Velocity Field in A Twin-Screw Extruder PDF

International Journal of Food Science and Technology 1997, 32, 241–253 241
Velocity field in a twin-screw extruder
Serafim Bakalis & Mukund V. Karwe*

Department of Food Science, The State University of New Jersey, US
Summary Velocity measurements inside the screw channels of a co-rotating, self-wiping, twin-screw
extruder were carried out. A laser Doppler anemometer (LDA) system was used to mea-
sure tangential and axial velocity components in the translational region of one of the
screws of a twin-screw extruder. The velocity measurements were carried out using heavy
corn syrup at different screw speeds and for two forwards conveying elements differing
in their pitch. It was found that velocity profiles in the translational region were sub-
stantially affected by the pitch of the screw element, however, screw speed (30–90 r.p.m.)
did not have a significant effect on the shape of the velocity profiles. The range of veloc-
ity values in a 14 mm pitch screw element was wider than that in a 28 mm pitch screw
element. Experimentally measured velocity profiles were compared with predictions of a
one-dimensional model, which showed that the model underpredicted the velocity values,
for the 14 mm pitch element.
Keywords Corn syrup, laser Doppler anemometry, Newtonian fluid.
The flow in a co-rotating twin-screw extruder

Introduction
has been simulated using numerical methods by
The use of the twin-screw extrusion process in the several investigators (Sastrohartono et al., 1992;
manufacture of food, plastics and pharmaceutical Tayeb et al., 1992; Yacu, 1985; Van Zuilichem et
products has increased substantially in the last al., 1992). It must be pointed out that all of the
decade. Substantial efforts have been devoted to models that are developed to describe the trans-
develop mathematical models of twin-screw port phenomena in an extruder must be validated
extrusion processes. The solution of these mathe- with experimental data. Very often the compari-
matical models is usually obtained using different son is made in terms of quantities such as die
discretizaton methods. Due to the complicated pressure and die temperature due to ease of mea-
geometry of a twin-screw extruder and complex surement of these quantities. Comparison of
rheological behaviour of many of the materials experimental results with numerical predictions
that are extruded, the numerical simulations of on a local scale has been attempted by only a few
three-dimensional flow in the screw channels is investigators. For example, Esseghir & Sernas
carried out using a variety of methods after (1994) measured the radial temperature distribu-
employing some simplifying assumptions. These tions in the screw channel of a single-screw
simulations are usually computational resource extruder, using a cam-driven thermocouple syn-
intensive. chronized with the screw shaft rotation. Another
way of making local comparisons would be in
terms of velocity distribution at various locations
*Correspondent: M. V. Karwe, Department of Food
in an extruder.
Science, PO Box 231, Cook College, Rutgers, The State
University of New Jersey, New Brunswick, New Jersey Most of the data for measured velocity distrib-
08903–0231, US. Fax: 1 1 732 932–8690. utions reported in the literature are for single-
e-mail: [email protected] screw extruders. Velocity distributions in the
© 1997 Blackwell Science Ltd

