earth orbital assembly of a few chunks.

[Rune]

11 hours ago, DBowman said:

Is there anything special about EML1/2 or high retro? Is what you suggest more or less the same thing as using cryo upper stage to get the highest energy orbit possible? i.e. could be a GTO or something higher - whatever the booster is capable of? and then follow up with the storable stage? I see that a high energy Earth orbit would let you cheaply fix inclination and do earth orbital assembly of a few chunks.

Not quite the same as using a higher orbit. Or rather, there is something special about EML1/2. They are the high points of Earth's Hill sphere, and they are close to weak boundaries, meaning from there to anywhere, it's always "down" (that is obviously not true, but it is the best simple way of putting it I know of). GEO and other high orbits do have energy in them (sometimes more energy, actually), but part of it is "locked", shall we say, or in other words you can't repurpose it at will. For example, to depart GEO you would first have to lower you Ap to get back low where the Oberth effect really works, while from the lagrange points, a minute burn would drop you as close to atmosphere as you want. A lunar Distant Retrograde Orbit is similar from what I hear, a weakly bound orbit that is somehow stable over long periods with minimal stationkeeping. From there, you can use the Moon to drop you pretty much anywhere from what I understand, but the truth is I don't really have a full understanding of it, I just trust NASA's word on that.

Of course all this stuff requires fancy orbital mechanics with chaotic behaviours due to the influence of more than one source of gravity. Only solvable numerically, but if you put a supercomputer to the task, they come up with some amazing stuff. Look up the wiki entry on the "Interplanetary Transport Network", for more info and references on the subject.

 

Rune. It's why the guys at JPL are known as dark wizards in the aerospace world, they are the guys that find these things in RL. Math geeks, the lot of them.

[fredinno]

19 hours ago, Rune said:

 

Want to get another great weight-saving measure? Depart form EML1/2, or a barely-bound distant retrograde orbit. Boosters can put around half their payload to LEO into such trajectories, but doing that would drop the dV budget to a pitiful 2 km/s, way less than half. Magic! And considering the rocket equation... Mass ratio for the departure/insertion stages ends up being around 1.9, meaning ~100mT to Mars can be launched for ~190mT in EML1, which works out to a ~380mT IMLEO equivalent or ~5 SLS (~40mT to EML1/2, IIRC).

Wouldn't using cryogens for the Earth Departure Burn also save mass? And using RP-1 for Mars insertion and departure is also likely going to save a lot of mass.

And sending a spacecraft to EML1/2 is a huge mistake- if you have the TWR, LEO is better- you can use the oberth effect. Build it up at 400 km, then bring the perpasis down to 100 km, then do the burn to Mars to make 100% use of the Oberth Effect.

[Rune]

46 minutes ago, fredinno said:

Wouldn't using cryogens for the Earth Departure Burn also save mass? And using RP-1 for Mars insertion and departure is also likely going to save a lot of mass.

And sending a spacecraft to EML1/2 is a huge mistake- if you have the TWR, LEO is better- you can use the oberth effect. Build it up at 400 km, then bring the perpasis down to 100 km, then do the burn to Mars to make 100% use of the Oberth Effect.

Sure, but in a ~2km/s burn, it won't be much. And needing to put ~90mT of propellant in EML1/2 just for the departure burn of the lightest element, you are either looking at a superbooster, of lots of losses, or a specialized fuel depot. Therefore, these things would be assembled throughout the whole 2 years between launch windows (even SLS wouldn't lift this in one go, or two) and, if I was czar of space, they would use an international standard of docking and fuel transfer, so everyone could contribute with fuel, or compete for the contracts, no matter with which rocket.

As to using RP1 or other things for the rest of the mission, obviating the obvious title of the thread, I will just say that I would rather trust my return to a hypergolic engine, especially if it has been sitting fueled in space for 2+ years.

