100% stock Mun mission designs and .craft files (version 0.13.x and 0.14)

[closette]

Glad to hear you had a good flight, tarspaceheel!

It will come as no surprise that my Munandback131A design was the result of many iterations (about 14) and that was after trying out some less-than-satisfactory stock Mun designs I found here on the message boards. Even then it has its shortcomings - if you try to pitchover too early in the atmosphere it will go haywire on you and you\'ll spend all your RCS fuel (needed for return from the Mun) trying to get back near vertical.

It sounds like many of us newbies have stability problems with the first 'lifting' stage. Symmetry is important (although the tool does have a few bugs, watch out for them), but here are some stability/controllability considerations I took into account:

- rotational inertia. a long skinny rocket is much more reluctant to change direction than a shorter fat one of the same mass, because in the former case much of the mass is far from the pivot point. (Grab a long stick at its center and wiggle it back and forth, then break the stick into 4-6 pieces and wiggle the bundle back and forth, it will feel much easier). Since v0.13.1 we can attach liquid fuel tanks 'sideways', and route fuel lines between them, which is a great help in this department, especially for orbital maneuvering.

- center of mass (CM) or center of gravity (CG). In free flight, a rocket will naturally rotate around its CG. (For example, throw a hammer a you\'ll see it rotate around its CG). For a Mun lander, a low CG will make it less likely to topple over, a good thing! On the other hand, for lifting the stack through the atmosphere, you don\'t want the combined CG to be too low, because for stable flight it should remain above the center of pressure (CP), see below. As liquid fuel tanks drain from top to bottom during the lift, the rocket gets less massive, but be aware that the CG gets lower due to the fuel remaining in the bottom of the tanks. That is one reason why a big liquid rocket can be stable at launch but then go nuts some time later. The other reason is

- center of Pressure (CP). This is the point on the rocket through which the sum of all the aerodynamic forces can be said to act. For a vertical rocket, if the CP is below the CG, the rocket will naturally want to climb nose up and will be stable to small deviations from vertical. Ever wonder why fins are placed on the tail of a rocket and not usually near the top? Adding more surface area lower down, causes the CP to stay below the CG even when fuel has almost run out. Also, if the fins are adjustable, placing them near the bottom (and far from the CG) will allow them to exert more torque (or turning effect) to control your rocket.

For a given geometry (i.e. not jettisoning tanks or other bits) the CP stays more or less in the same place as long as you stay nearly vertical. However, if your rocket ends up pitching over too far so that it\'s hitting the air flow sideways, the CP will move close to the rocket\'s mid-point, possibly above the rocket\'s CG. And you\'ll have one heck of an unstable beast (less ladylike words usually fly out of my mouth when this happens!).

A good explanation of rocket stability and CP and CG for model rockets can be found here: http://exploration.grc.nasa.gov/education/rocket/rktstab.html

Finally in KSP, if your rocket still needs a little help pointing straight, we have

- SAS modules. Throw a few into your lower stage 'stack'. I used three. if they max out during flight, and you already have fins and RCS, consider adding more. You only need one Advanced SAS (ASAS) module though, since it acts like a 'brain' to control the others. Conveniently, users have found that the ASAS module provides a pretty good crumple zone for surviving crashes, too, which is why I place mine under the command module.

As to how many fins/SAS modules/fuel tanks and where to put \'em - trial and error! LOTS of error!

Hope some of the above is useful in keeping the sharp end pointing upwards.

[Kosmo-not]

@tarspaceheel

Sounds like you need to put some struts on your rocket.

As far as fuel needed to reach orbit...

Reaching orbit with as little fuel as possible is a balancing game between thrust, gravity, and air resistance.

Gravity:

I\'m sure I don\'t need to tell you what gravity is. A significant amount of energy that your engines are producing goes to countering gravity in the initial phase of your flight to orbit. Take your craft on the launchpad and throttle up until you\'re hovering. That is how much fuel you\'re using per amount of time pointing straight up. If you point your nose level with the horizon, none of your fuel is being used up by gravity. It\'s called gravity drag, and you want to minimize it as much as possible.

