Gedanken experiment.

Consider a rectangular parallelepiped box. The box is closed and
isolated from the outside.

Let A be one of the faces (walls) of the box and let B be the opposite
face (wall). Let X be the direction that goes from A to B.

Suppose that, with the help of some device attached to wall A, a body of
mass m is thrown from this wall A towards wall B where it hits and
bounces back. The law of conservation of momentum predicts that the
"throw" will necessarily produce a recoil of the box
that will acquire a momentum increase in the semi direction =E2=80=93X
that equals (and cancels) the momentum along the semi direction +X
acquired by the body of mass m during the "throw".
Similarly when the body hits the wall B it will give to the box a
momentum increase (along the semi direction +X) that equals and cancels
the momentum change (along the semi direction -X) that the wall B gives
to the body making it bounce back. Therefore, on the whole, mainstream
physics predicts that the box (including all its content, i.e. the body
m and the device attached to the box that throws the mass along AB) will
not suffer any net impulse. Therefore the energy released by the device
that throws the mass cannot be used to propel the box (i.e. to change
its initial velocity relative to an inertial reference frame).

Suppose now that an intense source of neutrinos is located at the center
of the box. The source ejects neutrinos with an
"isotropy" such that averaging for the X_component of
the velocities (relative to the box) of the neutrinos, the result (i.e.
the average X_component) is null (zero). With such arrangement the
source of neutrinos does not exert reaction (recoil) forces along the
direction X.

When throwing now the body (of mass m) from wall A to wall B, so that it
passes near the source of neutrinos, it will cross a zone where there is
a big amount of flying neutrinos and some few of them will collide with
the body. In such collisions, as in all collisions, there will be
conservation of momentum. On the average of many collisions with
neutrinos, the body will receive a net impulse along the semidirection
BA, i.e. tending to decrease its speed. The collided neutrinos, on their
turn, will receive a net impulse whose average will be along the
direction AB and will exactly cancel the impulse (momentum change)
suffered by the body. But since the neutrinos are known to be able to
pass through big blocks of matter without interacting with it, there is
a very low probability that those neutrinos (those few that collided
with the body and suffered a momentum change) will collide with the
walls of the box and therefore will exit the box without transferring
any momentum to it.

On the whole, the box (including its content) will suffer a net (though
small) impulse along the direction BA.

If the body of mass m bounces (or is thrown) back from wall B to wall A
by a path that, in this case, passes far away from the neutrino source
(or passes when that source is "deactivated") and if the
process (throwing the body from A to B and back to A) is repeated many
times then it can be asserted that, in theory, it is possible to impulse
a closed, isolated system, free of external forces, using only internal
devices.

I'm not proposing a vehicle propelled by neutrinos. I'm
just asking if I got right the physics of the system described.

Carlos L

www.eterinica.net

[[Mod. note -- Your system isn't isolated -- it's emitting neutrinos
(which carry both energy and momentum). What you're doing is selectively
modulating the angular distribution of those neutrinos, so that the
outgoing neutrinos carry away a net linear momentum. (If a neutrino
collides with the body-of-mass-m and suffers a momentum change, that
means (assuming that neutrino doesn't scatter again) that that neutrino's
momentum carried out to infinity is different than it would have been
in the absence of the collision with the body-of-mass-m.)

In other words, your system is a "neutrino rocket". Like any rocket,
it can accelerate itself by ejecting stuff (in this case neutrinos)
which has a net linear momentum.
-- jt]]

El s=C3=A1bado, 14 de octubre de 2023 a las 3:48:10 UTC+2, asjch...@gmail.c=
om escribi=C3=B3:
[[Mod. note -- Quoted text trimmed. -J.T.]
> Carlos L
>
> www.eterinica.net
>
>
> [[Mod. note -- Your system isn't isolated -- it's emitting neutrinos
> (which carry both energy and momentum). What you're doing is selectively
> modulating the angular distribution of those neutrinos, so that the
> outgoing neutrinos carry away a net linear momentum. (If a neutrino
> collides with the body-of-mass-m and suffers a momentum change, that
> means (assuming that neutrino doesn't scatter again) that that neutrino's
> momentum carried out to infinity is different than it would have been
> in the absence of the collision with the body-of-mass-m.)
>
> In other words, your system is a "neutrino rocket". Like any rocket,
> it can accelerate itself by ejecting stuff (in this case neutrinos)
> which has a net linear momentum.
> -- jt]]

Thank you moderator for your comment and insight. I now understand that
the device that I have described amounts to a "neutrino rocket" and that
is not the subject of my interest.

My interest lies rather in analyzing if it is possible to give impulse to
an isolated box applying only internal physical phenomena and without
ejecting matter. Something like "paddling the vacuum". (The box should
contain of course some source of energy to sustain such hypothetical
physical phenomenon).

Carlos L

[[Mod. note -- You're basically asking if there's a way to violate
conservation of momentum. So far as we know, the answer is no.
-- jt]]