You are on page 1of 37

Lecture 15:

Design of paper and board


packaging
Advanced concepts:
FEM, Fracture Mechanics

After lecture 15 you should be able to


• Illustrate the finite element method and its use for paper-based
industry
• Illustrate how a finite element model is created
• Discuss the results from finite element analysis of:
– Compression test of a package
– Creasing and folding
• Discuss concepts such as the J-integral and the fracture toughness
• Account for the procedures of failure predictions in paper materials

1
Literature
• Pulp and Paper Chemistry and
T h l
Technology - Volume
V l 4
4, P
Paper P
Products
d t
Physics and Technology, Chapters 2, 11,
12.

Three-dimensional modelling and


analysis of paper and board, FEM

Adapted from
Mikael Nygårds

2
Why use numerical methods?
• Different types of loadings can be
investigated.
• The
Th effect
ff t off different
diff t material
t i l
properties can be investigated.
FEM
• More information about damage
and deformation mechanisms can Material
be gained. Performance
• Material behaviour and structural
behaviour can be predicted
Process
• Properties
P ti that
th t are important
i t t for
f
Solutions
the manufacturer can be linked to
properties that are important for the
user.
• Saves costs and enables more
efficient product development

Material performance
Establish the
important/critical
material properties

Converting

Packaging Design
Delamination

3
The finite element method
(FEM)

The finite element method is a…

• … method to solve partial differential


equations
ti
• … procedure to solve field problems with
engineering accuracy

4
Field problems
• In field problems the parameters of interest
are described byy continuous or discrete
field variables, e.g.
• Continuum mechanics
– Solid mechanics – stress, strain, temperature,
moisture etc.
– Fluid mechanics
– Heat transfer
• Electro magnetic field theory
•…

Finite element method


Assume a deformation (displacement field)

uN
Calculate displacements
u = NNuN
Calculate strain

ε = βNuN
Calculate stress
(in general non-linear
non linear relations)

σ = F (ε )
Principle of virtual work

∫ σ : εdV
V0
0 = ∫ t T ⋅ udS + ∫ f T ⋅ udV
S V

5
Way of working
1. Understand the problem (boundary conditions,
symmetries, material models, approximations)
2. Modelling, discretisation, input data, selection of
elements, Verification (PRE-PROCESSING)
3. Solving of systems of equations and calculation of
relevant properties e.g. stress and strain
4. Analysis of output, graphical presentation (POST-
PROCESSING)
5. Consequences of analysis

Mechanical models

Geometry
(discretisation)

Material model
Boundary conditions
σ = f(ε)

6
Challenges within paper and
paperboard research
• Constitutive models for paper that involve the through
thi k
thickness di ti as wellll as moisture
direction i t and
d ttemperature
t
dependence is still active research
• Incomplete knowledge of material data for many
materials
• Simulation tools for many applications are still missing
• Convergence
C problems
bl and
d llack
k off robustness
b t

Compression of boxes
Experiments

7
Compression of cylinders
Experiments

What is the difference compared


to rectangular boxes?

FE-analysis of compression of box


MATERIAL
PROPERTIES
USE OF ANALYSIS
PRO-
DUCT
LOADS
BOX-
MODEL
(FEM)
LOADS
ON
DETAILS DETAIL
MODEL
(FEM)
CRITICAL
LOADS

8
Static compressive load on box
FEM and experiment

Reference: L. Beldie, Lund University, 2001

Divide box into its different


segments

9
Top and bottom segments

Reference: L. Beldie, Lund University, 2001

Middle segment

Reference: L. Beldie, Lund University, 2001

10
Experimental characterization of
crease

Reference: L. Beldie, Lund University, 2001

Static compressive load on box


FEM with crease elements and experiment

Test of whole package

Reference: L. Beldie, Lund University, 2001

11
Development of a three dimensional
paperboard model
Real p
process/object
j

Experimental Model
verification formulation
Laws of
mechanics

Solution of
mathematical Mathematical model
problems Numerical
method

Mechanical behaviour of
paperboard

Top layer

Middle layer/s

Bottom
o o layer
ye

12
In-plane behaviour

MD
40
CD

Stress [MPa]
30

CD
20

10

MD
0
-1 0 2 4

-10

Strain [%]

In-plane behaviour
Summary

• In-plane • Anisotropic elasticity


CD The elastic modulus in MD are 2-
3 times larger than the elastic
modulus in CD
• Anisotropic initial yield stress
The initial yield stress in MD are
2-3 times greater then the initial
MD yield stress in CD
• Anisotropic plastic strain
hardening
Paper hardens more in MD than
in the CD

Reference: Q. Xia, MIT, Boston, 2002.

13
Out-of-plane behaviour

0.4 1.2
Norrmal Stress [MPa]

ear Stress [MPa]


0.3 0.9

0.2 0.6

0.

