automated fiber placement composites for improved ...arrow.utias.utoronto.ca/ncss13/ncsa...

PASINI LAB

Automated Fiber Placement Composites for

Improved Structural Efficiency of Aircrafts

Kazem Fayazbakhsh & Damiano Pasini

Pasini Lab

Mechanical Engineering

McGill University

May 16th, 2013

UTIAS National Colloquium on Sustainable Aviation

1

PASINI LAB

Outline

• Introduction

• A sustainable aircraft : Boeing 787

• Composite design concepts

• Automated Fiber Placement

• Talk objectives

• Research expertise of the Lab

• Concluding remarks

2

PASINI LAB

A sustainable aircraft : Boeing 787

• Fuel use reduced

• Automated manufacturing technologies

• Emissions cut

• Quieter take-offs and landings

• Point-to-point travel enabled

• End-of-life recycling

• A life cycle approach

3

PASINI LAB

Boeing 787 (a sustainable aircraft)

• Fuel use reduced

• Automated manufacturing technologies

• Emissions cut

• Quieter take-offs and landings

• Point-to-point travel enabled

• End-of-life recycling

• A life cycle approach

4

PASINI LAB


• Fuel use reduced • Increased use of light weight composite materials

• New engines

• More-efficient system applications

• Modern aerodynamics

• Advanced manufacturing technologies • Automated Fiber Placement

5

PASINI LAB


• Fuel use reduced • Increased use of light weight composite materials

• New engines

• More-efficient system applications

• Modern aerodynamics

• Advanced manufacturing technologies • Automated Fiber Placement

6

PASINI LAB

Composite design concepts

• Constant stiffness (CS)

• Traditional composite design

• Keeping the fiber angle constant within each layer

• Variable stiffness (VS)

• Allowing fibers to follow curvilinear paths

• More favorable stress distribution

7

Constant stiffness variable stiffness

PASINI LAB

Automated Fiber Placement machine (AFP)

8

• Robotic arm which places strips of material side-by-side to create a band

• Lays down bands to create the laminate

• Pros:

• High manufacturing flexibility

• Fully automated process

• Speeds up the layup time

• Ideal for large structures

• Cons:

• Defects produced during the manufacturing

Source: Coriolis website.

PASINI LAB

9

PASINI LAB

Automated Fiber Placement machine (AFP)

10

• Robotic arm which places strips of material side-by-side to create a band

• Lays down bands to create the laminate

• Pros:

• High manufacturing flexibility

• Fully automated process

• Speeds up the layup time

• Ideal for large structures

• Cons:

• Defects produced during the manufacturing

Source: Coriolis website.

PASINI LAB

Variable stiffness defects

Defects can be categorized as gaps and overlaps

11

Overlaps Gaps

PASINI LAB

Talk objectives

• Exploiting variable stiffness design to improve mechanical efficiency of lightweight laminate composites

• Development of a simulation toolbox to capture the mechanical impact of AFP defects

• Incorporating the effect of defects in the analysis and optimization of variable stiffness composite laminates

12

PASINI LAB

Lab Expertise

13

Microarchitectured Materials

PASINI LAB

Lab Expertise

14


Biomechanics

Multiscale Mechanics

Design Optimization

Biomimetics

PASINI LAB

Lab Expertise

15


Biomechanics

Multiscale Mechanics

Design Optimization

Biomimetics

Aerospace Structures

Biomedical Devices

PASINI LAB

Current Projects

16

Multiphysics of lattice materials

AFP variable stiffness laminate composites

Lattice stent-like devices

Cellular hip replacement implants

Ultralightweight lattice panels via additive

manufacturing

Plant cellular tissue inspired materials

PASINI LAB

Talk objectives




17

PASINI LAB

Talk objectives




18

PASINI LAB

Curvilinear fiber path

19

1 0 1[ | | ]T T T

• Constant curvature fiber path is used as the reference fiber path*

• The reference fiber path is shifted to manufacture the whole laminate

• 10 x 16 in plate is considered as a case study

T0

T1 y

x

O ρ

T1

*Blom et al., Journal of Composite Materials (2009).

Free Free

PASINI LAB

Multi-objective optimization

• Simultaneously optimization of [±<T1|T0|T1>]4s for

– In-plane Stiffness

– Buckling Load

– The effect of defects is ignored.

