Landing on Mars

GW Johnson · 2013-01-14 16:01:05

Russel:

That was an interesting NASA paper link in your post 349 just above. Thanks. One can certainly see the KE ratio-dependent intensity of the heating environment. Escape speed on Mars is just about 5 km/s. MSL was entering faster than that, at a bit over 100 kg/sq.m. My calcs were for interface at 3.5 km/sec, and showed "realistic" stagnation-point heating rates around 5-15 W/sq.cm at 100-400 kg/sq.m. Theirs are in the 20-100 W/sq.cm range. Doesn't surprise me at all, since the entry KE they had was around 4 times higher.

BTW, PICA is just a fancy name for a reformulation of our old friend carbon-phenolic. I had lots of experiences using carbon-, silica-, glass-, and paper-phenolics as ablative insulators in solid rocket motors, at chamber temperatures near 5000-6000 F (2760-3315 C). That environment has enormous scrubbing action, especially on the supersonic bell, due to all the solids in the stream, as well as the enormous near-chamber recovery temperatures. It's substantially more erosive than the reentry environment.

Steam sweat cooling was one of 3 proposed entry protection schemes they were going to experiment with on the old X-20 Dyna-Soar program. Too bad it was cancelled just as a couple of first articles were coming off the assembly line at Boeing. If you can use some sort of piping or passage arrangement, you can take advantage of both melting ice and boiling water in the process of making your steam, before you bleed it through the porous skin. No one has since done it, but it's still a good idea to try.

L/D is whatever it is throughout any entry trajectory. What modifies the trajectory shape is not L/D, it is L/W. That peaks the same place as D/W, which is where my crude model shows the pulse of deceleration (which is D/W).

If you want my 1956-vintage entry spreadsheet "model", I could send you a copy of the Excel spreadsheet file, or provide a list of the formulas that go in the cells.

GW

Russel · 2013-01-15 04:47:32

Because all of these quantities are inter-related its difficult to express them in clear English. Yes, its L/W directly, but at a fixed L/D its getting enough D that matters. Actually its worse because L/D degrades with mach. But yeah we're saying much the same thing.

On reflection, steam might be a useful coolant. Based on the sorts of figures here we're talking a very wide ballpark of 20-200MJ/sqm. 200MJ would turn about 70Kg of ice into steam. Less for hot steam.

As for spreadsheets, the best way to publish is probably google drive. If not a basic list of formulas would be fine.

Russel · 2013-01-21 10:48:35

GW,

I went back through your blog entries and pulled out the relevant formulas and had a play with them on a calculator

First, the density model - the simplified best fit one.

p(z) = p(0) * exp(-z/Hp)

p(z) is density at altitude z where z is in Km.
Using p(0) = 0.0302 Kg/m^3 and Hp = 8.757Km

That formula appears to be consistent with the data you've tabled and also makes good physical sense to me.

Next formula appears black magic at first, because the input is altitude. However, its the following

V(z) = Vatm * exp(-C * exp(-z/Hρ))

Where C = (ρ(0) * Hρ * 1000)/(2*β*sinθ)

V(z) is velocity at altitude z in Km
p(0) is the same constant as above
H(p) is the same constant as above
β is the ballistic coefficient
θ is the entry angle

The next thing I tried was running some real numbers through this. I used β = 1200 and θ = 1.63 and the numbers I got were entirely consistent with your tables.

Nevertheless I'm more comfortable working directly with Newton's Laws so I went looking for a way to relate drag force to velocity and density and came up with this.

Drag (Newtons) = Cd * A * V^2 * p * 0.5

Which in the case of β = 1200 reduces to

Drag (Newtons) = V^2 * p * 25 25 is 0.5 * 50, and 50 is 60,000/1200 or the effective area of your lander.

So then I was able to cross check.

From the formula relating velocity to altitude, its possible to do the calculation with a very small change in altitude - I used 1m. From there its straight forward to figure out the force needed.

For instance V(30) = 3056.8314 m/s and V(29.999) = 3056.7872 m/s .

And from the direct drag formula I calculated the drag at the same altitude using the same model for density.

You'd be happy to know they agree very well. For comparison I get 0.39 gees over this small interval.

Of course there's an embedded assumption in the original formula - that is the angle to the local horizontal doesn't change. I wanted to test that and see how good or bad an assumption it is.

