Jump to content

How do you design a nuclear tug?


How do you design a nuclear tug?  

13 members have voted

  1. 1. Nuclear is better for large payloads but there are no size large LF-only tanks. What do you do?

    • Mk1 tanks in line with engine(s)
      0
    • Radial Mk1 tanks
      0
    • Mk3 fuselage
      5
    • Regular tanks with oxidizer removed, despite poor mass ratio
      0
    • Mod with more appropriate tanks
      6
    • Other
      2
  2. 2. About what minimum Kerbin-relative TWR do you avoid going below with payload & full tanks?

    • ~.13 or lower
      4
    • ~.17
      5
    • ~.25
      2
    • ~.33
      0
    • ~.45 or higher
      2
  3. 3. What is your main way of dealing with low thrust and long burn times?

    • More NERVs or smaller payloads, higher TWR e.g. ~.5 , so not a concern in flight
      3
    • Liquid boosters for initial/ejection burn
      0
    • Lots of periapsis kicks
      7
    • Just overengineer the dV, point at maneuver, go grab a sandwich while waiting
      3
    • Other
      0


Recommended Posts

Anyone care to share their designs or thoughts on nuclear tugs?

(I found some old forum threads where people linked images of their tugs, but image links were dead. Some of the ones I found on KerbalX were clearly for show / concept art rather than for in-game practicality.)

Background:
I'm new to KSP; in sandbox mode I've landed on Mun and Minmus and orbited Duna, but I'm playing career mode now and haven't yet progressed beyond orbiting the Mun.

I hadn't thought I'd be in a position to research the nuclear tech anytime soon, but just got offered a contract to test the NERV on escape trajectory from the Mun, plus a contract to put a station in Mun orbit. Seems like a good opportunity to put the two together and then have the tug available in LKO. But I was scratching my head over what'd be a nice design.

One issue is the lack of size large LF-only tanks. The Mk1 tank is cost-effective, but a vertical stack of those is unwieldy at >3x the height things would be with size large. I don't have any of the spaceplane tech yet, so Mk3 tanks are out for this contract. Similarly, the stack bi/tri/quad adapters are out. I could use the bi- or tri-couplers:

fI60jFji_o.jpg

Oddly, counter to the surface area to volume scaling relation, Mk0 tanks have the best mass ratio, 11. Mk1 are 9, and Mk2/3 all ~8. So one can get a little more dV by clustering small tanks. Clustering Mk0 around a central Mk1 conveniently equals size large, or one can use radial Mk1:

YygFKCw8_o.jpg

But radial Mk0 leads to part count explosion (8 Mk0 = 1 Mk1) and inconvenience having to refuel all of them. Radial Mk1 is either very wide or involves part clipping. And if you do clip them inwards, it can be problematic to clip them all the way in to fit flush within 2.5m, as then whatever they're attached to will be fully occluded (so e.g. if it's a tank it cannot be refueled).

There's also navigating the TWR tradeoff. Station parts can be heavy enough that the fuel savings of nuclear make a difference even for Mun/Minmus trips, but seems like it needs more than one NERV to make reasonable burns. The above drafts with 3xNERV have just over 2000 m/s of dV when hauling a 36t Jumbo-64 tank, at 0.3 TWR. I'd be interested to hear what others have done.

Link to comment
Share on other sites

  • For tanks, I use Simple Fuel Switch, which allows all LFO tanks to be converted to LF for pretty much exactly this purpose. If that is not an option, I usually use Mk3 fuselages, because it lowers part count a lot and the lower mass ratio doesn't actually reduce Δv all that much.
     
  • I try to keep TWR in the range of 0.15 - 0.4 g (i.e. 7 - 20% of the craft's total mass is NERVs). 
     
  • I generally accept burn times up to about 1/6 of the orbital period at the current altitude. If the planned maneuver is larger than that, I try a few different approaches:
    • Initially, I try using multiple periapsis kicks. (This usually works.)
    • If the desired trajectory is so much faster than escape velocity that periapsis kicks are insufficient (e.g. Kerbin to Jool with low TWR), I attempt to find a closer large celestial body (e.g. Eve) and begin a gravity assist chain there.
    • If no such body exists, I eject as much as I can and then finish the rest outside the SOI. In my experience, this usually happens when around a relatively small object in a very fast orbit (e.g. Hale from OPM, or Ammenon from Whirligig World), in which case the losses from reduced Oberth effect are minimal.