242 Velocity field in a twin-screw extruder S. Bakalis & M. V. Karwe
screw channels of a single-screw extruder were transfer flow, was also detected with the LDA
obtained either by tracing particles or ink inject- technique.
ed in the flow (Choo et al., 1980; Eccher & This paper reports velocity measurements
Valentinotti, 1958) or by using a non-invasive under isothermal conditions for a Newtonian
technique such as NMR (McCarthy et al., 1992). fluid in the translational region of a co-rotating
Eccher & Valentinotti, using photographic tech- twin-screw extruder, for two different forwards-
niques, reported some cross-wise circulatory conveying screw elements, differing in their pitch.
motion within the screw channel. The correctness The effect of screw speed on the measured veloc-
of the shallow-channel theory in a single-screw ity distribution was also examined. Although
extruder was confirmed by Mohr et al. (1961). most food materials have a non-Newtonian
Numerical results obtained using a finite- behaviour, a Newtonian fluid was chosen as a
difference method were found to be in reasonably model fluid for the initial investigation reported
good agreement with velocity measurements per- in this paper. Future investigations will include
formed in deep-channel screw elements by Choo model non-Newtonian fluids.
et al. (1980).
Few data for experimentally measured velocity
patterns have been published for a twin-screw Materials and methods
extruder. Karwe & Sernas (1996) showed that
Experimental set up
laser Doppler anemometry (LDA) can be used to
measure velocity distributions in a twin-screw The experimental set up consisting of a twin-
extruder using model fluids. They measured axial screw extruder and a laser Doppler anemometer
and tangential velocity profiles in the translation- system is shown schematically in Fig. 1. A four-
al region of a twin-screw extruder using corn beam 300-mW argon-ion LDA system (Dantec
syrup for a 28 mm screw element at a fixed screw Measurement Technology, Inc., Mahwah, New
speed of 30 r.p.m. The tangential velocity pat- Jersey) was used to obtain the velocity data.
terns were compared with numerically predicted Using this apparatus two velocity components
profiles of Chiruvella et al. (1996). The measured can be measured simultaneously. The movement
values were higher than the numerically predicted of the laser probe, which was used to focus the
values. laser beams at a desired location in the screw
Karwe et al. (1997) have reported velocity mea- channel, was controlled by a 3–D traverse inter-
surements in the nip region of a twin-screw faced with a micro-computer. Details of the LDA
extruder. The measured velocity distributions measurement technique and the experimental
indicated a three dimensional flow. Leakage flow apparatus have been described by Karwe &
between the two screws, sometimes referred to as Sernas (1996).
Figure 1 Schematic diagram of

the extruder and the LDA system.
International Journal of Food Science and Technology 1997, 32, 241–253 © 1997 Blackwell Science Ltd
Velocity field in a twin-screw extruder S. Bakalis & M. V. Karwe 243
Velocity measurements were made in the screw

channels of a ZSK-30 twin-screw extruder
(Werner & Pfleiderer Corp., Ramsey, New Jersey).
This is a co-rotating, self-wiping, twin-screw
extruder. The internal diameter of each of the bar-
rel bores (D) was 30.85 mm and the L/D ratio was
29, where L is the total length of the screws. The
maximum channel depth for the screws is 4.77
mm. A Plexiglas window was constructed and Figure 2 Sketch of two types of screw elements:
fitted into the vent port of the extruder to access (a) 28/28 along with encoder positions (b) 14/14.
the flow in the screw channels for velocity mea-
surements (see Fig. 1). The screw configuration conveying screw elements having pitch(mm)/
used in these experiments was assembled by using length(mm) ratios of 28/28 and 14/14. These are
forwards-conveying elements (low shear) only. shown schematically in Figs 2(a) and 2(b), respec-
The die used had two 5-mm long holes of 3.6 mm tively. The shape of the screw channels is self wip-
diameter each. The hopper through which the ing and the equation describing the geometry is
screws were fed was always kept full. The screw given by Booy (1980).
elements in the feed section had wider screw chan-
nels as compared with the rest of the screw ele-
Region of measurement
ments. This ensured that the two screws were
completely filled with syrup. This also resulted in The region of velocity measurements is shown in
different mass flow rates at different screw speeds, Fig. 3. The front and top views of a section taken
like a single-screw extruder. normal to the axes of the twin-screw extruder are
LDA measurements require optically transpar- shown in Figs 3(a) and 3(b), respectively. The
ent materials for laser beams to penetrate and intermeshing (nip) region is the region where the
scattered light to come out. Therefore, heavy corn two screw channels meet. The translational region
syrup (Globe corn syrup 1132, Corn-Products is the region away from the intermeshing region.
International, Summit-Argo, Illinois), which is a Velocity measurements reported in this paper
Newtonian fluid of 1425 kg m23 density, 1.49 were carried out at the centre of the left screw as
refractive index and shear viscosities of 74.0, 17.5 seen in the top view. This region is about a quar-
and 0.54 Pas at 26.6, 37.8 and 48.9i8C, respec- ter turn away from the nip region.
tively, was selected as the working fluid.
Encoder
Screw elements
An encoder was used to keep track of the angular
Velocity profiles were measured for two forwards- position of the screw shaft. It was fitted on the
Figure 3 Measurement geometry:

