Is there any reason to not use nukes for interplanetary anything?

[Laie]

It doesn't make a good space tug either, cause hydrogen isn't storable.

People keep saying that, yet in that Mars Mission Proposal it was assumed that a nuclear tug would go to Mars and back without ISRU. I wonder how. Maybe I should take that to the science labs subforum.

[Brainlord Mesomorph]

Hi guys,

I’ve been doing some experiments in the game so we can put some numbers to these ideas.

As many of you know, I wrote this tutorial, where I describe how to align your perigee point to your ejection angle. And then a 9000 km Ap and 80 km Pe, made a pretty good departure orbit. I call that Kerbin Departure Orbit or KDO.

A 9000 km apogee is about a 10 hour orbit, so you’re never more than 5 hours away from your optimal departure moment. Which means no more and then 5 m/s from your optimal delta V tops. Also, you’re safely just shy of Munar influence.

That said, we could go higher.

The Kerbin SOI goes out to about 83,350 km (twice the orbit of Mimnus) with an 80 km perigee that is a 10 (Earth) day orbit. If you can avoid running into the Mun and Mimnus. Needless to say with a 10 day orbital period you will almost certainly miss your departure window.

But as one reader of my tutorial pointed out, there is a way to precisely time hitting your perigee/departure point at the optimum departure moment. It goes something like this:

You’re at your perigee/departure point several days before departure. First, you need to know precisely how long. For the sake of argument let’s say 72 hours (a nice round number). Now you need to select an orbital period that is evenly divisible by that time. For instance you could make three 24 hour orbits in precisely that time. Or a single 72 hour orbit.

At this web page, orbital period calculator, enter your perigee and the number of hours you want this orbit to be, and it will provide you with the appropriate apogee. Now you do exactly that burn, and you should be precisely at your ejection point precisely at your optimal departure moment.

Now that I’ve explained how to do that, my conclusion is that it really isn’t worth it.

The 9000 km apogee is exactly 98 m/s short of Kerbin escape velocity. You could do all of the work I just outlined, and at MOST all you’re going to be able to do is move 98 m/s from one burn to another. In my example of a 72 hour orbit you’re only moving about 80 m/s to the earlier burn. And crossing the orbit of the Mun twice (good luck!). And for all that work what are you saving, 3 m/s?

Like I said, just not worth it.

. But I personally have trouble planning maneuvers to other planets when in another planet's SOI, because in my experience I can't see my path around the Sun until I've made a burn that puts me past escape velocity, so I'll always make a transfer burn to another planet after I leave Kerbin's SOI. Hmm... maybe if I plan a second burn it will allow me to see my solar orbit?

I totally understand what you mean. Trying to move a maneuver node around the planet Kerbin while looking at the solar system is impossible. A difference of one pixel at that scale is halfway around the world. That’s another reason why it’s critical to make your perigee point your ejection angle. Then you know where the point is around Kerbin, and when you zoom out to the entire solar system all you have to adjust the delta V. It’s much easier!

Edited February 20, 2015 by Brainlord Mesomorph

[cantab]

Plus you have to carry a massive shadow shield if the spacecraft is piloted, and keep everything within the cone it casts, including another spacecraft if you are trying to dock.

I'm not sure I "buy" this. After all, nuclear submarines have their reactors on board, and typically located amidships. While weight is less of an issue for a sub it's still a factor.

[Brainlord Mesomorph]

I'm not sure I "buy" this. After all, nuclear submarines have their reactors on board, and typically located amidships. While weight is less of an issue for a sub it's still a factor.

A nuclear sub uses a contained nuclear reactor for electricity to power propellers.

A nuclear thruster expels fusioning and/or fissioning material into space.

That said, keeping the ships mass between the crew and the nukes should be enough, and keeping windows out of line of site with exhaust streams. or Kerbals are just immune to rads!