So, now I have another model (yet to be coded) that goes direct from the atmospheric model, to drag force, to changes in velocity. Its a simple simulator engine this time stepping through discrete quantities of time. In it I should be able to correctly model the curvature of the planet and the force of gravity.

I'll let you know how it goes

GW Johnson · 2013-01-21 14:45:05

Russel:

Glad to see my crude model is holding up reasonably well. I just posted a user's guide for the spreadsheet model over at "exrocketman", which should be more than sufficient for you to recreate my spreadsheet. I also tried out the Earth entry workbook on Apollo returning from the moon, and posted that over at "exrocketman", too. It looks pretty good, actually. The posted articles also cite references, including the NASA paper from which I extracted the the atmosphere data as well as the ancient entry model. http://exrocketman.blogspot.com

GW

Russel · 2013-01-21 20:18:44

I hope that including gravity won't be a showstopper, but we'll see..

Russel · 2013-01-22 09:42:41

Hi,

Well, I built a little simulator with Java which is basically 2 dimensional with a coordinate system centered on the center of the planet in question.

What it does is step through increments of time, taking into account at each step drag force, acceleration due to gravity and finally lift force.

Simple version of what I see.

Firstly on Mars, if you drop something from 135Km altittude straight down it will hit the surface at 991m/s (after 274 seconds)

Add in drag and it will hit the surface at 894m/s (275 seconds)

Just for fun I started an object off with a horizontal velocity of 3500m/s and sure enough, it orbits. Add drag and it hits the surface about 134 degrees around the planet.

Ok, for a 60 tonne craft, with a Beta of 400 velocity 3469 m/s and down angle 1.63 degrees...

Initial y velocity is -98.675m/s (negative is initially down in the cartesian plane)
Initial x velocity is 3467.596m/s (to the right, so we're going clockwise around the planet)

I've also told it to calculate the integrated range and local angle to horizontal.

Without lift effects, but including the effects of drag and gravity..

Hit the surface at 1061 seconds at 559 m/s with accumulated (integrated) range of 3,444Km

Passes through 700m/s at the 1049 second mark at 2.4Km to the surface. The angle to the local horizontal at this point is about 17 degrees.
Peak deceleration is about 2 gees at 12.4Km, when we're still going 1480m/s.

The same thing, done with Beta = 1200 hits the surface at 1235m/s.

Going back to the Beta = 400 case (again this is the 60 tonne lander) but now applying lift (in the software, lift is simply anti gravity).

With lift to drag of 0.1 we go through 700m/s at 5.25Km with an angle to the local horizontal at that point of about 13 degrees.

with lift to drag of 0.3 we go through 700m/s at nearly 9Km with an angle to the local horizontal at that point of about 10 degrees. The flight path actually flattens out to 0.7 degrees at 42Km.

And with lots of lift (the space shuttle type design) and a lift to drag of 1.0 we get to 700m/s at over 15Km. It actually bounces twice. First time it hits 56Km and bounces up to 68Km. Next time is around the 48 to 49Km mark.

I've not factored in heating as of yet.

Unless I'm seriously in error (and everything makes physical sense to me, when tested individually), what this points to is that the time involved is too long to ignore lost altitude due to gravity. With that lost altitude is less time to decelerate.

Russel · 2013-01-22 09:53:31

With the same simulator I tried a 5,000Kg capsule with Beta = 150 (roughly a 6m diameter heat shield.

Lift to drag 0.25.

Curiously it reaches a terminal velocity of about 200m/s in the last few Km.. but that's with the oversimplified density model.

Mach 3 at 16.7Km with a 11 degree down angle. Peak of about 0.9 gees at about 28Km.

Last edited by Russel (2013-01-22 10:02:11)

GW Johnson · 2013-01-22 11:17:46

Russel:

Looks like you have a pretty good first version of some kind of trajectory code. Small time step goes nicely with really-simple forward-stepping difference equations. No real need for that Runge-Kutta stuff with modern computers. Exactly what I would do. You will need some capsule drag coefficients vs Mach all the way from subsonic to hypersonic. I already have a little of that posted over at "exrocketman".

Amazing how close my crummy little model actually gets, isn't it? The idea was something cheap and simple to get one into the right ballpark. Once there, a real spherical trajectory code is the right tool. No doubt about that.

I'm guessing it worked better as a "straight line" approximation for warheads because those enter very steeply and remain very hypersonic to impact/detonation. Still, I find it useful as a first step screening tool, even for the landers we are talking about.

GW

Russel · 2013-01-23 05:43:55

Yep, I think the ballistic model does assume you're not going to take very long.