And of course you use Oberth's effect when departing from EML1/2. A minimal burn (on the order of tens of meters per second) puts you into an atmosphere-grazing trajectory. Probably already hyperbolic, if the Moon is handy, but here's where supercomputers come in, to find the best way to get to a given ejection, so since it's a chaotic problem I really can't tell you, it's an optimization problem. And then you add the rest of the energy to it at Pe, when Oberth is present at full. You know how this thing works, right? How it's cheaper to use the ITN than a normal transfer to the moon, or anywhere else?

 

Rune. Yup, more Isp is less weight. But is that what we are trying here?

[fredinno]

19 minutes ago, Rune said:

Sure, but in a ~2km/s burn, it won't be much. And needing to put ~90mT of propellant in EML1/2 just for the departure burn of the lightest element, you are either looking at a superbooster, of lots of losses, or a specialized fuel depot. Therefore, these things would be assembled throughout the whole 2 years between launch windows (even SLS wouldn't lift this in one go, or two) and, if I was czar of space, they would use an international standard of docking and fuel transfer, so everyone could contribute with fuel, or compete for the contracts, no matter with which rocket.

As to using RP1 or other things for the rest of the mission, obviating the obvious title of the thread, I will just say that I would rather trust my return to a hypergolic engine, especially if it has been sitting fueled in space for 2+ years.

And of course you use Oberth's effect when departing from EML1/2. A minimal burn (on the order of tens of meters per second) puts you into an atmosphere-grazing trajectory. Probably already hyperbolic, if the Moon is handy, but here's where supercomputers come in, to find the best way to get to a given ejection, so since it's a chaotic problem I really can't tell you, it's an optimization problem. And then you add the rest of the energy to it at Pe, when Oberth is present at full. You know how this thing works, right? How it's cheaper to use the ITN than a normal transfer to the moon, or anywhere else?

 

Rune. Yup, more Isp is less weight. But is that what we are trying here?

Yes, but the only problem is that hypergols have too low ISP to make a mars mission viable. At least the Mars orbit injection should be cryogen, as you can go in on a free-return trajectory and abort back to Earth. Mars Ascent and departure make more sense to use hpergol though.

Guess what? The increased ISP from cryogen H2/O2 actually DECREASES fuel requirements for the booster stage- ignoring boil-off, which depends on your launch rate (thus how much propellant will stay in LEO) and insulation. LC-39B is stated to be able to launch 3 SLS (similar in size to Saturn V), so 3 Saturn V or Saturn V derivatives (not sure about Super Saturn V (Saturn V with 4x Shuttle SRBs) but you should theoretically be able to launch 6 or so Saturn Vs a year from LC-39A and B- and use LC-34 for Saturn IB. Using 6 Super-Saturns a year, you can launch 223.5T/rocket x 6= 1341 T to LEO. That should be enough to launch a Mars mission using cryogenics as the first two burns (departure and Mars orbit insertion.)

http://www.astronautix.com/lvs/satv25sb.htm

This map shows you can do a 0.7 m/s burn back down to LEO transfer. That's still a lot of delta V. Also, doing this in EML points reduces the size of each odule and makes the mission FAR more complex as a result.

It's really not worth it, even if you use the moon's gravity to reduce the needed delta v (which is shown here to reduce delta V to ~0.6km/s. It's very doubtful increased stage efficiency from the upper stages will offset the 0.6 km/s burn and increased complexity.

 

 

[Rune]

But that's a huge number of SLS flights, and I was trying to trim this! Compare what you have described with this, and note I am restricting myself to the fuels that the OP states:

All big components (Lander, surface Hab, transit vehicle) are sized to around 40mT. That means anything form a Transhab, to a >400mT wet stage, but launched empty. Likewise, any lander can be launched empty. I think that is enough. Myself, I would launch them into LEO empty, fuel them, move them to L2 and refuel them, but hey, I believe in using the smallest possible hammer for a given job. If you have to use SLS...