She
0.1
01
3

0 0
0 0.5 1.0 1.5 0 0.5 1.0 1.5 2.0

Normal displacement [mm] Shear displacement [mm]

Reference: N. Stenberg, STFI, Stockholm, 2000.

Out-of-plane behaviour
Summary

• Small amount of non-linearity


before pre
pre-peak
peak load

• Dominating softening behavior after


the peak load

• Shear strength is pressure


dependent

• Shearing causes normal dilatation

• Residual shear-load remains under


normal compressive loading.

Reference: N. Stenberg, STFI, Stockholm, 2000. Q. Xia et. al., MIT, Boston, 2002.

14
Model
In-plane: elastic-plastic continuum
Constitutive model: • Anisotropic
p elasticity y
• Anisotropic initial yield
• Anisotropic plastic strain hardening

Out-of-plane: interfacial model


• Post-peak softening tensile and shear
behavior.
• Pressure dependent shear resistance
• Normal dilation under shearing
• Existence of shear friction
• History
Hi t d
dependent
d t

Reference: Q. Xia et. al., MIT, Boston, 2002.

Verification – Uniaxial tension


F F

References:
N. Stenberg, STFI, Stockholm, 2002
Q. Xia, MIT, Boston, 2002.

15
Verification – Biaxial, compression

Out-of-plane compression

Reference:
N. Stenberg, STFI, Stockholm, 2002
In-plane compression Q. Xia, MIT, Boston, 2002.

Delamination model – failure surface

Reference:
N. Stenberg, STFI, Stockholm, 2002
Q. Xia, MIT, Boston, 2002.

16
Delamination model – ZD tension

Reference:
N. Stenberg, STFI, Stockholm, 2002
Q. Xia, MIT, Boston, 2002.

Delamination model - shear

Thickness increase under shear

Reference:
N. Stenberg, STFI, Stockholm, 2002
Q. Xia, MIT, Boston, 2002.

17
Creasing and folding
Model setup

Top ply

Middle ply

Bottom ply

Creasing of Paperboard

Reference: H. Dunn, MIT, 2000. Reference: STFI-Packforsk/Tetra Pak 2004

18
Creasing of Paperboard

Reference: H. Dunn, MIT, 2000. Reference: STFI-Packforsk/Tetra Pak 2004

Creasing of Paperboard

Reference: H. Dunn, MIT, 2000. Reference: STFI-Packforsk/Tetra Pak 2004

19
Creasing of Paperboard

Reference: H. Dunn, MIT, 2000. Reference: STFI-Packforsk/Tetra Pak 2004

Verification – Creasing
Measurements

• Reaction force, F

• Displacement, u

20
Verification –Folding
Measurements
• Reaction force, F

Centre of rotation
• Rotation angle, θ

Specimen

Load cell

Clamps

Creasing and folding

Mikael Nygårds, STFI-Packforsk

21
Creasing and folding
Comparison with experiments

To improve, include for example:


• Through-thickness plasticity
• Friction between dies and paperboard

Application
Formning of aseptic packaging

Ulf Nyman, LTH, 2007

22
Fracture Mechanics

Adapted from
Petri Mäkelä
Innventia AB

The A4-example
Load: 1450 N Load: 750 N
Elongation: 5.2 mm Elongation: 1.4 mm

10 mm
edge crack

48 % reduction of load carrying capacity and


73 % reduction of strain at break

23
Potential applications
Punching of corrugated board Failure of sacks

Web breaks in paper machines (and printing presses) K-cracks

Reduced effective width of a paper web

24
Corresponding stress distribution

Reduced strength

Predictions of the behaviour of a


structure
1. Material behaviour σ
ε=
(
(e.g. tensile
t il stiffness)
tiff ) E
Material
testing

2.

Full-scale
predictions
of failure

25
The three modes of crack opening

The modes I and II are predominant under in-plane


loading and mode I is considered most severe.

Linear elastic fracture mechanics, LEFM


y
r

ϕ
x

Stress state in the crack tip region


KI
σ ij = kij (ϕ )
σ r
K I = f (material, geometry,
E
loading)
ε

26
Fracture criterion, LEFM
Stress
Tensile strength

Strain
KI
Critical KI

Strain

Linear elastic fracture mechanics


MD CD
12
1.2 12
1.2
σnom/σb, εnom/εb

σnom/σb, εnom/εb

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
Stress at break, exp. Stress at break, num. Stress at break, exp. Stress at break, num.
Strain at break
break, exp.
exp Strain at break
break, num
num. Strain at break, exp. Strain at break, num.
0.0 0.0
0 5 10 15 20 0 5 10 15 20

Crack size [mm] Crack size [mm]

27
Conclusions, LEFM

• Does not apply to paper materials in


general, but is better than not considering
defects at all.
• Large cracks in very large structures are
required in order for LEFM to be applicable
t paper and
to d board.
b d

Non-linear fracture mechanics, NLFM


y
r

ϕ
x

Stress state in the crack tip region

σ
1
⎛J⎞ N +1
σ ij = ⎜ ⎟ k ij (E , ϕ )
⎝r⎠
E , E0 , N
J = f (material,geometry,loading)
ε

28
The HRR crack tip fields

1
⎛J⎞ n +1
σ ij = a ⎜ ⎟ f ij ( n, ϕ )
⎝r⎠

Hutchinson 1968, Rice, Rosengren 1968

J is the energy release rate


in a non-linear “elastic” material

• In the special case of a linear elastic material,


J is proportional to KI2

• This means that LEFM is a special case of


NLFM.

