IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Preforming: Test and simulation methods

General introduction to preforming:
o Draping
o Thermoforming
o Braiding
Tutorial examples:
o Draping with PAM-QUIKFORM
o Draping with PAM-FORM

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Outline of Lecture
Composites Preforming:
General introduction:
o Pre-preg, infusion and braiding technologies
o Overview of simulation methods and their status
Draping of Preforms:
o Practical examples
o Methods for drape analysis
Kinematic mapping methods
Finite Element methods (Macro- and Meso-)
o Analysis examples
o Test methods for fabrics and yarns mechanical properties
Thermoforming
Braiding of Preforms:
o State-of-the-Art and some current research

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Manufacturing methods and resin types vary greatly

The two main manufacturing methods:
1. Pre-preg technologies: Mix
fibres and resin at the materials
stage

Toughened resins
Excellent
Shaping/fibre control
preforming
High Vf ratios
Low
shapeability/integration

2. Infusion technologies: Mix fibres and

resin at the part manufacturing stage

Low viscosity resins (low

ductility)
Less control over fibres and Vf
Parts integration is possible

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Some (main) composites processing methods: Prepregs, LRI and Braiding

Pre-pregs

Liquid Resin Infusion (LRI)

Dry fabrics

Braiding with LRI

Dry yarns

Infused with resin

> (Pre-pregs)
Stacking of plies

Curing of composite
(usually in an autoclave)

Stacking of fabrics
> Preform

Braid of yarns to
a Preform

Infusion of Preform
with a resin

Infusion of Preform
with a resin

Curing of composite
(usually in an oven)

Curing of composite
(usually in an oven)

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Some (main) simulation technologies for composites process and mechanical

analysis: Status of development and industrialisation
Technology

Problem

Liquid resin
infusion

Draping

Estimated level of maturity of simulation

Upstream research
Research
Industrial

Resin infusion

* (for complex textiles;

e.g. NCF)

* (for plain weave

type textiles)

* For VARI processes

* For RTM processes

Structural analysis:
Stress
Stress with failure
Impact and crash
Buckling, fatigue..
Residual stresses
Pre-preg

Tape laying

Little work to date

(consolidation)

Autoclave cure

Little work to date

(CFD codes)

Structural

Braiding

Preform simulation

Mainly industrial Prepregs are extensively used in Motorsport and

Aerospace; analysis methods are well developed, but failure, impact and
crash are still challenging (research)

Infusion
Structural

Research Extensive testing would be needed for mechanical properties;

especially for braids with varying architectures

12

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Outline of Lecture
Composites Preforming:
General introduction
o Pre-preg, infusion and braiding technologies
o Overview of simulation methods and their status
Draping of Preforms:
o Practical examples
o Methods for drape analysis
Kinematic mapping methods
Finite Element methods (Macro- and Meso-)
o Analysis examples
o Test methods for fabrics and yarns mechanical properties
Thermoforming
Braiding of Preforms:
o State-of-the-Art and some current research

14

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Preforming and fabric shaping (= Draping) and possible subsequent

steps (RTM infusion)
Preforming

Infusion

Typical LRI (Liquid Resin Infusion) processes

16

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Liquid resin infusion: The press forming (draping) step

Stacking the dry fabrics

The fabric is held in

pre-tensioners (= Blankholders) ready for press
forming (draping).
Fabrics may have a well
distributed
thermoplastic binder that
softens under heat and
temporarily glues the
layers together.

The preform is trimmed and

extracted ready for the next
resin infusion step
17

Courtesy BMW
http://www.e90post.com/forums /showthread.php?t=405334

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Liquid resin infusion: Infusion step

The infusions (RTM) tooling is prepared
with release agent

The infused part is

extracted from the
RTM tooling

View of the fabric

layers used and the
completed part

18

IFB

Courtesy BMW
http://www.e90post.com/forums /showthread.php?t=405334

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Liquid resin infusion: Final part

Further views of the

completed part

19

Courtesy BMW
http://www.e90post.com/forums /showthread.php?t=405334

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Liquid resin infusion: Draping of complex Non Crimp Fabrics

Plain and Twill type weaves are usually
preferred for outer plies as they give a
good surface and optical appearance.
Drapeable Non Crimp Fabrics are often
preferred for non visible plies; they
give good stiffness and strength for
relatively low cost.

