How to make a round-trip, HEAVY reusable interplanetary rocket?

[MerlinsMaster]

Hey Guys,

I'm trying to outfit a interplanetary mission to Eeloo, and I'm having the hardest time coming up with a design that has the delta v budget I need.

The payload is 120 tons (this is non-negotiable, I already shaved it down from 250), and I need 10,000 m/s delta v to get from LKO to Eeloo orbit and back again. And it needs to return intact.

The trouble is that, the more fuel I add, the more the law of diminishing returns rears its ugly head. I'm also unsure of how many engines is the optimal number to use for something like this.

So I'm laying the gauntlet down. Does anyone know of a way to do this? I don't mind if it's slightly cheaty (within reason). I also don't mind if the trip is slightly longer than the typical interplanetary jaunt.

Edited July 28, 2014 by MerlinsMaster

[Motokid600]

Hm... Well for starters grab some pictures of the craft. You say 120tons. Now is that 120 tons to Eeloo orbit? Or is that 120 tons part of the tug making the ejection burns?

[Wanderfound]

And, to the point: why are you taking 120 tons to Eeloo? Most of my spacecraft aren't that heavy in total, including launch vehicles.

As for engines: how patient are you? A single Ion would get you there eventually, but you'd chew your arms off in boredom during the hours-long burn. More realistically, you're probably looking at a bunch of Poodles. The optimum number would be as few as you can tolerate.

[MerlinsMaster]

Hm... Well for starters grab some pictures of the craft. You say 120tons. Now is that 120 tons to Eeloo orbit? Or is that 120 tons part of the tug making the ejection burns?

The payload mass by itself is 120 tons. The drive section would be additional.

And, to the point: why are you taking 120 tons to Eeloo? Most of my spacecraft aren't that heavy in total, including launch vehicles.

As for engines: how patient are you? A single Ion would get you there eventually, but you'd chew your arms off in boredom during the hours-long burn. More realistically, you're probably looking at a bunch of Poodles. The optimum number would be as few as you can tolerate.

Well, the ship has a crew of eight. It's a long trip, so they need a little space so they don't go space-crazy. So there's four hitchhiker containers and a mobile lab. Also, there is a docking module that holds all of the landing vehicles (three landers and two rovers). And there's a command section which consists of a command pod, two cryocabins, and ion propulsion. This way, if the mission goes pearshaped, the command section can detach from the rest of the ship, and the crew can go into cold sleep for the years-long trip home.

I like big ships, sue me. But seriously, it doesn't really make much sense to go way the heck out to Eeloo with little more than a Mk3 capsule and one hitchhiker container like most other people seem to do.

And you say Poodles are the way to go? Interesting, I hadn't thought of those for an interplanetary transfer. I would've thought it would involve mostly LV-Ns. Or maybe a combination of the two? How many engines total do you think would work best?

As far as the ion engines go, how long would that make the trip? I'm patient, but for story purposes I would prefer it didn't take more than two years each way.

[sojourner]

Ion engines would require injection burns measured in real life days for a ship that big.

[lincourtl]

As far as the ion engines go, how long would that make the trip? I'm patient, but for story purposes I would prefer it didn't take more than two years each way.

The time required to achieve the delta-v for ejection would be so great you'd need to spend years doing phased burns -- burn for a few minutes, orbit back around, burn for a few more minutes, orbit around again, repeat a few thousand times.

For 120 tons of payload, it makes much more sense to send things like landers separately, then dock in Eeloo orbit.

[Frank327]

Well, whatever you come up with, make sure to upload it. I'm pretty sure it would take a spectacular spacecraft to do it. Most people would be happy just to do a return mission to Eloo. What you're doing is like deciding to pack a refrigerator in your backpack so you can have a cold drink on the mount everest.

[lincourtl]

What you're doing is like deciding to pack a refrigerator in your backpack so you can have a cold drink on the mount everest.

Hah! I'm pretty sure there's a Kickstarter for that.

[cantab]

Well to start with you can shave off a fair chunk of dV by aerobraking at Kerbin.

Then if you want it fully reusable you will need to pack quite a lot of fuel along with a big LV-N cluster. I'd say six LV-N's minimum, and you will indeed need multiple orange or Kerbodyne tanks.

