Designing packaging systems for Strawberries for optimal cooling

Designing packaging systems for Strawberries for optimal cooling

Designing Retail Packages for
Fruits for Uniform Cooling
R. Paul Singh and M. J. Ferrua
University of California, Davis.
Davis, California
Overview
 Industrial practice in packaging of berry fruit
 Problem definition and rationale
 Engineering approach to address industrial problem
 Fluid flow in complex systems
 Use of a non-intrusive Particle Imaging Velocimetry
(PIV) to determine flow field
 Computational fluid dynamic modeling (CFD) and
validation of flow field
 Prediction and experimental validation of CFD
models of heat transfer
 Use of CFD models to design next generation
packaging systems
Industrial Practice
 Strawberries are field packed in individual containers.
Typical commercial clamshells used for
strawberries
(a)
(c)
(b)
(d)
Clamshell packages in cardboard trays
 Clamshells are placed into
open-top cardboard trays,
and trays are stacked onto
each other.
 Finally, they are rushed to the
cooling facility.
Forced air cooling of pallet loads
Cooler
wall
Suction Fan
Air
flow
Air
flow
Tarp
Airflow in Clamshells and Cardboard Trays
(a)
(c)
(b)
(d)
Industrial Needs
 Develop a scientific basis for the design of
clamshells and cardboard trays, as a system, to
promote optimum cooling and energy efficiency.
 The design of the system should include size and
location of vent openings in walls of clamshells and
cardboard trays, and related issues such as pallet
load assembly in creating cooling tunnels.
A computational model of the
physical system
Model
 System:





Strawberries,
clamshell,
tray package,
clamshells in a tray,
trays organized on a pallet.
Strawberry model
Model of a real strawberry
3.72
cm
4.67
cm
3.62
cm
Clamshell
Strawberry package
Pallet structure
Corrugated Tray
Corrugated Tray
Corrugated Tray
Corrugated Tray
Corrugated Tray
Corrugated Tray
1.218 m
1.626 m
1.0
1.016 m
16
8m
m
1
1.2
Fan
(suction)
t4
ll e
Pa
t3
lle
Pa
t4
lle
Pa
t2
lle
Pa
ll
Pa
Row 2
Row 1
t3
Tarp
et 1
ll
Pa
ll e
Pa
et 2
t1
lle
Pa
Tray to be
modeled
Final model
Symmetry boundary
Symmetry boundary
Cold air
Predicted Air flow inside the system
using CFD model
Simulation of the airflow through the
palletized structure
Cold air
Clamshell 1
1.30e+00
1.23e+00
1.17e+00
1.10e+00
1.04e+00
9.75e-01
9.10e-01
Fluid Flow
Fluid flow
8.46e-01
7.81e-01
7.17e-01
6.52e-01
5.88e-01
5.23e-01
4.59e-01
3.94e-01
3.30e-01
2.65e-01
2.00e-01
1.36e-01
7.14e-02
Y
ZX
6.91e-03
Velocity Vectors Colored By Velocity Magnitude (m/s) (Time=3.6300e+03)
Mar 01, 2007
FLUENT 6.3 (3d, pbns, lam, unsteady)
Vertical plane (first tray)
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
Fluid flow
7.00e-01
6.50e-01
6.00e-01
Cold air
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
2.00e-01
1.50e-01
Vertical plane (second tray)
Y
1.00e-01
5.00e-02
Z
X
0.00e+00
9.90e-01
9.40e-01
Contours of Velocity Magnitude (m/s) (Time=3.6300e+03)
8.91e-01
8.41e-01
7.92e-01
7.42e-01
6.93e-01
6.43e-01
5.94e-01
5.44e-01
4.95e-01
4.45e-01
3.96e-01
3.46e-01
2.97e-01
2.47e-01
1.98e-01
1.48e-01
Y
9.90e-02
4.95e-02
Z
X
0.00e+00
Contours of Velocity Magnitude (m/s) (Time=3.6300e+03)
Mar 01, 2007
FLUENT 6.3 (3d, pbns, lam, unsteady)
Mar 01, 2007
FLUENT 6.3 (3d, pbns, lam, unsteady)
Experimental Setup
Flow Field Studies
PARTICLE IMAGE VELOCIMETRY (PIV)
PARTICLE IMAGE VELOCIMETRY (PIV)
 PIV determines the flow fields
over global domains by optically
measuring the motion of small
markers seeded in the flow.


