Mun landing rocket design

[MPDerksen]

I have a combination of missions:
1. rescue a Kerbalnaut stranded on the backside of the Mun surface
2. gather science data from the Mun surface
3. dock 2 vessels in Mun orbit

My plan is to use the design below.  I'll have a pilot take the lander down, with an extra seat, pick up some science and the stranded peep, then come back into orbit and dock.  However, my rocket is SO topheavy that it tips over almost immediately on liftoff.  I had some fins on there, but removed them for the picture.  I think I have plenty of dV for the entire mission, but can't get it to fly straight.

Recommendations on a design change?

[bewing]

Suggestion #1 would be to put 4 control surfaces (tailfins, for example) at either the top or bottom of the rocket.

Suggestion #2 is that your goal #3 (docking to complete a contract) can't be done that way. KSP keeps track of which rockets came from which launches. If you take a rocket, undock 2 pieces, and redock them -- the game will know what you did and won't give you docking credit.

 

[Foxster]

Coupla things...

Get more drag at the back end. You have that big draggy lump at the front and not much at the back. Add fins at the back until it flies right. 

Use autostrut if you aren't already. A wobbly rocket can be a flippy rocket. 

Edited March 13, 2019 by Foxster

[Harry Rhodan]

3 hours ago, MPDerksen said:

However, my rocket is SO topheavy that it tips over almost immediately on liftoff.

Topheavy would actually be good, bottom heavy is less stable. You CoM is visibly not very high. Your main problem is that you have too much stuff that adds drag above your CoM.

Don't attach SRBs somewhere in the middle, but lower and hope that the CoM will rise above the tips of the SRBs. And try to attach them in the north-south axis if you are going for an equatorial launch. And like the others said: smaller fairing at the top and some fins at the bottom.

[Jan Kerjann]

bewing is right, you need to launch two vessels to get the docking thing done. Suggestion: One ship with a lander and enough fuel to get the stranded kerbal and then get to a rendevous on an equatorial münar orbit, and another ship with three seats (one for it's pilot, one for the lander's pilot, one for the kerbal that wanted to find out what Pink Floyd was singing about) and the gear to land back home on the bright blue pixel. And do not forget to transfer the science to the return ship.

[Tyko]

You can also set the fuel flow priority in your two big tanks so the bottom one drains first - i.e. set the bottom tank to a higher priority number. This will keep the weight up front for longer.

[Aegolius13]

14 hours ago, Harry Rhodan said:

Topheavy would actually be good, bottom heavy is less stable.

See also: https://en.wikipedia.org/wiki/Pendulum_rocket_fallacy.  Robert Goddard himself got tripped up by the same issue.

Lots of good advice above (move SRBs down, make the fairing narrower).   Along the same lines, if you need struts on your SRBs, put them at the bottom so they help move the center of drag down. (Struts are quite draggy).  It's also generally helpful to have your decoupler closer to the top of the SRB, as that makes them detach more cleanly.  

Using advanced nose cones on the SRBs might also help push the center of drag back a smidge -- at least if the top of the SRBs remains higher than the center of mass.

 

 

[MPDerksen]

right, so then change of plans, like Jan suggests.  This lander has a poodle engine under it, and a dV of 967 (according to the VAB readout).  I'm trying to work backwards on the dV map.  If I want to have this craft in orbit, I'll need 580 dV to get safely down, and another 580 to get back up, correct?  Plus, extra since I'm new-ish...
Would I add additional fuel?  A different engine?  

From there, again looking at the map, I'll need a second stage with roughly 1,200dV to leave Kerbin orbit, and enter Mun orbit.  And finally, I'll need to build a booster stage with 3,400 to get off the ground into orbit.  I'm thinking of this correct, yes?  This mission will end with the rescued Kerbal in the second seat, and in Mun orbit with my pilot.  I have RCS on it for docking with craft #2.  I know I'm reading this a bit wrong, since clearly it doesn't take 3,400 to come back down from orbit, so some explanation would be appreciated.

