The Mun and back Cheapskate Challenge

[camacju]

Another option would be to get into an almost-orbit that's high enough that you can orbit once around, before circularizing with the lander. Then glide back to KSC

[jinnantonix]

11 hours ago, ManEatingApe said:

You should definitely consider a low cost entry to the Ultimate Jool5 Challenge.

I would like to see that.  My Jool 5 "expendable" low cost entry (the only one so far) is documented here: 

I am not sure, but I believe there are no restrictions on mining (AFAIK), including no rules around mining for fuel on the launch pad, and so a re-suable entry could potentially be zero or negative cost.

 

Quote

I feel there's considerable potential in your design for further costs savings. For example it cost a lot of fuel to accelerate the whole SSTO to orbital velocity in order to go around to the KSC, but another approach would be to turn around and go back, SpaceX style and let the lander circularize around Kerbin. This would save LF/OX at the cost of some extra liquid fuel to fly back through the atmosphere, but the higher atmospheric ISP of the Rapier is in your favour here.

Agreed, and I am working on the larger lander for that option, need to add another fuel tank at least.  I think the new lander will also be suitable for a command chair entry.

Edited October 28, 2020 by jinnantonix

[camacju]

... I was trying to do a quintuple Mun flyby to save fuel, and I made KSP get a stack overflow. Maybe I'm putting too much effort into this

[camacju]

https://imgur.com/a/xpcgZPK

@ManEatingApe Here's my submission under the new rules (just because I don't feel comfortable having a top submission under old, easier rules), so I flew a combination of my 2 part and 3 part designs, spending 179.8 funds.

I'm honestly surprised that the rapier alone spends less fuel than the nuke and rapier. I guess the mass of the nuke changes the mass ratio too much in a 5 ton craft.

Edited October 28, 2020 by camacju

[jinnantonix]

20 minutes ago, camacju said:

https://imgur.com/a/xpcgZPK

@ManEatingApe Here's my submission under the new rules (just because I don't feel comfortable having a top submission under old, easier rules), so I flew a combination of my 2 part and 3 part designs, spending 179.8 funds.

I'm honestly surprised that the rapier alone spends less fuel than the nuke and rapier. I guess the mass of the nuke changes the mass ratio too much in a 5 ton craft.

@camacju - looks to me like your Kerbal is exposed during re-entry - how do you stop Val from exploding?  Is she hiding behind the battery?

Also, I thought Mechjeb was not allowed, as it affects gameplay.  @ManEatingApe Am I wrong about that?

Edited October 28, 2020 by jinnantonix

[jinnantonix]

My entry in the Command Chair category.  Total 2094 funds. 

Stage 1: Thumper with exploding fins.   Fairing and decoupler.  5 parts.
Stage 2: Command chair, Spark engine, 3x baguette tanks and adapter plate as a heat shield.  6 parts.

 

[ManEatingApe]

8 hours ago, camacju said:

Here's my submission under the new rules (just because I don't feel comfortable having a top submission under old, easier rules), so I flew a combination of my 2 part and 3 part designs, spending 179.8 funds.

I'm honestly surprised that the rapier alone spends less fuel than the nuke and rapier. I guess the mass of the nuke changes the mass ratio too much in a 5 ton craft.

While I approve of the effort to further drive down expenditure, there are a few issues with the entry:

• I don't see any screenshots in the VAB with the initial cost and no screenshots with the "Mission Summary for <craft>" dialog showing recovered cost
• The entry again violates the rule on excessive clipping - there is an engine clipped into the lander fuel tank and also (from what I can tell) a reaction wheel clipped into the fuel tank also. Since this is to allow the lander to fit within the drag occlusion of the service bay this gives an unfair advantage over entries that adhere to the rules.
• Was MechJeb autopilot/autoland functionality used? Having MechJeb/KER installed with informational only panels is fine and minor utilities like Kerbal Alarm Clock are ok, but part of this challenge is flying the mission manually.

8 hours ago, jinnantonix said:

Also, I thought Mechjeb was not allowed, as it affects gameplay.  @ManEatingApe Am I wrong about that?

When this challenge was created 4 years ago, many QOL features were not in the stock game, such as delta-v per stage information and precise maneuver node editing. For that reason I was ok with using common mods such as MechJeb and KER to display information only as a convenenience. I'm trying to strike a balance between allowing players to continue to use their favourite setup while keeping things fair to all entrants. Howvere the spirit of the challenge is that the mission is flown manually. Any autopilot assistance/autoland/auto-circularise would indeed by out of bounds.

