Intro fast
D - W7X, optimised for neoclassics, long pulse --> ECRH
D - Standard: Flat profiles, clamped Ti, low+flat impuritry profiles
D     --> ITG turbulence in ion+electron heat transport, impurity transport. 
D       Main ion/electron particle transport: Clearly requires pinch. Some turbulence models show a pinch, but no quantitive match with source profile.

D - Contrast: Various scenarios with density gradients. (manily Ti vs ne plot)
D    Pellets: Clear reduction of ion heat transport to neoclassical (but not electrons).
D    Same with TESPEL, boron powder and spontaneously in some low power shots.
?      (Ask Golo: Now hard evidence for same in high ne?)
 
D - NBI: Some hint at higher Ti with a little ECRH, but needed to look more closely at transport
D   - Couldn't do e/i separation, have to look at chi_eff.
D   - Yes, x4 lower chi_eff. With adding ECRH can take advantage.
ζχ   - As expected, hold so long as ne gradient.

D - Why ne gradient? Just fuelling?
D   - No, fuelling isn't enough to explain. Dynamic changes.
D   - Reduced flux kicks in at some timepoint.
D   - 2018 data: Use dynamics to see if we can see v/D trends - yes, distinct reduction in v,D at some time. Seems to be consistent with some a/Ln.
D         - No predictions for this yet! (Spanish work doesn't cover these gradient ranges)
D         - Er changes - nothing clear. Is always ion-root. Some indication of strong changes in v// in all 'high performance' scenarios - being investigated with CXRS.
D   - Carefully scanned initial density. Can well reproduce on-set times for whatever machine condition.

   - 2023: Tried to exploit to get higher Ti.
      - Introduce more ECRH.
D      - chi_effs look roughly the same story. Now with clear spontaneous back transistion! - definitely an internal change of heat transport.
      - In some cases indication of threshold behaviour on particle transport too.
      - Clearest in impurities (see Thilo).

D      - Too litte ECRH - density continues rising. Low P/n. Impurity accumulation.
D      - Too much ECRH - density collapses --> back transition.

      - Later reintroduction has less effect. (More power is still ok)
      - Configuration effects unclear - nothing much between EJM and KKM.


Q ~ ne. dT
Q = chi ne dT

chi_gb = rho^2 c / a
		rho = ion gyro tadius
		c = ion sound speed
		a = minor radius
chi_gb = sqrt(m/q) T^(3/2) / a / B^2

chi ~ chi_gb
chi = (chi/chi_gb) * chi_gb