sevenperforce

Assuming the stack is dropped in proper LEO, it needs 3.63 km/s to get all the way to LOP-G. My calculations give a naked FHeUS 44.8 tonnes of residuals in LEO, but FHe is quoted as being able to deliver up to 63.8 tonnes of payload to LEO, so my number may be low. Even with my numbers, however, a naked FHeUS can take 18.9 tonnes all the way to LOP-G. If a naked FHeUS has closer to 55 tonnes of residuals, on the other hand, it would be able to deliver about 25 tonnes to LOP-G. 10-20% margin on remaining propellant is way, way too high. You only need 3-5%, really. Tops. Oh but if you do that then there's no reason to use SLS in the first place, which means Boeing doesn't get as much money.

Very little can be apprised of the Boeing proposal from the images and presser alone, so I am hoping the full proposal leaks some time soon. A few things can be gleaned, though. It's possible that the apparent single engine on the EUS is just a trick of perspective and that the other three nozzles are simply obscured (and that the artist forgot to apply their firing layer). This is definitely the EUS, however. I assume these are the methalox engines already in dev by Intuitive Machines. Thus we see eight on the upper stage and likely twelve on the lower stage. Also, there is no central engine on either stage. The square print on the underside of the ascent stage in the landed image and the apparent descent ladder on the lower stage shows that there is an airlock in the lower compartment, reducing mass on the ascent vehicle and greatly simplifying egress. I had proposed this several times in the past. Additionally, the absence of engines underneath means cargo can be delivered in that way as well. If we go by very rough estimates and say that the EUS can throw 37 tonnes to TLI, and we assume methalox isp on the order of 370 seconds, and you need a touch over 3 km/s to get from TLI to Gateway to LLO to the lunar surface, then that landing stage will burn 21 tonnes of props getting to that point. Guesstimating lower stage dry mass is completely speculative, especially given that we have the lower hab module and any potential amount of cargo, but if we wildly guess that total landed mass on the descent stage is 30% of propellant mass (The Apollo descent stage dry mass was 25% the mass of the propellant it carried), then that puts it at 6.3 tonnes down, leaving an ascent stage at 9.7 tonnes. With the same 370-second isp (and that number could be off-base) and 2.6 km/s needed to get back to LOP-G, that gives us an ascent stage dry mass of 4.7 tonnes, which is quite respectable.

Does anyone have an actual copy of the Boeing proposal or has it not been leaked? I can FOIA it.

Well, not zero missions. It would have one mission: launch maximal money on escape velocity into minimal pockets. Playing around with possible mission configurations for Artemis has me pretty well convinced that distributed launch is hugely advantageous for almost anything BLEO. Going beyond LEO is significantly different than getting to LEO. Use an LEO-optimized LV like Falcon Heavy or Vulcan to get to LEO, and use a high-energy upper stage to go beyond.

I always forget how honking big those fairings really are.

You'd have to make the upper stage fatter to accommodate more propellant (methalox is not as dense as kerolox) and the larger engine bell. A sea level Raptor is only slightly smaller than a vacuum Merlin. And then it would make sense to make the lower stage fatter as well, so you have common tooling. And then you have New Glenn.

I'm thinking two tanks for the transfer burn and two tanks for the descent. If the dV requirements on the sortie vehicle are lowered, total throw is reduced, and so the Falcon upper stage has more propellant reserves. There's no hydrogen, of course, and while there's LOX boil-off it's not too bad. The higher specific impulse of kerolox makes it better to use kerolox until the dry mass of the MVac becomes problematic. You also reduce total throw by sending the reusable capsule ahead separately. You might be able to do this to help with an elliptical starting orbit before launching the TLI stage, but then you run into node phase issues, so it's better to just reserve performance and recover the side boosters. Reuse is a noble goal, but there's little reason to reuse tankage in cislunar space unless you have a way of refilling locally. This particular concept only reuses the crew capsule and RCS thrusters; a new descent/ascent vehicle is launched with each sortie. It mates to the reusable crew capsule. Part of that mating process includes a small tank that presses the crew capsule's RCS and carries enough propellant to get from LLO to LOP-G, and that's something that can be verified to work before leaving LOP-G.

I was surprised to see that they popped the chutes before dropping the service module. Also interesting that they launch without the skirt. I knew the skirt was there primarily for aerodynamic stability w/r/t the Centaur, but I had assumed it would come off with the vehicle on abort to aid in aerodynamic stability. CoM must be awfully close to CoP (compared to Dragon 2, where CoM is far ahead of CoP).

Everything else aside.... Is the mechanism for doing all of this going to be cheaper and lighter than simply lofting a rocket into the stratosphere with a bunch of helium balloons? Because if not, we'd already be using helium balloons for launch assist if such a thing was feasible.