You could reduce total mass by using drop tanks, then redock new ones at Kerbin for repeat trips.

[metaphor]

You shouldn't need 10,000 m/s of delta-v to get from Kerbin orbit to Eeloo orbit and back. It should only take about 2000 to leave + 500 for plane change + 1500 for capture, and another 1500 to get back, or 5500 total (6000 with contingency). If you use LV-Ns, a 120-ton payload can be pushed by a 240-ton stage (something like 6 orange tanks and 12 LV-Ns, to keep burn times reasonable). If you leave most of your payload at Eeloo, you can get by with less.

[MerlinsMaster]

The time required to achieve the delta-v for ejection would be so great you'd need to spend years doing phased burns -- burn for a few minutes, orbit back around, burn for a few more minutes, orbit around again, repeat a few thousand times.

For 120 tons of payload, it makes much more sense to send things like landers separately, then dock in Eeloo orbit.

That's a pretty good point. I never really considered ion to be an option anyway, but at this point I'm considering anything. When I try to pile on more orange fuel tanks, the delta v seems to reach a ceiling at around 8000 m/s that I can't seem to break through.

[MerlinsMaster]

What you're doing is like deciding to pack a refrigerator in your backpack so you can have a cold drink on the mount everest.

Maybe. But that cold drink will taste pretty effin good.

[Wanderfound]

And you say Poodles are the way to go? Interesting, I hadn't thought of those for an interplanetary transfer. I would've thought it would involve mostly LV-Ns. Or maybe a combination of the two? How many engines total do you think would work best?

Ah, I forgot about the nukes; I don't like them, so I never use them. For interplanetary stuff, I use LV909's for small craft and Poodles for large. Anything with a 390 vacuum ISP, basically.

In space, ISP is all that really matters; an LV909 will push just as much mass just as far for the same amount of fuel as a Poodle. It'll just take a lot longer to do it. Ditto for single vs multiple engines: regardless of the size of your craft, one engine will always be a smidge more efficient than two of the same engine (and two will be better than three, and so on), purely because it isn't pushing the extra mass of the second engine.

if you don't have my LV-N aversion (I'm not actually anti-nuke; I used to work at a reactor. I just think they're a bit unbalanced in-game; they make long trips too easy), then the nukes are certainly the way to go. Again, a single engine is most efficient, but would be painfully slow.

The limitation isn't anything in-game; the limiting factor is how long you're willing to sit there watching a burn happen.

[MerlinsMaster]

You shouldn't need 10,000 m/s of delta-v to get from Kerbin orbit to Eeloo orbit and back. It should only take about 2000 to leave + 500 for plane change + 1500 for capture, and another 1500 to get back, or 5500 total (6000 with contingency). If you use LV-Ns, a 120-ton payload can be pushed by a 240-ton stage (something like 6 orange tanks and 12 LV-Ns, to keep burn times reasonable). If you leave most of your payload at Eeloo, you can get by with less.

Are you saying that I need 4000 to get out there, but only 1500 to get back? If so, how does that work?

[MerlinsMaster]

if you don't have my LV-N aversion (I'm not actually anti-nuke; I used to work at a reactor. I just think they're a bit unbalanced in-game; they make long trips too easy), then the nukes are certainly the way to go. Again, a single engine is most efficient, but would be painfully slow.

The limitation isn't anything in-game; the limiting factor is how long you're willing to sit there watching a burn happen.

For story purposes, I'm restricting myself from using nukes in LKO, as they are extremely radioactive (although there are no ingame effects), and always planned on having couple conventional engines to get me out of that area before engaging the LV-Ns.

As far as running on a single engine, I would prefer not to have to do numerous phased burns, as that would be needlessly complicated and there are only so many hours in the day. I really want to just make a burn at the beginning, a mid course correction half way there, and then a burn at the end.

[lincourtl]

For story purposes, I'm restricting myself from using nukes in LKO, as they are extremely radioactive...

In real life the NERVA engine was to have had scrubbers to remove fission products (see http://www.osti.gov/scitech/servlets/purl/4510638) but Kerbals being Kerbals would probably never bother with such nonsense. Besides, given the daily radiation doses experienced by Kerbals in their jobs, they're probably not overly sensitive to radiation anyway.