Every big component makes it to Mars on its own, if necessary on different launch windows. That means a mission with lander, surface hab, and a crew flight, can take six years to launch. Honestly, that doesn't sound so bad. Launch cadence for SLS works out as that of the real world, i.e: one launch a year. If the US can afford to fly two any given year (doubtful), I would reckon they would use the second for something else.

All fuel needs are met by whoever offers the best price per kg to your awaiting empty stages. International cooperation and competition come into play here perfectly. The number of launches increases dramatically, so much so that you can actually do a statistical treatment of failures, and count on losing a few unmanned vehicles. That way you are even safer, since the whole program risk is lower (no single launch can doom your mission if it fails... or at least the smallest amount of launches possible). The crewed launch lost among the numbers involved here, and afterthought almost. Mind you, this are going to be a lot of launches. Using EELV equivalents (~20mT to LEO), you can get maybe a 10mT vehicle going to EML1/2, so in real life maybe ~5 or perhaps 7mT of propellant in the best case. Say five and go pessimistic, that would be about ~20 launches to fuel every ~100mT transfer stage (heavy rounding going on here, beware), so 2-6 billion in launch costs for fuel (100-300M$ per rocket, which is about the most accurate figure Ihave ever seen, and the reason I don't really care about precision when working these things out). Hefty price tag, yes, but let's remember at this point that at around four SLS's, you hit similar launch costs... And with such a launch market, I'm sure we could find ways to lower the costs of such a repeatable mission, couldn't we?

All in all, to me it sounds like a reasonable plan, certainly with one of the lowest technical risk and complexity of the many I have seen. Yes, it's not the lightest one out there, and it is not so by a good stretch. After all, other mission plans can use the same tricks, and many more. Heck, I would be the first to advocate the use of methane propulsion for Mars missions, given the huge ISRU potential. But, as this thread shows, IMHO, hypergols are perfectly viable to do this... and in fact, they are in many aspects easier, if all you care about are flags and footprints.

As to the chart, all I can guess is that Nyrath doesn't have access to the kind of data JPL has when calculating trajectories. After all, that chart shows clearly that Cassini did some black magic to circularize around Saturn and then visit each and every moon in there with a single puny hypergolic stage, doesn't it?  Then again, with good reason, since Nyrath is showing there general information about typical transfer orbits, with Hohmanns and such, intended as a resource for SF writers (as is the rest of his page), not to plan real life missions.

 

Rune. Highly tuned RL trajectories take into account things like the precise calendar date, for example...

Edited March 23, 2016 by Rune

[fredinno]

26 minutes ago, Rune said:

But that's a huge number of SLS flights, and I was trying to trim this! Compare what you have described with this, and note I am restricting myself to the fuels that the OP states:

All big components (Lander, surface Hab, transit vehicle) are sized to around 40mT. That means anything form a Transhab, to a >400mT wet stage, but launched empty. Likewise, any lander can be launched empty. I think that is enough. Myself, I would launch them into LEO empty, fuel them, move them to L2 and refuel them, but hey, I believe in using the smallest possible hammer for a given job. If you have to use SLS...

Every big component makes it to Mars on its own, if necessary on different launch windows. That means a mission with lander, surface hab, and a crew flight, can take six years to launch. Honestly, that doesn't sound so bad. Launch cadence for SLS works out as that of the real world, i.e: one launch a year. If the US can afford to fly two any given year (doubtful), I would reckon they would use the second for something else.

