1. Use the Rate_limit function in the reaction rate; 2. A portion of carbon and ash can be added to the initial bed material. Here is a detailed explanation:
There is one such function in the rate of chemical reaction generated by 1.c3m:
1 Double function rate_limit (x) 2 Double Precision x 3 Rate_limit = x/(x + 1e-)4 return5End
The explanation given in manual is that this function can alleviate the problem of difficulty convergence caused by the sudden decrease of some values. As an example:
1 RATES (moisture_release_1) = & 2 610592107819465d0*rop_s (IJK,1) *x_s (IJK,1, &3 coalmoisture _1) *rate_limit (x_s (IJK,1, &4 coalmoisture_1)) * Exp (-10668847726726d0/min (t_s (IJK, &5 1), 2500d0))
This is the reaction rate of water gasification, rate_limit () in the water quality fraction, that is, the change in water quality fraction caused by the instability will be limited by the function.
I looked at the image of this function, and found that when X was very large, the function was almost level 1, the special hour was 0, and in the vicinity of the 0~1, it was not quite understood how the function was stable. But this function does work, I write in my own NOx generation reaction rate, when the function is not added, the step to 0.1*e-5 is not convergent, the reactant as a parameter in the function of the ride up, to 0.1*e-3 can be very good convergence.
2. Bed material volume fraction is 0.5, of which 0.3 is sand, 0.2 is carbon and ash. This is my previous whim so set, did not expect to converge is very helpful, when I put all the bed material for the sand after to 0.1e-5 can barely converge. Think about it, it is estimated that because of the mixing of carbon in the bed material, due to the occurrence of the reaction, the void fraction of the bed material is reduced momentarily, is advantageous to feed the mouth feed; but just speculation.
Method of Mfix simulating fluidized bed combustion to help converge