Hi,
I ran into the issue of negative eccentricity with the BrowerLyddane Propagator when converting from osculating to mean orbit elements: From Osculating to Mean Elements with Brouwer-Lyddane Propagator, negative eccentricity (#947) · Issues · Orekit / Orekit · GitLab.
Following this advice , i attempted to use the EcksteinHechlerPropagator. However, in a test case where i start with a mean orbit → compute the corresponding osculating orbit → use computeMeanOrbit to the compute back the mean orbit, i realize that the initial mean orbit and the final computed mean orbit differ by ~300 meters in SMA.
Here is my test code:
a = 6978135
e = 0.001 # Eccentricity
i_rad = math.radians(95) # Inclination in rad
omega_rad = math.radians(100) # Argument of perigee in rad
raan_rad = math.radians(100) # Right ascension of ascending node in rad
anomaly_rad = math.radians(0) # True anomaly in rad
initial_meanOrbit = KeplerianOrbit(a, e, i_rad, omega_rad, raan_rad, anomaly_rad,
PositionAngleType.TRUE, inertialFrame, initialDate, mu)
gravity_field_degree = 10
gravity_field_order = 10
gravityProvider = GravityFieldFactory.getUnnormalizedProvider(gravity_field_degree, gravity_field_order)
propagator = EcksteinHechlerPropagator(initial_meanOrbit, gravityProvider, PropagationType.MEAN)
osc_State = propagator.propagate(initialDate)
osc_Orbit = osc_State.getOrbit()
harmonics = gravityProvider.onDate(osc_Orbit.getDate())
final_mean_orbit = propagator.computeMeanOrbit(osc_Orbit, gravityProvider, harmonics)
For this reason, I don’t feel very confident using the EcksteinHechlerPropagator.
Is there any other way of computing the mean orbit elements from the osculating elements? I am aware of the OsculatingToMeanElementsConverter. However looks like it works well with DSST propagator, but I am using a numerical propagator.