mockingbirdnest / Principia

𝑛-Body and Extended Body Gravitation for Kerbal Space Program
MIT License
769 stars 69 forks source link

Bug reported by mhoram-kerbin #631

Closed eggrobin closed 8 years ago

eggrobin commented 9 years ago

Bug report by @mhoram-kerbin, including INFO log. Stacktrace, to be decoded:

    @   64D19399    principia__SayHello [0x64D19398+360744]
    @   64D0456F    principia__SayHello [0x64D0456E+275198]
    @   64CF9EEA    principia__SayHello [0x64CF9EE9+232569]
    @   64CFB757    principia__SayHello [0x64CFB756+238822]
    @   64CF1F18    principia__SayHello [0x64CF1F17+199847]
    @   64CEF20C    principia__SayHello [0x64CEF20B+188315]
    @   64CEF31A    principia__SayHello [0x64CEF319+188585]
    @   64CE8ED3    principia__SayHello [0x64CE8ED2+162914]
    @   64CE33F5    principia__SayHello [0x64CE33F4+139652]
    @   64CBF891    principia__AdvanceTime [0x64CBF890+80]
F0827 20:37:08.580535 27716 continuous_trajectory_body.hpp:346] Check failed: adjusted_tolerance_ < 1E6 * previous_adjusted_tolerance (+2.07970750065819120e+04 m vs. +8.36760047436097011e+03 m) Apocalypse occurred at +3.69360002000000002e+06 s, displacements are: {+2.00014081136158829e+10 m, +6.22684278540717239e+10 m, +5.16749613391143858e+08 m} {+1.99879834252745056e+10 m, +6.22796599512423477e+10 m, +5.17017022493503511e+08 m} {+1.99739502453383179e+10 m, +6.22907864793080597e+10 m, +5.17284493047730446e+08 m} {+1.99593192770084915e+10 m, +6.23017325059046860e+10 m, +5.17551811048148572e+08 m} {+1.99441220020342484e+10 m, +6.23124374580687790e+10 m, +5.17818713204804718e+08 m} {+1.99284043447352982e+10 m, +6.23228730581115799e+10 m, +5.18084809048591316e+08 m} {+1.99122072635346756e+10 m, +6.23330816137687073e+10 m, +5.18349165468758583e+08 m} {+1.98954546813748512e+10 m, +6.23433683762633743e+10 m, +5.18605943136402369e+08 m} {+1.98700990129785919e+10 m, +6.23583748846784897e+10 m, +5.18330642157473266e+08 m}, velocities are: {-4.86010381727614094e+03 m s^-1, +4.16994503606722628e+03 m s^-1, +9.90059283069209357e+01 m s^-1} {-5.08477531697844461e+03 m s^-1, +4.14530726333875737e+03 m s^-1, +9.90638698926969283e+01 m s^-1} {-5.30951147166898136e+03 m s^-1, +4.09186631483063866e+03 m s^-1, +9.90493098011526172e+01 m s^-1} {-5.52626364447118794e+03 m s^-1, +4.01245881114019585e+03 m s^-1, +9.89481211891008172e+01 m s^-1} {-5.72804405192298418e+03 m s^-1, +3.91519902359158186e+03 m s^-1, +9.87347519128410767e+01 m s^-1} {-5.91179892346620181e+03 m s^-1, +3.81684176446339870e+03 m s^-1, +9.83259207618487068e+01 m s^-1} {-6.08704133771054694e+03 m s^-1, +3.75811478870737210e+03 m s^-1, +9.73005382853895640e+01 m s^-1} {-6.38075126868301686e+03 m s^-1, +3.96375968253334349e+03 m s^-1, +9.00137943389807020e+01 m s^-1} {-1.09163845788669951e+04 m s^-1, +6.28249524832451243e+03 m s^-1, -1.99108893076852610e+02 m s^-1}
pleroy commented 9 years ago

The positions look fine, but the last two velocities (in particular, the last one) seem to indicate that the celestial makes a tight right turn and puts the pedal to the metal. That makes it impossible to find a polynomial with a good fit. Either the integrator is getting mad or the system is seriously unstable. If your celestial accelerates at 1.7g, you are not going to space today.

pleroy commented 9 years ago

It turns out that Vall is very close to colliding with Tylo. Pictures below.

We want to know when this kind of apocalypse happens, as the polynomial fitting stops working and basically produces random junk. We'll fix the root of the problem by making sure that the Jool system is in a stable resonance and that the satellites don't go all over the place.

screenshot3 screenshot4 screenshot5 screenshot6

eggrobin commented 8 years ago

Another occurence of this reported by @Norgg. https://gist.github.com/Norgg/0689315214ffc4011a0a