New Mars Forums

I AM going to answer to this, GW (and I expect I'll say mostly good things), but I'm afraid it will have to wait until I get back to Madrid, my own computer, and maybe a couple of exams.

Rune. Just logged in to let you all know... I will be back!

Actually, on earth, Dragon turns subsonic with no parachutes involved just fine. Our atmosphere is just that great for aerodynamic deceleration. They will retain a single chute, but that is a backup in case either the engines fail, or you spend the landing fuel to escape an exploding rocket. Right now, the parachute system is triple-redundant, since just one of them would be enough for a safe terminal velocity. Again, and just to be crystal clear, a nominal land landing of a crewed Dragon would not involve deploying chutes at any point. Which is great in principle at least, because you don't have to repack it. So theoretically it's just a quick refuel, and you can stick the capsule on another booster and use it the same day.

Rune. I've seen that misinterpretation too often not to point it out.

Well, Twin beam and Russel, my aerodynamics teachers told me that Drag is a function of two things (apart from density and speed): Cd and area. Yes, Cd is a function of shape. But area is area, and even though vortex formation through shaping reasons can affect Cd, the area is what the area is. So it may sound way better than it is in real life, and you might find a birdie needs it's cone perforated to move the center of pressure away from the center of mass and thus exert more momentum with the same drag, not because the drag is greater that way. Also, if you are trying to increase area with the least amount of mass, you wouldn't pick one of the densest metals like tungsten is.

As to the 60mT lander described by GW: well, with MR >1.3, it may fit, payload-wise, riding empty on a Falcon Heavy launch to be fueled later... but that 7.1 diameter heatshield doesn't fit in any payload shroud currently in existence. You could design a bigger shroud, or make it ride "naked", but both things would need an aerodynamic analysis to make sure the launcher can handle it. Or you could use the hypothetical SLS and it's 10m humongous shroud. Or you could finish development of inflatable shields. Pick one, but note the cheapest (modify/recertify an existing launcher) is also the most limited, when you think of possibly even greater future payloads.

And another point, with >13 of those 60mT (MR>1.3) in the form of dense storable propellants, packaging density for the important part (pressurized capsule where humans and their supplies ride) looks much better. Same thing for the very dense rigid heatshield. What mass fraction would ypu assign to it anyway? 10-15% of braked mass like Zubrin?

As to LMO entry... well, as I said, even a flimsy unshielded orbiter with solar panels extended can lower it's orbit as much as it wants through aerobraking passes (from a suitably elliptic orbit) with minimal fuel expenditure. So starting entry from LMO does not necessarily exclude parking the return vehicle in a loosely bound elliptical capture orbit. The only extra requirement from that goes to the ascent vehicle, and it saves more same delta-v from the Earth return vehicle than it costs the ascent vehicle.

Rune. Summer! And here I am discussing geek stuff...

Good notion, we can move the discussion to where it's slightly less OT. When I finally stop partying like I am doing lately (holidays! ^^), I might take a stab at it and try to implement the various tricks to see the effects on each (by "tricks" I mean aerobrake, basing at high mars orbit instead of low, departing form a Lagrange point/HEO...).

Rune. I do have like 5 different threads open in tabs that I would like to write a reply to...

With 5% for one-of-a-kind stuff like MSL and JWST, right? I said there would be exceptions. I'm referring more to the real space economy drivers, the ones that have to make some economical sense. The constellations in LEO and GEO, which is what 99% of space really is, these days, are launched by rockets costing 50-150 million bucks (though some estimates on the big deltas are closer to 250). The big GEO birds cost about a few hundred million, and LEO constellations, like GPS, are slightly cheaper (the best I can come up with for the cost of a single GPS sat is around 100-150 million), but they use cheaper rockets too.

Rune. Never forget, there have been a shitload of launches to date, in the thousands. Manned space and science is almost anecdotal, if not for the share of funding it takes.

And yet, historically, the price of boosters is about 30-50% of the total launch cost, no matter what launcher or payload you are talking about (although I am sure you can find some exceptions, generally speaking it's like that). Even the shuttle lies somewhere in there, a you have shown (and I had my doubts, shuttle is such a unique case in almost every aspect). There is a reason, too: you are not going to waste ~150 million bucks to launch a cheap satellite built for a couple grand, you go back to the drawing board, and cram enough tech and engineering hours in there to see the price raise into the hundreds of millions.

Cheap launchers would be a very strong incentive to launch cheap payloads. There may not be any physical law that says so, and yet I believe this trend of boosters costing slightly less than payloads will be a constant as long as expendable rockets exist, just because "humans work that way", if you will. You could also argue it's an economics thing, probably, but that's not my field.

Of course, ground support is more or less a fixed cost, and it works quite differently compared with payloads or boosters. It is mostly a function of mission complexity, I would say, at least at first approximation. And yet I look at the "mission control in a trailer" SpaceX started with (before they tried to approach the ISS and their red tape shielding), compare it to routine shuttle operations, and I see that the overall... ¿cost climate, shall we say? also has a hand in there. If you don't a have billions of bucks to build a rocket, you don't bother designing mission controls employing hundreds of people.

