See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/251539939

Modelling of shear rate distribution in two

planetary mixtures for studying development
of cake batter structure. J Food Eng
ARTICLE in JOURNAL OF FOOD ENGINEERING JULY 2011
Impact Factor: 2.77 DOI: 10.1016/j.jfoodeng.2011.02.044

CITATIONS

READS

40

4 AUTHORS, INCLUDING:
Geoff D. Moggridge

Ian Wilson

University of Cambridge

University of Cambridge

101 PUBLICATIONS 1,331 CITATIONS

255 PUBLICATIONS 2,501 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

SEE PROFILE

Available from: Geoff D. Moggridge

Retrieved on: 17 February 2016

Journal of Food Engineering 105 (2011) 343350

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Modelling of shear rate distribution in two planetary mixtures for studying

development of cake batter structure
A.K.S. Chesterton a, G.D. Moggridge a, P.A. Sadd b, D.I. Wilson a,
a
b

Department of Chemical Engineering & Biotechnology, New Museums Site, Cambridge CB2 3RA, UK
Premier Foods, High Wycombe, Buckinghamshire HP12 3QR, UK

a r t i c l e

i n f o

Article history:
Received 23 November 2010
Received in revised form 14 February 2011
Accepted 20 February 2011
Available online 24 February 2011
Keywords:
Kenwood
Hobart
Planetary mixer
Shear rate

a b s t r a c t
The shear rate experienced by a uid near the wall of a planetary mixer when agitated by a wire whisk
tool has been estimated using a simple geometrical analysis. The bowl and whisk geometries were measured for a Kenwood KM250 and a Hobart N50 mixer which are in widespread use in domestic and laboratory installations. The shear rate is shown to be a maximum at the bowl wall. This value is relatively
uniform over a large fraction of the wall height, except for a small volume near the base and the region
above the maximum width of the mixer. The shear rate prole is sensitive to the vertical positioning of
the agitator within the bowl. For standard manufacturer speed settings, the range of maximum shear
rates was estimated to be 100-500 s1 in the Hobart and 20100 s1 in the Kenwood.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Mixing is a key step in many manufacturing processes as this is
used both to combine ingredients and to generate microstructure.
Understanding the impact of processes such as mixing on product
structure is essential for process improvement and the effect can
be quantied by determining rheological properties (Cullen and
ODonnell, 2009). Industrial processing can involve a wide range
of shear regimes (Table 1): it is therefore important to know the
magnitude and range of shear rates experienced by the material
during mixing for (i) relating results from different types of mixer
(e.g. in comparing results from different laboratories), (ii) scaling
up from laboratory to factory scales, and (iii) determining the relevant range of shear rates for rheological measurement.
Planetary mixers are often used in domestic, laboratory and
industrial applications to prepare food and other soft solid materials. To our knowledge the shear rates developed in these devices
have not been quantied. Rotational and beater tip speeds are
reported but these are not direct indicators of shear rate. We report
an approximate method for calculating representative shear
rates generated in two widely-used laboratory mixers (namely a
Kenwood KM250 and a Hobart N50 mixer).
Our particular interest lies in the preparation of cake batters,
where planetary mixers are widely used to combine solid and liquid ingredients to form a liquid matrix and then to incorporate
air by vigorous mixing. In a companion paper (Chesterton et al.,
Corresponding author. Tel.: +44 1223 334 791; fax: +44 1223 331 796.
E-mail address: Ian_wilson@cheng.cam.ac.uk (D.I. Wilson).
0260-8774/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jfoodeng.2011.02.044

2011) the information on shear rates is used to inform an investigation of the development of structure in aerated cake batters.
1.1. Planetary mixing
Planetary mixers are batch devices which use a characteristic
whisking motion created by two simultaneous rotations. This
arrangement allows the entire mixed volume to receive a vigorous
beating action and ensures efcient mixing whilst minimising agitator diameter, and avoiding high specic power consumption and
temperature rises (Niranjan et al., 1994). Planetary mixing is popular in many industrial applications, particularly within the pharmaceutical and food industries, as it is a reliable and robust way
of homogeneously mixing solidliquid systems (Hiseman et al.,
2002). It is also commonly used in laboratory studies and domestic
applications, as the volume of material can be varied readily and
the progress of mixing monitored visually. Bakeries often use planetary type vertical mixers for the preparation of sponge batters.
Many development or test bakeries use this type of machine to develop a product even though a continuous machine might be used
for production purposes (Cauvain and Cyster, 1996).
Scaling up mixing processes from small scale batch studies to
larger devices with different geometries remains a major challenge
in mixing and powder technology (Delaplace et al., 2007). Classical
measures of mixing performance such as modied power numbers
and Reynolds numbers have been reported (e.g. Delaplace et al.,
2005) but the difculty lies in linking mechanical action to
micro-scale attributes such as distribution of components (e.g. in
solids mixing) or development of microstructure in the materials