[Red Iron Crown]

But as one reader of my tutorial pointed out, there is a way to precisely time hitting your perigee/departure point at the optimum departure moment. It goes something like this:

You’re at your perigee/departure point several days before departure. First, you need to know precisely how long. For the sake of argument let’s say 72 hours (a nice round number). Now you need to select an orbital period that is evenly divisible by that time. For instance you could make three 24 hour orbits in precisely that time. Or a single 72 hour orbit.

At this web page, orbital period calculator, enter your perigee and the number of hours you want this orbit to be, and it will provide you with the appropriate apogee. Now you do exactly that burn, and you should be precisely at your ejection point precisely at your optimal departure moment.

Now that I’ve explained how to do that, my conclusion is that it really isn’t worth it.

Or they could just use a different method for periapsis kicking, like the precomputed burns linked in my signature. Start at 11 or 15 Kerbin-days before the ideal window (with the adjusted ejection angle) and you won't miss it by more than half the period of a 100km orbit (about 16 mins).

For most transfers I'd agree that a bit of time offset doesn't matter, but Moho is troublesome. The ideal window is very narrow, only about 45 minutes or so before dV requirements start rising rapidly.

[Kibble]

keeping the ships mass between the crew and the nukes should be enough

Most of a space vehicle's mass is in its rocket fuel. I don't think you could design a spacecraft that concentrates all or even most of its dry mass between the astronauts and the nuclear reactor - especially because most of the non-payload dry mass is the engine itself, and fuel tanks - which are, by necessity hollow.

[AlexinTokyo]

A nuclear thruster expels fusioning and/or fissioning material into space.

Not so. An NTR (like the IRL NERVA or the KSP LV-N) runs a working fluid (often liquid hydrogen) through a contained nuclear reactor, unless you're talking about a Nuclear Salt Water Rocket, which does indeed look like a nightmare.

You would need some shielding, of course, but real world designs with TWRs of the order of 7:1 (still an order of magnitude worse than chemical rockets) have been built, if not flown.

The radiation concern for a NERVA-type NTR would be from containment failure in a crash or collision, or a core meltdown due to runaway fission.

[WuphonsReach]

I totally understand what you mean. Trying to move a maneuver node around the planet Kerbin while looking at the solar system is impossible. A difference of one pixel at that scale is halfway around the world. That’s another reason why it’s critical to make your perigee point your ejection angle. Then you know where the point is around Kerbin, and when you zoom out to the entire solar system all you have to adjust the delta V. It’s much easier!

Precise Node mod fixes that issue as you can set the scale then use the "1" and "3" keys on the number pad to move your node date/time backwards and forwards.

[Kibble]

The radiation concern for a NERVA-type NTR would be from containment failure in a crash or collision, or a core meltdown due to runaway fission.

Nuclear thermal rocket engines still emit deadly radiation, even if the exhaust isn't radioactive.

[tater]

Nuclear thermal rocket engines still emit deadly radiation, even if the exhaust isn't radioactive.

Shielding is not a huge issue. If you look at reference designs (for solid core, vapor core, particle bed, whatever) the tankage in obviously in front, which helps provides distance so that shielding (a separate shield just forward of the actual reactor) can be minimal. There might also be structure to increase that distance.

I use NTRs in KSP, but I try to make them look reasonable and separate the living quarters.

I agree with them not being ideal ferries due to boil off (got to assume H2 propellant), I think of them as main propulsion for long planetary missions, primarily. Really, you'd want bimodal for power, too.

Some of the much more complex designs than Rover/NERVA offer substantial increases in Isp, such that more useful propellants can be considered while maintaining better Isp than chemical.

[Renegrade]

Nunless you're talking about a Nuclear Salt Water Rocket, which does indeed look like a nightmare.

That first design they mention about would have as much thrust as the space shuttle's SRB boosters, and specific impulse in the 6,700s range. 6,700s. Buh.

