Value of `dr_tf_inboard` inconsistent in MFILE

[jonmaddock]

When selected as an optimisation parameter, the solution value of dr_tf_inboard in the MFILE is inconsistent with its itvarXXX value.

 dr_tf_inboard____________________________________________________________ (itvar007)_____________________ 1.19999999999999996e+00
 dr_tf_inboard_(final_value/initial_value)________________________________ (xcm007)_______________________ 1.00000000000000000e+00
 dr_tf_inboard_(range_normalised)_________________________________________ (nitvar007)____________________ 1.16279069767441845e-01
...
 TF_coil_inboard_leg_radial_thickness_(m)_________________________________ (dr_tf_inboard)________________ 7.61474358840885657e-01

The optimisation parameter solution value (itvar007) appears to still be at its initial value (see xcm007 == 1.0). This needs to be reconciled.

[clmould]

PROCESS/process/build.py

Lines 1684 to 1704 in aac5f4d

	# Issue #514 Radial dimensions of inboard leg
	# Calculate build_variables.dr_tf_inboard if tfcoil_variables.dr_tf_wp is an iteration variable (140)
	if any(numerics.ixc[0 : numerics.nvar] == 140):
	# SC TF coil thickness defined using its maximum (diagonal)
	if tfcoil_variables.i_tf_sup == 1:
	build_variables.dr_tf_inboard = (
	build_variables.r_tf_inboard_in
	+ tfcoil_variables.dr_tf_wp
	+ tfcoil_variables.casthi
	+ tfcoil_variables.dr_tf_nose_case
	) / np.cos(
	np.pi / tfcoil_variables.n_tf_coils
	) - build_variables.r_tf_inboard_in
	# Rounded resistive TF geometry
	else:
	build_variables.dr_tf_inboard = (
	tfcoil_variables.dr_tf_wp
	+ tfcoil_variables.casthi
	+ tfcoil_variables.dr_tf_nose_case
	)

The large tokamak file has both ixc = 140 and ixc = 13. Having 140 enabled causes the value of dr_tf_inboard (which is ixc=13) to be overwritten.

[mkovari]

The docs say

Direct parametrization: the TF radial inboard thickness width is set as an input variable : dr_tf_inboard (iteration variable 57).

Iteration variable 57 is actually dr_tf_nose_case. This error in the docs may not be relevant to the issue, but it seems worth fixing.

The docs are very clear that

dr_tf_wp and dr_tf_inboard cannot be used as iteration variables simultaneously.

[mkovari]

@ajpearcey Alex, You might want to think about how you would like to set up the TF coil geometry in the DEMO and large tokamak input files.

[clmould]

@ajpearcey one idea we had is to remove ixc = 13 from the LT in.dat files since this variable is being overwritten, do you think this would make sense?

[jonmaddock]

Interestingly, @timothy-nunn also found this in #3654 . It is my view that the model should never be aware of what optimisation parameters are active; this means the model is dependent on the type of problem being solved.