344

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

Nomenclature
A
B
C
D
HA
Hh
Hk
r
R
t
VD
x, y
z

point on whisk perimeter of maximum diameter

attachment axis of rotation
centre of bowl
arbitrary point on whisk perimeter
height of maximum radius on whisk tool, mm
height of Hobart mixer bowl, mm
height of Kenwood mixer bowl, mm
radius of rotation of attachment shaft about the bowl
centre, mm
radius of rotation of attachment shaft about its own
axis, mm
time, s
velocity of point D, m s1
Cartesian co-ordinates, Fig. 3
height within the bowl, mm

Table 1
Typical food processes and their associated shear regimes.a

b
a

Source: Stern and Cmolik (1976).

Source: adapted from Steffe (1996).

employed in the food and pharmaceutical sectors. In the area of

cake batters, although planetary mixers have been used for many
years and it is known that the batter structure develops during extended mixing, published information on their operating characteristics is limited (Cauvain and Cyster, 1996).
Some studies of the ow patterns in planetary mixers have been
reported, for a range of materials. Clifford et al. (2004) investigated
liquid mixing in a simplied planetary mixer by visualisation of
dye placed near the surface of a stirred vat of transparent liquid
(golden syrup or glycerol). They observed complex, three dimensional ow patterns. Similarly complex patterns were observed
for granular materials by Hiseman et al. (2002), who studied the
ow of dry and wet (514 wt.% water) lactose powders in a planetary mixer using positron emission particle tracking (PEPT). They
reported two ow regions, independent of agitator speed and level
of ll: (i) a well-mixed central region characterised by rapid motion, and (ii) quasi-stagnant wall layers which were periodically
disturbed by the passage of the mixer blade. The PEPT tracer particles transferred regularly between the two regions, implying that,
averaged over time, all the material experienced similar shear histories. For both liquids and solids, it is clear that simulation of
these shear histories would require extensive computational effort.
Linking material distribution and structure development during
mixing is not straightforward for many food and pharmaceutical
materials as these are frequently soft-solids (Coussot, 2005) whose
structures and rheological behaviour can be stress or strain-rate
dependent. Correlating the shear experienced during mixing with
subsequent product properties has been attempted for granulation
of pharmaceutical products (Schmidt and Kleinebudde, 1999;
Visavarungroj and Remon, 1991) by comparing planetary mixers
(representing low shear devices) to different designs of higher
shear mixers. In such studies the shear regime was described
qualitatively as low and high, and quantitative information was

Xr
XR
hr
hR
d
c_ w

rotational speed of the shaft about the bowl centre,

rad s1
rotational speed of the whisk about its shaft, rad s1
angle created over time due to the revolution of the
attachment shaft about bowl centre
angle created over time due to the revolution of the
attachment about its own axis
gap between the whisk and wall, m
estimated peak shear rate experienced at the bowl wall,
s1

