The exact equations the solver works for a aluminium fluoride (alf3 from h2sif6) — the same math shown in the app's "Theory" panel, not a black box.
H2SiF6+2Al(OH)3→2AlF3+SiO2+4H2O(the wet route on scrubber liquor; throws a silica by-product)
3HF+Al(OH)3→AlF3+3H2O(the free-HF route; no silica)
ξH2SiF6=XH2SiF6n˙H2SiF6,ξHF=XHFn˙HF/3
2ξH2SiF6+ξHF≤n˙Al(OH)3(else both extents scale down: hydroxide-limited)
wAlF3=m˙AlF3+m˙SiO2+m˙Al(OH)3,unreacted+m˙H2Om˙AlF3(grade: unreacted hydroxide reports to the product)
ξH2SiF6
extent of the fluosilicic-acid route [mol/s]
ξHF
extent of the free-HF route [mol/s]
XH2SiF6
fluosilicic-acid conversion [-]
XHF
HF conversion [-]
n˙H2SiF6
fluosilicic acid fed [mol/s]
n˙HF
HF fed [mol/s]
n˙Al(OH)3
aluminium hydroxide fed [mol/s]
wAlF3
product grade, mass fraction AlF3 [-]
m˙AlF3
aluminium fluoride made [kg/s]
m˙SiO2
silica reporting to the product [kg/s]
m˙Al(OH)3,unreacted
unreacted hydroxide in the product [kg/s]
m˙H2O
moisture in the product cake [kg/s]
Parameters
h2sif6_conversion and/or hf_conversion [0..1, whichever acid the liquor carries], product_moisture [0..1), silica_to_product [0..1) -- ALL required, measured per plant, never defaulted. Two routes, run on what the acid carries: H2SiF6 + 2 Al(OH)3 -> 2 AlF3 + SiO2 + 4 H2O (throws a silica by-product), and 3 HF + Al(OH)3 -> AlF3 + 3 H2O. Inlets by port: acid, hydroxide (else the solid inlet is the hydroxide). Outlets product (solid) + liquor; unreacted Al(OH)3 reports to the product and lowers its grade. NO energy balance -- AlF3's own formation enthalpy is held but aluminium hydroxide's is in no source bundled here, so the reaction heat cannot be formed. No crystallisation, so the trihydrate and its calcination are a separate dryer the flowsheet draws; no cryolite route