[MerlinsMaster]

In real life the NERVA engine was to have had scrubbers to remove fission products (see http://www.osti.gov/scitech/servlets/purl/4510638) but Kerbals being Kerbals would probably never bother with such nonsense. Besides, given the daily radiation doses experienced by Kerbals in their jobs, they're probably not overly sensitive to radiation anyway.

Interesting. On the other hand, I've spent some time on the

Atomic Rockets website reading about how a nuclear spaceship has to be designed in a shape that tapers inward toward the engine so it doesn't fall outside of the engine radiation shield's shadow.

As I understood it, it was basically talking about nuclear thermal rockets in general, not specifically the NERVA, although the NERVA does fall under that catagory.

I wonder how effective those scrubbers would be. I don't know if I'd trust that they were.

Edited July 27, 2014 by MerlinsMaster

[cantab]

Are you saying that I need 4000 to get out there, but only 1500 to get back? If so, how does that work?

The delta-V for a trip is symmetric, the same outbound as return, but it's how you provide that delta-V that introduces the asymmetry. On the return trip you can capture into LKO using drag from Kerbin's atmosphere, meaning you don't need to use your engines for that.

And if you'd rather not use the LV-N, try the KR-2L. One should be more than enough thrust, and the superlative TWR may reduce your dry mass compared to a bunch of LV-909's or aerospikes.

[MerlinsMaster]

The delta-V for a trip is symmetric, the same outbound as return, but it's how you provide that delta-V that introduces the asymmetry. On the return trip you can capture into LKO using drag from Kerbin's atmosphere, meaning you don't need to use your engines for that.

Yeah, I plan on aerobraking wherever possible, but I don't want to have to depend on that. I'd rather have more than I need, that way I'm not screwed if the aerobraking goes badly.

And if you'd rather not use the LV-N, try the KR-2L. One should be more than enough thrust, and the superlative TWR may reduce your dry mass compared to a bunch of LV-909's or aerospikes.

Hmm...I never thought of using one of those in space, figured it be way too heavy. But if I just need one, that could work.

[r_rolo1]

Are you saying that I need 4000 to get out there, but only 1500 to get back? If so, how does that work?

Remember that you need far less energy to leave Eeloo SoI than to leave Kerbin SoI and that , given that Eeloo is much further away from Sol than Kerbin, you have far less velocity change to slow down enough to get to Kerbin orbit than the oposite way ( in the same way you need 900 ish dV to get to Minmus from LKO, but only need 200-300 dV to get back ). More, remember that it is assumed that you will aerobrake in Kerbin return

Anyway, you can give this webapp a look for the dV needs . If you plan well and are willing to wait for the ultimate best windows to get there and to get back you can do the whole trip for less than 5000 dV, but as a rough sketch, the numbers metaphor gives are correct ... You definitely don't need 10k dV, 6k should be enough.

Edited July 27, 2014 by r_rolo1

[LethalDose]

It's probably not what you want to hear, but I think the best advice is to reconsider your design strategy. It doesn't need to be totally reusable (rockets and be staged, and parts can be left on Eeloo), it doesn't need to be all sent together, and there are engines you're choosing not to use.

Breaking up the payload could help, instead of sending 120t all at once. I usually send mission payloads for interplanetary missions in parts, e.g. for my last Duna mission, I sent the surface base & rover on one rocket, a station, return fuel, and lander on a second rocket, and then crew and science on a third, modular vessel (parts went up on separate rockets). Breaking the payload up into sections makes design and flying the mission much easier than a monolithic payload.

Staging and drop tanks can also increase your dV for a similar payload fraction.

[MerlinsMaster]

Remember that you need far less energy to leave Eeloo SoI than to leave Kerbin SoI and that , given that Eeloo is much further away from Sol than Kerbin, you have far less velocity change to slow down enough to get to Kerbin orbit than the oposite way ( in the same way you need 900 ish dV to get to Minmus from LKO, but only need 200-300 dV to get back ). More, remember that it is assumed that you will aerobrake in Kerbin return

I did not know that. Live and learn.

Anyway, you can give this webapp a look for the dV needs .