Done some tidying up and made the lifting force more accurate so here's some interesting results.

Starting with a 60 tonne lander and no lift to drag with a Beta of 400 (effective drag area of 150 m^2). You go through 700m/s at 2.4Km altitude and a vertical descent of 200m/s.

If you add a lift/drag of 0.24 (MSL) then you go through 700m/s at 7.9Km with a vertical rate of 141m/s. Now that's quite doable.

As for me - I'm still more interested in a crew only vehicle. Picking a round number of 5 tonnes, the same lift to drag, and a heat shield of 3.7m (dragon) and a CoD of 1.2 then I get..

8.2Km and 141m/s. Again, workable but I'd prefer more margin for crew.

Now if I simply borrow the heat shield from MSL I get 11.6Km

What if I went a tad further and used a full 7m heat shield? Now its 19.3Km. That's more like it.

Of course that's a bigger, but its also a thinner heat shield. I've lost half the KE by 42Km up so less peak heating. At some point I'd like to wire that trade off into this.

Which brings me to where I started. Where, if anywhere, could fuel be used to slow either the absolute velocity or the rate of vertical descent such that two things are possible. One is to use a non-ablative/reusable heat shield. The other is get the thing down to Mach 3 with lots of margin. From the above I conclude that the latter is easier - its simply a matter of a big heat shield.

The bigger heat shield also reduces the rate of heating and spreads the heat over a larger area - that possibly helps.

Anyhow.. I need to sit back and think about where to go next.

Russel · 2013-01-23 05:57:38

Here's another wild/half baked idea. Think hovercraft here on Earth..

Imagine a large circular platform. Doesn't have to be flat. Ringed around the edge by a "skirt" or perhaps even a donut shape. Other better shapes are possible.

Important part is that instead of being convex to the oncoming stream, its concave.

In suitable places you place water or ice - within the skirt, or the platform, or both.

Steam is generated and its allowed to fill the volume contained under the platform. The pressure created keeps the heated gas from coming near the platform - at a generally greater standoff distance than with a conventional heat shield. You arrange the skirt to either contain, or be cooled by water or steam. Some steam will mix to cool the gas stream that goes up over the skirt in any case.

So basically you're talking about sub 500C temperatures on the platform and sub 1000C temperatures (remember the bigger it is, the cooler it is) at the skirt. So well within modest alloys, or even make some of that inflatable.

Some simple refinements include making the platform itself concave - rather like the bottom of a pressure cylinder.

But the pressures we're talking about aren't huge. A fraction of an atmosphere.

What do you think?

GW Johnson · 2013-01-23 15:11:34

Hi Russel:

The concave idea you describe is pretty much the same principle I invoke for doing supersonic or even hypersonic retro thrust through holes in a heat shield. If the room containing the engines is sealed, it stops through-flow through the holes in the heat shield (other than a tiny accumulation as pressure and density rises). Without throughflow, it never gets very hot in that room. No effective hot gas convection, and there is no better conduction insulator than a static gas column. So, there is no need to cap the holes for entry, and then try to open them in order to fire the engines. Sacrificial steam would help lower the engine room temperatures even more. There is some radiant heat getting in.

GW

Russel · 2013-01-23 19:55:58

Yes, its an ancestor of that idea, in this case the hole in the heat shield is extended to nearly the size of the heat shield. You could still put engine nozzles within that space. Main consideration would be keep the gas/steam mix within the space relatively stable.

In the sequence of ideas behind this, one of them is a delta winged lander. In the nose, pointed downwards is a thruster the original purpose of which was for standing the whole thing up for ascent, but then I realise the same thruster could be used to control the descent trajectory, both providing a little thrust where needed and controlling attitude and lift.

Then it occurred to me that in such a lander you could use steam cooling for the nose and it might be possible to give this forward "nozzle" a second purpose, injecting steam which would then flow out and provide some cooling to the nose and underbelly.

Yep, much testing required.

Russel · 2013-01-24 05:17:19

I decided to add the convection heating model to my simulation.

With a 60 tonne lander, Beta = 400, nose radius = 6.3 I get the following.

