Mixing Tank Design: A Systematic Treatment of Geometry, Hydrodynamics, Multiphase Duty, and Mechanical Integrity

Mixing tank design is not a matter of selecting a vessel and attaching an agitator. It is a coupled problem in which the process objective, whether blending, suspension, dispersion, heat transfer, or reaction, dictates the flow field; the flow field dictates the impeller and baffle configuration; and that configuration, once fixed, imposes mechanical loads that the shaft, bearings, and drive must carry without resonance or fatigue. A design that satisfies one of these requirements while neglecting another is not a design but a partial calculation. The discussion that follows treats the principal decisions in the order in which they constrain one another, moving from geometric framework and impeller selection through hydrodynamic characterization and multiphase duty to scale-up and mechanical verification.

Geometric framework and standard proportions

Industrial practice begins with a set of geometric ratios that establish a coherent, scalable baseline. The impeller-to-tank diameter ratio D/T typically falls between 0.3 and 0.5; the liquid fill height is usually set approximately equal to the tank diameter, so that H/T is about 1; baffle width is about one-tenth of the tank diameter; and impeller off-bottom clearance Cb/T lies in the range of 0.2 to 0.25. These ratios are not arbitrary conventions. They represent a compromise between top-to-bottom circulation, power efficiency, and the avoidance of stagnant regions, and they provide the geometric similarity that later permits scale-up correlations to be applied with reasonable confidence. Departures from these proportions are common, including high-aspect-ratio vessels for gas-liquid contacting, dished or profiled bottoms for solids suspension, and close-clearance impellers for viscous duty, but such departures are design choices that must be justified against the baseline rather than defaults.

Impeller selection as a flow-field decision

The impeller is the device that converts shaft power into fluid motion, and its selection is properly made on the basis of the desired flow pattern rather than on a generic notion of mixing intensity. Axial-flow impellers, including marine propellers and hydrofoils, generate a flow field directed along the axis of rotation. Down-pumping axial impellers produce a downward jet that turns radially at the tank floor and travels up the vessel wall, giving strong top-to-bottom turnover and making them well suited to blending, solids suspension, and heat transfer. Hydrofoils such as the A310 or HE3 are carefully profiled to produce a strong axial flow at low turbulence intensity, and their power numbers in the turbulent regime are correspondingly low, typically around 0.3 to 0.4, compared with values of five or more for a classical Rushton disc turbine. This low power number is the origin of their energy advantage in duties where bulk movement, rather than local shear, is the controlling requirement.

Radial-flow impellers, of which the six-blade Rushton disc turbine is the canonical example, discharge fluid tangentially outward from the impeller and generate high local shear and turbulence in the impeller zone. They produce relatively little axial top-to-bottom turnover and therefore require careful placement, often at lower clearance from the tank floor, to avoid stratification. Their high power number reflects the intensity of turbulence generation near the blade tips, and this is precisely what is needed when the process is controlled by turbulent mixing, gas dispersion, or liquid-liquid emulsification. Between these extremes lie the mixed-flow impellers, most commonly the pitched-blade turbine, whose blades are set at an angle, typically 45 degrees, so that the discharge has both axial and radial components. A 45-degree pitched-blade turbine is a versatile compromise, but the blade angle is a genuine design variable: shallower angles push the behavior toward axial flow, while steeper angles push it toward radial flow. The selection of an impeller class is therefore a statement about which flow characteristic the process cannot do without: axial turnover for uniformity and suspension, radial shear for dispersion and mass transfer, or a controlled balance of the two.

