Today (May 1, 2024) NASA Administrator Bill Nelson testified before the House Science, Space, and Technology Committee that nuclear thermal propulsion would enable us to go faster to Mars. He said:

If we can go fast, we don't have to stay on the surface for, on the first time, second time, a year or two, until the planets realign. ... we could go for a short visit.

My question is, how exactly does the more advanced propulsion enable a short visit? Ideally, I'd like to understand this in terms of a chart that shows mission duration versus some metric of the propulsion system's performance, such as its total delta-v or ISP. Also, is there a propulsion improvement threshold above which a whole new (and faster) approach to the mission becomes possible?

Update: The January 24, 2024 Braintruffle video Master the Complexity of Spaceflight provides some interesting background on the complexities of spaceflight.

  • 1
    $\begingroup$ related search term: porkchop plot $\endgroup$
    – Erin Anne
    Commented May 1 at 1:33
  • $\begingroup$ Linked, though not duplicate question space.stackexchange.com/q/57568/26356 since hopefully an answer here focuses on plots for low but continuous thrust rather than the classic porkchop instantaneous version. $\endgroup$ Commented May 1 at 7:55
  • $\begingroup$ It sounds like Bill Nelson is referring to a short stay mission profile, which can have a total trip duration of about a year and a half. The drawback being that most of the time is spent in space with as little as 30 days on Mars. And it takes a lot of energy, along with a Venus flyby. Mars enthusiasts routinely shout the idea down because of the limited time for exploration and because it requires new propulsion technology which they don't want to wait for. You may want to check if my suspicion is correct about what Nelson is proposing before trying to figure out performance thresholds etc. $\endgroup$ Commented May 1 at 20:06
  • $\begingroup$ Yes, I had the same thoughts. But your comment raises a good point. Venus flyby trajectories would need to be considered to properly answer this question. Do most online porkchop plot generators simplify to just the Sun, Earth, and Mars or do they also consider the influence of other planets such as Venus and Jupiter? $\endgroup$
    – phil1008
    Commented May 1 at 20:36
  • 1
    $\begingroup$ @ErinAnne and others: Pork chop plots are not appropriate for low but continuous thrust missions. JPL calls the corresponding concept for such missions a "bacon plot", and such plots look very different from pork chop plots. $\endgroup$ Commented May 2 at 15:15

3 Answers 3


My question is, how exactly does the more advanced propulsion enable a short visit?

By enabling shorter and more flexible trips to and from Mars.

With conventional chemical propulsion, the trips to and from have to be timed exactly right. There are only a few times every other year where a trip to Mars is feasible using conventional chemical propulsion, and only a few times where a trip from Mars is feasible. This means waiting on Mars for a while until the timing is right. Vehicles that use chemical propulsion must drift (not fire thrusters) for the vast majority of the time on trips to and from Mars. This long coast period is an inherent necessity due to the low specific impulse of chemistry-based thrusters.

Nuclear powered electric propulsion has the promise of a much, much higher specific impulse. The high specific impulse electric engines come at a cost, which is the need for a lot of electric power. Solar powered electric engines have a ridiculously low thrust but an extremely high specific impulse. That's okay for lightweight uncrewed vehicles, but is not good for the much more massive vehicles needed to take humans to and from Mars. Nuclear powered vehicles might overcome that final issue.


Various comments and answers have suggested pork chop plots. Those are completely inappropriate for low but continuous thrust missions. Pork chop plots are built by assuming an impulsive burns at the start and end, a ballistic coast in between, selecting departure and arrival dates, solving Lambert's for each choice, and then using a nice plotting tool to generate contours. The assumption of two impulsive burns does not apply to low but continuous thrust missions. Pork chop plots are relatively easy to create and are not vehicle specific. The solution to Lambert's problem for a selected pair of departure and arrival dates yields the energy per unit mass needed for the departure and arrival burns.

Bacon plots (JPL's term) for low but continuous thrust look very different from pork chop plots, are rather hard to construct, and are very vehicle-specific. Figure 4 in Woolley, Ryan C., et al. "Low-thrust trajectory Bacon plots for Mars mission design." (2019). shows a bacon plot overlaid with a pork chop plot. They are indeed completely different.

  • $\begingroup$ I don't disagree, but my question is more along the lines of "...but how exactly"? For example, could JEOD be used to generate plots like I described in this comment above and would those plots reveal how the shorter duration mission profiles (e.g. perhaps ones that use a Venus flyby) only become possible with higher ISP engines? $\endgroup$
    – phil1008
    Commented May 2 at 5:54
  • $\begingroup$ @phil1008 JEOD doesn't generate plots (Trick does, as does python). JEOD also does not do vehicle-specific modeling (nor does Trick). Since the optimization for low but continuous thrust missions is very vehicle specific, that would have to be a feature added by the simulation developer. Trick does have a nice Monte Carlo capability, and that combined with JEOD and the sim-specific vehicle thrust model could be used to optimize individual trajectories and then combined over multiple Monte Carlo analyses to generate bacon plots. This however seems like overkill to me. $\endgroup$ Commented May 2 at 15:51
  • 1
    $\begingroup$ @phil1008 With regard to "but exactly how" -- low but continuous thrust significantly broadens the time windows that dictate when transfer is feasible. The windows with a burn-coast-burn strategy (with the burns being impulsive, or nearly so) are rather narrow, less than a month wide every 25 months. With low but continuous thrust vehicles the feasible windows become several months wide (or even wider). $\endgroup$ Commented May 2 at 18:39
  • 1
    $\begingroup$ You make a good point that the duration of thrust is important for mission profile. You are correct that conventional chemical rockets, when used for interplanetary transfers, use short duration thrust and that SOLAR powered electric rockets use long, low-thrust burns. I’m not sure your assumption is correct that nuclear THERMAL rockets would also be used for low-trust trajectories. $\endgroup$
    – Woody
    Commented May 2 at 20:34
  • 1
    $\begingroup$ Just like there is no such thing as a “small” nuclear explosion, I think there is a lower limit to the size of a nuclear thermal rocket. The NERVA nuclear thermal rocket (1969) had a thrust of 246,663N, about 25% of a SpaceX Merlin. $\endgroup$
    – Woody
    Commented May 2 at 20:34