20

IFB

Courtesy BMW
http://www.e90post.com/forums /showthread.php?t=405334

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Possible fabric deformation mechanisms during draping

Ply principle deformation modes

Inter-ply deformation mode

c) Fibre buckling
b) Inter-fibre sliding
a) Inter-fibre shear

d) Fibre straightening (woven fabrics only)

There are two main shapes to consider

Fibre
buckling

21

Single curvature: Local fibre stretching/

compression-buckling and inter-ply slip

Interply
slip

Double curvature:
Generally large amounts of
inter-ply shearing occur

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

IFB

Preforming and fabric shaping

Preforming and prior to LRI will change the fabric architecture:

Fibre directions will change

throughout the structure (due
to draping) leading to
substantial local changes in
architecture. This will change:

Permeability distribution in
infusion

Final part mechanical

properties (stiffness,
failure, impact).

25mm

22

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Typical fabric deformations and buckling defects

E.G. Draping of a Non Crimp Fabric over a mould
Fibres directions will change
throughout the structure (due
to draping) leading to
substantial local changes in:

25mm

Stiffness distribution
Thickness distribution
Failure criteria distribution

If a critical shearing angle is reached the fabric

will buckle out-of-plane (it cannot shear any
more in-the-plane).
In a hemisphere lines of maximum shear occur
at 90 intervals (----) and zero shear between
these.
23

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Some factors that influence the fabric drapeability

Less interlocking therefore

greater drapeability

Woven Fabrics:
The architecture
The degree of compaction of yarns
Non Crimp Fabrics:
The architecture (Bi-axial good drapeability, but tri-axial
and Quadri-axial and nearly impossible for complex
double curvatures.
The degree of compaction
The Yarn and especially the stitching type
Some other factors:
Dry fabrics are easier to drape; hence (partly) their interest for Liquid Composite molding
of complex geometries; especially in the automotive industry.
Pre-impregnated thermosetting composites are more difficult due to the viscous matrix
resin. High temperature (Hot forming) lowers the matrix viscosity and aids drapeability.
Themoplastics are hot formed.
24

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Overview of possible analysis methods and analysis codes for draping

Mapping/Kinematic methods (PAMQUIKFORM, FibreSim, MSc Laminate
modeller + several others: Fast and
approximate with minimal input.

Finite Element methods (PAM-FORM,

but maybe others, e.g. Abaqus,
Dyna3D.Accurate, but CPU intensive
large input requirements

Research area only

Kinematic
methods

FE methods
(macro-level)

FE methods
(meso-level)

LOW (OLD
1950+)
25

HIGH (NEW 2000+)

Effort/complexity
Input requirements
Modelling effort
CPU costs
Accuracy

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Draping analysis methods

Mapping/Kinematic draping
analysis

26

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Mapping/Kinematic methods: Fabric deformations are computed using

the pin joint model
Main assumptions:
L1

Fibres are inextensible.

A

Trellis deformations only via shear (the

fibre crossovers act as pin joints).

L2

i-1, j

a
b

Initial starting point (e.g. A) and initial

fibre direction (warp L1, weft L2) must
be assumed to get the process started.

i, j-1

i, j

surface F(x ij , y ij , z ij )
A

Uses simple mathematics:

First the two geodesic lines L1 and L2 are
developed then new points xij are found from
old points xi-1, j and xi , j-1.

(x ij x i 1, j ) 2 + (y ij y i 1, j ) 2 + (zij z i 1, j ) 2 = a 2
2

(x ij x i , j1 ) + (y ij y i , j1 ) + (zij z i , j1 ) = b
27

L2

L1

EG ESI Product QUIK-FORM

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

General comments to the mapping/Kinematic methods

Main assumptions:
Fibres are inextensible.
Trellis deformations only via shear (the fibre crossovers act as pin joints).
Initial starting point (e.g. A) and initial fibre direction (warp L1, weft L2) must be
assumed to get the process started. Can be a big assumption!
Plus a few others:
No interaction between plies (e.g. friction effects). Some codes allow thicknesses to be
defined so the effect of stacking can be treated.
No chance to model tooling or blank-holders (which hold fabrics in tension).
No chance to model friction between tools and plies.
Any theoretical shear up to 90 could be predicted clearly impossible in practice.
Therefore compare predicted shear angles with allowable values measured in tests.
All fabric types produce the same results!!!
Where the mapping method is good:
For balanced (woven) fabrics formed over a symmetric tool with a clearly defined
starting point and yarn directions.
For a quick analysis (1-10 seconds of CPU).
28