If anyone is using other mods then please ask first - it's not possible to cover every single mod and situation.

2 hours ago, jinnantonix said:

My entry in the Command Chair category.  Total 2094 funds. 

Stage 1: Thumper with exploding fins.   Fairing and decoupler.  5 parts.
Stage 2: Command chair, Spark engine, 3x baguette tanks and adapter plate as a heat shield.  6 parts.

Very minimal design on both part count and cost and a stylish lander! I chuckled at the Kerbal chasing the rolling runaway after Mun touchdown. Added you to the leaderboard in first place in the command chair category, congrats.

Edited October 28, 2020 by ManEatingApe

[camacju]

@ManEatingApe The clipping isn't necessary for the lander to fit into the service bay; I just left the reaction wheel clipped because I forgot about it. Same with the engine.

Both sections of the craft landed at the KSC; all parts were recovered at the KSC. So I assumed that I would get a 100% recovery value. If this isn't reasonable then I'm happy to re-fly the mission.

I had Mechjeb landing autopilot enabled for the Mun landing but it wasn't managing vertical velocity correctly so I was manually flying for most of the landing. Only at the beginning, before I realized the autopilot wasn't working, and at the end, when I wanted a better suicide burn, did I use Mechjeb.

[ManEatingApe]

On 10/28/2020 at 12:56 PM, camacju said:

The clipping isn't necessary for the lander to fit into the service bay; I just left the reaction wheel clipped because I forgot about it. Same with the engine.

Then it sounds like the design can work without the clipping. Extract the tanks, reactions wheels, engines, Kerbals and whatnot from each other and you have a really solid entry.

On 10/28/2020 at 12:56 PM, camacju said:

Both sections of the craft landed at the KSC; all parts were recovered at the KSC. So I assumed that I would get a 100% recovery value. If this isn't reasonable then I'm happy to re-fly the mission.

Assumptions that you or I or anyone else make may or may not be correct. To remove any doubt, keep things fair for all contestants and to make it as straightforward for me to accept entries - the only figure that counts is what the game says. Include screenshots of the original cost in the VAB and recovered cost for each seperate component.

On 10/28/2020 at 12:56 PM, camacju said:

I had Mechjeb landing autopilot enabled for the Mun landing but it wasn't managing vertical velocity correctly so I was manually flying for most of the landing. Only at the beginning, before I realized the autopilot wasn't working, and at the end, when I wanted a better suicide burn, did I use Mechjeb.

The rule on mods was quite clear - MechJeb modifies gameplay by enabling capability that doesn't exist in the base game. The spirit of the challenge is to fly the mission manually.
If you want to have information only delta-v readout panels open or orbit information open then that's fine. Any other mods then please ask first.

Edited October 29, 2020 by ManEatingApe

[jinnantonix]

My second attempt in the re-usable category:  

Two stage craft: First stage using only an airbreathing engine achieves a high speed suborbital trajectory, reaching an altitude of about 75km.  A lightweight lunar lander is launched at Apoapsis, burns to LKO and completes the mission with Jeb planting a flag on the Mun and returning to the KSC.  The launch vehicle re-enters the atmosphere, reverses direction SpaceX style, and lands at the KSC.

Total cost in the SPH =65,436, with 62,708 and 1,211 recovered, leaving a total of 1,517   

 

 

 

 

 

 

 

 

 

[ManEatingApe]

Given the interest in the re-usable category I think a recap on how recovery works in KSP would be helpful.

This is the wiki page on recovery. Summary:

• On the runway or launchpad: 100%
• Not on the runway or launchpad but within KSC grounds: 98%
• Outside the KSC: Formula depending on distance, less the further away the craft is.

For example, here is the worst rover on all of Kerbin, costing 100,000 funds.

Halfway down the runway, I regret my life choices and recover the vessel.

Since the craft was on the runway, 100% of the costs is recovered.

In an alternate reality, I pull onto the taxiway, then decide to start over with a new life on Laythe and recover the craft.

Even though the craft is within Kerbal throwing distance of the runway, since it's not on the runway proper only 98% of the costs (2,000 funds difference in this case) is returned.

 

Edited October 29, 2020 by ManEatingApe

[ManEatingApe]

8 hours ago, jinnantonix said:

My second attempt in the re-usable category:  

Two stage craft: First stage using only an airbreathing engine achieves a high speed suborbital trajectory, reaching an altitude of about 75km.  A lightweight lunar lander is launched at Apoapsis, burns to LKO and completes the mission with Jeb planting a flag on the Mun and returning to the KSC.  The launch vehicle re-enters the atmosphere, reverses direction SpaceX style, and lands at the KSC.

