I am writing again to ask about the stellar half mass radius used in the IllustrisTNG simulation this time. I am trying to study the composition of the subhalos found at z=0 within twice the stellar half mass radius (2Rh) using the fields 'SubhaloMassInRad' and 'SubhaloMassInRadType', but I have had some problems because, for some subhalos, the amount of stellar or dark matter mass corresponding to these fields is equal to zero. Therefore, I would like to know how is this stellar half mass radius defined and if there is a way to know its value.
I have also checked how many of the subhalos found at z=0 have stellar mass using the 'SubhaloMass' field and I found out there are lots of them (approximately 4 million) that do not have any. Does this make any sense? Is it due to the simulation's resolution?
It should be true that all subhalos with nonzero stellar mass have a measured SubhaloHalfmassRadType. Only in this case can quantities be measured within "twice the stellar half mass radius". The values can then be anything, there is no guarantee that there are any stars or dark matter within this radius, so it is possible that those values would be zero.
If a subhalo has no stars, then SubhaloHalfmassRadType will be zero, and all fields such as SubhaloMassInRad should also be zero.
A subhalo with no stars i.e. a "dark subhalo" is extremely common at low masses. While essentially all massive halos will host a luminous galaxy, small halos need not. (The reason the "missing satellites problem" isn't actually a problem is the difference between these two populations.)