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Some variations possible with the kinematic mapping method

The initial projection line:
Many forms of initial projection lines are possible that could
help to fit to the actual process:
A vector direction
Normal to a plane
A user defined curve
Geodesic curves.....
Mapping sequence:
This is usually performed outwards from the initial
contact point. But some codes also allow a
sequence to be specified that can more closely
match the actual manufacturing sequence.

Numerous sequences are possible

Possible variations for other fabric types:

A stretch factor is possible (e.g. in PAM-QUIKFORM) for UD Non Crimp Fabrics
Non-orthogonal initial fabric shear
Non symmetric shear behaviour (e.g. For certain NCFs)....
29

Beware: All of these are based on mathematical algorithms that do not

represent physical (mechanical) fabric properties.

10

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

IFB

Composites modelling:
Test and simulation methods for preforming

Example industrial application using the kinematic mapping method

Draping Strategy 1

Draping Strategy 2

Seed point
Seed line

Catia V5 example study:

Two variations of seed lines and their influence on the fabric shear distributions

30

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

IFB

Composites modelling:
Test and simulation methods for preforming

Use of quick draping Simulation to optimize the process and Part

Performance

Use of Genetic Algorithms(GA)

(by A. Skordos, Cranfield
University)

The GA determines what should

be the seed line/point location in
order to:
Minimize the maximum shearing
Get the fibers in the wanted
direction in the wanted area

EC Project PRECARBI
(EUROCOPTER, Cranfield University, etc)

31

11

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming
Assumed seed
point and fabric
directions

Comparison of kinematic mapping

and Finite Elements

0
28.3
Mapping fabric shear plot

Mapping method fabric shear angle

results

wrinkling

bridging
28
14
32

IFB

Finite Element section view

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

FE simulation fabric shear angle results

Composites modelling:
Test and simulation methods for preforming

Example: Kinematic draping

Draping example using the

Mapping/Kinematic software
PAM-QUIKFORM

33

12

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Input and (some) output from a mapping analysis

A finite element mesh is used to define the

surface description: this allows any shape to be
defined (some codes are limited to simple shapes
only as they use analytical (mathematical)
descriptions for the shape

Results include contour plots of shear strain distribution, visualization of fibre directions
and the 2D flat cutting pattern.

34

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Main requited data for a woven fabric

The two main data inputs are:

1. A starting contact point from which the draping process starts
2. A vector (U axis) for the first draping direction at the point of contact.
3. A second vector (V axis) for the second draping direction at the point of contact (it
is usually orthogonal, but does not have to be).
35

13

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Understanding the PAM-QUIKFORM input

This example runs 4 (orthogonal) fabric plies having the orientations shown below.
The four plies are numbered 1, 2, 3, 4.
The dataset name for the tool mesh is hemisphere; the extension .ps or .dat is ignored

Node = 2

Typical
ply data

Ply No.
Ply No. and
contact node
on mesh
Ply No. and
Global
direction of
fibre vector

36

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Three kinematic draping: Tutorial examples:

1. Simple double hemisphere

2. Drape analysis of an aircraft spar

3. Coupled drape and infusion of a sikken

type form

Fabric permeabilities are

adjusted depending on
the shear angle (using a
simple sinus function)
37

14

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Draping analysis methods

Macro- Finite element draping

analysis

38

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Development of FE methods
Brite-Euram Project,
"Investigation for an Innovative
Methodology for Sheet Metal
Stamping", 1989 - 1992.

First (?) simple

metal stamping
simulation ca. 1988

PAM-STAMP: Metal stamping

and sheet forming

Brite-Euram Project:
"Industrial Press Forming of
Continuous Fibre Reinforced
Thermoplastic Sheets and the
Development of Numerical
Simulation Methods", 19921996.

