Raw calcium bentonite swells little. Soda ash replaces exchangeable Ca2+ on the montmorillonite layers with Na+; the Ca precipitates as CaCO3, which pulls the exchange to completion.
Na2CO3 + Ca-clay -> 2 Na-clay + CaCO3(s) 1 g Na2CO3 = 2000 / 105.99 = 18.87 meq X_inf = min( Ca_ex , eta * dose_meq ) X(t) = X_inf (1 - exp(-t/tau)), tau = 40 h * (12/w)^1.3 f_Na = (Na_0 + X) / CEC swell = 6 + 26 f_Na^1.2 - 1.2 * excess[wt%] (mL/2 g)
Assumed clay: CEC 85 meq/100 g, exchangeable Ca 55, initial Na 8, mixing efficiency eta = 0.85. The kinetics (tau vs moisture) and the swell correlation are illustrative empirical fits, not measured data; real plants calibrate on ASTM D5890 free-swell tests. Above the stoichiometric dose extra soda adds nothing and slightly depresses swelling.
Independent checks: