Say we have created a bounded propagator which is basically the ephemerides valid only during a time range for a specific satellite. I’ve been wondering if it is possible that this bounded propagator we created can be used to propagate “another” satellite that carries similar orbital elements (perhaps another satellite in the same constellation).

If this is not possible, is there a way of doing this with another method that is included in Orekit?

I am not sure I understand what you mean by “similar orbital elements”, so my answer is probably flawed here.

If you mean that most parameters are equal but one or two differ by a fair amount (like for example right ascension of ascending node for a satellite in a different plane of a constellation or phasing angle within the orbit), then the ephemerides can probably not be transformed accurately from one satellite to the other. Let’s just take the ascending node as an example. If you computed the ephemeris for one plane and just expect this ephemeris to be rotated about the Z axis to reproduce the orbit on another plane, you will only have the Keplerian part correct, but not the effect of tesseral terms in the Earth gravity filed (which depend on longitude) and not the Sun and Moon 3rd bidy effects as they depend on the relative position of perturbing body, Eart and satellite.

So if you want to have something more accuracte than just Kepler plus the zonal terms, you must probably compute each ephemeris precisely.

There is a way to build a propagator by modifying an existing propagator, it is the AdapterPropagator class, but it is not suited for large offsets. It is dedicated to small variations, and was initially developped for estimating small station keeping maneuvers, hence the reference propagator was an accurate ephemeris computed by numerical propagation, and then an estimation loop was set up to add a bunch of small maneuvers and check they achieved the station-keeping goal, until convergence was reached, thus avoid recomputing the costly numerical propagation. Only one final numerical propagation was performed once the maneuvers set was known. As you seen this use case was for close enough orbits as the maneuvers were small. This is different from using a reference from one satellite in a constellation to deduce the motion of another satellite far away.