Air Resistance:

I think when you double your speed, you quadruple the air resistance. The best way to counter this is to go slower.

Thrust and mass:

The more thrust, the higher the acceleration, but engines weigh as much as 6 and 2/3 empty fuel tanks. The less mass you have, the higher the acceleration. Fuel is spent lifting every engine you have on your rocket and the fuel tanks.

The balancing game:

You want to get out of the lower atmosphere quickly and economically. If you go too fast, you spend more fuel fighting air resistance than you have to. If you go too slow, you\'re losing a lot of fuel to gravity drag. If you pitch over too soon or too much to reduce gravity drag, you spend more fuel getting out of the lower atmosphere.

But as you go higher, air density becomes lower, which reduces air resistance increases your velocity. But gravity drag hardly changes at all. You need to adjust your pitch to reduce gravity drag while still ascending out of the atmosphere.

You can use staging to reduce the amount of mass you are carrying to orbit. Sometimes it is beneficial to drop a few engines. I\'ll let you play around with that one. You have to be sure that, if you drop some engines, the remaining engines will have enough time to accelerate into orbit before passing apoapsis.

It\'s a pretty tough game when you\'re trying to get something into orbit with the smallest rocket you can. If you want an example rocket that plays this balancing game, download and fly my Kosmo Zero Mk6. With as few fuel tanks as it uses, It has to play this balancing game to get to orbit. http://kerbalspaceprogram.com/forum/index.php?topic=6310.0

[closette]

Nice coverage of the fuel budget considerations, Kosmo. And yes I forgot to mention struts for structural stability (not quite the same as aerodynamic stability).

If I may comment on one statement you made in the 'balancing game':

If you go too fast, you spend more fuel fighting air resistance than you have to. If you go too slow, you\'re losing a lot of fuel to gravity drag.

From posts elsewhere, my understanding is that for Kerbin\'s gravity and atmosphere, the balance is tilted strongly in favor of 'get as high as you can as soon as you can'. Yes the higher speeds will increase drag, which is why I use SRBs to help push through it, and then jettison them asap.

I would love to say 'GO at throttle-up' after reaching some altitude, but every spacecraft I\'ve tried is most efficient if thrust is at a maximum from launch to at least 20K. (By which time you\'re above most of the atmosphere and deciding on pitch angles). That\'s true even for my fairly draggy Munandback131A craft, but I would be pleased to learn of counter-examples.

I read somewhere(*) that if/when a more accurate drag model is implemented in KSP, this will skew us even more towards 'power hard'.

(*) Found it, FWIW. Read Hypocee\'s replies in this thread: http://kerbalspaceprogram.com/forum/index.php?topic=6008.0

[Kosmo-not]

Actually, I have experimented with lifting off at full throttle vs throttling back a little bit. If you have a rocket with high enough thrust, throttling back will save fuel.

[closette]

Cool! Is this with your Mk6 Munar design, or a lightweight powerful orbiter sans lander?

Please report back with some comparisons and numbers (throttle settings, fuel consumption, screenshots?) if you have time. The cleanest way to test this would be with purely vertical launches up to say 50km, kind of boring though.

I updated my previous reply with a reference to a thread on thrust/weight ratios. Not much evidence in there but it would be interesting to see if your experience accords with the claims made.

[Kosmo-not]

I just tested it. Launched with capsule, asas, 3 tanks, tri-coupler, and 3 LFEs with no thrust vectoring.

First launch was full power all the way through, got to 83km.

Second launch I had throttled back a little for most of the thick part, then throttled up again. 93km. And that doesn\'t take into account the extra gravity drag because of the longer thrust time.

[closette]

Excellent test, and I stand corrected! My own spacecraft are so under-powered that this hasn\'t been an issue for me.