Fabric (Carpet)
stamping ca. 2000

PAM-FORM: Fabric
and composite draping
39

15

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Typical Finite Element mechanical law for a fabric model

S = f ( E S , I , II )
E1
E2

G12

11 E1 11

22 = E 2 22
G
12 12 12

Two stacked plies and

constitutive laws for NCF
Membrane FE could be used, but
shells are better to predict
potential fabric buckling

For simple fabrics (e.g. woven fabrics):

One simple ply is needed

The main (dry) fabric law has decoupled stretch moduli in directions 1 and
2 with a shear modulus in 12. These may be constants or curve functions.

Bending stiffness is decoupled from axial stiffness.

Added viscosity effects may be added for pre-impregnated (viscous) plies

For stitched (Non Crimp) fabrics:

40

IFB

Separate plies could be modelled for each NCF blanket

With stitch type elements for the NCF stitching

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

PAM-FORM: Example draping model and simulation

Test setup

Top tool: To help position the

fabric (no real influence)
Upper ring: compacts the fibres
Fabric (UD NCF, NCF)
Lower ring: applies pressure to
the fabric (friction)

Hemisphere: velocity of 0.5 mm/ms

41

16

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

IFB

Experiment, QuikForm (mapping) and PAM-FORM (FE) results

UDs1

UDc2
35

Quickform

30
25

Pamform

20
15

Experiment

10
5

Shear angle (degrees)

Shear angle (degrees)

35

0
0

20

40

60

80

100

15

5
0
20

40

10
5

Shear angle (degrees)

Experiment

60

80

100

Position angle (degrees)

35

Pamform

15

Pamform

10

UDc3

25
20

Experiment

20

Quickform

30

Shear angle (degrees)

25

Position angle (degrees)

NCF1

Quickform

30

Quickform

30
25

Pamform

20
15

Experiment

10
5
0

0
0

10

20

30

40

50

60

70

80

90

100

20

40

60

80

100

Position angle (degrees)

Position angle (degrees)

Mapping methods always over predict due to the assumption of shear trellis-only deformation.

FE methods show better agreement but there is room for improvement.
42

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Example NCF Finite Element drape simulation of an automotive floor

pan (courtesy ESI and Sotira)
Some features:
Blankholder
NCF fabric
Lower die
Punch

Shell elements with

stitching for NCF
Rigid body tooling
(note the meshing for
accurate geometry)
Pressure blankholders correctly
modelled
Contacts defined
between fabrics and
tooling / blankholders
Tooling is velocity
driven with final
pressure loads

43

17

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Example NCF draping: Before and after process improvements

Front inner preform Sotira

Front inner preform new

Courtesy the CEC

TECABS project

44

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Example NCF draping: The completed floor preform

Courtesy the CEC

TECABS project

45

18

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Example NCF Finite Element drape simulation: An automotive floor pan

Some features of model:

Blankholder
NCF fabric

Shell elements for plies with

interface stitching for NCF
Rigid body tooling

Lower die
Punch

Pressure blank-holders
Tooling is velocity driven
with final pressure loads

Courtesy CEC Tecabs project and ESI/Sotira

46

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Example NCF Finite Element drape simulation: An aerospace beam

47

Master thesis (IFB) and in cooperation with

Eurocopter (ESI Technical Forum 2010)

19

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Draping analysis analysis

Meso- Finite Element draping

analysis (research today)

48

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Meso- drape modeling: CPU expensive today, but a method for the future
The textile fabric is usually idealised as a set of shell (or membrane) finite elements.
The main deformation mechanisms can be realistically captured:

49

Shear
Fibre straightening
Bending

BUT assumptions of continuum shell elements inhibits inter-tow and tow-to-stitch

sliding mechanisms. Meso scale modelling at the tow level may be a better way
forward.