All fuel needs are met by whoever offers the best price per kg to your awaiting empty stages. International cooperation and competition come into play here perfectly. The number of launches increases dramatically, so much so that you can actually do a statistical treatment of failures, and count on losing a few unmanned vehicles. That way you are even safer, since the whole program risk is lower (no single launch can doom your mission if it fails... or at least the smallest amount of launches possible). The crewed launch lost among the numbers involved here, and afterthought almost. Mind you, this are going to be a lot of launches. Using EELV equivalents (~20mT to LEO), you can get maybe a 10mT vehicle going to EML1/2, so in real life maybe ~5 or perhaps 7mT of propellant in the best case. Say five and go pessimistic, that would be about ~20 launches to fuel every ~100mT transfer stage (heavy rounding going on here, beware), so 2-6 billion in launch costs for fuel (100-300M$ per rocket, which is about the most accurate figure Ihave ever seen, and the reason I don't really care about precision when working these things out). Hefty price tag, yes, but let's remember at this point that at around four SLS's, you hit similar launch costs... And with such a launch market, I'm sure we could find ways to lower the costs of such a repeatable mission, couldn't we?

All in all, to me it sounds like a reasonable plan, certainly with one of the lowest technical risk and complexity of the many I have seen. Yes, it's not the lightest one out there, and it is not so by a good stretch. After all, other mission plans can use the same tricks, and many more. Heck, I would be the first to advocate the use of methane propulsion for Mars missions, given the huge ISRU potential. But, as this thread shows, IMHO, hypergols are perfectly viable to do this... and in fact, they are in many aspects easier, if all you care about are flags and footprints.

As to the chart, all I can guess is that Nyrath doesn't have access to the kind of data JPL has when calculating trajectories. After all, that chart shows clearly that Cassini did some black magic to circularize around Saturn and then visit each and every moon in there with a single puny hypergolic stage, doesn't it?  Then again, with good reason, since Nyrath is showing there general information about typical transfer orbits, with Hohmanns and such, intended as a resource for SF writers (as is the rest of his page), not to plan real life missions.

 

Rune. Highly tuned RL trajectories take into account things like the precise calendar date, for example...

Talking about SLS is stupid, since the OP was talking about a 70s mars mission.

 And how is using Hypergol for earth ejection supposed to save fuel? I just stated the maximum launch cadence for LC-39, not the number of launches that would actually be used!

 Yes, GENERAL, but manned missions will use those numbers and less gravity assists, as those take longer.

Using the smallest hammer is a bad thing if it increases complexity. When it comes to fuel/propulsion (the most of the mission mass), you want the largest launch vehicle that you have available, as it means less refueling flights, and each refueling flight increases complexity. Designing it so everyone can contribute also increases complexity, and in any case, is pretty much impossible in the 70s. Increased complexity means higher cost.

 

Edited March 23, 2016 by fredinno

[lobe]

22 hours ago, Rune said:

To deliver 112.3mT to Mars, you should need, at most, 6.6km/s going by the crappy map on the wiki (using no aerobraking, if not, you can make do with 5.2)

Where is this wiki? As for the mass ratios, I was pulling numbers from imagination and what little experience I have with aluminum, so a stage holding a thousand tons of fuel would mass about 50 tons.

[Rune]

10 hours ago, lobe said:

Where is this wiki? As for the mass ratios, I was pulling numbers from imagination and what little experience I have with aluminum, so a stage holding a thousand tons of fuel would mass about 50 tons.

@fredinno posted a better map here, actually, but you could also google "dV map" or "dV budget", and pick the first wiki link.

10 hours ago, fredinno said:

Talking about SLS is stupid, since the OP was talking about a 70s mars mission.

SLS is pretty much like a superbooster from those suggested for the seventies, and I did get into the thread when it had already drifted there, but yeah, you are right. In any case, as I said, a plan relying on fuel transfer is kind of launcher-agnostic, as long gas you can lift the empty stages and payloads.

10 hours ago, fredinno said:

And how is using Hypergol for earth ejection supposed to save fuel? I just stated the maximum launch cadence for LC-39, not the number of launches that would actually be used!

 Yes, GENERAL, but manned missions will use those numbers and less gravity assists, as those take longer.

