Table of Contents#### Download Safari Books Online apps: Apple iOS | Android | BlackBerry

### 3.4. AGITATION AND MIXING OF FLUIDS AND POWER REQUIREMENTS

#### 3.4A. Purposes of Agitation

#### 3.4B. Equipment for Agitation

##### Figure 3.4-1. Baffled tank and three-blade propeller agitator with axial-flow pattern: (a) side view, (b) bottom view.

##### 1. Three-blade propeller agitator

##### 2. Paddle agitators

##### Figure 3.4-2. Various types of agitators: (a) four-blade paddle, (b) gate or anchor paddle, (c) six-blade open turbine, (d) pitched-blade (45°) turbine.

##### 3. Turbine agitators

##### Figure 3.4-3. Baffled tank with six-blade turbine agitator with disk showing flow patterns: (a) side view, (b) bottom view, (c) dimensions of turbine and tank.

##### Figure 3.4-4. Other types of agitators: (a) high-efficiency, three-blade impeller, (b) double-helical-ribbon, (c) helical-screw. [Reprinted with permission from André Bakker and Lewis E. Gates, Chem. Eng. Progr., 91 (Dec.), 25 (1995). Copyright by the American Institute of Chemical Engineers.]

##### 4. Helical-ribbon agitators

##### 5. Agitator selection and viscosity ranges

#### 3.4C. Flow Patterns in Agitation

#### 3.4D. Typical "Standard" Design of Turbine

#### 3.4E. Power Used in Agitated Vessels

##### Figure 3.4-5. Power correlations for various impellers and baffles (see Fig. 3.4-3c for dimension D_{a}, D_{t}, J, and W).

#### 3.4F. Agitator Scale-Up

##### 1. Introduction

##### 2. Scale-up procedure

#### 3.4G. Mixing Times of Miscible Liquids

##### Figure 3.4-6. Correlation of mixing time for miscible liquids using a turbine in a baffled tank (for a plain turbine, turbine with disk, and pitched-blade turbine). [From "Flow Patterns and Mixing Rates in Agitated Vessels" by K. W. Norwood and A. B. Metzner, A.I.Ch.E.J., 6, 432 (1960). Reproduced by permission of the American Institute of Chemical Engineers, 1960.]

#### 3.4H. Flow Number and Circulation Rate in Agitation

#### 3.4I. Special Agitation Systems

##### 1. Suspension of solids

##### 2. Dispersion of gases and liquids in liquids

##### 3. Motionless mixers

##### Figure 3.4-7. Two-element helical motionless mixer with element length L and pipe diameter D (M2, M6).

#### 3.4J. Mixing of Powders, Viscous Materials, and Pastes

##### 1. Powders

##### Figure 3.4-8. Mixers for powders and pastes: (a) double-cone powder mixer, (b) ribbon powder mixer with two ribbons, (c) kneader mixer for pastes.

##### 2. Dough, pastes, and viscous materials

Entire Site

Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

In the chemical and other processing industries, many operations are dependent to a great extent on effective agitation and mixing of fluids. Generally, agitation refers to forcing a fluid by mechanical means to flow in a circulatory or other pattern inside a vessel. Mixing usually implies the taking of two or more separate phases, such as a fluid and a powdered solid or two fluids, and causing them to be randomly distributed through one another.

There are a number of purposes for agitating fluids, some of which are briefly summarized:

Blending of two miscible liquids, such as ethyl alcohol and water.

Dissolving solids in liquids, such as salt in water.

Dispersing a gas in a liquid as fine bubbles, such as oxygen from air in a suspension of microorganisms for fermentation or for the activated sludge process in waste treatment.

Suspending of fine solid particles in a liquid, as in the catalytic hydrogenation of a liquid, where solid catalyst particles and hydrogen bubbles are dispersed in the liquid.

Agitation of the fluid to increase heat transfer between the fluid and a coil or jacket in the vessel wall.