Subscripts
k
Kenwood mixer
h
Hobart mixer

only presented to differentiate the product properties. Knowledge

of the range of shear rates experienced by a material during mixing
is critical to understanding how its microstructure is developed
during processing; to be meaningful, rheological data must be
collected over the shear rate range appropriate for the problem
in question (Steffe, 1996), which may vary widely in industrial processes. Given the complexity of planetary mixers, a reliable spatial
distribution of shear rates and therefore shear history of even a
Newtonian uid is unlikely to be readily calculable, but knowledge
of the range of shear rates and particularly the maximum value
experienced during processing would be useful for comparing
machines and methods.
To our knowledge the estimation of shear regimes during
mixing has not been attempted previously for standard planetary
mixers. Steffe (1996) reported (Table 1) the shear rate range 101
103 s1 for mixing and stirring without detail of the calculation
or geometries involved. Delaplace et al. (2005) studied a planetary
device geometrically different to those considered here a submerged, rotating impeller and correlated the rheological properties over the shear range 0500 s1, but again without justication.
The aim of this work is to characterise the maximum apparent
shear rate experienced by a liquid in commonly used laboratory
designs of planetary mixers. Particular attention was given to mixer speeds used in the preparation of cake batters. Extensional shear
rates were not estimated.
2. Materials and methods
2.1. Equipment
Two planetary mixers with similar mixing capacities and tted
with balloon-whisk agitators were studied: a Kenwood KM250
(Kenwood UK Ltd., Havant) and a Hobart N50-110 (Hobart UK, London). Both whisk attachments were fabricated from stainless steel
wire, around 2 mm in diameter, arranged in loops equidistantly
positioned around the agitator axis to create an open structure
(Fig. 1). The Hobart whisk was used as supplied, with 11 wire
loops, and able to produce cakes with similar volume to industrially baked cakes. The Kenwood whisk was modied; the original
tool was supplied with 5 wire loops, and 5 more were added inside
the whisk, leaving the agitator prole unaffected. The addition of
extra loops improved the volume of cakes produced from batters
prepared with this mixer (data not shown), giving volumes comparable to those produced by the Hobart.
Table 2 summarises the settings and range of speeds available
for each mixer. The rotational speed of a point on the whisk

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

345

Fig. 1. Photographs of (a) Kenwood and (b) Hobart planetary mixer balloon-whisk agitator tools.

was used to plot the bowl and whisk proles and used to measure
the minimum gap distances by use of the softwares circle tool and
radius calculator.

Table 2
Rotation and revolution rates for each mixer speed setting.
Setting

Min
1
2
3
4
5
6
Max
a
b
c

Kenwood
Rotation a Xr/
rpm

Revolutionb
XR/rpm

Hobart
Rotationc Xr/
rpm

Revolutiona
XR/rpm

32
53
66
84
99
128
154
165

105
175
220
280
330
425
515
550

61
125
259

200
410
849

Calculated.
Measured.
Provided by manufacturer.

perimeter (planetary motion, XR) for the Kenwood mixer was

measured using a tachometer and the rotational speed of the
whisk shaft (orbital motion, Xr) was calculated from the gearing
ratio (10/3). The orbital rotational speed, Xr, for the Hobart device
was provided by the manufacturer and the planetary rotational
speed, XR, calculated from the gearing ratio (59/18). For cake
batter production two speeds of rotation are required, one for
the homogenisation of ingredients (slow speed), and one for the
incorporation of air (fast speed). For these studies the optimum slow
speeds were speed setting 1 on each mixer (Table 2); and for the
fast speeds, the maximum speed for the Kenwood and speed setting
3 for the Hobart were used. All speeds were investigated, with
attention given to those used in preparing batter for cake baking.
The bowl and whisk proles (Fig. 2) were measured by tracing a
cast, moulded to the inside of the bowl and outside of the whisk
perimeter, respectively. The cast was fashioned from a length of
10 mm wide fusible metal alloy, which could be manipulated readily at room temperature. This cast was then traced onto 5 mm
graph paper. The proles were generated by plotting the measurements at 5 mm vertical intervals for the majority of the prole,
with the bottom 5 mm measured in 1 mm intervals, and interpolation between the points. AutoCAD 2000 software (Autodesk, Inc.)

2.2. Visualising the whisking pattern

The visualisation methods used for each mixer to conrm the
calculated whisk trajectories were determined by the individual
mixer geometries. Upon removal of the mixing bowl, the Hobart
mixer whisked in normal operation with a clearance of several millimetres between the attachment and mixer base. A at sheet of
paper was inserted into this gap, and a pen attached vertically to
the whisk perimeter (at Rh = 60 mm) which traced the whisked
pattern. The trace was scanned, the colours adjusted to greyscale,
and the levels adjusted for contrast.
The Kenwood mixer operated with the attachment tip submerged in a recess in the mixer base such that the above method
could not be used. The Kenwood mixer did, however, operate in
the lifted position as efciently as in the down position (a trait
not shown by the Hobart). The characteristic whisking action
was followed in the lifted position by using time-lapse photography following a small LED attached to the whisk perimeter. The
battery operated LED was attached to the perimeter at
Rk = 47.5 mm by use of adhesive putty and adhesive tape, and the
battery xed to the attachment shaft. The mixer was set into its
lifted position, and the camera held in place by a clamp stand. A
series of time-lapse photographs were taken over 15 s, in the absence of any light source other than the LED.
3. Analysis of wall shear rate
The proles of the two mixers are shown in Fig. 2. Although
they feature similar bowl volumes, the whisk dimensions differ
noticeably. The planetary action common to both has been analysed by a simple geometrical model for a point, D, on the whisk
perimeter, as shown in plan view in Fig. 3. The planetary motion
involves two opposing rotations. Firstly the whisk shaft (B) rotates
anticlockwise about the bowl centre (C), tracing a circle with radius
r. Simultaneously, the whisk itself rotates clockwise about its shaft,

