Apply angular momentum correction the right way round in BH feedback

This merge request fixes the angular momentum limiter to reduce, rather than increase, the Bondi rate... It simply replaces max(t_bondi/t_visc, 1) with min(t_bondi/t_visc, 1).

Whoops :p

I'm gonna start some test boxes with this branch unless you already have @dc-bahe1?

Hi @jborrow, I have six runs with (a version of) this branch currently ongoing, all in /cosma7/data/dp004/dc-bahe1/EXL/. Two are 12.5 Mpc boxes with the XL and more efficient repositioning schemes, two the same in 25 Mpc boxes, and two the same as the 12.5 Mpc boxes, but writing much more frequent outputs that only contain BH data. The 12.5 Mpc boxes are now at z ~ 2.4; I'll send some plots soon.

For reference, here is a more detailed summary of these runs:

Great, thanks.

@jborrow running Ref 12.5 and 25 in master and this branch for comparison would be good.

Agreed, it would be a good validation and also check on random vs. systematic differences. Thanks, @jborrow!

Alright, I have four boxes running with f_E min = 0.5 and f_E max = 5.0 (in an attempt to recover the EAGLE Ref) in 25s and 12s with and without the fix.

Thanks both!

Irrespective of the tests that are currently running this needs to be fixed so I'll merge it already.

merged

mentioned in commit bb31dffd

My 12 Mpc box has finished and shows little difference in physical galaxy properties between the two.

I will let Yannick present any analysis of the black hole property changes.

See the following figures:

Thanks. Some differences in the regime M_200=10^10 - 10^11 but hard to tell in a 12Mpc box.