"It's actually better that the parachute didn't deploy because then it would have, uh, well......." I mean, in a sense they're right. A parachute failure (improper reefing, collapse, separation, etc.) is more likely to speak to a fundamental design problem with the parachutes themselves, while a deployment failure is more likely to be a component fault. We'd rather find the component fault now than later. All good reasons why having a functional abort system can only reduce LOCV risk, not eliminate it.

Loved it! Nice job with figuring a way to dispose of the hab drop tank. I had struggled with this. Your solution permits a nice closed-loop sortie without necessarily needing the Canadarm, though having one is obviously a good idea. Did you consider reducing the number of drop tanks and using the Falcon upper stage to perform the powered flyby and LOP-G insertion? Landing with the drop tanks still attached makes for a rather surprisingly higher landed mass. The drop tanks will have 1-2% residual props and their added weight means the landing legs and structure needs to be heavier. Additionally, that extra weight during the landing means more propellant utilization for the ascent stage, which means a bigger ascent stage tank, which increases mass all the way up. Hovering is the most wasteful part of the whole event so you want absolute minimum mass for the hover. Using lateral velocity disposal seems ideal if it can be managed.

Just watched it. Cloud of toxic smoke looks lovely. I forgot the Bantam engines (originally designed for LOX+alcohol) had been retrofit for hypergols. 2 out of 3 parachutes is not great.

sea level versions producing thrust and has a specific impulse

You cannot exactly simulate NRHO but you can model it by placing your station in a circular orbit at 12 million km, slightly ahead of the Mun, with an inclination of about 12 degrees. You'll librate up and down in front of the Mun but never enter its SOI. Seen on Twitter -- new plan, each astronaut swallows 7 kg of moon rocks.....

The actual math is rather extraordinarily complicated, but a back-of-the-envelope approach suggests that even with any one control surface locked at max/min, the other three have sufficient authority for sustainable roll and pitch control, and yaw can be handled with thrusters. The only edge case might be an aft flap (or Plasma Deflector Shields as I prefer to call them) locked in the full-forward position...that's tricky. In such a situation (e.g., the port aft PDS locked in full-out position), the SS might need to roll the entire vehicle to starboard to effectively "flap" that fixed surface back. For aerodynamic purposes, having both port flaps full-out and both starboard flaps full-back is not very different from having all flaps feathered halfway back.

Having four means that if one becomes locked, the other three can compensate.

I was looking at RocketLab's website, and I am curious as to whether this is an error.... That's got to be a typo. The syntax of the sentence suggests that it's intended to be a different number. If the mass flow of the two engines is the same, which only makes sense, then the second stage should produce a max thrust of 26.5 kN.

Wait, did I miss the legs actually going on? I knew they installed the leg retraction/deployment mounts but I didn't realize the legs were actually installed.

If it's lunar ice, do we call it the LICE age?

It might be possible to create a very broad estimate of New Glenn performance based on information in the NG user's guide. For example, we know that the NG upper stage provides 1,060 kN of thrust in comparison to the 490 kN of the BE-3 on New Shepard. Naively, one might assume this 8.2% boost in thrust is the result of a better expansion ratio and solve the ideal gas expansion equation, using an estimated heat capacity for probable hydrolox mixture ratios, to guess at chamber pressure and do the math from there to get isp. But that assumes mass flow is the same for both engines, and ignores the fact that the BE-3 uses combustion tap-off while the BE-3U is an open expander cycle. We can, however, estimate the volume of the second stage tanks from imagery in the NG User's Guide, then divide by stage burn times to get mass flow. Thrust over mass flow gives specific impulse. Pixel counting from the New Glenn User's Guide suggests an LH2 tank approximating a cylinder 7 meters wide and 12.55 meters long. It suggests a LOX tank comprising an 0.72 m wide band with spherical caps (radius 3.96 m, height 2.51 m) on both ends. The volume of a spherical cap is pi*h^2(3R-h)/6 or in this case 30.32 cubic meters, so the total volume of the LOX tank is 30.87 * 2 + 27.71 = 89.45 m^3. Total volume of the LH2 tank is 483.15 m^3. This suggests 34.21 tonnes of LH2 and 102.1 tonnes of LOX, representing an O/F mixture ratio of 2.98:1, which doesn't seem it could possibly be correct. Maybe an open expander cycle dumps excess hydrogen into the exhaust like crazy? I don't know. But taking our numbers at face value, we end up with 136.31 tonnes of propellant. The NG User's Guide states that for nominal 250-km perigee GTO missions, the first stage burns for 618 seconds to provide LEO insertion followed by a 99-second burn for GTO injection to a 35,786 km apogee. However, it states that for a notional LEO mission, it burns for just 600 seconds with no restart. It states, "initial launches are planned to carry conservative flight performance reserves for enhanced service reliability and vehicle recoverability." So these numbers do not represent full stage performance. It's obvious that something is wrong when I continue through the rest of the math. At the 717-second burn of a GTO mission, I would be looking at a mass flow rate of 190.11 kg/s, which would suggest a wild specific impulse of 568 seconds. If I take the 102.1 estimated tonnes of LOX and just plug in ordinary numbers for LH2 (let's suppose a 5.5:1 O/F mixture ratio), I get 18.56 tonnes of LH2, for a total propellant load of 120.66 tonnes on the second stage. Burning all that in 717 seconds suggests 168.3 kg/s, which gives a specific impulse of 642 seconds, which is even worse. If I reverse it and take the 34.21 estimated tonnes of LH2 with a mixture ratio of 5.5, I get 188.155 tonnes of LOX (don't know where they'll put it, mind you, but that's another question) and a total of 222.37 tonnes of propellant. This equation suggests 310 kg/s, which comes to a specific impulse of 349 seconds, which is obviously erring in the opposite direction. It is more likely that the 717-second burn uses a lower throttle setting to reduce gee-loading on the lower-mass payload. If we take the 600-second burn of a higher-mass LEO mission as a full-throttle burn from start to finish, then we get a mass flow rate of 227.2 kg/s, which suggests a specific impulse of 475.6 seconds. This is much closer to reality (though obviously inflated, probably by my ignorance of a lower-throttle terminal burn period).