(a) Section taken normal to the
axes of the screws and (b) Top
view as seen through the optical
window showing the nip and the
translational regions.
© 1997 Blackwell Science Ltd International Journal of Food Science and Technology 1997, 32, 241–253
back end of the left screw shaft. The encoder gave Dz 1 2 sin2 u
a 2.5 VDC pulse for every 0.72 degrees of shaft Da
5m !
1 2 m2 sin2 u (1)
rotation (500 pulses per revolution) and a 5 VDC
pulse at every full rotation. This signal was where Da is the distance the laser probe is moved
received by the flow velocity analyser (FVA, in in the air; Dz: is the distance the intersection point
Fig. 1) and it was combined with the measured of the laser beams is moved in the medium hav-
velocity data. A reference position of 08 was set ing refractive index m; and u is the half angle
when the left screw lobe was vertical, as shown in between the two laser beams in air 5 98.
Figs 2(a) and 3(a). This was arbitrary. However, For sinu « 1
it was not important where the 08 position was Dz sin u
selected as it did not affect the results. From the <m5
Da sin b (2)
encoder position, the axial location in the screw
channel where velocity data were obtained was In the derivation of eqn 1 the refraction of the
known. The right screw shaft had a phase laser beams at the interface between the Plexiglas
difference of 908 with respect to the left shaft. and the corn syrup was neglected because their
Hence, the orientation of one screw lobe auto- refractive indices are very close to each other, i.e.
matically defined the orientation of the other mplexiglas 5 1.507, mcorn syrup 5 1.49.. The error due to
screw lobe. the approximation in eqn 2 was about 0.5%.
Correction for refraction Results and discussion

Due to refraction of the laser beams at the inter- Velocity measurements were carried out for two
face between air and the surface of the Plexiglas forwards-conveying screw elements differing in
window, and between the Plexiglas and the corn their pitch. They were 14 mm and 28 mm,
syrup in the screw channel, a correction factor respectively, as shown in Fig. 2. Velocity mea-
had to be applied to the apparent location where surements were carried out at three different
the laser beams meet inside the screw channels. A screw speeds of 30, 60 and 90 r.p.m. The LDA
schematic diagram to illustrate this is shown in system was used to measure tangential, Ux, and
Fig. 4. The correction factor for the location of axial, Uy, velocity components, as shown in Fig.
the point of intersection of the laser beams was 3. As the screw rotated, the relative position of
derived, which is as follows: the measurement volume (intersection of the four
beams) changed. Thus it was necessary to keep
track of the relative screw position, which was
achieved by using the encoder. When the screw
flights cut the measurement volume, no signal
was obtained and this appeared as a gap in the
data. When the measurement volume was moved
deeper in the screw channel (along -z direction) a
wider gap in the data was observed, because the
flight width was higher near the screw root. For
the coordinate system shown in Fig. 3, the tan-
gential velocity values are expected to be positive
and the axial velocity values to be negative.
Positive values of Uy imply flow towards the hop-
per. In this paper the velocity distributions are
presented in a dimensionless form. The velocity of
the screw root was chosen as the characteristic
velocity and the channel depth (4.77 mm) as the
Figure 4 Refraction of laser beams at the interfaces characteristic length.
between air and Plexiglas. The zero position along the z direction was
taken at the point where the four beams focused no-slip boundary condition. Thus, near the screw
on the inner surface of the barrel. The error in the flight, the only non-zero component of the fluid
determination of z 5 0 position was estimated to velocity is Ux*. As can be seen from Figs 2 and
be about 0.4 mm which is about half the size of 3, angular positions of 08 and 1808 correspond to
the measurement volume formed by the intersec- measurements close to the screw flight tip. This
tion of the four laser beams. This will cause some was confirmed by the results shown in Fig. 5.
uncertainty or shift in the z direction. The top of Close to the screw flight tip (08 and 1808) the
the screw flight has a finite width (0.5 mm) which experimentally measured Ux* value was very
is larger than the width of the intersection volume close to the calculated solid body velocity Us at
(≈ 0.2 mm). Therefore, the zero encoder position the screw tip (Ut 5 V Dt, V the angular speed of
may get slightly shifted to left or right within the screw, Dt the screw tip diameter, 30.7 mm).
different data sets. This was corrected by shifting Since the barrel of the extruder does not move,
all the data sets by an appropriate phase shift. To the velocity of the fluid close to the barrel should
calculate the phase shift, the position of the screw be zero, except when a screw flight passes by.
flight in each data set was identified at the encoder This was experimentally confirmed as shown in
angle where the first data point was obtained. This Fig. 5. Velocity values between the flights are
position was slightly different in different repeti- close to zero near z 5 0 (barrel).
tions. An average encoder position was taken as The data presented in Fig. 5 are scattered. This
the correct position of the screw flight and all the scatter was observed close to the barrel and the
data set were shifted so that the first data point screw root in a number of repetitions. The scatter
appeared at the same encoder position. in the data reduced substantially when the
In Fig. 5 the Ux* (5 Ux/Us) and Uy* (5 measurement location was moved away from the
Uy/Us) are plotted against the angular position of barrel wall and the screw root, as will be seen in
the left screw shaft, for one screw channel of a 14 the subsequent figures. The reasons for scatter
mm pitch screw element, near the extruder barrel near the walls are not clear. It could be due to
(z 5 0), at a screw speed of 90 r.p.m. Here Us is partial cutting of the measurement volume by the
the screw root velocity (Us 5 V Ds, V the angu- walls.
lar speed of the screw, Ds the screw root diame- In Fig. 6, Ux* and Uy* are plotted vs. angular
ter, 21.26 mm). For all the figures in this paper, position near the screw root, for a 14 mm pitch
data from three repetitions are superimposed. The screw element, at a screw speed of 90 r.p.m. Here,
screw rotates only along the x direction. Near the the velocity of the fluid should be close to the
screw flight, the velocity of the fluid should be velocity of the screw root due to the no-slip
close to the velocity of the screw flight due to the boundary condition. Since the screw rotates only
Figure 5 Axial and tangential

