Confinement scaling for kinetic energy in tokamak L-mode plasmas

Scaling for the global energy confinement in tokamak L-mode plasmas has been studied by many authors with the use of the ITER confinement database. Dimensionally correct analysis by Lackner and Gottardi showed the L-mode confinement determined a gyro-Bohm type transport; τE ∼ B0.8 (n/P)0.6, while the analysis by Christiansen et al. indicated that a high-β collisional model can describe the L-mode confinement; τE ∼ B1.14 n0.16/P0.6 (τE is the global energy confinement time, B the magnetic field, n the plasma density, and P the heating power). The dependence of τE on ne is weak, which has been explained to be a decreased beam penetration at higher density and an increased beam ion contribution at lower density. The parameter dependence of τE is found to be not unique among different tokamak devices. Even in the JT-60 tokamak, dependence of τE on such plasma current Ip is different between for limiter discharges and for divertor discharge. The cause of these facts have not been clarified yet, and should be explained physically in order to use scaling expressions accurately for designing future tokamak reactors. In the present paper, we analyze in detail the kinetic database of the JT-60 tokamak in order to identify a satisfactory model for L-mode confinement and to clarify the cause of the above mentioned difference in JT-60 limiter and divertor discharges. (author) 4 refs.; 4 figs