(509 seconds) Velocity peaks at 3525m/s at 70Km altittude. This is the point where the drag starts to catch up with gravity. Heating is almost 1W/cm^2 at this point.
(597 seconds) 52Km altitude. Heating now at 2W/cm^2. Drag at 0.07 gees.
(649 seconds) 49Km altitude. Heating now at 3W/cm^2. Drag at 0.17 gees.
(691 seconds) 42.9Km altitude. Heating now at 4W/cm^2. Drag at 0.33 gees. Velocity still at 3380m/s.
(767 seconds) 30Km altitude. Heating now at its peak at 5.4W/cm^2. Drag at 1.06 gees. Velocity down to 2912m/s. So nearly half of original energy lost.
(849 seconds) 13Km altitude. Heating now down to 2.2W/cm^2. Drag now at its peak at 2.1 gees. 1574m/s.

899 seconds we reach 700m/s at 2.4Km

Total integrated heat load is 1.3KJ/cm^2.

So basically similar total heat load, but since gravity makes it all happen faster. Total time to surface = 911 seconds.

Going back to a 0.25 lift to drag...

Now it takes 1159 seconds.

700m/s at 7.9Km
Peak gees is 0.87 at 16Km.
Peak heating is 4.2W/cm^2 at 36Km

And curiously the integrated heat load is now 1.58KJ/cm^2.

On a quick inspection I think that's because of lift keeping us longer at higher altitudes where the V^3 heating term wins over the ~p^0.5 density term. In other words, more time spent cooking without the benefit of more drag to slow down. As always, I could be wrong.

Going back to a small crew lander with mass circa 5 tonnes its rather interesting. All I have to do is to add more lift - 0.5+ and I get a higher integrated heat load than with the big lander - as much as 2KJ/cm^2. Mind you, one has to multiply that by the total heat shield area which in this case I'm making to be 30 (bit of a squat shuttle shape).

If the simulator is to be believed, one conclusion that could be drawn (apart from the number one priority being low Beta) is that so long as you can handle the peak heating (say with coolant) then its better to zip through until you get to about 25-35Km and then apply as much drag/lift as you can - presumably altering your attitude (which is something I can't simulate just now).

Going back to a classic capsule shape but a bit fatter than a dragon - say 25m^2 effective drag area and a modest 0.24 lift.. Here's what I get.

Start at the "drop out of orbit" condition and you get to Mach 3 at 14Km.

But...

Kill some velocity - say down to 2800m/s and allow for 2 degree angle of entry. Doesn't change the altitude at Mach 3. Slightly higher peak heating, but the fun part is only a third of the heat load.
Using this setting (admittedly a bit hand tuned) it zips through the period of peak heating at about 11 degrees down angle, and then flattens out to about 6 degrees as it hits peak drag. And that's a totally uncontrolled descent.

Russel · 2013-01-27 09:47:19

Some interesting trends are showing up in this simulator.

For one thing, if you use altitude at Mach 3 as your figure of merit, it seems that thrusting prior to interface creates very little improvement. It happens no matter what size or drag area I use. The reason is fairly obvious. If you slow down you end up going down at a high angle, and that gives you less atmosphere to brake with. That result is pretty clear.

Another thing. I tested the idea of being able to thrust vertically (towards the planet) whilst at the same time travelling more or less horizontally (1.63 degrees to the local horizon down). In fact the control algorithm simulated is simply apply a thrust equal to near local gravity whenever the angle to the local horizontal sinks below 1.63 degrees. Thus simulating what would happen without gravity.

Makes a huge difference. For a given craft the altitude at Mach 3 goes from 14Km to 26Km for that given craft.

Also totaled the delta-V and it came to over 1100m/s.

So, all else being equal, if you were able to thrust straight down (at almost 90 degrees to the stream) then there might be a useful trade off - but is it worth the weight of fuel?

Just for the sake of it, I told it to stop burning fuel at a delta V of 800m/s I got 19Km and at 500m/s I got 15.3Km

So all I can say is there's room to tweak that.. it might be better to avoid doing this at higher altitudes and then concentrate on lower altitudes but not sure.

Lift seems to be more effective. And a combination of some small lift and a bit of thrusting? Not up to that yet

I'm getting less bothered by peak heating. I think with materials advances, a bit of steam and clever engineering it won't be a problem. Indeed its better to sail on past the upper layers (>50Km) and not slow down any more than is guaranteed by drag until peak heating is over - or at least that's what the numbers say.

Real problem is getting maximum purchase on the atmosphere from 35Km on down. Which leads me back to thinking big - large/light heat shield, or else having some way to increase the drag once peak heating is over. Put it this way, if you had some device, say a ballute, it actually seems to make sense to deploy it after peak heating but before gravity kicks in.. say around 2200m/s.