Generally, liquids are agitated in a cylindrical vessel which can be closed or open to the air. The height of liquid is approximately equal to the tank diameter. An impeller mounted on a shaft is driven by an electric motor. A typical agitator assembly is shown in Fig. 3.4-1.

There are several types of agitators that are widely used. A common type, shown in Fig. 3.4-1, is a three-bladed marine-type propeller similar to the propeller blade used in driving boats. The propeller can be a side-entering type in a tank or be clamped on the side of an open vessel in an off-center position. These propellers turn at high speeds of 400 to 1750 rpm (revolutions per minute) and are used for liquids of low viscosity. The flow pattern in a baffled tank with a propeller positioned on the center of the tank is shown in Fig. 3.4-1. This type of flow pattern is called axial flow since the fluid flows axially down the center axis or propeller shaft and up on the sides of the tank as shown.

Various types of paddle agitators are often used at low speeds, between about 20 and 200 rpm. Two-bladed and four-bladed flat paddles are often used, as shown in Fig. 3.4-2a. The total length of the paddle impeller is usually 60–80% of the tank diameter and the width of the blade to of its length. At low speeds mild agitation is obtained in an unbaffled vessel. At higher speeds baffles are used, since, without baffles, the liquid is simply swirled around with little actual mixing. The paddle agitator is ineffective for suspending solids, since good radial flow is present but little vertical or axial flow. An anchor or gate paddle, shown in Fig. 3.4-2b, is often used. It sweeps or scrapes the tank walls and sometimes the tank bottom. It is used with viscous liquids where deposits on walls can occur and to improve heat transfer to the walls. However, it is a poor mixer. Paddle agitators are often used to process starch pastes, paints, adhesives, and cosmetics.

Turbines that resemble multibladed paddle agitators with shorter blades are used at high speeds for liquids with a very wide range of viscosities. The diameter of a turbine is normally between 30 and 50% of the tank diameter. The turbines usually have four or six blades. Figure 3.4-3 shows a flat six-blade turbine agitator with disk. In Fig. 3.4-2c a flat six-blade open turbine is shown. The turbines with flat blades give radial flow, as shown in Fig. 3.4-3. They are also useful for good gas dispersion; the gas is introduced just below the impeller at its axis and is drawn up to the blades and chopped into fine bubbles. In the pitched-blade turbine shown in Fig. 3.4-2d, with the blades at 45°, some axial flow is imparted so that a combination of axial and radial flow is present. This type is useful in suspending solids since the currents flow downward and then sweep up the solids.

Often a pitched-blade turbine with only four blades is used in suspension of solids. A high-efficiency, three-blade impeller (B6, F2) shown in Fig. 3.4-4a is similar to a four-blade pitched turbine; however, it features a larger pitch angle of 30–60° at the hub and a smaller angle of 10–30° at the tip. This axial-flow impeller produces more fluid motion and mixing per unit of power and is very useful in suspension of solids.

This type of agitator is used in highly viscous solutions and operates at a low RPM in the laminar region. The ribbon is formed in a helical path and is attached to a central shaft. The liquid moves in a tortuous flow path down the center and up along the sides in a twisting motion. Similar types are the double-helical-ribbon agitator shown in Fig. 3.4-4b and the helical-screw impeller shown in Fig. 3.4-4c.

The viscosity of the fluid is one of several factors affecting the selection of the type of agitator. Indications of the viscosity ranges of these agitators are as follows. Propellers are used for fluid viscosities below about 3 Pa · s (3000 cp); turbines can be used below about 100 Pa · s (100 000 cp); modified paddles such as anchor agitators can be used above 50 Pa · s to about 500 Pa · s (500 000 cp); helical and ribbon-type agitators are often used above this range to about 1000 Pa · s and have been used up to 25 000 Pa · s. For viscosities greater than about 2.5 to 5 Pa · s (5000 cp) and above, baffles are not needed since little swirling is present above these viscosities.