Baffles and the suppression of swirl

An unbaffled vertical cylindrical tank agitated by a centrally mounted impeller develops a strong tangential velocity field. The liquid body rotates as a solid mass, a free-surface vortex forms, and the impeller's relative velocity with respect to the fluid falls, so that power draw decreases and top-to-bottom exchange is severely impaired. Wall baffles convert this useless swirl into axial and radial flow. The standard fully baffled configuration consists of four vertical flat-plate baffles mounted at 90-degree intervals, each with a radial width of about one-tenth to one-twelfth of the tank diameter, extending over the full liquid height. With this arrangement, the gross rotation is broken, the vortex is suppressed, and the intended flow field develops: downward axial flow in the case of a hydrofoil or down-pumping pitched-blade turbine, or outward radial flow in the case of a Rushton turbine. Baffles are sometimes offset slightly from the wall, by roughly one-sixth of their width, to eliminate stagnant pockets behind them, and they impose additional fluid forces on the tank structure that must be considered in mechanical design. In unbaffled or partially baffled vessels, vortex formation can sometimes be mitigated by eccentric impeller mounting, but the fully baffled configuration remains the basis for most quantitative design correlations in turbulent duty.

Hydrodynamic characterization: dimensionless groups and power

Once the geometric and impeller configuration is fixed, performance is described through dimensionless groups that compress the variables of density, viscosity, impeller diameter, and rotational speed into quantities that can be compared across scales. The mixing Reynolds number, Re = rho N D^2 / mu, determines whether the flow is laminar, transitional, or turbulent. The power number, Po = P / (rho N^3 D^5), is a drag coefficient for the impeller. In the turbulent regime, generally for Re greater than about 10,000, it becomes approximately constant for a given impeller geometry, and it is the basis for calculating power draw and for scale-up. The flow number, Fl = Q / (N D^3), expresses the pumping capacity of the impeller per unit speed and diameter, and the Froude number, Fr = N^2 D / g, is relevant to vortex formation in unbaffled vessels and to gas dispersion duty. The power number is not a single universal property of an impeller family; it depends on blade detail, disc presence, blade angle, and the proximity of the impeller to the tank floor or to other impellers. An axial impeller placed very close to the floor can exhibit a mixed-flow character and a corresponding shift in power number, which is one reason why off-bottom clearance must be treated as a design variable rather than a fixed installation dimension.

The relationship between power number and Reynolds number carries a practical implication that is often underappreciated: in the turbulent regime, power is proportional to the fifth power of impeller diameter at constant speed. A modest increase in impeller diameter therefore produces a disproportionate increase in power draw, and the choice between a small impeller at high speed and a large impeller at low speed is fundamentally a choice between shear intensity and bulk flow efficiency. This is the core of impeller selection for energy-sensitive duties.

Multiphase duty: solids suspension and gas dispersion

When the tank must suspend solid particles, the controlling parameter is the just-suspended speed, the impeller speed at which no particle remains stationary on the tank bottom for more than a brief interval. The classical Zwietering correlation expresses the just-suspended speed as a function of particle properties, solids loading, liquid properties, and impeller diameter, with the correlation constant depending on impeller and tank geometry. Off-bottom clearance has a strong effect: reducing the clearance-to-tank-diameter ratio generally lowers the just-suspended speed for propellers, paddles, and disc turbines, although the original Zwietering correlation does not explicitly model this effect and modified forms have been proposed to account for it. In practice, a useful rule of thumb is to operate at an impeller speed 10 to 20 percent above the just-suspended speed to obtain a good distribution of solids rather than merely the onset of suspension. For buoyant solids that tend to accumulate at the surface, a second impeller near the liquid surface may be required to draw them into the bulk.

Gas-liquid duty imposes a different set of constraints. A well-designed gas-dispersion system should generate a large interfacial area between gas and liquid, exhibit minimal reduction in power draw upon gassing, and avoid impeller flooding, the condition in which gas rises through the impeller without being dispersed. Radial disc turbines have traditionally been the default choice for gas dispersion, but the power draw of a radial turbine under gassed conditions can fall substantially, and the design must balance agitator speed against gas flow rate. Increasing gas flow at constant speed degrades dispersion, while increasing speed at constant gas flow improves it; the just-dispersed and fully dispersed conditions can be estimated from empirical relationships involving the flow number, Froude number, and tank-to-impeller diameter ratio. Gas introduction close to the impeller, through a sparger, dip pipe, or hollow shaft, is generally more effective than surface entrainment, and multi-impeller combinations, for example a radial lower impeller for gas dispersion surmounted by axial hydrofoils for bulk circulation, are common in tall gas-liquid contactors. Three-phase systems, in which both gas dispersion and solids suspension must be achieved simultaneously, are inherently more demanding because the gas phase alters the liquid flow field that would otherwise suspend the solids, and the design must be validated against the specific gassing rate and particle characteristics of the process.