Answer: The higher specific impulse of nuclear propulsion, compared with chemical rockets, allows use of interplanetary transfers which are quicker (but which require greater characteristic energy). This allows larger launch windows and more flexible mission planning. As a result, shorter (and longer) visits to Mars are possible.

The energy requirements for different interplanetary transfers are usually visualized using a Porkchop Plot, two of which are illustrated in the figure below from https://mars.nasa.gov/spotlight/porkchop-image01.html

The horizontal scale is departure date and the vertical scale is arrival date. Red lines join points with equal journey duration. Blue curves join departure/arrival pairs requiring equal characteristic energy. The smallest blue curves enclose the optimum (lowest energy) departure/arrival local minima. This is often referred to as a "launch window".

Any deviation from this local minima requires the rocket to provide more energy, reducing payload. Nuclear rockets can meet this demand for more energy, allowing departure/arrival within a larger blue curve. In effect, a larger launch window.

The higher specific impulse of nuclear propulsion over chemical rockets could allow greater payload, faster transfers or some combination thereof.

Faster transfers allow shorter visits by giving a wider choice of arrival and departure times at the Mars end. The mission is not limited to the interval between local minima.

enter image description here

  • $\begingroup$ I think that porkchop plots for both directions with some annotations that show how the total mission duration can be shortened (including the stay on the surface of Mars) might help to answer my question. $\endgroup$
    – phil1008
    Commented May 1 at 18:26
  • $\begingroup$ @phil1008 ... you are correct. This was addressed in space.stackexchange.com/questions/6301/online-porkchop-plotter $\endgroup$
    – Woody
    Commented May 1 at 19:22
  • 2
    $\begingroup$ I agree with @Heopps that a pork chop plot may not be appropriate (or at least will look very different) for a continuous thrust trajectory. Solving this problem is very difficult as it's essentially an infinite dimensional nonlinear optimization problem, and is computationally expensive, even with discretization. Papers and books are still being written on this topic to this day. $\endgroup$ Commented May 2 at 13:57
  • 1
    $\begingroup$ In comparison, a single point on a pork chop plot is easy to solve for as these plots assume two impulsive burns, one at the start to place the vehicle on a conic section that takes the vehicle from Earth to Mars and another at the end. This essentially requires solving Lambert's problem (which is easy), and then using a nice plotting tool. $\endgroup$ Commented May 2 at 13:58
  • 1
    $\begingroup$ Who keeps upvoting this obviously wrong answer? $\endgroup$ Commented May 2 at 18:35

My tuppence; I think David Hammen's answer and the ensuing comments are in the right, I am only trying to add a different perspective. This isn't a full answer in its own right, in fact it is just a series of questions addressing part of the "How ..." concept.

It might be that "nuclear thermal propulsion would enable us to go faster" just means "a bit faster, depending on the mass of the vehicle".

Looking at this another way, this is a long way around but it is meant to appeal to non-astrodynamicists:

  1. what acceleration (not thrust) would be so high that the gravity well looks just like a ripple? Let's not bother to answer it yet, it just removes an important complication from the next bit of the thought experiment.

...which is

  1. what acceleration does our vehicle need to get to Mars if it were to accelerate for half the journey and decelerate for the next half of the journey, stay there for a bit and then come back, all in some arbitrary and convenient elapsed time? Its still not a back of the envelope calc as the positions of Earth and Mars are constantly changing, ignore that too if you want.
  2. what is that thrust in real world terms for our imaginary vehicle? Is it still huge? Can we close the design for an imaginary vehicle with the thrust and Isp touted for present day nuclear schemes?
  3. what happens when we factor back that the gravity well will not be just a ripple with current projected acceleration (i.e. nuclear-(something) thrust with plausible vehicle size plus rocket equation)?

Once again,sorry no actual answers. Where we are going is that as the vehicle scales up to manage the support services needed for both crew and the high-thrust/high Isp engine, the resulting acceleration might mean we are still in what we consider the low-thrust regime, a long spiral, however there "might" be some way to contrive for it to be "a bit" faster than an impulsive transfer.

  • 1
    $\begingroup$ Regarding your point #1, the Sun's gravity at 1 au is ~6 mm/s^2. $\endgroup$
    – PM 2Ring
    Commented May 3 at 23:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.