Recent content by Raccoon

Hm, RK find value between to point (interpolating) to get close to real result. Higher rank then higher accuracy. And my code not random, i guess you dont understand what I ask. I dont ask anyone to debug my code, i want explanation how to calculate solo perturbed effects.

I fix error. Solo atmospheric drag acceleration: Satellite acc(double const &t, Satellite &y) { Satellite f; f.loc = y.vel; double Cd = 2.2; QMatrix3x3 T = NutationMatrix(t) * PreccessionMatrix(MJD_J2000, t); f.vel = AccelDrag(t, y.loc, y.vel, T, y, y.size, y.mass, Cd)...

With RK4 implementation of atmospheric drag will be like this? void acc(Satellite &y, Satellite &f, double Mjd_TT,const QMatrix3x3 &T) { f.loc = y.vel; double Cd = 2.2; f.vel = AccelDrag(Mjd_TT, y.loc, y.vel, T, y.size, y.mass, Cd); } void RK4step(const double& h, Satellite& y...

I understand what you mean. With only "g" height changes. But with atmospheric drag its only rising, how long must be "run"? Your graph shows only decreasing, it is solo atmospheric drag?

so if my calculating formula for height is good. Why with this results height is rising?

I dont need result be absolutly the same. I am ok with float. I just want to know why atmospheric drag doesnt work alone. And also how correclty calculate heigh?

My result with meters and second and with h=10s. position: 3642902 7633650 260207.266 velocity: -5679.18213 5420.66357 -405.655853 Position without real part because QVector hold float type not double. So they are same as yours (not to much difference). I guess my...

This code work. y + k1 * h/2 mean your y[i] + stp * .5 * k1[i] Satellite is class, it has 2 vectors with length 3. First vector for location (loc) and second for velocity (vel). with RK4step(h, sat, Mjd_TT, T) I call RK4 method, I cant in good function names. Mjd_TT - it is modificated julian...

QVector3D partialStep(const Satellite &y, const QMatrix3x3 &T, const double& satSize, const double& satMass, const double &Mjd_TT ) { double r = Magnitude2(y.loc); // MU - Earth gravity = 398602 double temp = -MU / (r * r * r); QVector3D acceleration = y.loc *...

Do you mean to check if my RK works? But why another propagator show same results? its rather confusing. Simpler integrator - euler method?

martins, as I said don't have éxample to compare. But I try another propagator like this: https://www.mathworks.com/matlabcentral/fileexchange/55167-high-precision-orbit-propagator . I disable all perturbation effects except atmospheric drag, and I get same result as mine. So I came to...

martins, vector returned by AccelDrag is opposite to mine y.vel value but values are too small from E-20 to E-17. And their influence is small too. about matrix T i am not sure, but i take it from book "O.Montenbruck, E. Gill; Satellite Orbits - Models, Methods, and Applications", may be i miss...

martins, i understand that i can simulate whatever i want, but under incorrect i mean that result will be not "good" for understanding theme. I am programmer not physicist, i need to understand how propagation work on simple example. With 2-Body was simple but other perturbation effects are hard...

Hello! I’ve write a program for orbit propagation and choose numerical integration method runge-kutta 4. I wrote only 2 perturbation effect: 2-body and atmospheric drag. Here is the code of main cycle: Satellite sat(QVector3D(7000,0,500), QVector3D(0,9,0),0.0025, 3725.0); QVector< QVector3D...