I am assuming that you are using Flux Unity calculations (there are also other appoaches). That concept is a core of Herman Seger's original approach....... that the sum of the fluxes always adds up to one. The software will ALWAYS keep the fluxes set to one... no matter how you adjust the Al/Si ratio.
Some glaze software allows you to change the calculation approach to other options (like Insight).
There ARE "typical limits" for the amouints of Alumina and Silica that a given cone level glaze has in it. You should be able to find references for them. For example, Insight has a number of these "Target" formula tables in it for various cone ranges and types of glazes.