A pulsed sheet of laser light
illuminates the flow domain at
different instants of time.
The location of the markers at the
time of each pulse is recorded by the
light scattered by them into a camera
lens.
The requirement of optical access to the flow has limited
PIV application to simple geometries.
A transparent model for PIV validation
 PIV requires optical access to the flow field.
 Need a transparent model of the packed structure.
MODEL FOR PIV APPLICATIONS
 The optical access to the flow domain requires:

Transparent setup of the
individual package of produce.

Perfect refractive index (η) match between the transparent
model setup and the working fluid (to avoid distortions of
both the laser sheet and the scattered light as they pass
through the system).
Air cannot be used as a model
fluid, its η is of different order of
magnitude than solids’ η.
Transparent model
 An appropriate combination of solid/liquid.



Refractive index matching.
Low Fluid viscosity (easily pumped).
Cost.
Fused silica spheres.
Air
Fused silica spheres.
Johnson & Johnson baby oil-Drakesol
260 oil mixture.
Transparent model
 Transparent experimental set up.
Packed test section
1.5 m
0.04 m
Vented walls
Reservoir
Pump
Transparent model
 The packed test section design was based on the packed
structure and the smaller cross-section dimensions of a
typical 0.5 kg strawberry package.

Berries were modeled by spheres whose volume represent the
average volume of the fruit.
5.3 cm
4.3 cm
5.3 cm
PIV MEASUREMENTS
 The flow field in four horizontal planes within the
system was measured.
Camera
Flow with tracer particles
Laser
Source
Level 1
Fluid flow
direction
Camera
Flow with tracer particles
Laser
Source
RESULTS
Fluid flow
direction
Camera
Flow with tracer particles
Laser
Source
Top view
Fluid
flow
Side view
Computational Model (for
transparent system using quartz)
MATHEMATICAL MODEL
 Steady laminar flow of an incompressible, Newtonian fluid
with constant fluid viscosity.
Navier-Stokes equations:
Ui
0
 xi
U i
 2U i
1 P
Uj


 gi
 xj
 xi
x j x j
Where: Ui: velocity component in xi direction, m/s.
P: static pressure, Pa.
ρ: density of the fluid, kg/m3.
ν: kinematic viscosity of the fluid, m2/s.
gi: volume force per unit of mass, m/s2.
COMPUTATIONAL MODEL
 The first step in any CFD simulation is the creation
of a computational model that accurately reproduces
the experimental domain under study.
NUMERICAL RESULTS
 CFD analysis provide a detailed flow field description.
1. Velocity field within different planes
1.88e+01
1.79e+01
1.70e+01
1.60e+01
1.51e+01
1.41e+01
1.32e+01
1.22e+01
1.13e+01
1.04e+01
9.42e+00
8.48e+00
7.54e+00
6.60e+00
5.65e+00
4.71e+00
3.77e+00
2.83e+00
1.88e+00
Y
9.42e-01
Z
X
0.00e+00
Velocity Vectors Colored By Velocity Magnitude (cm/s)
Y
Z
X
Jul 26, 2006
FLUENT 6.2 (3d, segregated, lam)
NUMERICAL RESULTS
 CFD analysis provide a detailed flow field description.
3. Particle tracking.
2.10e+01
2.00e+01
1.89e+01
1.79e+01
1.68e+01
1.58e+01
1.47e+01
1.37e+01
1.26e+01
1.16e+01
1.05e+01
9.47e+00
8.42e+00
7.37e+00
6.31e+00
5.26e+00
4.21e+00
3.16e+00
2.10e+00
1.05e+00
Y
ZX
0.00e+00
Path Lines Colored by Velocity Magnitude (cm/s)
Aug 04, 2006
FLUENT 6.2 (3d, segregated, lam)
Validation of CFD model (of flow
around quartz spheres)
VALIDATION USING PIV MEASUREMENTS
 Plane through the bottom layer of
 Plane through the bottom and
spheres.
middle layer of spheres.
Y
Y
Z
X
Z
Grid
Aug 01, 2006
FLUENT 6.2 (3d, segregated, lam)
Aug 01, 2006
FLUENT 6.2 (3d, segregated, lam)
0.06
0.03
0.05
0.025
0.04
0.02
Fluent
PIV
0.03
Vxz (m/s)
Vxz (m/s)
Grid
X
Fluent
PIV
0.015
0.02
0.01
0.01
0.005
0
0
0
0.01
0.02
0.03
x (m)
0.04
0.05
0.06
0
0.01
0.02
0.03
x (m)
0.04
0.05
0.06
Level 1
PIV velocity profile within level 5 (1) (m/s)
1
0.02
0.9
0.8
0.015
z (m)
0.7
0.6
0.01
Fluent velocity pfrofile within plane 1 (m/s)
0.5
1
0.02
0.4
0.005
0.9
0.3
0.8
0
0.005
0.01
x (m)
0.015
PIV measurements
0.02
0.015
0.2
0.7
z (m)
0
0.6
0.01
0.5
0.4
0.005
0.3
0
0
0.005
0.01
x (m)
0.015
CFD Predictions
0.02
0.2
Heat Transfer in Clamshell/Tray
System
MATHEMATICAL MODEL
 Energy equations for the fluid and solid regions .