Sorry if this is a bit opaque or boring anyone

And if you want the code let me know.

GW Johnson · 2013-01-27 10:56:21

Russel:

You are getting some very interesting results. That's good work, and your modeling code seems to be getting realistic results. Bravo! I'd like to play with it. Home email "gwj5886@gmail.com".

The Mach 3 endpoint is more-or-less arbitrary, and I first encountered it studying hypersonic aerothermodynamics way back in engineering grad school decades ago. It's Mach 3 for blunt objects, and is closer to Mach 5 for "pointy" objects.

That's the speed at which drag coefficient varies very little with increasing Mach. Below that speed, the drag coefficient variation is strong (and constant ballistic coefficient becomes a very inappropriate assumption to make). That's inherently a fuzzy boundary. But for blunt objects, the Mach 3 point is not far from the max safe opening speed of Mach 2.5 that we get with ribbon chute designs. (And JPL's ringsail chute for the Mars probes is really just a variation on the ribbon chute.)

There's many empirical correlations for stagnation point heating, the one I used is just one of them. "Pointy" things have a small nose radius, which drives local heating way up. That's why sharp noses and aerosurface leading edges tend to burn away very quickly during Earth reentry.

Which in turn is why they never boosted an X-15 into orbit, although that had been proposed about 1960-ish. It couldn't come back nose first, and was too structurally weak to fly back at high AOA, which has the effect of moving the stagnation point to locations with larger radii that lower the heating.

The X-20 was a moderate AOA reentry, as was the shuttle. X-20 was to try ablatives, skin sweat cooling, and heat-sinking, separately and in combinations, as a experimental vehicle. Unfortunately, it never flew. Shuttle used a combination of low-density refractories and ablatives. The nose cap and aerosurface leading edges were ablative carbon-carbon, tough enough to use a few times before replacement. The refractories flew many more times than the carbon-carbon, except for tile damage.

The problem with aerosurfaces, so useful later for landing at 100-200 knots, is that it is impossible to take dead-broadside airloads and still be flightweight. Going dead broadside converts you to a very large effective "nose" radius, and very drastically reduces stagnation-point heating. That's why we quit the rocket planes and went space capsules for the moon program. The ablatives were heavy (and still are), but a blunt "nose" lets you use a lot less of them.

I suspect without proof (yet) that if you folded your aerosurfaces into the wake zone, you could reenter dead broadside, and still be flightweight in your heat protection, be it refractories, ablatives, sweat cooling, or heat-sinking, or any combination. The only "trick" would be going nose-streamline at Mach 3-ish and unfolding the wings. That's a serious airload-structural and aerodynamic-control problem in itself.

The only way I have come up with so far out of that dilemma might be to stay folded and dead-broadside, with a dorsal drogue chute, all the way down to subsonic speed, then release that drogue and go streamline with another drogue out the tail, and finally then unfold the wings at far lower subsonic airloads. Cut the second drogue, and start flying at that point, as a proper subsonic airplane.

I think that sequence might lead to an airframe design with approximately the volume density and wing loading of a WW2-vintage carrier fighter, or maybe a bit less. It might make a pretty good small spaceplane for a crew of 1-4 that could be boosted to orbit with current launchers, and still fly hundreds or thousands of times with a small logisical tail, like that carrier airplane, not the giant logistical tail of the shuttle.

I'm thinking a belly and sides of aluminum monocoque, coated with an inch or so layer of low-density alumino-silicate potting compound, and surfaced with a skin layer of alumino-silicate fire curtain cloth to provide positive retention. Same thing on the aerosurfaces, maybe not as thick, as these are immersed in the wake only, since they are folded.

I haven't run the numbers, but I know you want skin surfaces under 2300 F with this material, and aluminum substrate temperatures under 200 F (with a bit of adjacent fuel or water heat-sinking). The low-density refractory is capable of gradients like that, I've done it before. The "trick" is getting the surface stagnation temperature down under 2300 F by the lower heating dead broadside with the large effective nose radius.

Voila: light weight reentry vehicle. One with straight wings for good low-speed landing characteristics. And perhaps the space and weight allowance for go-around rocket fuel, eliminating the shuttle's dangerous deadstick restrictions.

Concept, anyway.

And, anything that works for reentry here would work on Mars, the heating there is less. I just don't think subsonic wings are much good on Mars. A capsule with low-density refractory covered with a fire curtain cloth skin, for a lightweight heat shield? Hmmmmm!