346

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

z
rk
Hk
Rk

A
HA

60 mm

z
rh
Hh

Rh

A
HA

60 mm

Fig. 2. Bowl and whisk proles: (a) Kenwood (Hk = 155 mm, HA = 70 mm,
Rk = 48.5 mm, rk = 28 mm) and (b) Hobart (Hh = 170 mm, HA = 90 mm, Rh = 64.5 mm,
rh = 30 mm). C denotes the orbital axis of rotation about the bowl, B the axis of
rotation of the revolving whisk and A is the point of maximum diameter on the
whisk prole, located at height HA.

The ratio XR/Xr remains constant for each speed setting, dependent
on the device gear ratio.
The whisk radius, R, varies with height, resulting in a range of
motion patterns and velocities within the bowl. The trajectory calculated for the Kenwood device at Rk = 37.5 mm (i.e. near the maximum, point A, Fig. 2) is illustrated in Fig. 4. The locus describes a
series of 10 loops due to the gearing ratio (10/3) after which the
pattern repeats itself. This result matches experimental observations exactly. Image Fig. 4d) was obtained by attaching an LED to
the whisk at Rk = 37.5 mm and photographing the moving device
in a darkened room with a 15 s extended exposure time.
The gearing ratio for the Hobart device (59/18) results in a 59
loop repeating pattern, as shown in Fig. 5. This calculated pattern
(Fig. 5a) also matches the experimental result (Fig. 5b), acquired
by tracing the trajectory by use of a pen adhered vertically to the
whisk perimeter.
Fig. 6 shows the calculated trajectories for 20 mm intervals between vertical heights z = 1090 mm for each mixer. At the base
(R = 0, data not reported) the whisk tip traces a circle as there is
no contribution from XR. As the whisk diameter, R, increases from
z = 0 with height, the contribution from XR becomes more pronounced causing larger loops in the trajectory. The loops intersect
more often as the height above the base increases. The pattern at
z = 70 mm corresponds to the maximum diameter of the Kenwood
attachment (Rk = 48.5 mm) after which the diameter reduces, and
similar trajectories to those observed at smaller diameters repeat.
The Hobart whisk maximum diameter is located at z = 90 mm
(Rh = 64.5 mm). These trajectories, which resemble Lissajous gures, are similar to those reported for PEPT studies by Hiseman
et al. (2002) and Laurent (2005).
The whisk perimeter experiences a non-constant velocity when
tracing the described patterns. The local linear speed can be obtained from differentiation of Eq. (1) with respect to time, which
yields:

q
V D 2p XR  Xr 2 R2 X2r r 2  2Xr XR  Xr rR cos2pXR t

where VD is the velocity of a point on the whisk perimeter, r and R

are the corresponding radii at that height. The lowest velocity occurs when the rotations (planetary and orbital) oppose one another
directly, which occurs when the point under consideration is closest
to the bowl wall. The velocity is at a maximum as the point passes
the centre of the bowl. For the Kenwood mixer, at maximum radius
(R = 48.5 mm), the velocity at the wall is 0.28 m s1 for speed setting 1 and 1.45 m s1 for the maximum speed. For the Hobart mixer
the corresponding speeds are 0.51 m s1 and 2.18 m s1 at
R = 48.5 mm for settings 1 and 3, respectively. Generally the speed
of the whisk at the wall is higher for the Hobart machine at comparable R values. Although the whisk speed is lowest at closest proximity to the bowl wall, the gap between the whisk and wall, d, is
also at a minimum at this point. The apparent shear rate, c_ w , calculated via
Fig. 3. Plan view of agitator motion. C is the bowl centre and orbital axis, dened as
the origin (see also Fig. 2). B is the whisk axis (Fig. 2) which revolves anti-clockwise
about point C. D is any point on the agitator perimeter, which revolves about B in a
clockwise direction.

