I have a simulation of glass dissolving in water with xaffin set to SiO2(am). The SiO2 is kinetic - has a rate equation associated with it. The simulation runs fine until Q/K approaches 1. It then slows dramatically and the slow down appears to be related to changes in Q/K. After letting the simulation run over night, I aborted it and looked at the values of Q/K. These were changing by less than 1 part in 10^7 (see the table below where the first col is the days of reaction, the second is the dissolution rate of the glass and the third is the Q/K.
d - 4174.975215 R - 1.33E-17 Q/K - 0.999999219
4199.950368 1.22E-17 0.999999283
4224.926398 1.12E-17 0.999999342
I then picked up the 'system' from this simulation and started a new simulation at 4000 d. Again the simulation slowed at about 4200 d. I then increased theta to 0.6 and ran into the problem outlined in my other recent post. With theta set to 0.6, the simulation ran to its end at 5000 d, but I was unable to look at the output so I can't say how the results looked. Is this the best approach for these simulations that appear to be hung with Q/K changing very slowly (larger theta, pickup system, start where the simulation left off)?