GW

Russel · 2013-01-28 04:07:25

Ok, I'll tidy that up and email.

I've been toying with the idea of parking some part of the structure out of the way. In that sense it doesn't have to be a wing (in terms of generating much lift) though that helps.

You can augment a standard capsule with "air brake" panels - basically perforated sheet metal - which also double as an outer aeroshell. After peak heating, they fold out like petals.

Hardest thing about re-configurable wings is the practical considerations. Here's a half-way house between a capsule and a "airplane".

Start with a tube with a blunt nose. This shape has been proposed for Mars landers. I think NASA are working with this shape too. So the heat shield area is not just the nose but the belly. And as it stands, it generates a bit of lift too.

Now imagine that the upper surface of this tube unfolds, just like the payload doors of the shuttle. Unfolded they nearly double the effective drag - and there is still a bit of lift.

One design I'm fond of is just a somewhat more elegant version of the shuttle. But with no moving surfaces. Just a rigid body. Unlike the shuttle its not able to come to a rolling landing. Instead it lands under propulsion.

In the nose is a set of thrusters firing downwards through a portal protected just as you suggested with steam. Minimal solution is 3 thrusters. One direct straight down. Two canted somewhat to the side. Towards the tail are two sets of thrusters, again pointed downwards. Again, with a protected hollow.

At the tail is a set of main thrusters, just as with the shuttle.

Landing is on skids belly first. Take off happens like this. The nose thrusters fire and rotate the whole craft towards the vertical with the rear skids design to allow this. As it nears vertical the main thrusters fire and thus you begin ascent.

Descent begins with a low angle of attack. On the most critical surfaces you have a refractory metal (either a foam, or structured to lower conductivity). Or you put a small thickness of thermal insulation behind the skin. The idea here is the skin is designed to heat and lose some energy radiatively. Behind this is a tank of slush/water. Steam generated fills the hollows in which the thrusters sit, and selectively on certain areas such as the leading edges. A low angle of attack takes you through peak heating quickly with minimal usage of water.

As you descend past peak heating and lose velocity you raise the angle of attack, increasing drag and lift. You then use a small amount of thrust if necessary (I don't know if it is) to lower your rate of descent. Then as one last maneuver you broadside to get as much drag as possible before going into retro thrust. Then its mainly the belly thrusters that do the work of positioning for landing.

Ok, that's my dream craft for the moment. I'm still not convinced its necessary to go to that much trouble when for the same effort you can always start with a regular capsule shape and instead just put the effort into making the heat shield-aeroshell as light as possible.

Returning to where I originally came into this. I'm still wary of landing crews along with large payloads. And looking at the numbers I can see there are still advantages to landing the large mass payloads and the crew separately because that leads to a more optimal design for each.

With a large payload but un-crewed you can afford to have smaller margins in terms of time and altitude. Also to some extent its not important that the habitat is a particularly accurate landing if instead the crewed landing can be more accurate as a trade off. If you do split the two then a standard capsule shape works well for the large mass. Since the heat shield is by design expendable you can oversize it (built in space in segments). Your 60 tonne lander with a 13m heat shield and a Beta of 400 now becomes a lander with an 18m heat shied and a Beta of 250 (even allowing for more mass in the heat shield).

Edit: And a 60 tonne craft with a Beta of 250 and no lift still gets you to Mach 3 at 8.7Km.. with 0.1 lift its 11Km.

What this also does is allow you to build a more squat habitat with more floor area. Or alternately the habitat ships to Earth orbit as a series of modules that interlock - a bit like a mini "space station" concept. Onto that structure you then bolt the heat shield segments. The overall structure provides some of the stiffness for the heat shield itself on descent. During landing the heat shield pieces are shed and you're basically landing a "station".. and that could extend past 20m in size. Even greater reductions in Beta are possible if you land certain other things separately like consumables.

And since its unmanned you can afford to lower margins. Ok, so you crater one now and then but noone gets hurt. But the crewed lander has a lot more margin - and that's where lift, and other fancy features, I think, comes into play.

Last edited by Russel (2013-01-28 04:18:32)

Russel · 2013-01-30 02:58:45

Thinking of winged things...

