EXTRUDER DESIGN AND SCALING

M. Dhanasekharan1 and J. L. Kokini1

1
Department of Food Science, NJ-NSCORT, Rutgers, The State University of New Jersey, New Brunswick, NJ
08901

INTRODUCTION
Designing an extruder, with minimum weight and energy consumption is very crucial when resources are available
in limited quantities in space. Also, for space applications, it is important to scale-down an extruder to produce
consistent products which have the same characteristics as those developed in a laboratory extruder. Most of the
successful applications involving extrusion design and scaling are based on dimensional analysis coupled with trial
and error, and a keen sense of observation. When designing a system for space applications, a trial and error
approach would need extensive experimentation and may not allow for an optimum, efficient extruder. Therefore,
understanding the basic transport phenomena in an extruder to match key operating parameters linked to quality
such as Residence Time Distribution (RTD) and Specific Mechanical Energy (SME) is important for successful
design and scale-down. This project studies the flow and heat transfer characteristics of the extruder using numerical
simulation and develops ‘Design Charts’ which predict the relationship between independent variables such as
screw design, temperature, moisture content and throughput rate with the design variables such as residence time
and specific mechanical energy.

CURRENT STATUS OF RESEARCH

Methods
This project can be thought of as divided into three phases. Phase 1: Conduct rheological modeling to
choose appropriate constitutive models and obtain model parameters for wheat flour doughs. Phase 2: Incorporate
the rheological model along with the continuum equations and conduct numerical simulation of flow and heat
transfer using a finite element code, Polyflow (Fluent Inc., New Hampshire). Phase 3: Develop Design and Scaling
Charts from numerical simulation results for extruder design and scaling.

Results
During phase 1, three differential viscoelastic models, the Phan-Thien Tanner (PTT) model, the White-Metzner
model, and the Giesekus model were considered. Of these models, the Phan-Thien Tanner model was concluded to
be the best suited for the simulations. For Phase 2 of the work, a 3D viscoelastic flow simulation was conducted for
a single-screw geometry of a lab-scale extruder. Complete fluid flow profiles were obtained for Newtonian and PTT
fluids. The viscoelastic simulations take around 10 days of computational time. To conduct a wide range of
simulations for developing design and scaling charts, the Morgan et al (1989) model, modified by Mackey et al
(1990), was used. This model provides reliable results with lesser computational time. As part of phase 3, a
preliminary design chart has been developed which matches the SME (figure 2) and RTD (figure 3) across extruders
of different sizes (figure 1).

Conclusion
Numerical simulation approach to develop design and scaling charts for dough extrusion is partially
developed in this project. The incorporation of viscoelastic effects into the modeling were not possible at this time
due to limitations in the non availability of robust numerical methods and also high computational times. However,
the use of Mackey at al (1990) rheological model is quite realistic as it accounts for shear rate dependence,
temperature dependence, moisture dependence and time-temperature history on the rheology of wheat flour doughs.
Preliminary design and scaling charts have been successfully developed. Some of the significant contributions from
this work lie in the fact that we have conducted a fully three-dimensional analysis of the simultaneous flow and heat
transfer without neglecting leakage flows using a practically realistic rheological model.

FUTURE PLANS
Future work will focus on testing of scaling concepts with key products. Simulations with twin-
screw extruder geometry will also be conducted. There is also a need for understanding dough rheology and develop
constitutive models that describe the complex rheology of dough materials to include effects of composition, strain
history etc. such that they are amenable to use with numerical simulation.

INDEX TERMS
Extrusion, scaling, viscoelastic, RTD, SME, design charts, rheological modeling, numerical simulations.

Small
extruder

Big
extruder

Figure 1.

SME TREND CHART MATCHING RTD CURVES BETWEEN BIG

450 AND SMALL EXTRUDERS
Helix angle = 17.66 0.2
Helix angle = 40
400
H=1 cm 0.18 big
e=0
small
350 L/D = 3 0.16

0.14
300
SME (KJ/ Kg)

0.12
250
E(t)

0.1

200
0.08

150 0.06

0.04
100

0.02
50
3 4 5 6 7 8 9 0
3.5 4.2 D/H 0 20 40 60 80
t (s)

Figure 2. Figure 3.