The flow patterns in an agitated tank depend upon the fluid properties, the geometry of the tank, the types of baffles in the tank, and the agitator itself. If a propeller or other agitator is mounted vertically in the center of a tank with no baffles, a swirling flow pattern usually develops. Generally, this is undesirable, because of excessive air entrainment, development of a large vortex, surging, and the like, especially at high speeds. To prevent this, an angular off-center position can be used with propellors with small horsepower. However, for vigorous agitation at higher power, unbalanced forces can become severe and limit the use of higher power.

For vigorous agitation with vertical agitators, baffles are generally used to reduce swirling and still promote good mixing. Baffles installed vertically on the walls of the tank are shown in Fig. 3.4-3. Usually four baffles are sufficient, with their width being about of the tank diameter for turbines and propellers. The turbine impeller drives the liquid radially against the wall, where it divides with one portion flowing upward near the surface and back to the impeller from above and the other flowing downward. Sometimes, in tanks with large liquid depths much greater than the tank diameter, two or three impellers are mounted on the same shaft, each acting as a separate mixer. The bottom impeller is about 1.0 impeller diameter above the tank bottom.

In an agitation system, the volume flow rate of fluid moved by the impeller, or circulation rate, is important in sweeping out the whole volume of the mixer in a reasonable time. Also, turbulence in the moving stream is important for mixing, since it entrains the material from the bulk liquid in the tank into the flowing stream. Some agitation systems require high turbulence with low circulation rates, others low turbulence with high circulation rates. This often depends on the types of fluids being mixed and on the amount of mixing needed.

The turbine agitator shown in Fig. 3.4-3 is the most commonly used agitator in the process industries. For design of an ordinary agitation system, this type of agitator is often used in the initial design. The geometric proportions of the agitation system which are considered as a typical "standard" design are given in Table 3.4-1. These relative proportions are the basis for the major correlations of agitator performance in numerous publications. (See Fig. 3.4-3c for nomenclature.)

In some cases W/D_{a} = for agitator correlations. The number of baffles is four in most uses. The clearance or gap between the baffles and the wall is usually 0.10–0.15 J to ensure that liquid does not form stagnant pockets next to the baffle and wall. In a few correlations the ratio of baffle to tank diameter is J/D_{t}
=
instead of .

In the design of an agitated vessel, an important factor is the power required to drive the impeller. Since the power required for a given system cannot be predicted theoretically, empirical correlations have been developed to predict the power required. The presence or absence of turbulence can be correlated with the impeller Reynolds number , defined as

Equation 3.4-1

where D_{a} is the impeller (agitator) diameter in m, N is rotational speed in rev/s, ρ is fluid density in kg/m^{3}, and μ is viscosity in kg/m · s. The flow is laminar in the tank for
< 10, turbulent for > 10^{4}, and for a range between 10 and 10^{4}, the flow is transitional, being turbulent at the impeller and laminar in remote parts of the vessel.

Power consumption is related to fluid density ρ, fluid viscosity μ, rotational speed N, and impeller diameter D_{a} by plots of power number N_{p} versus . The power number is

Equation 3.4-2

where P = power in J/s or W. In English units, P = ft · lb_{f}/s.

Figure 3.4-5 is a correlation (B3, R1) for frequently used impellers with Newtonian liquids contained in baffled, cylindrical vessels. Dimensional measurements of baffle, tank, and impeller sizes are given in Fig. 3.4-3c. These curves may also be used for the same impellers in unbaffled tanks when is 300 or less (B3, R1). When is above 300, the power consumption for an unbaffled vessel is considerably less than for a baffled vessel. Curves for other impellers are also available (B3, R1).

Curve 1. Flat six-blade turbine with disk (like Fig. 3.4-3 but six blades); D_{a}/W = 5; four baffles each D_{t}/J = 12.
Curve 2. Flat six-blade open turbine (like Fig. 3.4-2c); D_{a}/W = 8; four baffles each D_{t}/J = 12.
Curve 3. Six-blade open turbine (pitched-blade) but blades at 45° (like Fig. 3.4-2d); D_{a}/W = 8; four baffles each D_{t}/J = 12.
Curve 4. Propeller (like Fig. 3.4-1); pitch = 2D_{a}; four baffles each D_{t}/J = 10; also holds for same propeller in angular off-center position with no baffles.
Curve 5. Propeller; pitch = D_{a}; four baffles each D_{t}/J = 10; also holds for same propeller in angular off-center position with no baffles.
Curve 6. High-efficiency impeller (like Fig. 3-4-4a); four baffles each D_{t}/J = 12.
[Curves 1, 2, and 3 reprinted with permission from R. L. Bates. P. L. Fondy, and R. R. Corpstein, Ind. Eng. Chem. Proc. Des. Dev.,
2, 310 (1963). Copyright by the American Chemical Society. Curves 4 and 5 from J. H. Rushton, E. W. Costich, and H. J. Everett, Chem. Eng. Progr.,
46, 395, 467 (1950). With permission.]

The power-number curve for N_{p} for the high-efficiency, three-blade impeller is shown as curve 6 in Fig. 3-4-5.

## EXAMPLE 3.4-1. Power Consumption in an AgitatorA flat-blade turbine agitator with disk having six blades is installed in a tank similar to Fig. 3.4-3. The tank diameter D Calculate the required kW of the mixer. For the same conditions, except for the solution having a viscosity of 100 000 cp, calculate the required kW.
Solution:
For part (a) the following data are given: D Using Eq. (3.4-1), the Reynolds number is Using curve 1 in Fig. 3.4-5, since D This is in the laminar flow region. From Fig. 3.4-5, N Hence, a 10 000-fold increase in viscosity only increases the power from 1.324 to 3.71 kW. |

Variations of various geometric ratios from the "standard" design can have different effects on the power number N_{P} in the turbulent region of the various turbine agitators as follows (B3):

For the flat six-blade open turbine, N

_{P}∝ (W/D_{a})^{1.0}.For the flat six-blade open turbine, varying D

_{a}/D_{t}from 0.25 to 0.50 has practically no effect on N_{P}.With two six-blade open turbines installed on the same shaft and the spacing between the two impellers (vertical distance between the bottom edges of the two turbines) being at least equal to D

_{a}, the total power is 1.9 times a single flat-blade impeller. For two six-blade pitched-blade (45°) turbines, the power is also about 1.9 times that of a single pitched-blade impeller.A baffled, vertical square tank or a horizontal cylindrical tank has the same power number as a vertical cylindrical tank. However, marked changes in the flow patterns occur.

The power number for a plain anchor-type agitator similar to Fig. 3.4-2b but without the two horizontal crossbars is as follows for < 100 (H2):

Equation 3.4-3

where D_{a}/D_{t}
= 0.90, W/D_{t} = 0.10, and C/D_{t}
= 0.05.

The power number for a helical-ribbon agitator for very viscous liquids for < 20 is as follows (H2, P3):

Equation 3.4-4

Equation 3.4-5

The typical dimensional ratios used are D_{a}/D_{t}
= 0.95, with some ratios as low as 0.75, and W/D_{t} = 0.095. The agitator pitch is the vertical distance of a single flight of the helix in a 360° rotation (B6).

In the process industries, experimental data are often available for a laboratory-size or pilot-unit-size agitation system, and it is desired to scale up the results to design a full-scale unit. Since the processes to be scaled up are very diverse, no single method can handle all types of scale-up problems, and many approaches to scale-up exist. Geometric similarity is, of course, important and simplest to achieve. Kinematic similarity can be defined in terms of ratios of velocities or of times (R2). Dynamic similarity requires fixed ratios of viscous, inertial, or gravitational forces. Even if geometric similarity is achieved, dynamic and kinematic similarity often cannot be obtained at the same time. Hence, it is frequently up to the designer to rely on judgment and experience in the scale-up.

In many cases, the main objectives usually present in an agitation process are as follows: equal liquid motion, such as in liquid blending, where the liquid motion or corresponding velocities are approximately the same in both cases; equal suspension of solids, where the levels of suspension are the same; and equal rates of mass transfer, where mass transfer is occurring between a liquid and a solid phase, liquid-liquid phases, and so on, and the rates are the same.

A suggested step-by-step procedure to follow in the scale-up is detailed as follows for scaling up from the initial conditions, where the geometric sizes given in Table 3.4-1 are D_{a}_{1}, D_{T}_{1}, H_{1}, W_{1}, and so on, to the final conditions of D_{a}_{2}, D_{T}_{2}, and so on.

Then the ratio of the volumes is

Equation 3.4-7

The scale-up ratio is then

Equation 3.4-8

Using this value of R, apply it to all of the dimensions in Table 3.4-1 to calculate the new dimensions. For example,

Equation 3.4-9

Then a scale-up rule must be selected and applied to determine the agitator speed N

_{2}to be used to duplicate the small-scale results using N_{1}. This equation is as follows (R2):Equation 3.4-10

where n = 1 for equal liquid motion, n = for equal suspension of solids, and n = for equal rates of mass transfer (which is equivalent to equal power per unit volume). This value of n is based on empirical and theoretical considerations.

Knowing N

_{2}, the power required can be determined using Eq. (3.4-2) and Fig. 3.4-5.

## EXAMPLE 3.4-2. Derivation of Scale-Up Rule ExponentFor the scale-up-rule exponent n in Eq. (3.4-10), show the following for turbulent agitation: That when n = , the power per unit volume is constant in the scale-up. That when n = 1.0, the tip speed is constant in the scale-up.
Solution:
For part (a), from Fig. 3.4-5, N Equation 3.4-11 Then for equal power per unit volume, P Equation 3.4-12 Substituting P Equation 3.4-13 For part (b), using Eq. (3.4-10) with n = 1.0, rearranging, and multiplying by π, Equation 3.4-14 Equation 3.4-15 where πD |

To aid the designer of new agitation systems as well as serve as a guide for evaluating existing systems, some approximate guidelines are given as follows for liquids of normal viscosities (M2): for mild agitation and blending, 0.1 to 0.2 kW/m^{3} of fluid (0.0005 to 0.001 hp/gal); for vigorous agitation, 0.4 to 0.6 kW/m^{3} (0.002 to 0.003 hp/gal); for intense agitation or where mass transfer is important, 0.8 to 2.0 kW/m^{3} (0.004 to 0.010 hp/gal). This power in kW is the actual power delivered to the fluid as given in Fig. 3.4-5 and Eq. (3.4-2). This does not include the power used in the gear boxes and bearings. Typical efficiencies of electric motors are given in Section 3.3B. As an approximation, the power lost in the gear boxes and bearings and in inefficiency of the electric motor is about 30 to 40% of P, the actual power input to the fluid.

## EXAMPLE 3.4-3. Scale-Up of Turbine Agitation SystemAn existing agitation system is the same as given in Example 3.4-1a for a flat-blade turbine with a disk and six blades. The given conditions and sizes are D Where equal rate of mass transfer is desired. Where equal liquid motion is needed.
Solution:
Since H The dimensions of the larger agitation system are as follows: D For part (a), for equal mass transfer, n = in Eq. (3.4-10). Using Eq. (3.4-1), Using N The power per unit volume is The value of 0.2752 kW/m For part (b), for equal liquid motion, n = 1.0. |

In one method used to study the blending or mixing time for two miscible liquids, an amount of HCl acid is added to an equivalent of NaOH and the time required for the indicator to change color is noted. This is a measure of molecule-molecule mixing. Other experimental methods are also used. Rapid mixing takes place near the impeller, with slower mixing, which depends on the pumping circulation rate, in the outer zones.