I know it well. But so far I've been more likely to just check out one of the delta v maps instead, because I'm still having trouble wrapping my head around porkchop plots. But I'll keep at it anyway.

If you plan well and are willing to wait for the ultimate best windows to get there and to get back you can do the whole trip for less than 5000 dV, but as a rough sketch, the numbers metaphor gives are correct ... You definitely don't need 10k dV, 6k should be enough.

Well, that should make it a bit easier.

[Kryxal]

I'd definitely consider at least a few drop tanks if you're having that much trouble ... toss a controller on, a few solar panels, and a couple of separatrons, drop the first few tanks before you hit Kerbin SoI and just deorbit them. Enjoy the "free" fuel, that should get you around 800 extra delta-v (more if you're willing to clutter things up)

[lincourtl]

Yeah, I plan on aerobraking wherever possible, but I don't want to have to depend on that.

Just remember that 1 pound second equals 4.45 newton seconds and you'll be fine.

[Jodo42]

When it comes down to it, you're going to need a big rocket. However, there's lots of ways to reduce the rocket's size.

If you wanted, you could simply muscle the thing around with a Kerbodyne tank or two and the KR-2L. It's actually pretty efficient, at a vac. ISP of 380, nearly on par with Lv909s or Poodles. This might actually be your best bet. Or, you could use clustered LV-N engines on the transfer stage. The burn times with nukes will be significantly longer, however. Plus, you've stated that you're RPing and not using nukes in Kerbin orbit; getting that much stuff on a Kerbin escape trajectory without nukes might be tricky.

What you're looking for is delta-v. The amount of delta-v to Eeloo remains constant for your given Hohmann transfer; you need as much to get a 20 ton payload to Eeloo as you do for a 120 ton one. It's just much harder to get that delta-v with the 120 ton payload, because you're pushing a lot more. But there's actually a way to reduce the overall amount of delta-v needed to get places. You've stated you basically want to do a simple Hohmann transfer with a mid-course correction, but for something that size, you may want to look into gravity assists. That's a big-bad phrase around here, even for some of the more experienced users, but they're honestly not all that difficult. Allmhuran has a

explaining how they work, but getting them to line up in space and time can be tricky. There's a few tools out there to help you get timing right, like this one or this one that's specifically oriented towards finding flyby series. Again, these programs look daunting, but they're relatively easy to read and use once you've looked them over. On the topic of cutting delta-v, as others have mentioned, aerocapture at Kerbin on the return trip is going to save you a lot and won't take hardly any delta-v from Eeloo's altitude.

You say that 120 tons is set in stone. Is it? You'd likely be able to cut a significant amount of delta-v off the return trip if you ditch those landers and rovers. Why would you want to bring the rovers back anyways? Besides, Eeloo landers need to be little more than Mün landers, which should be cheaply and easily replaceable once you've returned to Kerbin's SOI. You might also consider ditching the command section once you're ready to return, if you know you have enough delta-v (I'll get to how to know that now).

Getting 120 tons to Eeloo, and back, would be a significant challenge even for some of the best players. Take a look at the staging article on the KSP wiki if you're not sure how serial/parallel/asparagus staging works. Cut your ship design down as much as you can. Something that might be very helpful for you would be Kerbal Engineering Redux, a mod that tells you exactly how much delta-v you have to start out with in the VAB, and gives you a bunch more information in flight, including your sea-level and ground-reference altitude to a body, remaining delta-v in your rocket, remaining burn time, and much more. Mechjeb does something similar, but also has some nice autopilot features for your rovers and the transfer burns to and from Eeloo. If you decide to go with a basic Hohmann transfer, Olexander Savchuk has a wonderful site that gives you a graphical representation of burn positions, http://ksp.olex.biz/. Alternatively, Alex Moon has a more number-y site that'll tell you an exact time and date that you should leave on to get the best possible transfer, as well as letting you pick earlier or later launch dates and slower or faster transfers.

I wish you the best of luck in this very Kerbal venture. Consider posting your progress in a thread Mission Reports in an Imgur album, or make some videos and post them in the Live From Mission Control subforum. I'm sure there's many users, including myself, that would like to see this pulled off.

Welcome to the forums.

Edited July 27, 2014 by Jodo42
Additional transfer tools