Nice entry. I like the neat spaceplane first stage and SpaceX recovery strategy.

The combination of 2 Spiders + 1 Ant on the lander is a nice touch. I was curious and checked this out in the VAB with the delta-v readout. For small landers this combination gives more delta-v than a Spark and also more delta-v than 2 Spiders alone despite the slightly increased mass, while retaining gimbal control.

Out of curiosity did you consider tweaking the thrust of the Spiders lower? This would increase the relative contribution from the Ant and overall ISP of the combination, perhaps enough to shrink or ditch a fuel tank.

8 hours ago, jinnantonix said:

Total cost in the SPH =65,436, with 62,708 and 1,211 recovered, leaving a total of 1,517  

So I have some good news...and some bad news (with a silver lining).

Good news: I've added you to the leaderboard with a score of 1,517 based on the recovery figures.
Bad news: At the end of the landing of the SSTO first stage, you did a little swoop to park front of the SPH. However this moved the craft from the runway (with 100% recovery costs) to the KSC grounds (with 98% recovery costs) thereby swindling yourself out of some recovery bonus.
Silver lining: Your score could be lower...considerably lower,  even with the same design.

Here's a still from your video, for one last suggestion...
It would be very Kerbal and perfectly in keeping with the spirit of the challenge to roll your lander on its side and have the Kerbal push it the rest of the way to the KSC for full cost recovery (if you have the patience)

 

Edited October 29, 2020 by ManEatingApe

[jinnantonix]

13 hours ago, ManEatingApe said:

Out of curiosity did you consider tweaking the thrust of the Spiders lower? This would increase the relative contribution from the Ant and overall ISP of the combination, perhaps enough to shrink or ditch a fuel tank.

So I have some good news...and some bad news (with a silver lining).

[snipped]

 

Wow, thanks for the tips, appreciate your experience.  I might do this mission again taking your advice.  

Edited October 30, 2020 by jinnantonix

[camacju]

Honestly a pretty good lander for this challenge, in terms of cost and ease of flying, consists of one ant engine, two oscar tanks, a reaction wheel and a battery.

From LKO you can complete a Mun landing and return with about 300 m/s left. The electrical systems aren't strictly necessary but they make the flying considerably easier.

Unrelated question - If I recover all parts of my craft on the runway, will I be able to use only fuel costs? There won't be any difference and the fuel cost will be an accurate measure; I just like keeping all my ksp forum challenges in the same save. @ManEatingApe

Edited November 1, 2020 by camacju

[ManEatingApe]

10 hours ago, camacju said:

<snip>... Unrelated question - If I recover all parts of my craft on the runway, will I be able to use only fuel costs? There won't be any difference and the fuel cost will be an accurate measure; I just like keeping all my ksp forum challenges in the same save. ...</snip>

As I've said above : To remove any doubt, keep things fair for all contestants and to make it as straightforward for me to accept entries - the only figure that counts is what the game says. Include screenshots of the original cost in the VAB and recovered cost for each separate component.

I'm always happy to accept entries that adhere to the rules of the challenge.

10 hours ago, camacju said:

Honestly a pretty good lander for this challenge, in terms of cost and ease of flying, consists of one ant engine, two oscar tanks, a reaction wheel and a battery.

From LKO you can complete a Mun landing and return with about 300 m/s left. The electrical systems aren't strictly necessary but they make the flying considerably easier.

Agree, and a it's also a solid design for a general purpose low gravity lander that could come in useful for an Gilly jaunt or a Jool 5 expedition.
When derping around the Kerbol system I like the convenience of reaction wheels.

Here's an interesting question - on paper (figuratively speaking, using the delta-v figures from the VAB) replacing the Ant+Reaction Wheel+Battery of your lander design with 2 Spiders increases the delta-v by only a modest amount (~20 m/s). However I wonder if the increased TWR would make the long Kerbin=>Mun transfer burn more efficient and also reduce gravity losses during the Mun landing by allowing a more last minute suicide burn. This could mean that in practice the savings are somewhat higher.