tracing a circle with radius R. The trajectory of point D is described

by (Hiseman, 1995).

x r cos2pt Xr R cos2ptXR  2pt Xr

1a

y r sin2pt Xr  R sin2ptXR  2pt Xr

1b

where Xr is the whisk shaft rotational speed about the bowl centre,
XR is the rotational speed of the whisk about its shaft, and t is time.

c_ w

VD
d

is expected to be greatest at this point. The assumption that shear

rate is maximum at the bowl wall is justied later. This shear rate
is used to characterise the mixer action.
The estimated wall shear rate distributions are presented in
Fig. 7. The shear rate distribution is not uniform with height, being
dependent on the whisk speed and gap dimensions, which both
vary with height. There is a signicantly wide range of shear rates
in the region between the base (z = 0) and z  20 mm, but above
z = 20 mm the shear rate is quite uniform, at 30 s1 for the
Kenwood and 110 s1 for the Hobart at speed setting 1 (Fig. 7a),
until z > 80 mm. Whether the mixer and agitator geometries were

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

347

Fig. 4. Trajectories calculated for a point on the circumference of the Kenwood whisk attachment at Rk = 47.5 mm and Xr = 165 rpm after (a) 0.2 s, (b) 0.4 s, and (c) 1.1 s.
Numbers indicate the order of loops traced. (d) Time-lapse photograph collected over 15 s of a small LED attached to the whisk perimeter. Axes scales in mm.

(a)

(b)

100

50

0
-100

-50

50

100

-50

-100

Fig. 5. Hobart mixer trajectory: (a) Plan view of the calculated trajectory for a point on the agitator perimeter at R = 60.0 mm showing a cycle of 59 loops. Axis dimensions in
mm. (b) Trace of corresponding point obtained by attaching a pen to the whisk perimeter.

designed or evolved with experience to deliver this even distribution is not known, but the result explains why these devices have
proven to deliver robust performance.
The form of the wall shear rate distributions do not alter with
rotational speed (Fig. 7b). The wall shear rate values experienced
at most heights of the mixer are 100 s1 for the Kenwood at its

maximum speed and 500 s1 for the Hobart at speed setting 3.
Specic attention is given to these speeds (1, 3, max) due to their
relevance to cake baking, where speed setting 1 is used by both
mixers to homogenise the ingredients, and speed setting 3
(Hobart) and maximum speed (Kenwood) to aerate the batter.
These values lie comfortably within the range reported for mixing

348

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

(b)

(a)

80

80

R = 62.8 mm

70

R = 48.5 mm

70

60

60

50

40

30

R = 56.5 mm

50

R = 45.3 mm

z / mm

z / mm

R = 64.5 mm

90

R = 46 mm

90

40

R = 45 mm

30

R = 39 mm

20

20

10

R = 24.3 mm

10

R = 27 mm

0
100

100
50

50
0

0
-50

50

50
0

0
-50

-50
-100

y / mm

100

100

-50
-100

-100
x / mm

y / mm

-100
x / mm

Fig. 6. Trajectories of ve points on the whisk perimeter for (a) Kenwood and (b) Hobart devices at z = 10, 30, 50, 70 and 90 mm (corresponding R values marked). The
location of largest R is shown for each mixer; Rk = 48.5 mm and Rh = 64.5 mm.

Fig. 7. Estimated wall shear rate at (a) speed setting 1, and (b) maximum speed, for Kenwood (solid grey symbols) and Hobart (open black symbols) mixers with balloon
whisk attachment. Interpolated lines between data points are added to guide the eye. One noticeable outlier exists for the Hobart mixer, at z = 25 mm: this is a measurement
artefact. Horizontal lines show ll levels when the bowl is lled with the homogenised ingredients (slurry) and aerated batter (foam) for the Kenwood mixer (grey) and the
Hobart mixer (black). Clearance at base = 2 mm.

