Characterisation and Development

1
Characterisation and Development of Open
3D-Knitted Composites
Dirk Philips Advisor Ignaas Verpoest Department
of Metallurgy and Materials Engineering Katholieke
Universiteit Leuven, Belgium
2
Situating 3D-Knitted Composites
Related research
3
Introduction
Before this PhD.-research started
- experimental results - modelling - applications

• 3D-woven sandwich structures
• 2D-knitted composites
• 3D-knitted composites
• - preliminary tests to try out concept
• - patent application

During this PhD.-research

• Optimisation of 3D-knits for composites
  production
• Processing
• Measure the mechanical properties of the
  composite
• Develop mechanical models and design tools
• Applications

4
Contents

• 1. Open 3D-knitted composites
• Production
• Optimisation of 3D-knits for composites

• 2. Skin stiffness
• Geometric description of the skins
• Experimental results (bending tensile tests)
• Modelled skin stiffness

• 3. Core compression
• Geometric description of the core
• Experimental results
• Modelled compression stiffness

4. Applications
5. Conclusions
5
1. Open 3D-knitted composites Production
Double-bed Rashell Knitting Machine
Guiding bars
Skin fibers
Pile fibers
Skin fibers
Pile yarns Adjustable length
Skins Open or Closed
Interconnect Integral Double Layer Fabric
6
1. Open 3D-knitted composites Production
Close-up of 3D-knitted textile
Pile fiber
Interconnection between skin core
Skin yarn
7
1. Open 3D-knitted composites Production
Type 1 Closed skin 3D-knit
Type 2 Open skin 3D-knit (expanded)
Plush fabric
8
1. Open 3D-knitted composites Production
0. Textile
1. Impregnation

• by hand
• 3 types of resin
• - low viscosity epoxy resin (1)
• - high viscosity epoxy resin (2)
• - low viscosity epoxy resin (3)

2. Curing

• in hot press under displacement control used as
  oven

9
1. Open 3D-knitted composites Optimisation
Originally

• Existing PET monofilament 3D-knits
• used for textile applications
• Impregnation problems
• Low mechanical properties

Adapt for composites

• 1. Use complex yarns
• Easier to knit
• Easy impregnation
• Increase mechanical properties
• 2. Introduce glass fibers inside textile
• 3. Stabilise the knits for handling

10
1. Open 3D-knitted composites Optimisation
Problems with plain PET-monofilament knittings

• Contradiction

A flexible yarn is needed for the knitting
process
A stiff yarn is needed to obtain a 3D-knit with
stable geometry
Solution Multifilament yarn
Solution Monofilament yarn
n x r
1 x R

• Impregnation quality

for multifilament OK
for monofilament very poor
Homogenous Impregnation of Pile Fibers
Poor Impregnation of Pile Fibers
11
1. Open 3D-knitted composites Optimisation
Improvement 1 Combined yarns
Solution
Combined Yarns - One central thin (flexibel)
PET-monofilament gt geometrical stability -
Wrapped with many fibers (i.e. ramie-viscose,
polyamide, ...) gt impregnation quality
reinforcing effect on composite
PET-monofilament with polyamide multifilament
spun around (100X)
Homogenous impregnation with PET / ramie-viscose
yarns (7X)
12
1. Open 3D-knitted composites Optimisation
Improvement 2 Glass in skins for stiffer
composites
Monofilament piles monofilament skins
Complex piles glass inserted in skins
Complex piles glass knitted in skins
different impregnation quality, roughness,
stiffness, ...
13
1. Open 3D-knitted composites Optimisation
Improvement 3 Increase knit stability
Increased stability
Low stability
Top
Bottom
14
1. Open 3D-knitted composites Optimisation
Improvement 4 Better surface quality
Rough
Smooth
Smooth flat
15
2. Skins Geometric description
0. Parameters
c Rhombic degree of expansion (experiments) b
Hexagonal degree of expansion (models)
16
2. Skins Geometric description
1. Regular unit cell
Complex cell Curved beams
Simple cell Straight beams
straight beams 1 parameter curved beams 2
parameters
17
2. Skins Geometric description
Relative areal density versus degree of expansion
Hexagonal model Straight beams
Pores
Complex model Curved beams
Relative areal density ()
Material
Hexagonal degree of expansion b ()
many properties are proportional to relative
areal density (weight, cost, stiffness, ...)
18
2. Skins Geometric description
Comparison of geometric model with experimental
data
at large deformations also beams elongate many
properties are proportional to relative areal
density (weight, cost, stiffness, ...)
19
2. Skins Geometric description
2. Deformed unit cells
Undeformed Shear a Tear g
20
2. Skins Geometric description
Deformed unit cells are needed for draping
complex surfaces
21
2. Skin stiffness Introduction
Practical interest bending stiffness of full
sandwich structure