Like Hypocee\'s posts which I referenced above, I\'d like to know if there\'s a universal rule-of-thumb for optimal thrust, but your demonstration at least shows that an optimum does exist between low thrust and maximum, with the current drag model.

[Kosmo-not]

I did the algebra myself and plotted it in excel.

The best TWR is 2.

[closette]

Could you please share the formula you used for 'fuel efficiency', or even if you don\'t mind attachthe spreadsheet. I trust your analysis 100%, I\'m hungry for more details!

[Kosmo-not]

Formula is:

(1-1/TWR)/sqrt(TWR-1)

[closette]

Excuse me being dense, but where does that formula come from as a measure of fuel efficiency?

[Kosmo-not]

Good question to ask. The formula I used to derive this from is the equation for drag resistance. With leaving air density as a constant, it\'s just comes out to be a function of the square of velocity. The other forces to consider are thrust and weight of the rocket. The measure of fuel economy is velocity divided by thrust (basically like miles per gallon, since fuel flow rate is directly proportional to thrust).

[closette]

Glad we\'re in the same time zone, thanks for the fast response. I think I can derive that formula, but if you care to provide some of the steps I won\'t mind, and nor will others I\'m sure, since this question has come up quite a few times.

Keeping the atmospheric density constant is good enough for now. Not sure whether Kerbin\'s atmosphere is exponential, and the drag model I am told isn\'t quite right (it has a mass term that should not be there), but these are minor details.

[Kosmo-not]

Here you go.

[closette]

Thanks!

[Kosmo-not]

The next thing I\'ll be doing is getting air densities and plotting those... not tonight, though. Gotta do this through experiments.

[closette]

Hi Kosmo,

Reverse engineering atmospheric density sounds like a great idea, I had thought of using video capture software to figure out thrust, mass, acceleration (and hence the missing drag term), but was dissuaded from doing so by this slightly confusing thread:

http://kerbalspaceprogram.com/forum/index.php?topic=5623.msg73154#msg73154

which might (or might not) imply that the spacecraft mass appears incorrectly in the drag model.

Also, the current version of KSP apparently just adds up the (drag coefficient x frontal area) of all parts together, regardless of how they are arranged, so a single stack of 4 tandem fuel tanks ends up with the same number as 4 tanks mounted side-by-side. So perhaps I should enjoy using my 'hip tank' design while I can...

In the meantime, I would genuinely not want you to waste your time on reverse engineering an incorrect model, and one which might soon be changed.

[Kosmo-not]

That\'s exactly how I figured out air density. I launched a rocket straight up, ejected my capsule and shot a video of the descent. Here is the graph I came up with.

Now that we have the density (and I don\'t think that will change anytime soon), we can go on to calculate other things when the new drag model arrives.

I\'ve included the excel file too.

Too bad I only have the starter version of excel. It doesn\'t let me do a power function for trendline. Maybe someone else can do that and see what equation it gives.

It seems to be logarithmic in nature until about 40km, then looks like it\'s about constant the rest of the way up.

[closette]

You are a scholar of the highest calibre!

I think you meant exponential, not logarithmic, because your data from 2500 to 38 000m almost exactly fits this equation with correlation >0.99:

density = 0.01008 kg/m³ e ^{-(height / 4907 m)}

Above 40km where you are moving slowly through thin air(*) your derived density data gets 'noisy' so it may be constant, or a continuation of the exponential, or l (my guess) inear, but it\'s not a big deal.

This is assuming that masses are given in kg (and thrusts given would then have to be in Newtons).

I think it is more likely that masses given are close to metric tons and thrusts in kN, based on comparisons with real world command module, parachute masses, and densities of the SRBs, as discussed in this thread:

http://kerbalspaceprogram.com/forum/index.php?topic=1769.0. That would scale the surface density up by a factor of 1000, but the exponential scale height would not be affected.