velocity variation with angular
position in a 14 mm pitch screw
element at a screw speed of 90
r.p.m., near the surface of the
extruder barrel (z 5 0) for corn
syrup at 27i8C and mass flow rate
0.0108 kg s21.
Figure 6 Axial and tangential

velocity variation with angular
position in a 14 mm pitch screw
element, at a screw speed of 90
r.p.m., near the screw root
(z 5 24.77 mm), for corn syrup
at 27i8C, and mass flow rate
0.0108 kg s21.
along the x direction, Uy* should be zero and In Fig. 8, variations of Uy* vs. angular posi-
Ux* should be equal to the calculated screw root tion for 14 and 28 mm screw elements are plotted
velocity, as seen from the experimental results. at a screw speed of 90 r.p.m. and at z 5 –2.32
In Fig. 7, the variations of Ux* with angular mm. Since the screw rotates only along the x
position are plotted for 14 and 28 mm pitch screw direction, close to the screw flight the Uy* values
elements, at a screw speed of 90 r.p.m. and at z were close to zero. The magnitude of Uy*
5 –2.32 mm, which is approximately half-way smoothly increased and became maximum in the
between the barrel and the screw root. The max- middle of the channel. No significant difference
imum for the tangential velocity is seen to occur between the Uy* velocity profiles in 14 and 28
closer to the screw flights. This velocity value mm pitch screw elements could be observed.
could be determined by the no-slip condition and Since the mass flow rate and the cross-sectional
thus is the same for both screw elements, since areas are the same for both screw elements, it is
both rotate at the same angular speed. The not surprising that the Uy* velocity profiles for
velocity values, for both screw elements, smooth- both screw elementa are similar, especially in the
ly decreased towards the middle of the channel, middle of the screw channel.
where the value of the Ux* was minimum, but The variations of dimensionless velocity, for
not zero. In the case of the 14 mm pitch screw the 14 mm pitch screw element are plotted in Fig.
element, the velocity profiles were steeper, when 9, at screw speeds of 60 and 90 r.p.m., at a fixed
compared to the 28 mm pitch screw element. depth (z 5 – 2.32 mm). It can be seen from the
Figure 7 Tangential (Ux*) velocity

variation with angular position in
a 14 and a 28 mm pitch screw
element, at a screw speed of
90 r.p.m., in the middle of the
screw channel (z 5 22.32 mm),
for corn syrup at 27i8C, and mass
flow rate 0.0108 kg s21.
Figure 8 Axial (Uy*) velocity

