Retuning of GLF23

In the process of studying the effect of Shafranov shift stabilization in DIII-D and JET ITB discharges, an error was found in the n=0 radial mode damping rate which enters in the mixing length formula for the fluxes in the GLF23 model. An unphysical dependence on the MHD in through the diamagnetic drift frequency which resulted in the quasi-linear transport being periodic in . Once this was fixed it was discovered that the linear growth rates did not compare well with the linear gyrokinetic growth rates from GKS for weak or reversed magnetic shear parameters typically found within the deep core region of ITB discharges.

Recently, the fit formula for the extended ballooning mode angle has been retuned to give a better fit to the linear gyrokinetic growth rates for simulations including electrons. In particular, the magnetic shear and safety factor tuning parameters were changed and a new parameter for the MHD was added. See the PDF file gam_vs_shat-std,NCS.pdf for a comparison of the growth rates versus magnet shear at safety factor values ranging from 1-5 for the standard test case at and for the NCS case at and . The NCS parameters are the same as the standard parameters but have , , and . For the standard parameters, the retuned model gives roughly the same level of agreement as the original model. However, the retuned model yields much better agreement at weak and reversed magnetic shear for the NCS parameters. Near agreement with the GKS growth rates begins to be rapidly lost, so we have placed a limit on within the code. If exceeds 4.0, then is set to 4.0.

Once the retuning of the growth rates was completed, the normalization of the diffusivities was performed using nonlinear GYRO ITG simulations assuming adiabatic electrons. See the accompanying PDF file chii_vs_a/Lti-std.pdf which compares versus normalized ion temperature gradient using the original and retuned GLF23 models against the GYRO results. The renormalization results in a decrease in the coefficient for the ITG/TEM modes by a factor of 2 and in increase in the coefficient for the ETG modes of a factor of 3 with respect to the original GLF23 model. The norm factor for the ETG modes was determined using a small subset of DIII-D discharges. The radial mode damping exponent adamp_gf was changed from 0.5 to 0.7 to give a better fit to the saturation at higher normalized temperature gradients. The shear multiplier is . To use the retuned version of the model, set iglf=1.