search

moble / scri

Python/numba code for manipulating time-dependent functions of spin-weighted spherical harmonics on future null infinity
MIT License
18 stars 20 forks source link

Treat the Christodoulou mass as Lorentz invariant #46

Closed 10220 closed 3 years ago

10220 commented 3 years ago

The numerical waveforms produced by SpEC are made dimensionless by scaling by the Christodoulou mass computed early in the simulation. This value should be treated as Lorentz invariant. This PR only needs to change the transformation for WaveformModes and WaveformGrid objects because the AsymptoticBondiData interface assumes the quantites are dimensionful (or that the Christodoulou mass is Lorentz invariant).

The primary argument for this is from @moble. The following plot shows the value of the Christodoulou mass over the course of the simulation along with a prediction from arxiv:gr-qc/0107080. Here we have the Lev6 mass value (averaged between T=2000 and 5000), and added the integrated change in mass. image

10220 commented 3 years ago

Comparing transformed waveforms with and without treating the Christodoulou mass as Lorentz invariant shows that the difference is equal to applied gamma factor appropriate to each waveform quantity.