operations in Table 1. The shear rates estimated for all other

available speed settings can be found in summary in Table 3.
The shear rates imposed by the two devices are different; the
Hobart gives approximately four times higher shear rates than
the Kenwood at speed setting 1, and ve times greater at the
maximum speed. This is due to (i) the differences in tool width,
the Hobarts being generally larger, and (ii) the gap size between
the whisk and bowl, which is generally smaller in the Hobart

machine although both approach 2 mm near the base of the

bowl. The Hobart device consistently imposes substantially higher
shear rates on the material being mixed in the peak shear zone
than the Kenwood.
Fig. 7 also shows the measured ll levels for the homogenised
ingredients (labelled slurry) and aerated batter (labelled foam)
stages of cake batter production, when the batch volume is 1 L.
When this batter level is not exceeded (which is the case with

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

the experiments reported in the companion paper), the lower shear

rates associated with the top of the bowl are not relevant. The
shear rate, however, also decreases below z = 20 mm approaching
zero and then rises dramatically at the whisk tip. The volume of
mixed material associated with this region of higher shear rate is
relatively small, at 12 ml for the Kenwood mixer and 4 ml for the
Hobart. Since the volume involved is small and the gap size is of
the order of 2 mm or less, it is assumed these shear rates will
not be experienced by the majority of the material at any given
time.
The calculated shear rate distribution presented in Fig. 7 is sensitive to the height of the whisk attachment relative to the bowl,
which can be modied by changing the vertical position of the
whisk within the bowl (for the Kenwood only; the Hobart whisk
location could not be altered). The effect of varying the clearance
on the magnitude and distribution of wall shear rates was investigated and the results are summarised in Fig. 8. Values of 2 mm (the
minimum practicable for this bowl design, and that for which
Figs. 6 and 7 were derived), 3, 5, 10 and 14 mm were considered.
The Figure shows that the shear rate distribution is most uniform
over the volume of interest at a clearance of 5 mm. A clearance
of 5 mm was therefore employed when studying batter properties.
These estimates of the shear rate at the wall make several
assumptions. Firstly, it is assumed that the maximum shear rate
occurs at the wall, because the smallest distance between moving
parts is located there. As the whisk approaches the wall, its speed
decreases but the gap size also decreases. Calculation of the change
in shear rate when the whisk approaches and then leaves the bowl
wall is a complicated exercise in 3D geometry, but analyses can be
attempted for points near the top of the bowl (where the wall and
whisk are approximately vertical). Calculations of this quasi 2D region (Fig. 9) conrm that the apparent shear rate is a maximum
when the whisk perimeter is closest to the bowl wall, and decreases as the tool moves away. The minimum shear rate is found
at the bowl centre, which then rises at the next wall apogee. The
extent of change in shear rate on approach to the bowl wall varies
with rotational speed and bowl height but the maximum does not
shift from the wall position.
Secondly, the whole charge of material is assumed to be subject
to the estimated wall shear rate for some fraction of its time in the
mixer. Hiseman et al. (2002) found that tracer particles regularly
transfer between the wall and bulk regions, when studying granulation using positron emission particle tracking (PEPT). The residence time of material in the wall zone has not been calculated,
but a perfect mixing argument would suggest that this would be
related to the volume of the wall region relative to the lled volume of the bowl (around 30%). Tracer studies would be needed
to conrm this.
In these studies the majority of material was subject to a uniform peak shear rate during mixing, namely 30 s1 during homogenisation (lower speed) and 100 s1 during aeration (high speed)
for the Kenwood mixer, and 100 s1 during homogenisation
and 500 s1 during aeration for the Hobart mixer. Only a small
Table 3
Average maximum wall shear rate for 1 L volume charge. Values are reported for each
mixer speed setting.
Setting

Kenwood c_ w /s1

Hobart c_ w /s1

Min
1
2
3
4
5
6
Max

20
30
3540
4050
5070
7090
80120
90120

100120
210250
450500

349

Fig. 8. Calculated wall shear rate proles for the Kenwood mixer for different
values of whisk-bowl clearance, at the maximum rotational speed. Horizontal lines
indicate ll level for slurry (dotted line) and for aerated batter (dashed line).