Scale-up as a deliberate choice of similarity criterion

Scale-up is the point at which the designer must decide which physical quantity is to be preserved, because geometric similarity alone does not preserve all mixing characteristics simultaneously. The two most widely used criteria are equal impeller tip speed and equal power per unit volume. Equal tip speed tends to preserve maximum shear rate and is therefore often selected for dispersion and emulsification duties. Equal power per unit volume tends to preserve overall turbulence intensity and bulk blending performance and is often selected for homogeneous mixing and suspension duties. These two criteria do not lead to the same impeller speed at the large scale, and most practical scale-up rules for geometrically similar vessels, including equal solids suspension, fall in the speed range between them. The choice is not merely academic: selecting equal power per unit volume can produce a large increase in tip speed at the large scale, with consequences for shear-sensitive materials, while selecting equal tip speed can underpower the large vessel with respect to bulk circulation and solids suspension. A defensible scale-up strategy states explicitly which criterion is being held and why, and checks the resulting large-scale impeller speed against the constraints of the other criteria.

Mechanical design: the shaft as a rotor-dynamic system

The agitator shaft is not a passive torque transmitter; it is a rotating beam submerged in a liquid, subject to bending from unbalanced hydraulic forces, torque fluctuations from blade-baffle interaction, and the possibility of resonant vibration. The shaft is considered stiff or rigid when its rotational frequency is below the first critical, or resonance, frequency of lateral vibration. A common criterion is that the operating speed should remain below 70 percent of the calculated critical speed for general mixing duty, and below about 50 percent for gas-liquid systems where the added mass and damping effects of the two-phase mixture are less predictable. The shaft must also be checked for torsional shear stress under the maximum torque transmitted by the drive, including the acceleration torque of the motor, and for fatigue at stress concentrations such as impeller hubs and bearing transitions. Baffles and the tank wall experience fluctuating fluid forces that must be accounted for in the structural design, and the entire support structure should be checked for resonance near the shaft speed and its harmonics. A mixing tank that is hydrodynamically well designed but mechanically under-supported will fail at the shaft, the bearing, or the seal, and the failure will often occur at a speed that the process does not require but that the structure cannot tolerate.

Computational design and modern practice

Computational fluid dynamics has become a routine design aid for mixing tanks, not as a replacement for correlations and pilot-scale testing but as a means of interrogating flow fields that correlations cannot resolve. CFD can predict power draw, mixing time, dead-volume fraction, and local shear rate for candidate geometries, and industrial case studies have demonstrated full-scale validation against measured power draw, liquid level, and mixing time with errors of a few percent. Optimization frameworks that couple CFD with surrogate models or design-of-experiments methods can explore impeller geometry and operating conditions against multiple objectives, including minimizing power consumption, dead volume, and excessive shear, including for non-Newtonian shear-thinning fluids where conventional turbulent correlations do not apply. CFD-DEM coupling extends this capability to solid-liquid suspension, allowing the designer to examine how bottom shape, baffle form, and impeller inlet geometry interact to eliminate dead zones and improve particle distribution. The appropriate role of CFD in a design sequence is to test and refine a configuration that has already been established on the basis of process requirements, geometric proportions, and dimensionless characterization, not to substitute for that foundation.

A concluding note on the design sequence

A coherent mixing tank design proceeds from the process objective to the flow field, from the flow field to the impeller and baffle configuration, from that configuration to the hydrodynamic operating point, and finally to the mechanical verification of the shaft and support system. The geometric ratios, dimensionless groups, and correlation constants that fill the literature are tools for executing these steps, not specifications that can be assembled independently. The most common failure mode in mixing tank design is not a shortage of correlations but a mismatch between the flow field that the impeller actually produces and the flow field that the process actually requires. A design that keeps that correspondence in view, and that treats scale-up and mechanical integrity as parts of the same problem rather than as afterthoughts, will be both hydraulically effective and mechanically durable.