Here are chi's vs. time for the R/L_T scan, using the thesis closure coefficients. The new curve is R/L_T=4.0. Since the growth rates are so small, this run takes a long time to settle down, and hasn't reached a nice steady state yet. Running at lower R/L_T would take even longer. The time averaged chi is between 3 and 4, which puts it near Curt's linear extrapolation.
Just to be clear, this R/L_T=4 simulation is in the regime where our linear errors are large (near or below threshold), therefore, we have never claimed accuracy in this regime. This chi should actually be zero. This result does not address the issue of whether or not the errors near threshold affect the simulations well above threshold.
Back to index of cyclone pages
Back to Mike Beer's home page