Not considering friction, the torque you will have to produce comes from the fact that the mixing action requires the mass inside the drum to rotate to an off-center position before it falls back on itself. That torque would be the product of the weight of the internal mixture and the HORIZONTAL distance to the vertical line along which the CG of the mass lies. It is the same effect you see after you stop mixing. The drum tries to rotate backward until the CG of the internal mixture is at bottom dead center.
If you know the density of the mixture, careful observations during mixing should enable you make a good estimate of the approximate cross sectional shape of the mass, and from that you could calculate the CG. Be sure to consider effects of maximum and minimum charge loads. Both the weight and the torque arm length will be affected by that. Add in a factor for friction, safety, (and then double it for good measure) and you should be good to go.