Re: [Orekit Users] Time of propagation DSSTPropagator

[Luc Maisonobe <Luc.Maisonobe@c-s.fr>]

Le 13/04/2016 15:02, Elisabet a écrit :
> Hello,

Hi Elisabet,

> 
> I have another question about the DSST and I would be grateful for your
> answers.
> 
> I used the DSSTPropagator to propagate an orbit with perturbations due to
> Earth, Moon and Sun and the drag force. For propagating an orbit during 1
> year, I spent 45s which seems quite a lot. Whereas, whit only Earth’s
> perturbations, I spent 6s. I also tested the propagation without drag which
> gave me also 45s and only with Erath and drag which gave me 34s. I used the
> RangeKutta integrator and the DormandPrince integrator without founding
> noticeable differences between them.  The minimum step of time used was 1h and
> maximum 1 day.

You probably could use steps even larger than one day with
Formand-Prince 8(5, 3). We did some studies one or two years ago on very
long term propogations for reentry and it appeared the propagator selected
quite large steps.

> 
>  I am wondering if it is not too much time and why there is such a big
> difference when there are Moon and Sun perturbations.

The current implementation of DSST in Orekit has not yet been optimized.
We have focused (and are still working) on accuracy and stability. The
performances could certainly be improved a lot.

If someone wants to give it a try, contributions are welcome!

best regards,
Luc

> 
> Best regards,
> 
> Elisabet
>