20

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Fabric deformations at the Tow level: Simple hemisphere example

Test set-up

Local tow and

stitch
deformations

Ply deformations
50

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Mesoscopic model: Construction and Finite Element idealisation

Material:
Fabric:
Weight:
Fibres:
Stitch:

Fabric
cell)

tows

stitching

Saertex V93813
Biaxial 45 NCF
321gm-2
12K TORAY T700
Tricot (PES 76 dtex)

(stitching unit-

Stitching bars
Solid tow layer 1
Solid tow layer 2

PhD thesis (Cranfield)

51 Gavin Creech 2006

Deformed fabric

21

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Meso- drape modelling: Draping of a hemisphere

Test setup
120,900 solid element
70,000 bar elements

Model correctly
captures shear bias

35

Test

30

Simulation

Shear Angle (deg)

25

20

15

10

Simulation model: Fully deformed

90

52

IFB

180

270

360

Hemisphere position (deg)

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Meso- drape modelling offer advantages for coupling to failure/impact

and crash simulation

The draped hemisphere problem

Stress

1. The draped solid mesh is used

directly (stitching removed)

initiation

44,100 solid elements

propagation

2. Inter-tow friction is replace by the

delamination model
()
3. The deformable tow model is

replaced by an impregnated elastic- 1

i
damage model

E()

const

u
i

53

22

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Outline of Lecture
Composites Preforming:
General introduction
o Pre-preg, infusion and braiding technologies
o Overview of simulation methods and their status
Draping of Preforms:
o Practical examples
o Methods for drape analysis
Kinematic mapping methods
Finite Element methods (Macro- and Meso-)
o Analysis examples
o Test methods for fabrics and yarns mechanical properties
Thermoforming
Braiding of Preforms:
o State-of-the-Art and some current research

54

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Shear data: Using either the bias extension test or picture frame test

Bias extension test: Convenient and easy test, but

beware only a limited region has pure shear and
special calculations are needed to extract the
55 required shear stress versus shear strain data.

The Picture Frame test requires

a special rig but creates a
uniform shear stress versus
shear strain distribution.
However its not that easy to
setup or perform (main problem
is get the initial fibre alignment
right).

23

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

IFB

Typical Force-Shear angle results for a woven fabric

30

25

Force (N)

20

Fabric
wrinkling

15

10

Yarn locally
compressed

Interlocking angle

Yarn rigid rotation

0
0

10

20

30

40

50

60

70

80

Shear Angle (degrees)

56

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Other tests that may be needed for a Finite Element model of draping

Picture frame test

Shear stiffness
Interlocking angle

Tensile test

E
E

fibre
stitch

Friction test

Friction coeff
Mould / fabric

57

24

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Typical test and simulation results to calibrate a fabric model

Example testing methods for fabric drape
properties:
a) In plane tension
b) Shear (Picture frame)
c) Friction (for tool-plies and plies-plies)
d) Bending properties

58

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Outline of Lecture
Composites Preforming:
General introduction
o Pre-preg, infusion and braiding technologies
o Overview of simulation methods and their status
Draping of Preforms:
o Practical examples
o Methods for drape analysis
Kinematic mapping methods
Finite Element methods (Macro- and Meso-)
o Analysis examples
o Test methods for fabrics and yarns mechanical properties
Thermoforming
Braiding of Preforms:
o State-of-the-Art and some current research

59

25

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Theromoforming: Brite-Euram project (1992-96)!

INDUSTRIAL PRESS FORMING OF CONTINUOUS FIBRE REINFORCED THERMOPLASTIC SHEETS AND THE DEVELOPMENT OF NUMERICAL SIMULATION TOOLS

Project: European Commission BE-5092

Partners: 9 software, industrial and research partners
Pre-heating
- Prediction of 3D temperature
- Optimization of heat sources
- Prediction of heating duration

Stamping
- Prediction of forming
- Temperature and pressure
- Part and process optimisation

Cooling
- Dimensional stability
- Residual stresses
- Process induces defects

Materials and processes considered

Materials
APC2-AS4
PEI-CETEX
PA6 and PA12

UD Carbon fibres in Peek resin

Plain weave Carbon fibres in PEI resin
Woven Glass fibres in PA resin

Pre-Heating
Infra-red
Contact platen
Convection oven
Forming
Matched metal tooling
Hydroforming with membranes
Rubber pad forming

60

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Forming of a Sikken: 8 plies QI Cetex

Upper punch

8 plies (Cetex)
Blank holder

0/90 plies

Lower tool

45 plies

Fabric
shear

61

26

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Sikken part: Temperature and Consolidation evolution

Intermediate
time

Temperature

320.0

Final time
Consolidation (Process window)

