2-body propagation of hyperbolic orbits

[Robert Jedicke]

Aloha oorb users.

I hope somebody can solve this issue for me.  I am developing a population of synthetic interstellar objects but I'm getting errors with 2-body propagation for the hyperbolic objects.  It does not happen for all objects.  n-body propagation works fine.

I've attached the sample input des file that fails and the config file I'm using to generate the following

(base) iso-generator$ oorb  --conf=oorb-iso-generator.conf  --task=propagation  --orb-in=iso-fail-example.des  --epoch-mjd-tt=60675.500800740738   --orb-out=iso-fail-example-propagated.des

***ERROR*** 14 Aug 2019 21:57:21UTC (Orbit / solveKeplerEquation) Newton's and Laguerre's methods were unsuccessful.

***ERROR*** 14 Aug 2019 21:57:21UTC (Orbit / propagate (multiple)) 2-body propagation using Stumpff-functions was unsuccessful.

***ERROR*** 14 Aug 2019 21:57:21UTC (Orbit / propagate (single)) TRACE BACK (single orbit)

***ERROR*** 14 Aug 2019 21:57:21UTC (oorb / propagation) TRACE BACK (165)

Thanks in advance,

Robert Jedicke

[Bryce Bolin]

Hello Robert,

I have experienced trouble propagating hyperbolic orbits with oorb
before. I found that by setting the orbit element type to cartesian
during computations fixed the issue. Have you already tried this
option? It is found under "GENERAL INVERSION PARAMETERS" around line
92 in the *.conf file.

If this is still giving you problems, you may send me some of your
problem examples and I can try propagating them on my set up.

Cheers,

Bryce

> --
> You received this message because you are subscribed to the Google Groups "OOrb" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to oorb+uns...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/oorb/972e589a-0317-4dd9-b0cd-f24b563ab7f6%40googlegroups.com.

[Bryce Bolin]

Hello Rob,

Sorry I did not see your attachments in your previous message. I ran
your sample command with your example orbit and configuration file and
got the same errors. I modified your command slightly with the
following:

oorb --conf=oorb-iso-generator.conf --task=propagation
--orb-in=iso-fail-example.des --epoch-mjd-tt=60675.500700740738
--orb-out=iso-fail-example-propagated.des

using --epoch-mjd-tt=60675.500700740738 instead of
--epoch-mjd-tt=60675.500800740738 which worked. I don't know what is
causing the problem when using --epoch-mjd-tt=60675.500800740738, but
you could run your command in n-body mode with all the perturbers
turned off except Pluto which is more or less equivalent to 2-body
mode if you absolutely must have --epoch-mjd-tt=60675.500800740738.

Cheers,

Bryce

On Thu, Aug 15, 2019 at 12:41 AM Robert Jedicke <rjed...@gmail.com> wrote:
>
>

[Mikael Granvik]

Hi Robert, Bryce,

I'm working on a (hopefully) permanent fix to this issue which stems from the
use of Stumpff functions for solving the Kepler equation.

Cheers,
Mikael

> To view this discussion on the web visit https://groups.google.com/d/msgid/oorb/CAE5xZ6pUZbJdESkZfYC%2BhcJPmR92JorXnOSP9WUoWyRfWEn5pg%40mail.gmail.com.
>

[Mikael Granvik]

Robert, Bryce,

the fix to the issue you're seeing has just been uploaded to the repo.

Let me know if you encounter a similar problem again.

Mikael

> https://groups.google.com/d/msgid/oorb/alpine.OSX.2.21.1908161104310.88572%40dhcp-fysatm-1.mobile.helsinki.fi.
>

[Bryce Bolin]

Hello Mikael,

Thank you for making this fix. I am able to run Robert's command above
that was previously failing:

oorb --conf=oorb-iso-generator.conf --task=propagation
--orb-in=iso-fail-example.des --epoch-mjd-tt=60675.500800740738

--orb-out=iso-fail-example-propagated-now-works.des

The output seems reasonable:

!!OID FORMAT q e i node argperi t_p H t_0 INDEX N_PAR MOID COMPCODE
ISO000000000001 COM 0.803027093825013E+01 0.942484259153820E+01
0.564780897437188E+02 0.254354387680145E+03 0.182058456850348E+03
0.572991075825276E+05 0.000000000000000E+00 0.60675500800741E+05 1
6 -0.100000000000000E+01 OPENORB

Thank you,

Bryce Bolin

> To view this discussion on the web visit https://groups.google.com/d/msgid/oorb/alpine.OSX.2.21.1908240016300.96019%40lm5-flafo-03.

[Robert Jedicke]

Aloha and mahalo Mikael.

I tested the new code on my synthetic ISOs.  In this particular test the old routine failed about 30% of the time on about 10,000 objects that are down-selected from a billion input objects.  The new routine has ZERO failures.  Wonderful!

Thanks again,

Robert

>> To unsubscribe from this group and stop receiving emails from it, send an email to oo...@googlegroups.com.

>> To view this discussion on the web visit https://groups.google.com/d/msgid/oorb/972e589a-0317-4dd9-b0cd-f24b563ab7f6%40googlegroups.com.
>
> --
> You received this message because you are subscribed to the Google Groups "OOrb" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to oo...@googlegroups.com.