SRB thrust at ion levels of specific impulse.... :S

The design sounds insane, but it's something Jeb would definitely like.

[Laie]

As to shielding, distance is probably your best bet. Having a grain of radioactive matter on your skin (or even inhaling/imbibing it) is a problem; being a hundred meters from an operating 1GW reactor is unremarkable (OK, I'd certainly find it remarkable if I came so close, and the security guards as well, I presume -- but that's not the kind trouble we're talking about).

@Red, Brainlord: how do you determine where your PE has to be days in advance? The best method I found is to plan an actual maneuver. But you keep talking about phase angles -- do you hold a protractor to the screen or what?

Edited February 20, 2015 by Laie

[tater]

^^^ Particularly since you need shielding for ambient radiation in space, anyway (not to mention a storm cellar).

[Kibble]

As to shielding, distance is probably your best bet. Having a grain of radioactive matter on your skin (or even inhaling/imbibing it) is a problem; being a hundred meters from an operating 1GW reactor

I just wish we had extensible hundred-meter booms!

[tater]

I just wish we had extensible hundred-meter booms!

So true. More stuff in general that folds up or telescopes.

[Red Iron Crown]

Red, Brainlord: how do you determine where your PE has to be days in advance? The best method I found is to plan an actual maneuver. But you keep talking about phase angles -- do you hold a protractor to the screen or what?

I use mods to determine the ejection angle of the node, and I calculate how much offset based on how fast Kerbin is revolving around the sun and the periods of the interim orbits. I suppose if one wanted to stay pure stock a physical protractor on the screen would do.

I believe Brainlord's method is to use Kerbin's speed of revolution to calculate the number of days ahead that the node should be placed at 90 or -90 degrees, which is much easier to gauge by eye or by using a ruler. This method works better in stock but has a longer wait time before completing the final ejection burn.

[Laie]

Can precise node place a node at a certain angle? I don't use it, under the assumption that MJ's node editor has basically the same functionality as precise node.

MJ/KER will display the current angle to prograde; I tried doing a normal burn as I pass it, then using the ascending node as the X that marks the spot. Results were underwhelming.

[Red Iron Crown]

Can precise node place a node at a certain angle? I don't use it, under the assumption that MJ's node editor has basically the same functionality as precise node.

Yes, PreciseNode gives the ejection angle for a node.

[John FX]

That first design they mention about would have as much thrust as the space shuttle's SRB boosters, and specific impulse in the 6,700s range. 6,700s. Buh.

SRB thrust at ion levels of specific impulse.... :S

The design sounds insane, but it's something Jeb would definitely like.

There's just a little issue with them...

"Whether fast criticality can be controlled in a rocket engine remains an open question."

[Brainlord Mesomorph]

@Red, Brainlord: how do you determine where your PE has to be days in advance? The best method I found is to plan an actual maneuver. But you keep talking

about phase angles -- do you hold a protractor to the screen or what?

Laie,

Didn’t you read my tutorial? There is a date prior to your launch when your departure angle, kerbin and the sun form a straight line, I have math to find that date, and all you have to do is line up your perigee point with the sun.

Red, I finally understand what you’ve done!

I couldn’t figure out what exactly it was you are pre- calculating. It’s the timing! Now I understand your method. Your method provides precise timing of the ejection, but requires a mod to find your ejection angle. My method works in a stock game to find the ejection angle, but is sloppy with the ejection timing (+ or -5 hours).

But if you want precise timing, I’ve now figured out and optional step in my method that will give you a precise timing:

Assuming you’re in KDO (9000km x 80 km) we wait until T-35 hours. You’ll have slightly more than three orbits before departure, and are coming up on perigee in about five hours.

You need to compare the time of perigee to your desired (optimal) time of departure. It will be a number between 25 and 35 hours. Whenever it is, divide that number in half. Then go to his web page, orbital period calculator, and enter that is your desired orbital period, and your current perigee. It will give you at apogee. Plot that burn as one last perigee kick.