In Fig. 3.4-6, a correlation of mixing time is given for a turbine agitator (B5, M5, N1). The dimensionless mixing factor f_{t} is defined as

Equation 3.4-16

where t_{T} is the mixing time in seconds. For > 1000, since f_{t} is approximately constant, then t_{T}N^{2/3} is constant. For some other mixers it has been shown that t_{T}N is approximately constant. For scaling up from vessel 1 to another size vessel 2 with similar geometry and with the same power/unit volume in the turbulent region, the mixing times are related by

Equation 3.4-17

Hence, the mixing time increases for the larger vessel. For scaling up while keeping the same mixing time, the power/unit volume P/V increases markedly:

Equation 3.4-18

Usually, in scaling up to large-size vessels, a somewhat larger mixing time is used so that the power/unit volume does not increase markedly.

The mixing time for a helical-ribbon agitator is as follows for < 20 (H2):

Equation 3.4-19

Equation 3.4-20

For very viscous liquids the helical-ribbon mixer gives a much smaller mixing time than a turbine for the same power/unit volume (M5). For nonviscous liquids, however, it gives longer times.

For a propellor agitator in a baffled tank, a mixing-time correlation is given by Biggs (B5), and that for an unbaffled tank by Fox and Gex (F1).

For a high-efficiency impeller in a baffled tank, mixing-time correlations are given by reference (F2), which shows that mixing times are lower than for pitched-blade agitators.

## EXAMPLE 3.4-4. Scale-Up of Mixing Time in a Turbine Agitation System.Using the existing conditions for the turbine with a disk as in Example 3.4-1, part (a), do as follows: Calculate the mixing time. Calculate the mixing time for a smaller vessel with a similar geometric ratio, where D _{t}is 0.30 m instead of 1.83 m. Do this for the same power per unit volume as used in part (a).Using the same mixing time calculated for the smaller vessel in part (b), calculate the new power per unit volume for the larger vessel in part (a).
Solution:
In part (a), D The power per unit volume is From Fig. 3.4-6 for = 5.185 × 10 For part (b), the scale-down ratio R from Eq. (3.4-8) is Also, H Hence, t For part (c), using the same mixing time of 5.73 s for the smaller vessel, the power per unit volume of the larger vessel is calculated from Eq. (3.4-18) for equal mixing times: Solving, P |

An agitator acts like a centrifugal pump impeller without a casing and gives a flow at a certain pressure head. This circulation rate Q in m^{3}/s from the edge of the impeller is the flow rate perpendicular to the impeller discharge area. Fluid velocities have been measured in mixers and have been used to calculate the circulation rates. Data for baffled vessels have been correlated using the dimensionless flow number N_{Q} (U1):

Equation 3.4-21

In some agitation systems a solid is suspended in the agitated liquid. Examples are when a finely dispersed solid is to be dissolved in the liquid, microorganisms are suspended in fermentation, a homogeneous liquid-solid mixture is to be produced for feed to a process, and a suspended solid is used as a catalyst to speed up a reaction. The suspension of solids is somewhat similar to a fluidized bed. In the agitated system, circulation currents of the liquid keep the particles in suspension. The amount and type of agitation needed depend mainly on the terminal settling velocity of the particles, which can be calculated using the equations in Section 14.3. Empirical equations for predicting the power required to suspend particles are given in references (M2, W1). Equations for pitched-blade turbines and high-efficiency impellers are given by Corpstein et al. (C4).

In gas-liquid dispersion processes, the gas is introduced below the impeller, which chops the gas into very fine bubbles. The type and degree of agitation affect the size of the bubbles and the total interfacial area. Typical of such processes are aeration in sewage treatment plants, hydrogenation of liquids by hydrogen gas in the presence of a catalyst, absorption of a solute from the gas by the liquid, and fermentation. Correlations are available for predicting the bubble size, holdup, and kW power needed (C3, L1, Z1). For liquids dispersed in immiscible liquids, see reference (T1). The power required for the agitator in gas-liquid dispersion systems can be as much as 10 to 50% less than that required when no gas is present (C3, T2).