F
Skin
Compression
Skin
Core
F
F
L
Tension

Bending test - on full sandwich structure - easy
to carry out - results can be converted to
properties of components (skins core)
Tensile test - on skins - takes more time to do
test - interesting for modelling - direct
measurement of skin stiffness
22
2. Skin stiffness Test samples
Knit type k20 k19 k21 Thickness
5.5 7 8 (mm) Resin content
56-63 50-52 45-49 (w) Glass
content 1.2 4.6 4.3
(m/m) Glass in skins Inserted Knitted
Knitted
m glass / m knitted beam
k21
k20
k19
23
2. Skin stiffness Tensile tests
Warp
stiffness decrease as degree of expansion goes
up more glass stiffer
24
2. Skin stiffness Tensile tests
- Same geometry - Different glass content
k19 4.6m glass
k20 1.2m glass
Relative areal density ()
k21 4.3m glass
- Different geometry - Same glass content
Rhombic degree of expansion c ()
25
2. Skin stiffness Tensile tests
Weft
stiffness very low in weft k20 (lowest glass
content) has highest stiffness due to resin walls
26
2. Skin stiffness Comparison
Bending

-20
-30
Tensile

skin stiffness from bending tests always lower
than for tensile tests
27
2. Skin stiffness Comparison

• k19 close to theoretical value
• - behaves the most like sandwich material
  (if only skin properties are considered)
• k20 deviation of 20
• - interaction between skins and core due to
  resin distribution (calculations will be less
  correct)
• - thin material with relatively thick skins
• k21 deviation of 30
• - sheared sample (top and bottom skin moved
  to each other)
• - asymmetric core
• gt interference with the calculations
  for symmetric sandwich

28
2. Skin stiffness Modelling
Model description
Type 3 Connection elements
Type 2 Cross-over elements
Type 1 Knitted beam elements
Warp
Weft
Diagonal
29
2. Skin stiffness Modelling
Parameter variation 1 Dimensions l and h
Parameter variation 2 Degree of expansion b
4
5
6
7
8 mm
Depends on h, l and b
Reference

h
5
B Variation in h
6
l
7
8
A Variation in l
30
2. Skin stiffness Modelling
Warp direction
Reference
Warp Stiffness (N/mm²)
Degree of expansion b ()
influence of h and l on stiffness stiffness
decreases rapidly as b goes up practical
stiffness is quite low (30 lt b lt 60)
31
2. Skin stiffness Modelling
Warp direction
For practical cases (b lt 70) Skin stiffness is
linearly dependent on relative areal density
Reference
b 20
Warp Stiffness (N/mm²)
30
40
50
60
Relative Areal Density ()
32
2. Skin stiffness Modelling
Warp, weft, diagonal directions
b 20
Warp
30
Diagonal
40
Warp, Weft Diag. Stiffness (N/mm²)
50
Weft
Relative Areal Density ()
for all directions similar conclusions interestin
g for rapid estimations (master curves)
33
2. Skin stiffness Modelling
Warp, weft, diagonal directions
5 6 7 8
If properties in 1 direction (e.g. warp) are
known, properties in other directions can be
deduced from diagramme