What you’ve just done, is turn your last three 10 hour orbits, into two longer orbits of a precise length of time (between 12.5 and 17.5 hours) . At the end of the second orbit, perigee, you will be precisely the right ejection angle and precisely the right time. (There I fixed it.)

What about the Mun? Yes, that’s a problem. This new orbit will certainly cross, if not be in, the path of the Mun. It’s possible that th Mun is going to be in your way, you could divide the time left by four, and create four slightly smaller orbits. but that will be a retrograde burn at a slight waste of fuel. But if don't have a choice...

I’ll be editing my tutorial this weekend’s to include that.

Edited February 21, 2015 by Brainlord Mesomorph

[Red Iron Crown]

I couldn’t figure out what exactly it was you are pre- calculating. It’s the timing! Now I understand your method. Your method provides precise timing of the ejection, but requires a mod to find your ejection angle. My method works in a stock game to find the ejection angle, but is sloppy with the ejection timing (+ or -5 hours).

Honestly, it's not 100% precise timing, it is +/- 16 minutes depending on the mean anomaly of the initial circular orbit. The ejection angle should be pretty precise, though.

My method was meant to give the player a whole number of days to offset the start of the maneuver, and a whole number of kilometers for the apoapses of the interim orbits, to make the maneuvers a little easier to plan. I also shared the spreadsheet used for my calculations for people who want to mess around with it or do their own calculations for other bodies, it would certainly be possible to tweak the final interim orbit to get the final burn to occur right at the ideal time.

[Brainlord Mesomorph]

Honestly, it's not 100% precise timing, it is +/- 16 minutes depending on the mean anomaly of the initial circular orbit. .

Well, your 16 minutes is precise compared to my 10-hours-is-close-enough method

But actually my goal was to make a parking orbit. I plan to put 10 ships in that orbit, 1 hour apart. And launch all ten (with long burns) in a single ten-hour window.

[Laie]

... all you have to do is line up your perigee point with the sun.

That is, eyeballing it. I'd not have accepted that just a few days ago, but...

...after having had ample opportunity to play with direct Laythe->Kerbin transfers in the past few days (where all planning problems are much more involved), I notice that give or take 5 degrees on periapsis placement are not a problem. As long as the eventual transfer maneuver has it's node somewhere within a minute of your eventual periapsis, the dv expenditure will remain very much the same (on the order of a tenth of a percent).

More precision is always good if you can do it, but it's not worth putting too much effort into it.

But if you want precise timing, I’ve now figured out and optional step in my method that will give you a precise timing:

No need to do pre-calculations for that; no need to risk a munar encounter either. Just don't raise the apoapsis quite as much as you could, go away, and come back a few days prior to the transfer.

Kicking by 780m/s: apo is halfway to the Mun, period ~6hrs

Kicking by 830m/s: apo is still short of the Mun, period ~11hrs

Those last 50m/s provide a lot of wiggle room even if your apoapsis stays well inside of the Mun.

Plan a maneuver at your periapsis; plan a second maneuver and shift it until it is at your desired time of transfer (ETA=08:15:23). Now add prograde to the first maneuver until the second one sits squarely on the periapsis.

[Starhawk]

Plan a maneuver at your periapsis; plan a second maneuver and shift it until it is at your desired time of transfer (ETA=08:15:23). Now add prograde to the first maneuver until the second one sits squarely on the periapsis.

You can do that!? I clearly haven't played around with the maneuver node system enough. I finally just executed my first successful gravity assist yesterday.

[Brainlord Mesomorph]

Plan a maneuver at your periapsis; plan a second maneuver and shift it until it is at your desired time of transfer (ETA=08:15:23). Now add prograde to the first maneuver until the second one sits squarely on the periapsis.

You'll have to write a much more lengthy detailed tutorial (with pictures) if you want us to understand that. .