A scaling law for energy confinement time, including radiated fraction, from Murari et al

Realistic estimation of energy confinement when radiation is important

Robust scaling laws for energy confinement time, including radiated fraction, in Tokamaks
A. Murari, E. Peluso, P. Gaudio and M. Gelfusa, 2017

If we use this, there should be no need to mess around with "cores" and "mantles".

This needs clarification, since the total input power should obviously include the fraction of the fusion power released in charged particles (alpha power in the DT case).

Two high-quality scaling laws, including the radiated fraction, have been identified by the technique; they are indicated as H1 and H2:

Murari also pastes in the IPB98(y,2), so we can assume that all the symbols have the same meaning. To find the meaning we should go back to the source used by Murari. The reference given is
McDonald D.C. et al 2007, although the original source is
ITER Physics Expert Group on Confinement and Transport et al, 1999, Nucl. Fusion 39 2175
See also Corrections to a sequence of papers in Nuclear Fusion.

Definitions
W = The total plasma energy. ...the effects of the energy in the fast, non-thermal, particles, $W_{FAST}$, must be calculated and removed.
P = loss power across the last closed flux surface (LCFS). The symbol $P_{L,Th}$ is also used. Units are MW.
$τ_E=W/P$ energy confinement time

The fact that the radiated power itself is not removed
from the estimate of the loss power also adds a further hidden
variable into the analysis. This is largely driven by the practical
consideration that estimates of the radiated power from the bulk
plasma are not available for the majority of discharges in the
database. The assumption is that this power is generally small
and/or located at the plasma edge where it does not have a large
impact on global confinement.

What does this mean? The phrase "not removed" seems to indicate that the loss power P is the total power lost from the plasma (excluding neutrons), including radiation. But this is not how the energy confinement time is usually understood. It wouldn't matter if all the pulses in the database had negligible radiation (which is how I always imagined it), but this is clearly not the case, since one of the selection criteria was that

Total radiated power is less than 60% of injected power.

@CoronelBuendia

No source files, tests, or entry points are identified in the issue. Start by locating the existing energy-confinement scaling implementation, then read the Murari et al. paper and its IPB98 references to resolve the definitions of input power, radiation, and confinement time. Done means the intended scaling law and treatment of radiated and charged-particle power are agreed and specified for implementation.