[camacju]

With the current lander it took me 623 m/s to land from a low Mun orbit, and the theoretical minimum (assuming infinite TWR and perfect timing) is 574. So I'm losing about 50 m/s to gravity losses. If you double the TWR then you'd expect to halve the gravity losses to save 25 m/s total. And I would have enough margin in this kind of mission to prefer the ease of the reaction wheels over a 25 m/s savings which represents less than 10% of my margin

[jinnantonix]

On 10/30/2020 at 3:51 AM, ManEatingApe said:

Nice entry. I like the neat spaceplane first stage and SpaceX recovery strategy.
[...]

Bad news: At the end of the landing of the SSTO first stage, you did a little swoop to park front of the SPH. However this moved the craft from the runway (with 100% recovery costs) to the KSC grounds (with 98% recovery costs) thereby swindling yourself out of some recovery bonus.
Silver lining: Your score could be lower...considerably lower,  even with the same design.

It would be very Kerbal and perfectly in keeping with the spirit of the challenge to roll your lander on its side and have the Kerbal push it the rest of the way to the KSC for full cost recovery (if you have the patience)

Such good advice, I decided to take it!  I reran the mission with a slightly lighter Munar Lander.  This time I made sure I got 100% recovery at the KSC for both craft, and what a difference it makes to the score!

Total cost in the SPH =65,398, with 63,958 and 1,219 recovered, leaving a total of 221  

 

Although I worked hard at optimising both the craft and the flight, I still feel there is room for improvement (I may run this again). 

Notes:

• The Kerbin launch to a suborbital  trajectory is definitely more efficient that using the Rapier closed cycle mode to get to LKO
• I tried continuing on around Kerbin from west to east, and it was less efficient than the SpaceX style return
• The Ant + 2 x Spider combination worked well, by reducing the thrust on the Spiders to 1% I could get gimballed direction control and an ISP = ~315 for most burns.
• The full thrust of Ant+ Spiders was just enough for the Kerbin propulsive landing (7m/s impact speed).  I also used for the initial Kerbin orbital burn, the suicide burn landing on the Mun and initial launch on the Mun.
• The Dv for the Mun landing and launch were not well optimised, room for improvement
• I had about 80 m/s Dv spare prior to landing on Kerbin, and had to burn the fuel to be light enough for the propulsive touch down
• On the LV return, I could have glided for longer, and saved a tad more fuel for recovery.

 

 

 

 

Edited November 3, 2020 by jinnantonix

[ManEatingApe]

22 hours ago, camacju said:

With the current lander it took me 623 m/s to land from a low Mun orbit, and the theoretical minimum (assuming infinite TWR and perfect timing) is 574. So I'm losing about 50 m/s to gravity losses. If you double the TWR then you'd expect to halve the gravity losses to save 25 m/s total. And I would have enough margin in this kind of mission to prefer the ease of the reaction wheels over a 25 m/s savings which represents less than 10% of my margin

Great detail on the delta-v figures. It seems the convenience of reaction wheels is worth it for 99% of cases, especially regular day to day play. Only in the rare case where the absolute maximum margins of performance need to be eked out (for example, challenges) does the balance shift.

14 hours ago, jinnantonix said:

Such good advice, I decided to take it!  I reran the mission with a slightly lighter Munar Lander.  This time I made sure I got 100% recovery at the KSC for both craft, and what a difference it makes to the score!

Total cost in the SPH =65,398, with 63,958 and 1,219 recovered, leaving a total of 221 

Excellent entry, you take a well deserved top spot on the leaderboard in the renewable category.
I laughed out loud at the Kerbal rolling the lander onto the runway.

Your mission notes are great reading and full of tips and tricks for anyone interested in maximising cost efficiency.  One suggestion for an area where you may be able to make some further savings: You could consider replacing some of the donuts with a 1.25m tank, for example an FL-T100. This should shield your Kerbal from re-entry heat without needing the extra dry mass of a structural part. Of course the tradeoff is that it would be a more functional asesthetic that isn't as shiny as your current design. 

Edited November 3, 2020 by ManEatingApe

[camacju]

Another entry, this time showing the true costs instead of using fuel cost (spoiler alert: it's the same number lol)

https://imgur.com/a/Vmb2H8f

Cost is 180 funds - 18250 in VAB and 18070 recovered

The goal here was to have a SSTO and have the lander be able to fit into a service bay. I unclipped the parts from one another as you suggested.