variation with angular position in
a 14 and a 28 mm pitch screw
elements, at a screw speed of 90
r.p.m., in the middle of the screw
channel (z 5 22.32 mm), for
corn syrup at 27i8C and mass
figures that no significant difference could be In Fig. 11 Uy* velocity values are plotted along
observed between the velocity profiles for this the z direction for a 14 mm pitch screw element
range of screw speeds. This implies that the axial at an encoder position of 90 degrees, at different
pressure gradient in the 14 mm pitch screw ele- screw speeds. Since the boundaries (screw root
ment was small when compared to a 28 mm pitch and barrel) do not move along the y direction
screw element. Uy* is zero.
In Fig. 10, the Ux* profiles are plotted along The velocity field in the translational region in
the z direction, within the screw channel at the a co-rotating twin-screw extruder is similar to the
encoder position of 90, which corresponds to the flow in the screw channels of a single-screw
middle of screw channel, for the 14 mm pitch extruder (Chiruvella et al., 1996). Therefore, the
screw element, at different screw speeds. The lines flow can be considered as the superposition of a
in the graph represent third order polynomials drag and a pressure flow. The drag flow is pro-
obtained from a least square fit to the experi- portional to the screw speed. The pressure flow is
mental data. The purpose of these lines is just to a function of the die opening. Since the extruder
show the trends of the three velocity profiles. is fully filled, the net flow rate depends on the
Near the barrel (z/H 5 0) Ux* is zero, since the screw speed for a fixed die. Therefore both com-
barrel is stationary. Near the screw root (z/H 5 ponents of the flow depend on the screw speed,
–1) Ux* takes the value of the screw root veloci- but in a different way. This implies that the con-
ty, which is close to one. The velocity profiles at tribution of the pressure flow, for the examined
different screw speeds are not significantly range of flow rates was negligible in the 14 mm
different from each other. pitch screw element.
Figure 9 Ux* and Uy* velocity

variations with angular position
in a 14 mm pitch screw element,
at screw speeds of 60 and 90
r.p.m., in the middle of the screw
channel (z 5 22.32 mm), for
corn syrup at 27i8C.
Figure 10 Ux* velocity profiles in

a 14 mm pitch screw element, for
different screw speeds, in the
middle of the screw channel
(90 degrees), for corn syrup at
27i8C. Lines represent third order
polynomials, obtained using the
least squares method.
Figure 11 Uy* velocity profiles in

different screw speed, in the
In Fig. 12, Ux* velocity profiles are plotted ment are higher than those generated in a 14 mm
along the z direction for a 28 mm pitch screw ele- pitch screw element, for the range of the screw
ment, at an encoder angle of 90 degrees, for speeds considered here.
different screw speeds. In this case the velocity In Fig. 13, the Uy* velocity profiles are plotted
profiles seem to be significantly different when against z/H for a 28 mm screw element for
compared with those for the 14 mm pitch screw different screw speeds. As expected the velocity
element. The maximum velocity seems to occur values are zero at the boundary and maximum in
away from the screw root. The effect of the pres- the middle. The screw speed did not significantly
sure flow is evident in this case. This implies that affect the velocity profiles, though velocity magni-
the pressure gradients in a 28 mm pitch screw ele- tudes were different because of different flow rates.
Figure 12 Ux* velocity profiles in

Figure 13 Uy* velocity profiles in

for 2H #_z # 0. All symbols are defined in the

nomenclature.
In order to compare the results derived from
this simplified analysis with the experimental
results, the experimental results were transformed
to down channel and barrel moving formulation
(Karwe et al., 1995) as follows:
U§ experimental 5
2 Ux cos f 2 Uy sin f 1 Us cos f (5)
Uj experimental 5
2 Ux sin f 1 Uy cos f 2 Us sin f (6)
Figure 14 Downchannel and axial coordinate systems.
The experimentally obtained velocity profiles at

