Modeling satellite attitude for propagation of orbit

I recently obtained good results with a high-precision numerical orbital propagator using a constant effective area for drag and radiation pressure. I have now become interested in checking if the accuracy can be improved by using a variable effective area, i.e., values that vary along propagation to provide a more accurate calculation of the drag and radiation pressure to which a satellite is subject.

My understanding is that this requires knowing what is known as satellite attitude, i.e., its orientation in 3D, in addition to a model of the shape of the satellite. By combining these, we can get an estimate of the cross-section of the satellite at a given time, and an estimate of the effective surface of the satellite subject to radiation pressure.

To simplify things, let's assume a simple box-wing model, with a central body with 2 flat panels attached. Is there any simplified model that allows predicting the satellite attitude as a function of the position, speed, acceleration and forces at a given instant?

Is there a way to model such improving the accuracy of a high-precision numerical propagator by using a variable

• box-wing and attitude for modeling solar radiation pressure are pretty standard, but most drag models are so uncertain themselves that most propagators I can think of assume spherical satellites for that portion of the calculation. If you get it to work, that would be really nifty, but getting good truth data is going to be difficult, because most things with well-known orbits are well away from drag. The ISS is probably your best bet at a target to compare with, and it is also very different from a sphere. Dec 21, 2021 at 5:53
• Thanks a lot! Then, for the case of atmospheric drag, most propagators use a constant effective area, right? Since treating the satellite as a sphere, would mean that its attitude does not affect the cross section. Would you have by any chance any source for reading on how attitude modeling for solar radiation pressure is performed in the simpler case of a box-wing model? Thanks!
– Rafa
Dec 21, 2021 at 9:23