b 20
Diagonal
30
40
50
Weft Diag. Stiffness (N/mm²)
60
l
Weft
Warp Stiffness (N/mm²)
34
2. Skin stiffness Modelling
Influence of complex deformations on warp
stiffness
Unsheared
Shear
35
2. Skin stiffness Modelling
Influence of complex deformations on warp
stiffness
Unteared
Tear
36
2. Skin stiffness Modelling
Reference undeformed material
37
2. Skin stiffness Modelling
Comparison with experiments
38
3. Core Geometric description
Regular pile
Warp sheared pile
Weft sheared pile
fwarp
Warp
Skin
Core
weft shear reduces significantly the composite
properties quality weft shear can be reduced by
designing better x-knits
39
3. Core Geometric description
Distorted piles
Resin distribution
Special morphologies
so far textile, now composite (resin distribution)
40
3. Core compression Tests
c1
c2
Rhombic test samples same of unit cells and
piles
Easy to convert all data to properties for one
pile fiber (modelling)
easier to compare with modelled results less
damage to samples
41
3. Core compression Test results
1. Influence of pile fiber density ( degree of
expansion)
Compression Stiffness
N/cell
Hexagonal model Straight beams
Complex model Curved beams
Relative areal density ()
Degree of expansion
Degree of expansion b ()
42
3. Core compression Test results
2. Influence of pile fiber length
k20, 5.5mm
Compression strength (N/mm²)
k24, 6.2mm
k25, 7mm
H10 6mm
k26, 10mm
Resin content (w)
k24, k25, k25 similar knits except for height k20
has more vertical piles
43
3. Core compression Test results
3. Influence of resin content
Increasing resin content
k24, y x4.6
k25, y x4.1
Compression strength (N/mm²)
k26, y x5.9
H10, y x6.7
Resin content (w)
44
3. Core compression Modelling
Types of elements
MODE I
MODE II
Type 3 Resin1
Type 2 Skin
Type 4 Resin2
Type 1 Pile
Flatwise compression
Indentation
model for pile 30 beam elements reality lies
between two compression modes
45
3. Core compression Modelling
MODE I Compressive shear allowed
1. Influence of warp shear (fwarp) and pile
curvature (a/b)
Warp shear ( Tan fwarp)
0
Pile stiffness (N/mm²)
3
5
7
10
14
18
.
Ratio a/b
46
3. Core compression Modelling
MODE I Warp shear 0
2. Influence of resin feet
47
3. Core compression Comparison
3A. Comparison of experiments and model 2 main
effects
Combination of effects - Increasing amount of
resin at pile feet - Increasing radius
Effect 2 Increasing amount of resin at pile
feet
Pile stiffness (N/pile)
Effect 1 Increasing radius
x7
Number of connection elements
x6
x5
x4
x3
x2
x1
x
x
x
Resin content (w)
Radius ro
48
3. Core compression Comparison
Pile length 6mm Comp. shear allowed Pile radius
at 20w resin 0.13mm Warp shear 12 a/b 0.4
3B. Comparison for k25-composite
49
4. Applications
Development of orthopaedic casts
Current cast materials
- plaster of Paris reinforced with woven cotton
bandages (excellent properties cheap) - glass
or polyester knitted bandage impregnated with
water hardenable PU-prepolymer Main
disadvantage - completely closed (ventilation
restricted)
Requirements for cast materials
- certain strength / stiffness - handleability
(applicability) - conformability - breathable -
easy removable
50
4. Applications
Excellent drapability of 3D-knitted preform!
51
4. Applications
Mechanical requirements
F
F
1cm
Crush strength force at 1cm compression
speed 300 mm/min
tubular cast
anvil
Commercial materials reference tests
52
4. Applications
Some test results
Length reduced to 2/3 density properties
increase
Weft preform, Ø 70mm k21-knit, 20 unit cells
c 52
Crush strength (N)
Glass
c 45
Polyester
Curing Time (Min)
requirements can be reached - slightly lower
weight - better ventilation - faster contact
with water / curing
53
5. Conclusions

• 3D-Knitted textiles
• Existing 3D-knitted fabrics were adapted for
  composites production
• - to obtain a better impregnation
• - to optimise the mechanical properties of the
  composites
• 3D-Knitted Composites
• Full geometric description
• Mechanical properties were determined
  experimentally
• (bending stiffness, skin stiffness, core
  compression)
• Models were set up to predict these mechanical
  properties
• - FEM analytical models
• - final goal develop a design tool
• Applications
• Development of 3D-knitted orthopaedic casts
• - 3D-knitted fabrics were specially developed
  for this application
• - mechanical requirements could be reached
• - some basic modelling was carried out

54
4. Applications
Other advantage of 3D-knitted casts Happy
people!