Last, per Harry's comment, I'm not sure which direction is North when I'm in the VAB.  and I'm having some trouble getting the SRB to attach to the clamps anywhere but in the middle.  They seem to snap in close if I try to move them from center, and then they don't later detach.  Not sure what I'm doing incorrectly.  This is happening with both the short and long clamps.

[FloppyRocket]

1 hour ago, MPDerksen said:

 

right, so then change of plans, like Jan suggests.  This lander has a poodle engine under it, and a dV of 967 (according to the VAB readout).  I'm trying to work backwards on the dV map.  If I want to have this craft in orbit, I'll need 580 dV to get safely down, and another 580 to get back up, correct?  Plus, extra since I'm new-ish...
Would I add additional fuel?  A different engine?  

From there, again looking at the map, I'll need a second stage with roughly 1,200dV to leave Kerbin orbit, and enter Mun orbit.  And finally, I'll need to build a booster stage with 3,400 to get off the ground into orbit.  I'm thinking of this correct, yes?  This mission will end with the rescued Kerbal in the second seat, and in Mun orbit with my pilot.  I have RCS on it for docking with craft #2.  I know I'm reading this a bit wrong, since clearly it doesn't take 3,400 to come back down from orbit, so some explanation would be appreciated.

Last, per Harry's comment, I'm not sure which direction is North when I'm in the VAB.  and I'm having some trouble getting the SRB to attach to the clamps anywhere but in the middle.  They seem to snap in close if I try to move them from center, and then they don't later detach.  Not sure what I'm doing incorrectly.  This is happening with both the short and long clamps.

I totally agree with starting from the lander and working backwards.

A Poodle is excessively large engine-age for this task. Since you have to boost the lander into Mun orbit from Kerbin, you want the mass to be as small as possible.

Assuming the "payload" part of the lander is the 2-Kerbal pod (dry), plus science parts, plus the docking port, batteries, solar panels, and landing legs amount to a total of 1.6 tons.

Using a rocket design tool (like "KSP Optimal Rocket Calculator"), plug in 1400 m/s for min delta v, lower the min TWR to 0.3, and select "Kerbin" for the local gravity (even though you're landing on the Mun, the Isp of the engines is measured on Kerbin), disable the "add decoupler" option, and it coughs-up this design:

• Tanks: six Oscar-B
• Engine: one Spark

Total mass: 3.05 tons, dv = 1569 m/s, TWR = 0.67 (that's plenty for launching from the Mun)

That total mass then becomes the "payload" mass for designing the rocket that gets the lander up to orbit. Its delta-v requirement is 3400+860+310=4600 m/s (approx), plus whatever margin you need for pilotage skills. I usually add 10% for that. So maybe design the booster for 5,000 m/s total.

Note that when you build the rocket in the VAB and attach the lander to it, the built-in DV tool is going to add the lander's delta-v to the total values displayed. That can be misleading. Also, set the VAB's DV tool to display "vacuum" values - the 3400 m/s standard for orbit is in vacuum values.

At least, it works for me. Others have much more experience than me.

You repeat the whole process for the "orbiter" that you need to get up there in order to bring the Kerbals back. Start from the pod that splashes down, that's your final "payload" mass. For launch you add up all the delta-v values to get into low Mun orbit. On the return trip, you need the same delta-v used to get there, except do not include the launch 3400 m/s, because you aerobrake for landing.

 

Edited March 14, 2019 by FloppyRocket

[MPDerksen]

15 minutes ago, FloppyRocket said:

Assuming the "payload" part of the lander is the 2-Kerbal pod (dry), plus science parts, plus the docking port, batteries, solar panels, and landing legs amount to a total of 1.6 tons.

Using a rocket design tool (like "KSP Optimal Rocket Calculator"), plug in 1400 m/s for min delta v, lower the min TWR to 0.3, and select "Kerbin" for the local gravity (even though you're landing on the Mun, the Isp of the engines is measured on Kerbin), disable the "add decoupler" option, and it coughs-up this design:

• Tanks: six Oscar-B
• Engine: one Spark

Total mass: 3.05 tons

That total mass then becomes the "payload" mass for designing the rocket that gets the lander up to orbit. Its delta-v requirement is 3400+860+310=4600 m/s (approx), plus whatever margin you need for pilotage skills. I usually add 10% for that. So maybe design the booster for 5,000 m/s total.