Fluid phase:   C p  f
 T f

   uT f      k f T f 

 t

Solid phase:   C p s
Ts
   ksTs
t
 Boundary conditions at the fluid-solid interfacial area Afs
are given by the continuity of temperature and heat flux
Tf  Ts
on Afs
nfs  k f Tf  nfs  ksTs
on Afs
Air and Product Temperature in Tray 1
1.08e+01
1.03e+01
9.72e+00
Fluid flow
9.18e+00
8.64e+00
8.10e+00
7.56e+00
7.02e+00
6.48e+00
1.08e+01
5.94e+00
1.03e+01
5.40e+00
9.72e+00
4.86e+00
9.18e+00
4.32e+00
8.64e+00
3.78e+00
8.10e+00
3.24e+00
7.56e+00
2.70e+00
7.02e+00
2.16e+00
6.48e+00
1.62e+00
5.94e+00
1.08e+00
5.40e+00
5.40e-01 X
4.86e+00
Y
Z
0.00e+00
4.32e+00
3.78e+00
3.24e+00
Contours
of Static Temperature (c) (Time=3.5900e+03)
2.70e+00
Jun 01, 2007
FLUENT 6.2 (3d, segregated, lam, unsteady)
Air and Product Temperature Profile in Tray 2
1.08e+01
Fluid flow
1.03e+01
9.72e+00
9.18e+00
8.64e+00
8.10e+00
7.56e+00
7.02e+00
1.08e+01
6.48e+00
1.03e+01
5.94e+00
9.72e+00
5.40e+00
9.18e+00
4.86e+00
8.64e+00
4.32e+00
8.10e+00
3.78e+00
7.56e+00
3.24e+00
7.02e+00
2.70e+00
6.48e+00
2.16e+00
5.94e+00
1.62e+00
5.40e+00
1.08e+00
4.86e+00
5.40e-01 X
4.32e+00
0.00e+00
Y
Z
3.78e+00
3.24e+00
Contours
of Static Temperature (c) (Time=3.5900e+03)
2.70e+00
2.16e+00
Jun 01, 2007
FLUENT 6.2 (3d, segregated, lam, unsteady)
Validation of Heat Transfer in
Clamshell/Tray System
Average fruit temperature
8
7 6
in Clamshells
- measured
Experimental fruit average temperature within
Airflow
5
clamshells
18
16
Clamshell 1
14
Clamshell 4
Clamshell 5
Clamshell 8
T (°C)
12
4 3
2 1
10
8
6
4
2
0
0
20
40
60
Time (min)
80
100
120
Average fruit temperature
in Clamshell 1 – measured
8
7 6
and predicted
Fruit average temperature within clamshell 1
18
Airflow
5
4 3
2 1
16
14
Predicted
Experimental
T (°C)
12
10
8
6
4
2
0
0
20
40
60
Time (min)
80
100
120
Strawberry temperatures
in Clamshell 8
Airflow
8
Strawberries within clamshell 8
7 6
5
18
Exp S7
Exp S1
Exp S8
Exp S3
Exp S9
Exp S12
Exp S11
Exp S4
16
14
T (°C)
12
10
8
6
4
2
0
0
20
40
60
Time (min)
80
100
120
4 3
2 1
Strawberry temperatures in Clamshell 1 –
Strawberries
within clamshell 1
measured and
predicted
18
16
14
T (°C)
12
Pred S1
Pred S3
Pred S7
Pred S8
Pred S9
Pred S11
Pred S12
Pred S15
Exp S7
Exp S1
Exp S8
Exp S3
Exp S9
Exp S12
Exp S11
Exp S15
10
8
6
4
2
0
0
20
40
60
Time (min)
80
100
120