And I am still concerned that the presence of an erroneous mass term in KSP\'s drag equation (mentioned in the thread I referred to in my previous post) might mean you\'ll get the same freefall data but different derived drag force and density for, say, a capsule with some additional mass attached (such as an empty RCS tank).

But the robust result is that the scale height of Kerbin\'s atmosphere is 4900+/-80m. Sounds about right, but now we KNOW!

(*) No idea what the composition of Kerbin\'s atmosphere is, but I use 'air' anyway.

[BillWiskins]

This is entirely a marvellous thread, this is. I feel like I\'ve tripped and fallen down a science lesson.

And while I hate to interrupt all the science, I did want to briefly pose a question to some other stock-rocket fans. It would be unnecessary, but I am away from my copy of KSP and as such all your lovely .craft files are useless to me.

The question is this one: closette, with your fancy munandback, are you leaving the Mun powered by RCS only? I ask because, having finally made successful Mun landings a thing that I can do with relative ease, I\'d like now to be able to get some of the little fellows home. With the command modular population of the Mun rapidly increasing (and my poor laptop trembling at the prospect of persistent CMs littering the surface), I am constantly getting closer, but ultimately failing, to a proper return journey.

Of course, it might well be that my design is in need of a massive update, but the lander stage in your pictures seems to resemble mine and I wonder what it is I\'m doing wrong. My lander looks like this:

Parachute (for safety)

CM (for being in)

Decoupler (for decoupling)

3 RCS tanks (with 12 blocks in three sets of four)

4 of those wobbly winglets (for landing on).

Having landed, I\'ve usually used only half to two-thirds of a tank of RCS. When leaving, I boost straight up on the H key and slowly turn towards 270 degrees, in an attempt to achieve some kind of useful speed. Trouble is, I always run out of fuel long before I break free of the Mun. So far my best is 78,000 metres and around 3/400 m/s.

So before I try six more times, get angry, try six more times and climb inside the washing machine to go slowly mad, what\'s glaringly wrong with my plan?

Any advice gratefully received. I did do a quick search for this kind of thing, but I didn\'t find much of use. If this is a thing that has been covered a hundred and seventy-nine times already, then I apologise. Just leave a message saying \'sod off\' if this is the case.

Good day to you!

[tarspaceheel]

Thanks for the replies. The first few will be invaluable when I go back to the drawing board later today, and though the remainder of the conversation is a smidge above my pay grade, I\'m still looking on it with stars in my eyes, wishing I\'d stuck with the sciences rather than running off to hide in law school.

Bill, I\'m probably not the one you should be listening to, but at least the way that I used it, Closette\'s lander did return from the Mun using just RCS. I can\'t really tell you whether your procedure for getting back to Kerbin is correct, as I\'ve only ever successfully landed on the Mun once, and am pretty sure that I only got back through sheer luck.

What I can suggest is that if you\'re running out of fuel, perhaps your lander is too massive. Try putting a decoupler between the 2nd and 3rd RCS tanks. Shortly after getting off of the Munar surface (and before you start using your 2nd tank), detach the 3rd RCS and the winglets. What you have left will be difficult to control, but should be fuel-efficient enough to set up a transfer into Kerbal (Kerban? Kerbinian?) orbit, and from there, it should just be a matter of finding a good place to land.

[closette]

Hi Bill,

Yes, Kosmo-not\'s work should be published in a more obvious place, it\'s so good.

Your Mun experiences are only a few weeks behind mine, and your craft sounds similar, but by all means post the .craft file when you get a chance.

Yes, I use RCS to return from the Mun, and usually start with close to 2 tanks left (either a little more or a little less). The 3rd (empty or nearly empty tank) is attached to an ASAS module and the fins, all below the stack decoupler. Shortly after taking off from the surface, I hit the spacebar and jettison these items, leaving me with the command module+parachute on top of 2 RCS tanks. A bit squirrely to maneuver but OK for crude pointing.