Potential areas for improvement:

-I could take a more efficient ascent profile - during previous runs I was able to orbit and land the SSTO with more than twice as much oxidizer remaining at the end. Flatter ascent will probably be better in this case
-Time the Mun descent better and take a flatter descent - landing on the crater rim instead of in the crater will probably save about 25 m/s both ways
-Use the EVA pack more. For example the kerbal and lander could brake into Mun orbit separately and rejoin each other in low Mun orbit. Similarly, the kerbal's EVA pack could be used for small course corrections as well as the Mun ejection. This could be a source of significant delta-v savings since the kerbal is a large portion of the lander's dry mass.
-Figure out how to go without reaction wheels and use a spider engine instead of ant engine - but the higher vacuum ISP of the ant will counteract that. I wish ksp had smaller reaction wheels as the tiny size reaction wheel is way overkill for such a small lander

@ManEatingApeis this one acceptable?

Edited November 3, 2020 by camacju

[camacju]

Also - the series of gravity assists to get to the Mun is ripped off from inspired by a post by PLAD, but I use fewer gravity assists. An 843 m/s burn gets me an orbit that just reaches Mun's orbit and pulls me up into a higher orbit. After 7 orbits I get another Mun assist that pulls me up into an orbit higher than Mun. This gives me a Mun encounter after one more orbit that is pretty much an ellipse and only requires 205 m/s to circularize instead of 250 m/s. I had enough fuel that I didn't need to do this but it's useful for saving those last droplets of fuel.

Some further thoughts:
The lander consumes 34 funds worth of fuel and uses 2543 m/s of delta-v. I bet I could get delta-v usage under 2500 with more careful flying.
The SSTO consumes 146 funds worth of fuel. 88 in rocket mode and 58 in jet mode. Since rocket mode is the biggest part of the mission, it could potentially be useful to add a spark engine or two onto the ship and use that for circularizing the orbit. It'll barely affect dry mass but it will increase specific impulse.
Another possible way to increase savings is to use the lander for the final circularization and fly an orbit-once-around with the SSTO, saving the fuel needed to boost the heavy SSTO up to orbit and then pull it back down again.

Edited November 3, 2020 by camacju

[ManEatingApe]

3 hours ago, camacju said:

<snip> Cost is 180 funds - 18250 in VAB and 18070 recovered    </snip>

Normally bargain basement space travel should be viewed with suspicion, but in this challenge it's a badge of honour!

Question because I noticed many non-informational only MechJeb panels: Did you use MechJeb assistance (or any other mod) with any part of the mission? (for example manuever creation or execution, landing, ascent, circularisation)
 

Edited November 3, 2020 by ManEatingApe

[camacju]

I only used it for plotting maneuvers and displaying delta-v @ManEatingApe

Edited November 3, 2020 by camacju

[ManEatingApe]

@camacju Added you to the leaderboard in first place.

@camacju @jinnantonix  I really like how you've taken different approaches on the spaceplane + launcher theme, constantly iterating, refining and improving with each attempt. Both designs are really ridiculously cheap!

Loads of great ideas from you both on how to further reduce cost and I feel that both approaches still have scope for further reductions.

Edited November 3, 2020 by ManEatingApe

[camacju]

Summary of a couple optimizations for the booster:
Remove one of the elevons of the SSTO and one of the intakes; add a spark engine - cost of orbiting is now 131 funds compared to 146
Remove one of the liquid fuel tanks and replace the wings with basic fins, and remove 22 oxidizer from the fuel tank - cost of orbiting is now 127 funds with a *tiny* margin of fuel left. This is probably close to the cheapest SSTO booster possible
In total I've shaved off 19 funds from the SSTO phase. Any more and I'll probably have to significantly change the design and/or the flight plan

[camacju]

https://imgur.com/a/CMVSlfQ

Okay, so...

I have no idea what happened between Mun insertion and low Mun orbit - I obviously didn't have infinite fuel enabled but I somehow gained fuel.

This mission cost me 159 funds but I feel bad about that glitch so I'll subtract another 200 m/s worth of fuel from my score, bringing mission costs to 160 funds even.

Some more further thoughts:
The lander uses 33 funds worth of fuel. If the remaining 300 m/s of margin is used for circularization, it would use 35 funds of fuel. Meanwhile the SSTO's rocket phase uses ~70 funds, and ~40 of that goes to circularization. I might be able to save a significant amount of funds by doing something like jinnantonix and not pushing all the way to orbit on the first stage. I will have to use some rocket on the booster however - otherwise the lander won't fit in the service bay anymore.

Another possible way to absolutely crash the challenge is to make a magic wing heat shield prop plane. Then the entire cost of the mission would be 35-36 funds. Since I can't make a good stock prop for the life of me I'll leave that opening to someone else.

@ManEatingApe

Edited November 6, 2020 by camacju