As to the later comments on "saddles", totally in agreement, only I would argue the saddle function changes with the market size: when you are launching tens of tons a year, a 10mT launcher makes sense. When you are launching millions of tons a year... see where I'm going? There is an alternate universe where Sea Dragon is a standard-size rocket.

And a shame what happened to the CAM. Not that that experiment couldn't have been performed much cheaper, and probably better, with an independent free flyer, like the one the Mars society wanted to try with tethers and small animals. The micro-gravity scientists complain enough of vibrations on ISS already.

Rune. Is that thing still going forward, BTW?

Exactly. It may be hell to model by hand, being a vary elastic system, but that is one of the reasons we invented fancy computers, and have plenty of geeks trained to come up with fancy realtime control systems. The part I don't get is how expended expendables shield you while coming back to earth.

I think he blundered here. Terminal velocity assumes an infinite atmosphere of uniform density. The atmosphere of Mars is anything but. Can you brake fast enough so you hit terminal speed before you hit the ground? The charts I've seen seem to indicate you really can't with a beta anywhere close to what a 11mT Dragon would have, but then again 340m/s is quite supersonic on Mars, so he seems to be obviating the issue of supersonic propulsion too.

IMO, if you double the injected payload (assemble payloads with their propulsion stages in high earth orbit from pairs of FH), then the margin stops being such a ridiculously low one, and the whole thing looks way more feasible.

Rune. Other than that, no problems with the idea

Exactly, it's the internet address for that file/page. You know, what appears in your address bar when you are looking at an image. For the images in your blog, it goes something like this:

http://1.bp.blogspot.com/-Cwfu4JCrSDA/T … rofile.bmp

... but it looks better when you put it between tags like this (check it out by quoting this post if you want to):

You can get the address of any picture on the internet, not just yours, by right-clicking on it, clicking "see image" (or whatever looks similar in your browser) and copy-pasting the address of the picture that comes up. Oh, and .bmp is a very wasteful file archive. Paint or any other image program can change them into .jpg (or save them as such from the beggining), which takes much less data for each image and looks just as good.

Yup, you either can get your beta in the vicinity of 150-200, or you can forget about chutes, that's also my take. Inflatable heatshields might take us there, but then you can get very low betas, and then you don't need chutes. If you have to take a very big ballistic coefficient... yes, then you just have to go with rockets all the way. But of course, I want to see numbers before I say some method is the best!

So int he end the lesson I take from all of this is that supersonic retropropulsion and inflatable heatshield should both be two very high-priority technology developments. Which they aren't right now.

Rune. Methinks that, at some level, someone doesn't want to run out of excuses not to go.

I was going to reply to say that, but it would be repetitive now. So +1. As to flying fast, why would you want scramjets if you have a nuclear thermal engine: run it with atmospheric air until there is either not enough or it's coming too fast (/too hot) if you wanna get fancy, then switch to onboard propellant. Or just use a VTOVL NTR, Isp is enough. As to arcjets, they run on anything but have very bulky power needs. Nowhere near T/W>1.

Rune. I doubt safe NTR's can approach chemical engines' power to weight ratios anyway.

Just one quick comment more... that is an awesome idea! Not many people look into that kind of problems, and having to wait every time you do an EVA like in ISS is horribly time-consuming, and risky.

It's cool weight works the other direction and you can balance both, then. Imagine an arbitrarily big hole sealed with a 25mT/m2 (10/0.4=25) homogeneous cover, pressurized at 1atm and... net load on the structure is zero and you can use bubblegum to build it, no matter the size.

Rune. Totally off-topic, too, but I'll get back on it on the next post... I've got a lot of stuff to reply to.

First, to vindicate GW and me a little bit, I'll say your idea is not really at odds with our comments: you turn on the engines when you can't get enough "umph" form the air alone. Having said that, I love it when someone shows me numbers to support their idea, and the result is not what I expected. Going by gut feeling, I would have said you were just paying bigger gravity loses, but apparently I would have been wrong.

As to the scenario you have painted in particular, that is close to 600 beta, so the descent trajectory is going to look different than the ones I got my hands at. I assume you got the terminal velocity from drag and density? If so, what altitude-density is that? 'Cause you have to have enough space and G's to slow down from there, and if you don't have the room to do it in vertical, that is more gravity losses ('cause a vertical descent takes the least time, and gravity losses are a function of time, but I have a feeling you know all this already). Also, consider that if the landing craft is meant to be refueled and launched (it's a classic MAV with ISRU on the surface), asking 10G's or even more out of it when it is mostly empty of fuel is not really that much. Changing the G's changes the analysis you did back there a bit, but I don't think much.

Rune. The more I think about it, the more useless the chutes look.