I don't quite get what you are attacking here. Using hypergols for Earth ejection will of course not save fuel. However, launching empty stages with not enough mass ratio to make it to mars, but instead raising them to L2 where they do (when refueled), will definitely guarantee you the smallest required launcher for the job. It will mean that you have lower design requirements across the whole thing, and a higher fraction of your budget will be launch costs. But it doesn't mean that your overall budget has to be higher!

10 hours ago, fredinno said:

Yes, GENERAL, but manned missions will use those numbers and less gravity assists, as those take longer.

Seriously, learn to copy/paste quotes, or hit the quote button and edit accordingly. Following you is weird and difficult, I don't know what "general" you are talking about here. If you are saying that a departure from L2 is longer, yeah, it is. By about a couple of days, in a multi-year mission. So... yeah. I get the feeling you don't play much KSP

The trick of aerobraking at the end does add a month or two to total mission time, yes. But nothing is stopping the ground form launching a capsule to rendezvous on the high orbit to pick the astronauts, and then bring the Hab down unmanned at leisure.

10 hours ago, fredinno said:

Using the smallest hammer is a bad thing if it increases complexity. When it comes to fuel/propulsion (the most of the mission mass), you want the largest launch vehicle that you have available, as it means less refueling flights, and each refueling flight increases complexity. Designing it so everyone can contribute also increases complexity, and in any case, is pretty much impossible in the 70s. Increased complexity means higher cost.

I'm hearing the spirit of Michael Griffin in those comments. Nope, a higher number of mission flights does not necessarily increase complexity or mission risk, it just means you have a higher probability of losing one flight of the many. That may sound like I have gone bonkers, but on every plan where you count your launches in dozens, you count on losing at least one vehicle, without jeopardizing your overall mission. In fact, it can enhance mission assurance, if you know you can lose any one flight and still accomplish your objective, plus crewed flights are lost in a sea of unmanned launches, enhancing reliability through statistics alone. And even if it did increase cost, it would do so much less than designing a bunch of unique vehicles each with a different propulsion system.

How many docking accidents have there been, in the history of ever, anyhow? I can only remember the one off the top of my head, and it didn't result in LOC, or loss of the ship. Did we have to abandon ISS when not one but three vehicles in a row had launch failures last year?

And of course, mass production of common elements will very much not increase costs, if anything it will lower them. On the other hand, a bigger booster can very well end up being more expensive per kg, as SLS or Saturn V prove.

 

Rune. If weight equaled cost, Delta IV would be the cheapest EELV-class launcher out there.

Edited March 24, 2016 by Rune

[fredinno]

 

8 hours ago, Rune said:

18 hours ago, fredinno said:

 

Seriously, learn to copy/paste quotes, or hit the quote button and edit accordingly. Following you is weird and difficult, I don't know what "general" you are talking about here. If you are saying that a departure from L2 is longer, yeah, it is. By about a couple of days, in a multi-year mission. So... yeah. I get the feeling you don't play much KSP

The trick of aerobraking at the end does add a month or two to total mission time, yes. But nothing is stopping the ground form launching a capsule to rendezvous on the high orbit to pick the astronauts, and then bring the Hab down unmanned at leisure.

Yes, but I DID account for lunar gravity assists. From EM-1 to LEO transfer, it takes 0.6 km/s (with a lunar gravity assist- 0.7 km/s without). No amount of aerobraking will help you if you can't get down to the atmosphere in the first place.

You need to burn 0.6 km/s to be able to use the oberth effect from a staging in EM-1. At that point, you destroy any gains you might gain from greater launcher efficiency.

 

8 hours ago, Rune said:

How many docking accidents have there been, in the history of ever, anyhow? I can only remember the one off the top of my head, and it didn't result in LOC, or loss of the ship. Did we have to abandon ISS when not one but three vehicles in a row had launch failures last year?

It's not about docking, it's about the complexity of the modules. If you use only small launchers, you're going to use smaller tanks, which are also less mass-efficient, and connecting them all together is more complex than using only a few pieces (ie Mir vs Skylab).