I guess the closest I get to a vehicle with light weight "wings" - wings in the sense of generating lift - is like this. Start with a cylindrical shape as your "fuselage". Maybe shape the nose a little to shape the shock waves. But basically nothing fancy, and you can get away with a regular cylinder shape. Ok, now give it wings. Now you don't have the give the wings true aerfoil shape, but you still need need a counter-force to give the wings an angle of attack. So what I come up with is basically two very long slender "glider" wings projecting outwards from the body of the cylinder. I'd give the center spar a layered structure with a core giving it strength at extremes of temperature inside a good insulator - even an old fashioned dewar structure. These wings are fixed. No flaps, just structure. Ok, further aft you have two much smaller wings. Similar basic structure but this time able to be rotated - acting like an elevator.

The principle here is just to do the long glide using lift in the upper atmosphere. Much of the heat load is experienced by the wing, but also re-radiated. Now, as you approach denser layers it might be necessary to shed the wings, or even shed portions of the wings in stages. Anyhow that's the nearest I can get to a "tiger moth" idea.

I guess that if you've gone to the trouble of developing the materials for that, it might still make sense to simply adopt the "shuttlecock" shape, extending a conventional heat shield outwards. Gut feeling says that that would work with a progressively less dense, more porous surface towards the edges. Again, expendable.

Now, going back to my obsession - which is a lander/ascent vehicle. And thinking about a conventional capsule shape. One of my issues is minimizing the extra mass that would be needed to turn a lander into an ascent vehicle. Here is one such idea. Following on from the discussion above, suppose I build a torus in the order of 7-9m in major diameter made out of something tough and moderately temperature resistant. Attached to that, and sealing the inside of the torus, a dome structure providing the platform for the engines. Now, on the bottom surface of the torus is a thin skin of a refractory material with some degree of insulation. Purpose here is to make the best use of radiative cooling.

On descent you use water as a coolant, generating steam as above. The water is stored inside the torus, keeping the center of mass low as possible.

On the surface, the inside of the torus is a convenient volume which then forms the CO tank for the ascent. Which means the metal used is doing double duty and you're keeping tankage mass down.

Obviously you need either a separate smaller tank for the fuel for the descent. And I've yet to consider the oxygen but the oxygen demands about a third of the volume of the CO tank.

Russel · 2013-01-30 09:06:17

Suppose a lander/ascent vehicle weighing 7.9 tonnes without landing/ascent fuel but including structure/crew/RCS mass.

The base of the lander is a torus 7m outside diameter, 3.4m inner diameter, 1.8m high. That would contain 32.8 tonnes of CO. To burn this at stoichiometric ratio we need 11.92 tonnes of oxygen. That fits into a sphere 2.7m in diameter. With a mass ratio of 6.6 the craft is capable of 5.5Km/s Delta V. Split that into 4.5Km/s going up and then 1Km/s going down.

On return from orbit it contains 3.25 tonnes of fuel. To that I'll add 1 tonne of water ice stolen from the transit vehicle. So prior to entry it has a mass of 12.15 tonnes.

Allowing for the water to be consumed and reading between the lines, we get to hit the retros at about 7Km. More like 10Km if a small amount of additional drag is added - such as that "air brake" idea.

Still would like some more altitude but I think that could be accomplished by making the base closer to 9m in diameter.

Fairly happy that its not impossible. But the more shared structure the better.

Edit:Typo

Last edited by Russel (2013-02-01 19:31:43)

Russel · 2013-02-01 02:47:46

I'm falling in love with the "air brake" idea.

Basic idea is a standard capsule shape but the upper conical aeroshell is not a pressure vessel. Rather the pressure vessel that contains the crew is much smaller and there is a large space between the inner pressure vessel containing the crew and the outer aeroshell.

The outer aeroshell may not even be a solid surface. What is sufficient is that it is enough to provide stability.

Its divided into a number of separate sections (at least 4). Each can independently fold outwards. The best analogy I can think of is petals.

Now prior to entry they're folded up into the classic conical shape. They're left that way through peak heating. As heating subsides they're folded out.

Having played with this idea a bit more I've figured out a few things. First, the net effect is to easily double the effective drag area of the whole vehicle. Second, when you simulate it, the real benefit is in the denser air under 25Km. I get good benefit making it kick in at about 1700m/s which is at about 20Km. Under those circumstances 700m/s is reached at 13Km. That compares to half that altitude without the device.

Indeed the limitation might be how many gees you wish to inflict on your crew. In the simulator it peaks out at over 4 gees but only for 10 seconds or so.

