Comment (by Roland Haas):

Since the part in the ET that typically usually has “scale” is the fluid code, we document the conventional units in HydroBase’s interface.ccl: https://bitbucket.org/einsteintoolkit/einsteinbase/src/master/HydroBase/interface.ccl

Outputting this (static) conversion each run seems overkill. Some thorns (eg the TOVSolver or LORENE reader) will output some overview information about the fluid body they constructed in SI units as well as “Cactus” units.

Note that it is perfectly possible to run simulations with different units, eg a mass unit of 1e6Msun when simulating accretion onto supermassive black holes. One may have to adjust some numerical parameters here and there (eg “epsilon” for WENO schemes).

The other assumption that is present is that, eg TwoPunctures makes the implicit assumption that the mass of the black holes is ~1 when it constructs its grid. Very different black hole masses may lead to sub-optimal grids.

--
Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2350/conversion-of-units