Certainly not all of these approaches will work for everyone, but I would highly recommend Simple Fuel Switch.

 

Link to comment
Share on other sites

I also use simple fuel switch, it's a small mod that makes a big difference when it comes to using nuclear. without that, I used Mk1 fuel tanks.

8hnmJrw.jpg

this is the Marco Polonium, a nuclear mothership I made for a challenge to land on as many planets on limited funds. Modular design because I was limited to 30 parts by the level 1 VAB. Today I can do better, but it was a great accomplishment at the time. twr was around 0.12 fully loaded

I once tried to use Mk0, but the game lagged a lot, and I didn't gain much anyway. in the same model, I tried to make something that could land on vall, but trying to get high thrust from nervs is too inefficient.

1lP7t4T.png

wzjgmLz.png

this is flying christmas tree. It wasn't well designed. with the mk0 tanks it had 500 parts, and lagged significantly. giving it enough engines to land on vall increased dry mass so much, I calculatged I'd have been better off getting the same thrust with wolfhounds. but it does look cool

ultimately, I found a twr between 0.1 and 0.2 is a good compromise between engine mass and practicality.

when i accepted modding, I started using simple fuel switch, much better for dealing with tanks. if I needed higher thrust, I used some additional chemical engines

CXuPZr5.png

A'Tuin is one of my best motherships. to enable it to fly in space with lots of deltaV but also land on duna, I gave A'Tuin a mixed propulsion system: 54 nervs, modded as a bigger version to reduce part count, for a 5300-ton ship to give a twr somewhat lower than 0.1. Quite uncomfortable for long burns, but we're already talking 160 tons of nuclear engines here. A'Tuin already had to carry a lot of extra mass to deal with the additional challenges introduced by the kerbalism mod, I cut where I could.

for higher thrust it had 54 wolfhounds, also modded as a bigger version for reduced part count, granting a twr of 0.5, just enough to take off from duna - to land you'd want better, but on landing the fuel tanks are mostly empty. I would have used rhinos, for better thrust, but they broke too easily with kerbalism. on the downside, the game cannot calculate deltaV with the different propulsion systems, I had to make my own datasheet. And I ended up giving A'Tuin way more oxidizer than it needed, I could have gained an additonal 1 km/s of vacuum deltaV while still being able to land on Duna

qwuC0p3.png

A more simple model, the Trucker, was attached to A'Tuin to carry around landers. A mothership has to carry everything, it's very big and expensive to move, and it's better left in high orbit to depart cheaply. A lander needs high thrust, so increasing deltaV is really problematic past a certain point. Trucker is meant to carry a lander from the mothership to low orbit. it doesn't have to do anything fancy, so i used nuclear engines alone, with twr around 0.2

 

 

Link to comment
Share on other sites

I have a standardised nuclear pusher stage,  that I've used to send stuff all over the Kerbol system.

KATHVHw.jpg

That's a Rockomax 64 converted to hold just liquid fuel (I use Configurable Containers, but other mods can do the same thing) with 3 x Mk1 tanks as outriggers and holding the engines. The service bay has monoprop tanks, probe core, batteries, reaction wheel and the large solar cells. A large Clamp-o-Tron both ends allows it to be docked to the payload.  The small docking ports were planned for refuelling, but they've never been used.

On its own, it has over 9000m/s. With a crew module / lander (hitchhikers + Mk2 lander can +fuel / engines for Mun landing and takeoff) it has over 6000m/s.

The advantage of this design is that I can dock two  together (180 degrees rotated) to have twice the fuel and 6 Nervs, so long as I remember to close the antennas of the bottom stage. I also have an inline version:

SJK697d.jpg

This is a Rockomax 64 + Rockmax 8 for more or less the same amount of fuel. This allows me to stack 3 together for more push and the inline ones un-dock after the Kerbin ejection burn for a slow return (AP about the level of Minmus) to the Mun for refuelling from the ISRU plant there.

Newer build versions forego the large solar panels and have four RTGs, where as most of the old build versions have been retrofitted with three RTGs clipped into the nosecones of the outriggers.