a screw speed of 90 r.p.m. in the middle of the
screw channel for a 28 mm pitch screw element,
Theoretical considerations
and the theoretically derived velocity profiles
For isothermal flow of a Newtonian material, in shown in Fig. 15. As one can see, the experimen-
a single-screw channel, assuming no-slip at walls tal values were reasonably close to the theoreti-
and ignoring inertia and body forces, and for cally obtained velocity values. In the case of the
W/D » 1, H/D « 1, the velocity profiles along the 14 mm pitch screw element, the deviation from the
down channel (z) and cross channel (j) directions, predicted velocity profile was substantial as shown
in barrel moving formulation (see Fig. 14) are in Fig. 16. This deviation can be explained by the
given in several textbooks, and are as follows: fact that the W/H ratio is 2.3 for a 28 mm pitch
screw element, and for a 14 mm screw element the
1 211H2 1 1H22 1
6Q Z 2 Z W/H ratio is 1.18. Thus, it is not surprising that
U§ theoretical 5 3Us cos f 2
WH the one dimensional model which is valid for large
values of W/H did not accurately predict the
1H2 1 U cos f
Z
Us cos f s velocity distributions in a 14 mm pitch screw ele-
(3)
ment. Also it should be noted that the compar-
isons reported here are in the middle of the screw
11H2 1 1H22 2
Z 2 Z
Uj theoretical 5 s2 3Us sin fd channel. As the W/H ratio increases the wall
effects on the velocity profile at the centre of the
1H2 2 U sin f
Z channel diminish, and the assumptions employed
Us sin f s
(4) in the one dimensional model become valid.
Figure 15 Theoretical and

experimental down and cross
channel dimensionless velocity
profiles for a 28/28 screw element,
at a screw speed of 90 r.p.m., for
corn syrup at 27i8C, and mass
Figure 16 Theoretical and

experimental down and cross
channel dimensionless velocity
profiles for a 14/14 screw element,
at a screw speed of 90 r.p.m., for
corn syrup at 27i8C, and mass
Velocity contour plots for different experimen- Conclusions

tal conditions were generated from the velocity
The axial and tangential velocity profiles in the
data. In Fig. 17 contours of the dimensionless
screw channel of a co-rotating twin-screw extrud-
tangential velocity Ux* for a 14 mm pitch screw
er were significantly affected by the pitch of the
element at a screw speed of 90 r.p.m. are shown.
screw element, and indicated that in a larger pitch
As expected, velocity values close to the barrel
screw element higher tangential pressure gradients
were zero. Maximum Ux* velocity values were
are produced. The measured velocity profiles in
seen close to the screw flight. Velocity values
the middle of the screw channel of a 28 mm pitch
smoothly decreased towards the centre of the
screw element could be predicted by a one dimen-
screw channel. A slight asymmetry was observed
sional model, but this model substantially under-
in the contour. This probably was due to experi-
predicted the velocity values for a 14 mm pitch
mental error.
screw element.
In Fig. 18 contours for the dimensionless axial
velocity Uy* are plotted. The velocity values close
to the barrel were zero. Also close to the screw
Acknowledgements
root, axial velocity values were zero. Maximum
axial velocity values are observed in the middle of This is manuscript D-10127–1-96 of the New
the screw channel, but closer to the barrel. Jersey Agricultural Experimental Station. This
Figure 17 Tangential velocity

(Ux*) contours in a 14 mm pitch
screw element at a screw speed of
90 r.p.m., for corn syrup at
27i8C, and mass flow rate
0.0108 kg s21.
research was supported by state appropriations, s Screw root

U.S. Hatch Act Funds. Equipment support was t Screw tip
x Tangential direction
provided by the Centre of Advanced Food
y Axial direction
Technology, Rutgers University, New Brunswick, b Angle between the refracted laser beam and the
New Jersey. The authors would like to thank radius vector (Fig. 4)
Dr V. Sernas for useful discussions on the Dz Distance the intersection point of the laser beams is
subject. moved in the fluid
Da Distance the laser probe is moved in the air
z Down channel direction
u Half the angle between the two laser beams
Appendix
m Refractive index
Nomenclature j Cross channel direction
f Screw helix angle
D Diameter V Angular speed of the screw
H Maximum channel depth
L Axial length of the screws
Q Volumetric flow rate Superscript
U Velocity
W Maximum channel width * Dimensionless velocity
Figure 18 Axial velocity (Uy*)

contours in a 14 mm pitch screw
element at a screw speed of
90 r.p.m., for corn syrup at
27i8C, and mass flow rate
0.0108 kg s21.
References measurements in the nip region of a co-rotating twin-

