An LIF technique was used to obtain unobtrusive measurements of scalar concentration as a function of time and mixing times in a filly baffled batch operated mixing vessel agitated by five forms of impeller (Rushton, 'bucket', six bladed 4 5 Oand 60"-pitched blade, hyperboloid).The mixing time was comparable to the time required for the dyekcalar to be transported from the top of the vessel to the bottom part plus the time required for a fiuther two rotations of the bulk flow in the circumferential direction. At constant power input the mixing times for similar impellers were similar, although different for different types.On a utilise une technique LIF afin d'obtenir des mesures non intrusives de concentration scalaire en fonction du temps et des temps de melange dans une cuve agitee munie de contrepales pour cinq types de turbines (Rushton, "a abues", a 6 pales inclinees a 45' et 60°, hyperbololdes) en regime discontinu.Le temps de melange est comparable au temps requis pour le transport du colorant/scalaire depuis le sommet du reservoir jusqu'a-sa base plus le temps requis pour encore deux rotations du coeur de I'ecoulement dans la direction circonferentielle. A puissance constante, les temps de melange pour des turbines semblables sont similaires, mais ils dimrent lorsque les turbines sont de differents types.Keywords: stirred tank, batch mixing, LIF, scalar concentration, mixing time.tirred tanks are widely used by the chemical, pharma-S ceutical, food and water treatment industries during the manufacture of paints, polymers, detergents, drugs and foods& and during the purification of industrial and household waste. The optimum design of a stirred tank depends on the requirements in terms of parameters like mixing time, mixing 'strength', mixing scales and other quantities which may be lMed and is achieved by the correct choice of geometry, rotational speed and location of fluid addition and subtraction. Therefore, a detailed knowledge of the power, velocity and scalar characteristics is required.Previous investigations of the present flow configuration and impellers included flow visualisation, power measurements and velocity measurements. Yianneskis et al. (1987) studied the effect of impeller diameter and clearance on the flow generated by the Rushton turbine and Yianneskis and Whitelaw (1 993) mapped the mean and turbulence structure of the trailing vortices produced by the Rushton turbine blades. N o d (1988) compared single and two phase flows with solid particles driven by a Rushton turbine and studied the effect of rotational speed and impeller size on the mean and turbulence velocities with the latter extended by Nouri and Whitelaw (1990) for different geometrical sizes of the flow configuration. Hockey (1 990) contrasted turbulent Newtonian and non-Newtonian shear thinning flows generated by a six bladed 60O-pitched blade impeller and a Rushton turbine with respect to velocity and turbulence structure, energy dissipation and shear rates. Nouri and Whitelaw (1994) investigated the flow field of a hy...