You can see the full setup in this image of a couple of fuel tankers (two on the front with orange outriggers) being sent to Moho.

WynUNGC.jpg

To your other questions:

  • I try to keep my TWR about Kerbin above 0.2 (about 2m/s)
  • For long ejection burns, I either do a couple of Pe kicks (never quite to the Mun's orbit) for Moho or push everything to Mun orbit and refuel from the ISRU station there, then drop to Kerbin and burn at Pe for ejection.
Link to comment
Share on other sites

10 hours ago, Robin Patenall said:

I have a standardised nuclear pusher stage

I found this very interesting by comparison.  It's similar to my Mule, which is based around the Mk3 Liquid Fuel Fuselage and is powered by 4 NERV.

It lifts off with some unnecessary drag incurred by its format and with a light load of LF (3000) because it only lifts off once("No one hears your finger nails drag across the black board in space.")

mo29TWs.png

Mule carries only a small amount of OX (for RCS and to supply limited amounts to other craft.

Once the decision to tolerate atmospheric drag was initially made[1], no service bay was required and the only part that might have really benefited from one would be the RTG (80kg) which is an adequate, reliable, lightweight power supply.  (And no need to worry about its radiation levels under the circumstances!!)

The particular detail I liked about your tug is the 180-degree alignment doppel-ganging with three engines providing an anti-collision phase shift!

There is a subtle advantage in keeping three similar axis dimensions in terms of rotational inertia (Mule) but I doubt it really provides significant advantage in rotation-for-docking maneuvers.

Here's another detail I do really like about your design: A Jumbo filled with 6400 LF exceeds the capacity of the Mk3 Liquid Fuselage.  That's a tie-breaker, I think!!  You win.

You didn't say what your tug is named so, unless you tell the name, my rev'eng will be called Patenall...  :)

[1] done again, I'd put a fairing around the whole thing for launch.

 

Edited by Hotel26
Link to comment
Share on other sites

Thanks everyone for your input so far! I've been playing stock but this may be what pushes me to try a mod for the first time.  I suspected people were tolerating lower TWR than the .3 I was looking at, but was a bit surprised to find the consensus is close to half that.

Haven't gotten to launching anything yet; I've been busy with life, and still have a couple other missions to finish before I do this contract. Would still be interested in seeing other designs if anyone would like to show something off.

On 8/28/2023 at 7:52 AM, king of nowhere said:

giving it enough engines to land on vall increased dry mass so much, I calculatged I'd have been better off getting the same thrust with wolfhounds.

I started using simple fuel switch... if I needed higher thrust, I used some additional chemical engines

While cobbling together the drafts above I did have the thought that, given NERVs' ISP advantage and the countervailing fact that you can get >33x the thrust of a NERV for 3x the weight, there was room for designs that relied on NERVs for most of the dV but relied on LF+OX engines for brief thrust at key moments; it's interesting to see a design that did just that.

Edited by Anaxagoras
Link to comment
Share on other sites

18 hours ago, Anaxagoras said:

While cobbling together the drafts above I did have the thought that, given NERVs' ISP advantage and the countervailing fact that you can get >33x the thrust of a NERV for 3x the weight, there was room for designs that relied on NERVs for most of the dV but relied on LF+OX engines for brief thrust at key moments; it's interesting to see a design that did just that.

do keep in mind, however, that doing mixed thrusters forces you to use twice the engines; your engine pack will get heavier, cutting into your deltaV.

however, to have something that could have 6 km/s of vacuum deltaV while also being able to land on duna, while carrying significant payload, there was no other choice. I could have had almost 9 km/s of deltaV if I had used nuclear propulsion alone. of course, 6 km/s are enough to go everywhere with smart gravity assists, and there aren't many places where you can find water and nitrogen, so it was an obligated choice. but it was a choice that sacrificed almost 3 km/s of deltaV - and that's calculated with empty oxidizer tanks

Link to comment
Share on other sites

I tend to use Mk3 fuselages for larger payloads or longer distances (especially when I was building a Vall colony) and clusters of Mk1s for smaller stuff. For TWR, I sorta have two limits; 0.25 TWR is my "soft" limit where I'll generally try not to go any lower, and if it's under my "hard" limit of 0.2 I won't launch the craft. Depending on the destination I either use periapsis kicks or just point and go.

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...