The problem with your request is the high viscosity and, as I suspect, the high specific gravity of the solution when mixed. I suggest a ribbon blender or high shear mixer such as one provided by Arde-barinco:
I assume that the viscosity mentioned is the solution or solvent. The viscosity of the final blend, whether slurry or solution will increase. If the solids loading is above 20%, or so, you can bet on a very high final viscosity and this will be higher if there is a reaction.
Most laboratories have small mixers on which you can test the solution. You will want a scale, an amp probe, and a good, calibrated ball and cup viscometer. The choice may change depending on viscosity of the final solution. I suggest a meter that works between 1 cP and 100,000 cP. Anyway, take the ingredients to a lab and run your test. The idea is to add the ingredients slowly and agitate at a constant speed. Later, you may want to experiment with speed but run the first trial at a continual speed. Ideally you will want a baffled tank but the baffles should be cut back: textbook recommendation would be T/24 in width with cleanout gaps of a couple of inches. T = tank ID.
A conventional impeller and design are out of the question. Have a vendor provide the constants needed for your calculation but do them yourself first after looking at the test data. (Watch the mixer carefully! If the motor gets too hot stop the mixer and get one with a bigger impeller and motor.)
Here's sizing information:
P = 6.566 X 10^-14(NpN^3D^5(S.G.)
P = BHP; Np = vendor data on impeller; N = speed, rpm;
D = impeller diameter, inches;
For slurries you can expect a high draw:
10 hp per 1000 gallon of solution.
If you are looking for ways to reduce power draw here's a few ideas:
* inject a gas (25-50% reduction); and
* add high flow (axial) impeller nearby high shear mixer.
With a high-shear mixer, a propeller mixer proped nearby can greatly improve turnover and feed the shear head.
Injecting gas is an idea if solvent is not flammable. Otherwise, injecting nitrogen can be expensive and risk asphyxiation. You may want to try this in your experiment. It is a good idea to have a vendor run a lab test on you solution too: the raw ingredients and the final solution. Lately, the idea of CFDs have been tried. Maybe, there's a solution there if you have the bucks and a lab won't do it.
Dirk Willard