Obligatory:

Upon review, that looks like four AJ-10s on the ascent stage and 6-7 on the descent stage. If you can refuel the ascent stage, great, but you need to master microgravity propellant transfer. And pressurant transfer, if you're using pressure-fed engines, which is challenging at best. I maintain that using thrusty engines for the ascent stage is a waste because a quarter of your dV requirement is not thrust-dependent.

I went ahead and threw together some references numbers for what kind of residuals can be expected if someone wanted to use a "naked" HLV upper stage to perform TLI or any number of BLEO ops. To calculate, I've taken published or estimated GTO performance figures for various vehicle configurations and done the math in reverse to calculate staging velocity, then "removed" the payload and calculated forward to get to LEO. As such, these estimates contain some degree of conservatism; the absence of a payload during primary ascent will improve performance marginally. Buoyed by this, I've also ignored the necessary addition of a docking ring: this is likely a safe assumption given that the upper stage will need no payload adapter or decoupling mechanism. By the time Artemis is realized, Atlas V will likely no longer be flying, but I included it for reference anyway. Assumptions were 2270 m/s from LEO to GTO, 3200 m/s from LEO to TLI, 430 m/s from TLI to NRHO. I estimated the propellant load of Centaur V by comparing the volume of Centaur V to the volume of the Atlas V Centaur; I estimated the dry mass of Centaur V by subtracting the mass of the RL-10 from the dry mass of the Atlas V Centaur, increasing mass in proportion to fractional surface area, and adding two RL-10s back in.

Exactly. From the link: So, let's just state up front: this drive won't work. The problem is that, even though the author does a very nice simulation, he has left out the fields that do the accelerating. When we accelerate ions using a magnetic or electric field, the ions push back on the field. There is an equal and opposite force exerted on the electrodes and coils that produce the fields, and those just happen to be in the spaceship, too. In the first step, where we accelerate the mass to a high relativistic speed, we also accelerate the cylinder in the opposite direction. Now, in special relativity, we don't conserve energy and momentum separately. Instead, they are conserved together. If you only consider momentum (and not energy), then you will find net forces everywhere due to inertial mass changes—things get heavier as they approach the speed of light. This is exactly what the author has found. If you consider energy and momentum simultaneously, those forces will suddenly disappear. This is where the increase in inertial mass comes from in the first place: energy is sucked out of the field and turned into mass. When the particles are slowed, that mass is given up as photons in the field, which slow the cylinder as they are absorbed. What is the net force? Zero, 0N of force.

Why? Gravity drag on descent is fairly low if you come in at a sharp angle. After a lot of iteration, I'm fairly convinced that the ideal configuration is going to have the LAM launched to LEO with a tank for the crew taxi and a single drop tank to handle descent only. It loiters in LEO while SLS sends Orion to LOP-G, then a Centaur and docking ring is launched as soon as possible thereafter to push the LAM to TLI and perform the rendezvous. The vehicle stacks at LOP-G and Centaur performs the transfer to LLO and initiation of descent; the single engine on the LAM uses its drop tank props to complete all but the last 300 m/s of descent with a better T/W ratio. The primary reason to go with a pressure-fed engine in the first place is safety first so that you never need engine-out capability. If you are going for multiple engines for engine-out, then you should go with pump-fed engines and balloon tanks, which saves way more dry mass and cranks up specific impulse.