250.0

Increased pressure

high pressure
uncontrolled flow

0.0

morphological
defects

Solidification rate
(faster cooling)

ic

a b 1
w + b a
o

+ 5

0.642
dT/dt
Papp( )(bo + wo )
d
[T ( )](aobo) 3

1
5

Final time

0.785
0.857
0.928
1.0

62

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Thermoforming
Diaphragm and sheet

vacuum

63

27

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Outline of Lecture
Composites Preforming:
General introduction
o Pre-preg, infusion and braiding technologies
o Overview of simulation methods and their status
Draping of Preforms:
o Practical examples
o Methods for drape analysis
Kinematic mapping methods
Finite Element methods (Macro- and Meso-)
o Analysis examples
o Test methods for fabrics and yarns mechanical properties
Thermoforming
Braiding of Preforms:
o State-of-the-Art and some current research

64

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

2D braiding for closed form sectional preforms

A typical 2D industrial
scale braider

Close-up view of
the bobbins

A typical braided
preform
65

28

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Sectional preforms: An example that could be made via braiding or

layered plies and RTM type infusion

Fairly heavy tow

yarns for fast
manufacturing

BMW
http://www.e90post.com/forums/showt
hread.php?t=405334
66

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Braiding process simulation and prediction of mechanical properties

Previous research on 3D braiding:

BMBF project "Faserverbundwerkstoffe mit

dreidimensionaler Geflechtverstrkung", 1991-1995.

BMBF project: "Rechnergesttzte, automatisierte

Herstellung von 3D-Geflecten", 1996 -1999.

Current work on 2D braiding:

Momentum project 2007-2010

Internal IFB research

67

29

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Early work (1991-99): The prototype 3D- rotary braiding machine

Example T profile
68

IFB

12*12 horngear 3D braider

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

FE braid simulation: Model details

Braiding
ring
Braiding
mandrel

Standing
yarns

Braiding
take-up
direction

Braiding yarns
Yarn paths

Bobbin elements

Models typically 50,000+

elements.
CPU: hrs to days on high
end PCs.
69

30

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

2D braiding: Example simulation and test

70

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Example rotary braiding over a generic mandrel

Simulation
model

Generic mandrel

Test results

71

31

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Comparing analytical and FE results for yarn paths

Analytical
equations

x ( t ) = v tu t

(t ) = c t
r ( t ) = r ( x ( t ))

A comparison of analytical and Finite

Element simulation of 2D braiding
Authors:
Anthony Pickett, Andreas Erber,
Tjark von Reden and Klaus Drechsler

72

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

An approach to structural meso-modelling for 2D braiding

Overbraiding
3-layers

Extracted yarn
model
74

32

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Structural meso-modeling: Merging braid and solid models

Options for merging (matrix-yarns):
1. Remeshing yarns (beams) and constraint
equations to tie to matrix solids.
2. SPH elements for matrix and FE beam
yarns.
3. Remeshing matrix (Solid FE) and tying
matrix/yarn nodes.
Option 3

Braid E
Yarn
Area
Resin E

Extracted
yarn model

230GPa
12000 * 7m
filaments=0.46mm2
E= 4GPa

Overlayed and connected

Matrix model

75

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

Structural meso-modelling: Analysis results

Torsion Moment@10
Test

300 kNmm

Simulation

283 KNmm

Linear FE analysis
Tests

Principle stresses matrix (GPa)

Yarn forces (kN)

Ongoing work: Failure/max load prediction and

(maybe) crash

76

33

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Building the FE models

Performing
the braid
simulation

Overlaying,
nesting and
compaction
Simulation

Extracting
the Braid
(coupon)

Overlaying
and
merging
solids

Materials,
BCs and
loading

77

IFB

Composites modelling:
Test and simulation methods for preforming

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Bi-axial braids: Stiffness studies and validations

Engineering constant Ex [MPa]

60.000
50.000
40.000
30.000
20.000
10.000
0
25

78

30

35

Test data

40
45
Braiding angle []
CLT

50

Wisetex real geometry

55

60

FE Implicit

34

IFB

A.K. Pickett, 2013-2014

Institut fr Flugzeugbau, University Stuttgart

Composites modelling:
Test and simulation methods for preforming

stress

Meso-failure modelling of braids: Example Bi-axial 55

Yield
75 MPa
plastic strain
= 0.005

Matrix dominated
failure
Initiation of local
first matrix failure

Residual fibre strength

79

35