Note that when you build the rocket in the VAB and attach the lander to it, the built-in DV tool is going to add the lander's delta-v to the total values displayed. That can be misleading. Also, set the VAB's DV tool to display "vacuum" values - the 3400 m/s standard for orbit is in vacuum values.

I was incorrect.  I had a Terrier engine, sorry for that.

I'm going to slow you down just a bit so I make sure I completely understand.  When I pulled the Terrier off, and put on a Spark, my dV jumped to several thousand.  In fact, I half empty the tank, and still have 2,122m/s.  I'm not sure how to make it show vac values, like you say, so a little help there?
I'm very unclear how you would attach 6 of the little Oscar-B tanks.  I tried that below the MK2 can, in a pattern, then put the Spark, and ended up with around 1,200m/s.  Enough, probably.  Preferred?

But why would the Spark give me SO much more dV than the Terrier?  When I look at the stats, the Spark says 20kN of thrust in vac, and the Terrier is 60kN.  I think that this is a fundamental part of me learning to properly build rockets.

Last, what direction am I facing if I look directly out of the VAB?  Is that North or East?

[bewing]

The crawler that puts your rocket on the launchpad does not rotate it. So looking straight out the door is East.

And you are looking at atmospheric dV numbers. Since the Mun has no atmosphere, those are the wrong numbers to use. You need to click the button that shows vacuum dV numbers instead.

 

[MPDerksen]

56 minutes ago, bewing said:

The crawler that puts your rocket on the launchpad does not rotate it. So looking straight out the door is East.

And you are looking at atmospheric dV numbers. Since the Mun has no atmosphere, those are the wrong numbers to use. You need to click the button that shows vacuum dV numbers instead.

 

even better.  Now, lowering my fuel from the tank I have a dV of 2,150m/s but it's about the same for the Terrier and the Spark.  So the Terrier uses 3x the fuel, but produces 3x the thrust?  So how would I ever decide which to use?

I get the feeling that something here will click in my brain, the way matching orbits did, and then I'll take another big step in my ability, but right now I just scratch my head at selecting the right parts to build with.

[FloppyRocket]

@MPDerksen

In the VAB, in the lower right corner of the screen, there is a little toolbar. One of the buttons is labeled "DV". Click on it, that opens the DV options menu. There you've got three buttons that are preset for different air pressure values - sea level (the default), an altitude you can set, and "vacuum".

It always defaults to "sea level", so whenever you enter the VAB, you have to re-set it to "vacuum".

I'll take a look at your questions about the relative delta-v values shortly.

[MPDerksen]

2 minutes ago, FloppyRocket said:

@MPDerksen

In the VAB, in the lower right corner of the screen, there is a little toolbar. One of the buttons is labeled "DV". Click on it, that opens the DV options menu. There you've got three buttons that are preset for different air pressure values - sea level (the default), an altitude you can set, and "vacuum".

It always defaults to "sea level", so whenever you enter the VAB, you have to re-set it to "vacuum".

I'll take a look at your questions about the relative delta-v values shortly.

Found it and posted the corrected values above.  One reason I want to have a larger (if mostly empty) tank is my longer term plan is to leave this lander in orbit, and refuel when the rescue craft comes.  Then I'll bounce down for more science.  But I do still need to understand the difference between this design and SIX of the smaller tanks, as it relates to function.

[Plusck]

2 hours ago, MPDerksen said:

When I pulled the Terrier off, and put on a Spark, my dV jumped to several thousand.  In fact, I half empty the tank, and still have 2,122m/s.  I'm not sure how to make it show vac values, like you say, so a little help there?

Along the bottom, there is a dV symbol (next to the engineer button that turns orange when you have problems).
Click on that and switch to vacuum (button on the right).

