As far as I know, for light and particles taking special relativity into account:

\begin{align} E^2 &= (T + m_0c^2)^2\\ &= p^2 c^2 + m_0^2 c^4 &\text{ (particles)}\\[1.5em] E &= p c & \text{ (photons)}\\[1.5em] F &= \frac{d\,p}{dt\phantom\,} \ne ma.^† \end{align}

If I have a bottle of hydrogen or xenon and 100% efficient and massless ion engine and light to electricity converters, I can accelerate away from a laser beam both by absorbing their momentum and by using their energy to accelerate ions back towards the source of the laser.

I think but am not sure that it is difficult to impossible to accelerate directly into the beam because 1) this comment and 2) a given amount of energy imparts more momentum to a photon than to a particle with nonzero rest mass $m_0$.


  1. Is that right? Even with 100% efficient and massless light to energy converters and ion engines, I can never accelerate directly into a beam of light?

  2. If so, for a given particle energy $T$ and rest mass $m_0$ what is the highest angle at which I can accelerate in the half-space (hemisphere) towards the laser beam, if any? Or can I only accelerate into the half-space away from it?


  • $\begingroup$ no prob! Sadly, I find the moderation here quite objectable, so to avoid me wasting emotional energy in the future, I'll delete my account on this site; it's been an honor :) $\endgroup$
    – user17550
    Commented Oct 28, 2020 at 15:40
  • $\begingroup$ @MarcusMüller thanks for you candor. I find the moderation here to be excellent, even exemplary Stack Exchange moderation, and fully adapted to the site's dedication to being welcoming to new users balanced with keeping things cordial and focused on the exploration of aspects of space exploration as well as on space itself. $\endgroup$
    – uhoh
    Commented Oct 28, 2020 at 20:03
  • $\begingroup$ @MarcusMüller If you feel the site could be improved why not express something in meta? When humans are involved there's always room for improvement, $\endgroup$
    – uhoh
    Commented Oct 28, 2020 at 20:07
  • $\begingroup$ the moderator in question just changed the wording of a question, thus clearly invalidating my answer, then without as much as a warning deleted my answer. Then I asked in a comment to the question whether, if he was as unsure about the change as he expressed to be, it was a careful moderation choice to just delete my answer instead of just downvoting it. And he just kept deleting my comments. So, nope, sorry, not wasting any effort on that one – he's the elected moderator, not gonna change him, so leaving this community :) $\endgroup$
    – user17550
    Commented Oct 28, 2020 at 20:17
  • 1
    $\begingroup$ not angry, just not willing to invest effort $\endgroup$
    – user17550
    Commented Oct 28, 2020 at 21:33

1 Answer 1


Because you are making use of fuel (reaction mass), this can only work for a finite time, but it must be possible for a while. We could imagine that we have a "largish" ship so the energy we get from the source imparts a negligible momentum. Then we could use that energy to push off from the bulk of the ship to accelerate toward the energy source. Done.

But it should work well for more realistic setups as well. As an example, I assume the ship is "large" enough that the xenon exhaust is a minor fraction of the total mass.

I receive 1MJ of photons. How hard does it "push" the ship?

$$dp = \frac{E}{c} = 0.0033 \text{kg m/s}$$

An engine can generate a relative velocity with xenon of 20 km/s. Most of the energy will go to the exhaust. How much Xenon can be accelerated from that energy with 100% efficiency?

$$m = \frac{2E}{v^2} = 0.005\text{kg}$$ $$dp = 0.005\text{kg}\ 20\text{km/s} = 100\text{kg m/s}$$

Since this is bigger, we can go in any direction.

The numbers aren't exact even with perfect efficiency (100% of the energy won't go to the exhaust, the engine isn't impulsive so some losses with $dp$, etc.) but the difference easily covers those caveats.


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.