Rheology can play an important part in residence time in a mill. Trying to classify a thixotropic slurry is most difficult since the particles are no longer acting as 'individuals' but rather as a mass that entraps the fine particles, preventing them from moving on to the next stage.

Why not re-phrase your question a bit more clearly. Are you trying to push the density to such a limit that you are grinding a paste which essentially means that you are no longer grinding! It would seem obvious that grinding in this condition is no longer grinding. Maybe your question is more how far can you push the density and still achieve acceptable performance. Note that this viscosity that you refer to may in fact change from day to day as a function of feed material such as clays.

You have a good point in re-phrasing your question. Certainly, as the ore mineralogy changes, so will the viscosity. Adding more water can help if the viscosity is high but then your mill capacity will decrease. There is always a price to pay when changes are made to a process circuit.

Ore characterisation is core to my work and one of the 'great unknowns' is what role mineralogy plays and what to extent, it is all part of the scope so thanks. You make a good point with regard to viscosity/density, I do need to address the question of limits to which densities may be pushed especially when dealing with ores that have a tendency to assume a paste-like consistency when ground.

I have seen a ball mill circuit where a cyclone cluster (instead of a thickener) was employed for dewatering of the mill product stream upstream from gold leaching, and the cyclone overflow stream (i.e. water with a low % ultra-fines) was used as mill dilution. Needless to say, a few months down the line the mill started showing serious throughput capacity problems, and this was due to a build-up of ultra-fines in the mill circuit over time i.e. the pulp rheology had gone toxic. The pulp inside the mill was so viscous, I could walk on it (my feet did not make much of an impression on the pulp surface).

So yes, maintaining proper pulp rheology (the combination of pulp SG and viscosity) in the mill is extremely important. However, if the ore characteristics dictate otherwise (such as a high clay content in the ore), it may be best to scrub & classify the ore upstream from passing the coarse size fraction to the mill.