So if I understand you have a little more than 2 RCS tanks left and follow a similar plan, but somehow don\'t succeed.

My first suggestion - move your stack decoupler down so that when tank #3 runs out after take-off, you can get rid of it along with the fins.

(You can still land safely back on Kerbin with a couple of nearly empty RCS tanks under your parachute, especially on water. On land, they may explode but your Kerbals will be all right.)

My second suggestion: get into Mun orbit in stages, as opposed to one continuous burn. After take off and pitchover to 270 degrees heading at about 15 degrees above the horizon, keep boosting while watching the map view. After a short while you\'ll see your Apoapsis (apomun?) move up and away from the spacecraft. Stop boosting and coast for a minute or so (you can accelerate time if you want). Then when you get close to apoapsis (at about 80% of its altitude), point along your velocity vector and hit 'H' again for 5-10 seconds. The apoapsis will climb up away from your spacecraft again and further around the Mun.

If you 'chase' the apoapsis a few times in this manner with small bursts, you will find yourself in a low retrograde orbit around the Mun (12 000 m is typical for me).

If you need to, depending on where you started from, orbit the Mun once completely, coasting (switch off RCS so that it won\'t leak on you if you have SAS engaged). Let the orbit take you to just before the 'leading edge' of the Mun\'s grey orbit line around Kerbin. At that point, do a long boost to produce an escape trajectory that should take you around the back of the Mun at about 850 m/s, and will spit you out of the Mun\'s sphere of influence going at about 500 m/s opposite to its orbital direction around Kerbin.

If you do this, you will have canceled out most of the Mun\'s motion around Kerbin, and you should fall almost straight down back home. If not, use whatever fuel remains to boost against your velocity and lower your periapsis below 30 000m into Kerbin\'s atmosphere.

I usually have ~ 1/2 RCS tank remaining when I come in to land. If more, I just burn it off in the atmosphere. But I save some RCS fuel so that I can reduce my landing speed under the parachute, using the 'H' key for the last 100 m or so. (The Soyuz capsule uses retro-rockets to brake just before landing also).

Hope this is clear. If not let me know and I\'ll add some screenshots next time I do it, and good luck!

[Darkshadow]

Too bad I only have the starter version of excel. It doesn\'t let me do a power function for trendline. Maybe someone else can do that and see what equation it gives.

It seems to be logarithmic in nature until about 40km, then looks like it\'s about constant the rest of the way up.

Wow just wow, you\'ve just plotted a perfect exponentiation decay exactly following the standard atmospheric model. Also does this mean that from 40,000m and higher it should be possible to get an orbit that slowly decays?

So before I try six more times, get angry, try six more times and climb inside the washing machine to go slowly mad, what\'s glaringly wrong with my plan?

My lander design uses the remainder of the fuel from the landing stage to get into orbit, before separating to 3 RCS tanks for the TKI and that is just enough.

Finally Closette you should edit the title of this topic, as it now includes all sorts of interest posts.

[closette]

Good idea Darkshadow, and done.

Oh, and yes you will have a very slowly decaying orbit between 40 and 70 km. I\'ve been 'stuck' with a periapsis in that range a number of times and no fuel remaining, and it takes forever to come back down.

[Kosmo-not]

I\'ve done some orbits where I\'m still in the upper atmosphere. I\'ve read somewhere that this is what the space shuttle did. The reason being that if something were to go wrong with the engines, they would be able to return to Earth before oxygen ran out and not be stuck in orbit.

My method of leaving the moon involves a long burn. A few seconds after leaving the surface, I nose down close to the horizon to rid myself of gravity drag. The Mun\'s gravity isn\'t much and is easy to counter if the velocity vector goes below the horizon. I keep burning until I\'m in orbit.

And about the graph: the density is probably not correct, due to certain assumptions to simplify calculations. But it does show us how density changes with altitude.