It jumped hugely going from a Terrier to a Spark because the Spark is not all that bad in an atmosphere. The Terrier and the Poodle, on the other hand, are really optimised for vacuum and are next to useless down in the lower atmosphere.
It isn't the thrust that changes anything for dV - it's purely the rocket's ISP.
From one atmosphere to vacuum, the Spark goes from 270s to 320s ISP. Meanwhile the Terrier goes from 85s to 345s, and the Poodle from 90s to 350s, which is enormous.

edit: ok, looks like I'm about 3 posts too late with all this. Dang.

Edited March 14, 2019 by Plusck

[FloppyRocket]

@MPDerksen

In modeling the delta-v performance, the important engine factors are their mass and ISP.

(Here I'm neglecting things like efficient launch trajectories within an atmosphere)

Thrust is only really a major issue for a booster that lifts mass off of a planet. To get off the surface, the total TWR has be be > 1.0 for that body's gravity.

Once you're in orbit, one downside to low trust is the increased burn time. Long burn time makes maneuvers a little more tricky to conduct accurately, because the maneuver node system assumes the delta-v is applied instantaneously. 

Comparing the Terrier and Spark:

	Terrier	Spark
Mass (tons)	0.5	0.1
Isp (vacuum - seconds)	345	320
Thrust (KN)	60	20

So the Spark is much lighter, but has only slightly lower ISP.  It has much lower thrust, but for the mass of your lander module and the job of getting off the the Mun's surface (with its low gravity), it is sufficient.

The mass of the engine plays a big role in the amount of delta-v that is created, because the engine counts in the dry mass. For higher delta-v, you want a high wet mass and low dry mass.

The total dry mass is the payload, the mass of the engine, and the empty mass of the fuel tanks.

Example numbers to follow...

 

 

Edited March 14, 2019 by FloppyRocket
cleaning up a few over-stated details

[Plusck]

10 minutes ago, MPDerksen said:

even better.  Now, lowering my fuel from the tank I have a dV of 2,150m/s but it's about the same for the Terrier and the Spark.  So the Terrier uses 3x the fuel, but produces 3x the thrust?  So how would I ever decide which to use?

I get the feeling that something here will click in my brain, the way matching orbits did, and then I'll take another big step in my ability, but right now I just scratch my head at selecting the right parts to build with.

Quite simply, each rocket engine is "sized" for the type of loads that fit its profile. The only real exception is the Vector, which is hugely overpowered (it's heavy, but arguably not heavy enough for what it can do), basically because it's designed to fit as one of 3 on the back of a Shuttle-sized vehicle.

Mostly, though, the bigger the engine, the heavier. Also (generally): the bigger, the more efficient. Also, bigger engines generate some electricity, which can be a life-saver, but none of the smaller engines (Terrier and smaller) do.
So it makes sense to build bigger rockets, but they become expensive and harder to control. And a bigger rocket in space means a much bigger rocket on the launchpad. The idea is therefore to reduce the size until you start losing more in inefficiency than you're gaining in ease of use and ease of getting the thing into orbit.

 

So if you have 60kN thrust from a Terrier, you can produce 1g of acceleration on 6 tonnes of craft (roughly). Since the Terrier weighs half a tonne already, that means the engine is 1/12th of the total craft mass if you want a TWR of one, under Kerbin's gravity.

Since the Mun has a sixth the gravity of Kerbin, then that 6-tonne, one-Terrier craft has a local TWR of 6. That's much more than you need to land comfortably, so you could easily double the size of the craft if you have to, but I wouldn't drop the TWR any further down than that (remember 0.5g acceleration means that you need to burn for 2 minutes to slow down before you hit the ground). That means that on the Mun, a single Terrier can happily carry about 3 or 4 FL-T400 tanks, plus a couple of tonnes of lander can, batteries etc.  If you're using a Rockomax X200-16, that's 9 tonnes of fuel tank already, meaning your craft is going to be at the limit of what is comfortable with a single Terrier.
Therefore, a single Terrier can feasibly power a 2-man landing craft for the Mun using 2.5m parts... but it's about at the limit. Two Terriers would be much more comfortable. The Poodle, with 250kN thrust, is arguably a bit too much engine for this purpose, but it would give you much more comfort while carrying a very decent fuel load.