Fig. 9. Variation in shear rate for one revolution (one rotation of the whisk about
axis B (Fig. 1) calculated for a point on the Hobart whisk perimeter (D, Fig. 1) at
z = 60 mm. The peak shear rate is located at the origin (0 rad), at which point D is at
its closest to the bowl wall. Conditions: clearance 2 mm, maximum speed.

fraction of the material experienced higher shear rates (at the

base), corresponding to a maximum of 1.2% of the charge in the
high speed case.
4. Conclusions
The geometrical model of the motion within a planetary mixer
reported by Hiseman et al. (2002) has been employed to estimate
the wall shear rate generated in planetary mixers. Experimental
visualisations of the whisk motion conrmed the validity of the
trajectory model. The peak shear rate was assumed to occur at
the bowl wall, where the whisk speed is lowest but the gap between whisk and bowl wall is at a minimum. This assumption
has been conrmed for the case where the bowl wall is vertical,

350

A.K.S. Chesterton et al. / Journal of Food Engineering 105 (2011) 343350

but its validity is not certain near the base where the curvature is
greater (and the 3D analysis was not performed). The latter region
represents a relatively small volume if the bowl is lled to a height
of 50 mm or greater: the latter value is suggested as a minimal level of ll for the devices employed here.
The distributions of shear rate with height obtained for two
commonly used laboratory scale devices, namely a Kenwood
KM250 and a Hobart N60 mixer, were found have a similar form.
The shear rates imposed at the wall were near uniform for the
majority of the wetted bowl region, despite the whisk speed and
wall gap both being a function of height within the bowl. The range
in shear rates achievable with the Kenwood was 20100 s1, while
the corresponding range for the Hobart was 100500 s1.
Acknowledgements
A Food Processing Faraday studentship for AKSC is gratefully
acknowledged, as is funding and technical support from Premier
Foods.
References
Cauvain, S.P., Cyster, J., 1996. Sponge cake technology. Campden and Chorleywood
Food Research Association, Chipping Campden.
Chesterton, A.K.S., Meza, B.E., Moggridge, G.D., Sadd, P.A., Wilson, D.I., (2011).
Rheological characterisation of cake batters generated by planetary mixing:
elastic versus viscous effects. Journal of Food Engineering 105 (2), 332342.

Clifford, M.J., Cox, S.M., Finn, M.D., 2004. Reynolds number effects in a simple
planetary mixer. Chemical Engineering Science 59 (16), 33713379.
Coussot, P., 2005. Rheometry of pastes, suspensions, and granular materials:
applications in industry and environment. John Wiley & Sons Inc..
Cullen, P.J., ODonnell, P., 2009. Mixing in the food industry: trends and challenges.
In: Cullen, P.J. (Ed.), Food Mixing Principles and Applications. Wiley-Blackwell.
Delaplace, G., Guerin, R., Leuliet, J.C., 2005. Dimensional analysis for planetary
mixer: modied power and Reynolds numbers. AIChE Journal 51 (12), 3094
3100.
Delaplace, G., Thakur, R.K., Bouvier, L., Andre, C., Torrez, C., 2007. Dimensional
analysis for planetary mixer: mixing time and Reynolds numbers. Chemical
Engineering Science 62 (5), 14421447.
Hiseman, M.J.P., 1995. Positron emission particle tracking analysis of granulation in
a planetary mixer. Department of Chemical Engineering, University of
Cambridge, Cambridge.
Hiseman, M.J.P., Laurent, B.F.C., Bridgwater, J., Wilson, D.I., Parker, D.J., North, N.,
Merrield, D.R., 2002. Granular ow in a planetary mixer. Chemical Engineering
Research & Design 80 (A5), 432440.
Laurent, B.F.C., 2005. Structure of powder ow in a planetary mixer during wetmass granulation. Chemical Engineering Science 60 (14), 38053816.
Niranjan, K., Smith, D.L.O., Rielly, C.D., Lindley, J.A., Phillips, V.R., 1994. Mixing
processes for agricultural and food materials. 5. Review of mixer types. Journal
of Agricultural Engineering Research 59 (3), 145161.
Schmidt, C., Kleinebudde, P., 1999. Inuence of the granulation step on pellets
prepared by extrusion/spheronization. Chemical & Pharmaceutical Bulletin 47
(3), 405412.
Steffe, J.F., 1996. Rheological Methods in Food Process Engineering. Freeman Press.
Stern, P., Cmolik, J., 1976. Study of rheological properties of margarine. Journal of
the American Oil Chemists Society 53 (10), 644647.
Visavarungroj, N., Remon, J.P., 1991. Cross-linked starch as binding-agent. 3.
Granulation of an insoluble ller. International Journal of Pharmaceutics 69 (1),
4351.