Mixing of two fluids can be accomplished in motionless mixers in a pipe with no moving parts. In such commercial devices, stationary elements inside a pipe successively divide portions of the stream and then recombine these portions.

Laminar-flow mixers are used to mix highly viscous mixtures. One type of static mixer has a series of fixed helical elements as shown in Fig. 3.4-7. (Note that most mixers have from six to 20 elements.) In the first element, the flow is split into two semicircular channels and the helical shape gives the streams a 180° twist. The next and successive elements are placed at 90° relative to each other and split the flows into two for each element. Each split in flow creates more interfacial area between layers. When these layers become sufficiently thin, molecular diffusion will eliminate concentration differences remaining. When each element divides the flow into two flow channels (M6),

Equation 3.4-22

where n is the number of elements in series, d is the maximum striation thickness, and D is pipe diameter. When n = 20, then about 10^{6} divisions occur and d is a very small thickness, which enhances the rate of diffusion.

In another commercial type of static mixer (K3), the stream is divided four times for each element. Each element has interacting bars or corrugated sheets placed lengthwise in the pipe and at a 45° angle to the pipe axis. The lengths L of the various types of elements vary from about 1.0 to 1.5 times the pipe diameter. These mixers are also used in turbulent-flow mixing.

In laminar flow with helical mixers the pressure drop (and, hence, power required) is approximately six times as large as that in the empty pipe. In turbulent flow, because of energy losses due to changes of direction, the pressure drop can be up to several hundred times as large (M6, P1). The power loss is typically about 10% of the power of a dynamic mixer (K3). Motionless mixers are also used for heat transfer, chemical reactions, and dispersion of gases in liquids.

In mixing of solid particles or powders it is necessary to displace parts of the powder mixture with respect to other parts. The simplest class of devices suitable for gentle blending is the tumbler. However, it is not usually used for breaking up agglomerates. A common type of tumbler is the double-cone blender, in which two cones are mounted with their open ends fastened together and rotated, as shown in Fig. 3.4-8a. Baffles can also be used internally. If an internal rotating device is also used in the double cone, agglomerates can also be broken up. Other geometries used are a cylindrical drum with internal baffles or twin-shell V type. Tumblers used specifically for breaking up agglomerates are rotating cylindrical or conical shells charged with metal or porcelain steel balls or rods.

Another class of devices for blending solids is the stationary shell device, in which the container is stationary and the material displacement is accomplished by single or multiple rotating inner devices. In the ribbon mixer in Fig. 3.4-8b, a shaft with two open helical screws numbers 1 and 2 attached to it rotates. One screw is left-handed and one right-handed. As the shaft rotates, sections of powder move in opposite directions and mixing occurs. Other types of internal rotating devices are available for special situations (P1). Also, in some devices both the shell and the internal device rotate.

In the mixing of dough, pastes, and viscous materials, large amounts of powder are required as the material is divided, folded, or recombined, and as different parts of the material are displaced relative to each other so that fresh surfaces recombine as often as possible. Some machines may require jacketed cooling to remove the heat generated.

The first type of device for this purpose is somewhat similar to those for agitating fluids, with an impeller slowly rotating in a tank. The impeller can be a close-fitting anchor agitator as in Fig. 3.4-2b, where the outer sweep assembly may have scraper blades. A gate impeller can also be used which has horizontal and vertical bars that cut the paste at various levels and at the wall, which may have stationary bars. A modified gate mixer is the shear-bar mixer, which contains vertical rotating bars or paddles passing between vertical stationary fingers. Other modifications of these types are those where the can or container will rotate as well as the bars and scrapers. These are called change-can mixers.

The most commonly used mixer for heavy pastes and dough is the double-arm kneader mixer. The mixing action is bulk movement, smearing, stretching, dividing, folding, and re-combining. The most widely used design employs two contrarotating arms of sigmoid shape which may rotate at different speeds, as shown in Fig. 3.4-8c.