[Rune]

6 minutes ago, fredinno said:

Yes, but I DID account for lunar gravity assists. From EM-1 to LEO transfer, it takes 0.6 km/s (with a lunar gravity assist- 0.7 km/s without). No amount of aerobraking will help you if you can't get down to the atmosphere in the first place.

You need to burn 0.6 km/s to be able to use the oberth effect from a staging in EM-1. At that point, you destroy any gains you might gain from greater launcher efficiency.

I thought you meant at the start of the mission, but that aerobraking comment has really thrown me off. If you mean the end of mission, the ship would capture into an elliptical, Moon-crossing barely bound orbit, from where it could try and navigate to EML 1/2 (or somewhere else), or continue with aerobraking all the way down over a few months. That would be the only sensible Earth entry, unless you could snag a free capture by lunar assist, because it is the only way you bring to bear Earth's Oberth effect on the capture burn. I highly doubt it's where those 600m/s you talk about come into play... where did you get those, BTW, if it's not too much to ask?

If you meant at the start of the mission, then you are misunderstanding other things. Leaving EML1/2 is free, mathematically speaking. You will leave it in a random trajectory if you do nothing, actually, and some of those random trajectories become actually hyperbolic, and thus escape trajectories already. Now if you want to leave it quickly, along a defined trajectory, yes, you do need some dV to control things. If you were to use EML 1/2 as a staging point, you wold do just that, letting the ship fall in a slow, chaotic trajectory over a few weeks into a low perigee, high apogee orbit with a semi-major axis very similar to the Moon's (the requirement to be synchronous with it, to have the option to keep changing your orbit "for free" every 28 days). Then, at your closest approach, you would kick yourself into your chosen hyperbolic escape trajectory, and be on your way with a very small Pe quick. But I am not equipped to calculate the dV requirements of such a maneuver, and I can almost guarantee you aren't either. You would need one of the very custom programs JPL uses to handle complex integrations of the n-body problem around boundary condition, and run them in a very beefy computer for a few weeks to optimize the results.

Let me be clear, the point of going to an Earth-Moon lagrange point before heading into interplanetary space is not to lower the total amount of dV (even tough it can do just that). It seems you keep on believing that's what I'm trying to do here. Instead, the point is to break the ejection into smaller burns, that allow a better stage efficiency and a smaller ejection burn. The benefit is felt on any mission that employs limited Isp or fuel transfer, dramatically decreasing the requirements of the escape maneuver and thus decreasing the size of the ejection stage(s). It also goes very well with reusability, bringing many roundtrips into the realm of chemical propulsion, thanks to the diminished ejection requirements, and is the reason EML 1/2 always crop up in discussions about fuel depots. They are short of another halfway point, between LEO ad the rest of the solar system.

6 minutes ago, fredinno said:

t's not about docking, it's about the complexity of the modules. If you use only small launchers, you're going to use smaller tanks, which are also less mass-efficient, and connecting them all together is more complex than using only a few pieces (ie Mir vs Skylab).

That is easily solvable by launching big stages empty, as is indeed the scenario I was talking about (I did write about <40mT stages carrying >400mT of propellant, I'm sure of that!). As I also showed above, even very inefficient fuel delivery vehicles can compete on price with superboosters, mostly because they launch on rockets with much better price per kg.

And yeah, re-reading a few posts up, I was also talking about using two identical stages around ~100mT wet, one for the outbound trip (pitiful ejection dV allows MR ~1.9), and another for the return trip, with a higher dV requirement due to the capture requirement I threw on top to avoid the need for an Orion-like return capsule (worked out to MR ~3). Those are the numbers I worked out earlier for a ~40mT payload (the transit hab) of course, but it's a good starting point to evaluate the transport architecture IMO, since the transfer stages can be made MUCH bigger and the refueling scales just by adding whichever-size-you-want launches.

 

Rune. Yup, that's definitely how I would go "to Mars on hypergolics".