screw extruder using laser Doppler anemometry.
Booy, M.L. (1980). Isothermal flow of viscous liquids in Advances in Polymer Processing, in press.
co-rotating twin screw devices. Polymer Engineering and Karwe, M.V., Chiruvella, R.V. & Jaluria, Y. (1995).
Science, 20, 1220–1228. Coordinate system independence of shear rate during
Chiruvella, R.V., Jaluria, Y., Karwe, M.V. & Sernas, V. isothermal single-screw extrusion of a Newtonian fluid.
(1996). Transport in a twin-screw extruder for the Journal of Food Process Engineering, 18, 55–69.
processing of polymers. Polymer Engineer Science, 36, McCarthy, K.L., Kauten, R.J. & Agemura, C.K. (1992).
1531–1540. Application of NMR imaging to the study of velocity
Choo, K.P., Neelakantan, N.R. & Pittman, J.F.T. (1980). profiles during extrusion processing. Trends in Food
Experimental deep channel velocity profiles and Science and Technology, 3, 167–190.
operating characteristics for a single screw extruder. Mohr, W.D., Clapp, J.B. & Starr, F.C. (1961). Flow
Polymer Engineering and Science, 20, 349–365. patterns in a non-Newtonian fluid in a single-screw
Eccher, S. & Valentinotti, A. (1958). Experimental extruder. Society of Plastic Engineers Transactions, 1,
determination of velocity profiles in an extruder screw. 113–120.
Industrial and Engineering Chemistry, 50, 829–836. Sastrohartono, T., Jaluria, Y. & Karwe, M.V. (1992).
Esseghir, M. & Sernas, V. (1994). On the measurements Numerical simulation of fluid flow and heat transport
of the radial temperature distribution in an extruder in twin-screw extruders for non-Newtonian fluids.
channel. Advances in Polymer Technology, 13, 133–139. Polymer Engineering and Science, 35, 1213–1221.
Karwe, M.V. & Sernas, V. (1996). Application of laser Tayeb, J., Della Valle, G., Barres, C. & Vergnes, B.
Doppler anemometry to measure velocity distribution (1992). Simulation of transport phenomena in twin-
inside the screw channel of a twin-screw extruder, screw extruders. In: Food Extrusion Science and
Journal of Food Process Engineering, 19, 135–152. Technology (edited by J. L. Kokini., C. T. Ho. & M.
Karwe, M.V., Bakalis, S. & Sernas, V. (1997). Velocity V. Karwe). Pp. 41–70. New York: Marcel Dekker.
Van Zuilichem, D.J., Van Der Laan, E., Stolp, W. & Ho & M. V. Karwe). Pp. 149–164. New York: Marcel
Van’t Riet, K. (1992). Modelling of heat transfer in a Dekker.
co-rotating twin-screw-extruder. In: Food Extrusion Yacu, W.A. (1985). Modelling a twin screw co-rotating
Science and Technology (edited by J. L. Kokini, C. T. extruder. Journal of Food Engineering, 8, 1–21.
Received 24 August 1996, revised and accepted 25 June 1997

Velocity Field in A Twin-Screw Extruder PDF

Uploaded by

Copyright:

Available Formats

Velocity Field in A Twin-Screw Extruder PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Velocity Field in A Twin-Screw Extruder PDF

Uploaded by

Copyright:

Available Formats

International Journal of Food Science and Technology 1997, 32, 241–253 241

Velocity field in a twin-screw extruder

Serafim Bakalis & Mukund V. Karwe*

The flow in a co-rotating twin-screw extruder

© 1997 Blackwell Science Ltd

Figure 1 Schematic diagram of

Velocity measurements were made in the screw

Figure 3 Measurement geometry:

Correction for refraction Results and discussion

Figure 5 Axial and tangential

Figure 6 Axial and tangential

Figure 7 Tangential (Ux*) velocity

Figure 8 Axial (Uy*) velocity

Figure 9 Ux* and Uy* velocity

Figure 10 Ux* velocity profiles in

Figure 11 Uy* velocity profiles in

Figure 12 Ux* velocity profiles in

Figure 13 Uy* velocity profiles in

for 2H #_z # 0. All symbols are defined in the

The experimentally obtained velocity profiles at

Figure 15 Theoretical and

Figure 16 Theoretical and

Velocity contour plots for different experimen- Conclusions

Figure 17 Tangential velocity

research was supported by state appropriations, s Screw root

Figure 18 Axial velocity (Uy*)

References measurements in the nip region of a co-rotating twin-

Received 24 August 1996, revised and accepted 25 June 1997

You might also like