GW Johnson · 2013-02-01 12:44:58

Lots of people are scared by 4+ gees, even in a short transient. But they needn't be. There are many roller coasters that pull 4 to 5 gee transients on ordinary members of the public. We do just fine. Even 6 gees is feasible that way, as a transient. That's sitting bolt upright and taking the gees eyeballs-down. Steady state, you pass out in several seconds. Short transient, no worries. Tolerance windows-of-time are very much longer if you recline the seats to on-your-back, taking gees eyeballs-back. Minutes-to-steady-state.

GW

Russel · 2013-02-01 17:56:34

GW,

Just wondering how the simulation code is working (or not) on your end?

Last edited by Russel (2013-02-02 08:07:28)

Russel · 2013-02-03 07:24:34

Rather interesting article.

http://www.wired.com/wiredscience/2012/ … nder-1966/

GW Johnson · 2013-02-03 09:47:31

Hi Russel:

I kinda like your air brake idea for capsule-like vehicles. It makes real sense and would be pretty easy to do with metallic structures and ordinary hydraulics. I had thought about using aeroshell panels as unload ramps. It would seem they could serve as decelerators, too. Triple duty. I like that.

Your code might be able to distinguish whether using the airbrakes earlier in an entry trajectory would do much good. If they have some sort of heat protection on them, they could be used that way. The effective shape with panels deployed does not have to be round.

I did not recognize what language that code is written in, so I don't know what software might run it. Too obsolete I guess. The last languages I knew were Fortran-IV and some QuickBASIC4.5, both vintage 1980-ish or earlier.

GW

GW Johnson · 2013-02-03 10:02:07

The article, that the link in Russel's post 372 above leads to, is very interesting indeed. The chemical lander designs I was looking at used about the same shape and a very similar internal layout. I used the descent engines as the ascent engines, to save a bit of hardware weight. My mission architecture staged from LMO to eliminate the narrow entry corridor problem inherently. I think the sense of the referenced article was direct entry from the interplanetary transfer orbit.

I probably used a more generous inert fraction allowance, but still, his weight (56 tons) vs mine (60 tons) is remarkably close. Great minds tend to think alike, because logical problem-solving leads different folks down pretty much the same paths. My design rough-outs are posted over at http://exrocketman.blogspot.com, if anybody wants to compare.

I was considering looking at a reusable form of the chemical lander to take advantage of surface propellant "refining" of some kind. The one-shot designs I posted were based on MMH-NTO for excellent long-term storability. I was also considering looking at a one-stage nuke "landing boat" capsule, that could make the trip two-way on one fueling with substantial cargo, or one-way with surface fueling and even more cargo. Nothing done along those lines yet.

GW

Russel · 2013-02-03 21:52:32

Oh ok.. its done in Java and requires two downloads. If you like, email me and I'll walk you through it.

New Mars Forums

Announcement

#351 2013-01-14 16:01:05

Re: Landing on Mars

#352 2013-01-15 04:47:32

Re: Landing on Mars

#353 2013-01-21 10:48:35

Re: Landing on Mars

#354 2013-01-21 14:45:05

Re: Landing on Mars

#355 2013-01-21 20:18:44

Re: Landing on Mars

#356 2013-01-22 09:42:41

Re: Landing on Mars

#357 2013-01-22 09:53:31

Re: Landing on Mars

#358 2013-01-22 11:17:46

Re: Landing on Mars

#359 2013-01-23 05:43:55

Re: Landing on Mars

#360 2013-01-23 05:57:38

Re: Landing on Mars

#361 2013-01-23 15:11:34

Re: Landing on Mars

#362 2013-01-23 19:55:58

Re: Landing on Mars

#363 2013-01-24 05:17:19

Re: Landing on Mars

#364 2013-01-27 09:47:19

Re: Landing on Mars

#365 2013-01-27 10:56:21

Re: Landing on Mars

#366 2013-01-28 04:07:25

Re: Landing on Mars

#367 2013-01-30 02:58:45

Re: Landing on Mars

#368 2013-01-30 09:06:17

Re: Landing on Mars

#369 2013-02-01 02:47:46

Re: Landing on Mars

#370 2013-02-01 12:44:58

Re: Landing on Mars

#371 2013-02-01 17:56:34

Re: Landing on Mars

#372 2013-02-03 07:24:34

Re: Landing on Mars

#373 2013-02-03 09:47:31

Re: Landing on Mars

#374 2013-02-03 10:02:07

Re: Landing on Mars

#375 2013-02-03 21:52:32

Re: Landing on Mars

Board footer