• J is defined as a path-independent line-integral


around the crack-tip involving expressions
containing stress, strain and displacement.

29
Fracture criterion, NLFM
Stress
Tensile strength

Strain
J
C iti l J
Critical

Strain

The J-integral method


MD CD
12
1.2 12
1.2
σnom/σb, εnom/εb

σnom/σb, εnom/εb

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
Stress at break, exp. Stress at break, num. Stress at break, exp. Stress at break, num.
Strain at break
break, exp.
exp Strain at break
break, num
num. Strain at break
break, exp.
exp Strain at break
break, num
num.
0.0 0.0
0 5 10 15 20 0 5 10 15 20

Crack size [mm] Crack size [mm]

60

30
Conclusions, NLFM
(J-integral method)

• Does
D predict
di t ffailure
il quantitatively
tit ti l ffor llarge
cracks and qualitatively for short cracks

• Numerically “cheap”

• Easy calibration

What information does J carry?

• J is a loading parameter

• J expresses the severity of the stresses at


the crack-tip

• When J reaches a critical value, the crack


starts to grow

31
Critical value of J (or K) is called
fracture toughness
1. In order to formulate a fracture criterion, we need to
know how severe stress states the material is able to
withstand.

2. We need to know the critical value of J for the material,


i.e. the fracture toughness (Jc) of the material.

3. The information from material testing on a test piece


with a man-made crack is required to evaluate the
fracture toughness.

Predictions using non-linear


fracture mechanics
Material
b h i
behaviour Geometry

Fracture
toughness
Loading

When two of these three factors are known for a


specific material, the third can be calculated!

32
Predictions of failure
Generally requires numerical methods
1. Material
behaviour 3.
Full-scale
predictions
2. Fracture of failure
toughness

FE-analysis
FE-analysis

Non-linear fracture mechanics…


1
⎛ J ⎞ N +1
σ ij ∝ ⎜ ⎟ Prescribes a singular stress
⎝r⎠ state in the crack tip region

Process zone
Damage will
D ill occur iin the
h crack-tip
k i region
i
within a process zone.

J-dominated zone

33
Micrograph of crack tip region

Crack Tip Damage

Stress state in vicinity of crack tip

Model Real material

34
Damage behaviour
Tensile testing
Elastic unloading
60
supports energy

Tensile Stress / MPa


50

40

Damage evolution 30
consumes energy
20

10
Elastic unloading
supports energy 0
0.0 0.5 1.0 1.5

Strain / %

Instability when rate of supported energy from


elastic unloading exceeds consumed energy
during damage evolution

Tensile test results


Long and short test pieces

70
Ordinary
Teensile Stress [MPa]

tensile test piece 60 Short strip


Long strip
50

40

30

20

10

0
“Short” 0 2 4 6 8 10
tensile test piece
Apparant strain [%]

35
Modelling of damage – cohesive zone
w

Stress ( σ ) σ0
σ (w) σ (w)

L+δ
δ0 Elongation ( δ )

Stress ( σ )
σ 0 −σ β
w = δ −δ0 + L σ (w ) = σ0 e−αw
E

Widening (w)

Elastic-plastic material + cohesive zone

MD CD
12
1.2 12
1.2
σnom/σb, εnom/εb

σnom/σb, εnom/εb

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
Stress at break, exp. Stress at break, num. Stress at break, exp. Stress at break, num.
Strain at break
break, exp.
exp Strain at break
break, num
num. Strain at break
break, exp
exp. Strain at break
break, num
num.
0.0 0.0
0 5 10 15 20 0 5 10 15 20

Crack size [mm] Crack size [mm]

72

36
Conclusions, Cohesive crack model

• Accurate p
predictions of failure for all crack sizes

• Numerically expensive

• Expensive, cumbersome and


time-consuming “calibration”

• No explicit fracture criterion needed

After lecture 15 you should be able to


• Illustrate the finite element method and its use for paper-based
industry
• Illustrate how a finite element model is created
• Discuss the results from finite element analysis of:
– Compression test of a package
– Creasing and folding
• Discuss concepts such as the J-integral and the fracture toughness
• Account for the procedures of failure predictions in paper materials

37

You might also like