Implications from dimensionless parameter scaling experiments

The dimensionless parameter scaling approach is increasingly useful for predicting future tokamak performance and guiding theoretical models of energy transport. Experiments to determine the {rho}* (gyroradius normalized to plasma size) scaling have been carried out in many regimes. The electron {rho}* scaling is always ``gyro-Bohm``, while the ion {rho}* scaling varied with regime. The ion variation is correlated with both density scale length (L mode, H mode) and current profile. The ion {rho}* scaling in the low-q, H-mode regime is gyro-Bohm, which is the most favorable confinement scaling observed. New experiments in {beta} scaling and collisionality scaling have been carried out in low-q discharges in both L mode and H mode. In L mode, global analysis shows that there is a slightly unfavorable {beta} dependence ({beta}{sup {minus}0.1}) and no {nu}* dependence. In H-mode, global analysis finds a weak {beta} dependence ({beta}{sup 0.1}) and an unfavorable dependence on {nu}*. The lack of significant {beta} scaling spans the range of {beta}{sub N} from 0.25 to 2.0. The very small {beta} dependence in L mode and H mode is in contradiction with the standard global scaling relations. This contradiction in H mode may be indicative of the impact on the H-mode database of low-n tearing instabilities which are observed at slightly higher {beta}{sub N} in the {beta} scaling experiments. The measured {beta} and {nu}* scalings explain the weak density dependence observed in engineering parameter scans. It also points to the power of the dimensionless parameter approach, since it is possible to obtain a definitive size scaling from experiments on a single tokamak