Add Heat Flux values to some regressions

[WallyMaier]

Proposed Changes

This PR will add heat flux to the values that are checked in the regression testing. This should help capture more of the code performance.

Related Work

@jtneedels noticed a possible issue of changes made in #1417. This will help avoid any potential unwanted changes.

[WallyMaier]

I removed the TIME_ITER to help with the regression scripts. Though the "SIM_VALUES" that are used in the regression don't seem to be the final simulation residual/coefficient values.

[pcarruscag]

That may not be ideal, without time iter the script will look at time step 0, we should test more than that.
With time iter it will look at the first inner iteration for the timestep you want.

[WallyMaier]

@pcarruscag @martin-g I'm not sure how the regressions should run using the AARCH64 tests. I made some changes, but everything is still passing.

[WallyMaier]

@pcarruscag I'm noticing some strange behavior with regressions. Mainly the dicrete_adjoint_axisymmetric_rans case, which I only added a screen output field and the discrete_adj_transonic_rotor which I didnt touch.

[pcarruscag]

Looks like the stator was just the intermittent hybrid AD problem that is being worked on.
As for the nozzle... it did not have an objective function specified, therefore it must have been defaulting to drag, which was zero because MARKER_MONITORING was empty, now you added something to monitor and puff, results changed.

So before the case was not testing anything 🤷 you have my blessing to do whatever you want, check if the new adjoints make some sense and update (good samaritan option), or blindly update, or don't modify the case, or delete it, up to you :)

[WallyMaier]

I have updated the discrete adjoint axisymmetric rans case regression values to match and pass. After trying to run the case, I noticed two things:
1.) Cd is extremely sensitive to both AoA and Mach (1e+13 and 1e+23). This is unrealistic, but may be due to the reservoir of the case being at Mach = 1e-9.
2.) The adjoint does not converge very well => ~4 orders of magnitude. with an ending RMS_ADJ_DENSITY of -3.