Orbit Determination Validation against GMAT

[scooterjp]

Hi, I have some questions on the propagator validation / accuracy.

I have the following Input Vector:

{
epoch: 1672562471200, // 2023-01-01T08:41:11.200Z
px: 4023.3885631619813,
py: 2295.3829421046115,
pz: -5048.048507712205,
vx: 5.505020806246804,
vy: 1.4725984148977468,
vz: 5.061557060430501
}
And this code:

let j2k = cosm.frame("EME2000");
// Set Propogator Settings
let step_size = 1000 * Unit::Millisecond;
let opts = PropOpts::with_fixed_step(14642397);
let orbital_dyn = OrbitalDynamics::two_body();
let setup = Propagator::new::<RK89>(orbital_dyn, opts);
// Propogate
let mut prop_est;
let nyx_epoch = Epoch::from_unix_milliseconds(time);
let nyx_state: Vector6<f64>= Vector6::new(
    4023.3885631619813, 
    2295.3829421046115, 
   -5048.048507712205, 
    5.505020806246804,
    1.4725984148977468,
    5.061557060430501);
let nyx_state = Orbit::cartesian_vec(
                                &nyx_state, 
                                nyx_epoch, 
                                j2k
                            );
prop_est = setup.with(nyx_state);
prop_est.for_duration(prop_time * Unit::Millisecond).unwrap();

Nyx prorogaties this over 14642397 milliseconds. It returns this Vector:

┌ ┐
│ -6157.447144534038 │
│ -2627.8513244002433 │
│ 1375.397964296156 │
│ -1.803803312859318 │
│ .3201275241241115 │
│ -7.417566789943529 │
└ ┘
When I run this same state vector through RK89 Propogation in GMAT, with, what I believe are the same settings (attached). I get this Vector:
Screenshot 2023-06-26 at 9 51 47 PM

┌ ┐
│ -6120.747335306143 │
│ -2656.585734678817 │
│ 1539.019862109160 │
│ -1.968968660807117 │
│ .2413706536038857 │
│ -7.368889935403013 │
│ -6120.747335306419 │
│ -2656.585734677576 │
│ 1539.019862111372 │
│ -1.968968660809157 │
│ .2413706536033870 │
│ -7.368889935402240 │
└ ┘
Am I setting up my NYX propagator wrong? What could be generating these deviations?

[ChristopherRabotin]

Hi there,

The main difference you're seeing is because the GMAT setup includes the JGM2 gravity field with a 4x4 but the setup un Nyx does not.

You'll want to setup the propagator similar to how it's done here:

nyx/tests/mission_design/orbitaldyn.rs

Line 893 in 2c4facb

let earth_sph_harm = HarmonicsMem::from_cof("data/JGM3.cof.gz", 12, 12, true).unwrap();

.

I hope this helps.
Chris