If you drop down to the Spark, you get 20kN thrust. Therefore a 2-tonne craft for 1g acceleration. The Spark only masses 0.1t, so therefore the engine only counts for 1/20th of the total mass of the vehicle (the Spark is arguably a touch too light, compared to the Terrier, and should instead mass closer to 0.2t).
Again, double the mass for 0.5g acceleration, that gives you 4 tonnes to play with. Considering that the lightest 1-man lander can is two thirds of a tonne, you can see that a single Spark is going to be about right for a one-man craft and a reasonable amount of fuel. But again, it's nearing the limit. A pair of Sparks, however, would be perfect for a one-man pod.

 

 

[FloppyRocket]

@MPDerksen

Quote

But I do still need to understand the difference between this design and SIX of the smaller tanks, as it relates to function.

Here are some numbers comparing the (Terrier + X200 tank) vs. (Spark + 6x Oscar-B tanks) in your example:

Assumptions:

• all tanks are full
• The mass of the "payload" (crew pod + science parts + batteries + legs)  is 1.6 tons
• delta-v = 9.81 * Isp(vac) * ln(wet mass/dry mass)
• assumes I've done the math right with pencil/paper and a calculator - I'm away from KSP so math and factual errors are possible. I'm using the KSP wiki for the data, and it may or may not be up-to-date.

	Terrier w/ X200 tank	Spark w/ Oscar-B tanks
Engine mass (t)	0.5	0.1
Engine Isp(vac)	345	320
Tank mass (t)	wet: 9.0 dry: 1.0	per tank :  six tanks wet: 0.23 : 1.38 dry: 0.03 : 0.18
Total mass (t)	wet: 1.6 + 0.5 + 9.0 = 11.1 dry: 1.6 + 0.5 + 1.0 = 3.1	wet: 1.6 + 0.1 + 1.38 = 3.08 dry: 1.6 + 0.1 + 0.18 = 1.88
delta-v	9.8 * 345 * ln(11.1/3.1) = 4,312 m/s	9.8 * 320 * ln(3.08/1.88) = 1,548 m/s

Conclusion:

The Spark w/6x  Oscar-B tanks gives less delta-v than the Terrier w/ X200 tank, but it has the required delta-v and is substantially lighter (about 1/4 of the mass). Lighter total mass makes it easier to launch the lander from Kerbin and put it in Mun orbit.

=================

Regarding how to install six Oscar-B tanks onto the lander: That is another question entirely, and a very good one. I can't answer it without getting on KSP and trying a few things.

Quote

So the Terrier uses 3x the fuel, but produces 3x the thrust?  So how would I ever decide which to use?

The Terrier is also 5x the mass of the Spark. Mass is always an important factor.

There are always tradeoffs. Sometimes there is more than one choice that is good enough.

 

Edited March 14, 2019 by FloppyRocket

[MPDerksen]

how about this?  Actually 7 Oscar-B tanks, and 1618m/s in vacuum.  

[MPDerksen]

Rescue completed.  Sort of.  I think I put the RCS on incorrectly, and I couldn't move forward/backward with the H/N keys.  So I was not able to actually DOCK.  To get what I could from the mission, I just EVA'd over, and left Jeb (and all my science) in LMO.  I guess I could have EVA'd the Scientist over to at least grab the data, but it was 2:00am.  I'm modifying my standard Rescue craft that I've used a dozen times, with a different/better RCS layout, and I'll practice in LKO before trying to go line up with my little lander again.  

One question:  Once I match orbits pretty close, can't I select Jeb in the lander, Target my rescue craft and turn retrograde to line up?  Then switch back to the rescue craft for the final docking?  (my RCS on the lander is also wonky).

[GoatRider]

1 hour ago, MPDerksen said:

One question:  Once I match orbits pretty close, can't I select Jeb in the lander, Target my rescue craft and turn retrograde to line up?  Then switch back to the rescue craft for the final docking?  (my RCS on the lander is also wonky).

Yes. You can dock without RCS that way, by switching back and forth to line up the docking ports on both ships, then give a light thrust on one of them to get them together.

[MPDerksen]

http://i63.tinypic.com/2nw1h5z.jpg

that was actually pretty easy!  I was even able to refuel so I can make another trip down.