20

0 1

0 1

min ; ,

. . , [0 ,90 ]

1 (

&

), ( )

25

1

,

T

eq crE N T T

s t T T R in

xx x x

Eeq

/EQuasi-isotropic

0.5 1.0 1.5 2.0 2.5 3.0

Ncr/N

Qu

asi-

iso

tro

pic

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

PASINI LAB

Performance of VS design without defects

21

Normalized stiffness (Eeq)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Norm

aliz

ed b

ucklin

g load (

Ncr)

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2

1

3

4

5

Design QI (baseline)

Design 3: Same Stiffness and Higher Buckling Load

Design 5: Max. Stiffness

Design 4: Same Buckling Load and Higher Stiffness

Design 2: Maximum Buckling Load 2 3

4 5

Ncr = 1.56

Eeq = 0.60 Ncr = 1.33

Eeq = 1

Ncr = 1

Eeq = 1.30

Ncr = 0.42

Eeq = 2.62

PASINI LAB

Variable stiffness defects

Defects can be categorized as gaps and overlaps

22

Overlaps Gaps

PASINI LAB

Talk objectives




23

PASINI LAB

AFP defects analysis toolbox

24

A lamina with gaps

0

0.257

0.386

0.515

0.644

0.772

0.901

1.030

1.158

0.129

Mesh generation Stress in y-direction

PASINI LAB

25

Normalized Stiffness

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Norm

aliz

ed b

ucklin

g load

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Design 2

Design 3

Ignoring gaps/overlaps

Overlaps

Gaps

Normalized buckling load



Design 2 1.56 1.40 1.88

Design 3 1.33 1.16 1.67

The effect of defects on VS laminates performance

PASINI LAB

Talk objectives




26

PASINI LAB

Multi-objective optimization including defects

• Simultaneously maximize objectives of [±<T1|T0|T1>]4s:

• In-plane Stiffness.

• Buckling Load.

• The effect of defects is considered during the optimization process.

• Defect layer method is used.

27

0 1

0 1

min ; ,

. . , [0 ,90 ]

1 (

&

), ( )

25

1

,

T

eq crE N T T

s t T T R in

xx x x

PASINI LAB

Impact on the mechanical properties

28

Normalized in-plane stiffness

0.0 0.5 1.0 1.5 2.0 2.5 3.0

No

rma

lize

d b

ucklin

g lo

ad

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Ignoring defects

Gaps

Overlaps

PASINI LAB

Impact on the optimum fiber paths

29

Normalized in-plane stiffness

0.0 0.5 1.0 1.5 2.0 2.5 3.0

No

rma

lize

d b

ucklin

g lo

ad

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Ignoring defects

Gaps

Overlaps

1

2

3

# Design

1 [±<14|36|14>]4s

2 [±<17|39|17>]4s

3 [±<19|41|19>]4s

PASINI LAB

Work underway: manufacturing and testing

30

PASINI LAB

Work underway: manufacturing and testing

31

Design 2: highest buckling load

compared to the baseline

PASINI LAB

Concluding remarks

• Exploiting variable stiffness design to improve mechanical efficiency : 56% improvement in buckling load

• Development of a simulation toolbox to capture the mechanical impact of AFP defects: • 88% improvement in buckling load for laminates with overlaps

• 40% improvement in buckling load for laminates with gaps

• Optimization of variable stiffness composite laminates including defects

32

PASINI LAB

Concluding remarks

• Exploiting variable stiffness design to improve mechanical efficiency : 56% improvement in buckling load

• Development of a simulation toolbox to capture the mechanical impact of AFP defects: • 88% improvement in buckling load for laminates with overlaps

• 40% improvement in buckling load for laminates with gaps

• Optimization of variable stiffness composite laminates including defects

A lighter structure, more fuel efficient and sustainable

33

PASINI LAB

Questions ?

34

PASINI LAB

Current Projects

35

Multiphysics of lattice materials

AFP variable stiffness laminate composites

Lattice stent-like devices

Cellular hip replacement implants

Ultralightweight lattice panels via additive

manufacturing

Plant cellular tissue inspired materials

PASINI LAB

Overlap-modified defect element

• A single layer [0]T laminate.

• An overlap is at the plate center and along fiber direction.

• Material and strength properties are the same as regular composite material.

• The effective element thickness is the average of the thickness in the element.

36

y

x

y

x

PASINI LAB

Superiority of defect element approach

• The element length is half of the tow width.

• The FE model for [+<26|45|26>] layer:

37

Real gap distribution

Defect element approach

Existing approach in the literature

Gap distribution in FE model

PASINI LAB

FE model for capturing defects (module 3)

• A novel approach, defect layer, is proposed to capture defects precisely.

– Gap-modified defect layer.

– Overlap-modified defect layer.

• Each element may contain any defect area percentage.

• Fiber orientation at the element midpoint is calculated and used as the element fiber orientation.

38

PASINI LAB

Gap-modified defect element

• A single layer [0]T laminate is considered.

• Gap is at the plate center and along the fiber

direction.

39

y

x

1

2

3

1

2

3

1

2

3

Ex Ey G

• Test simulations are used to find material and strength properties.

PASINI LAB

Modified material and strength properties

40

Gap area percentage (%)

0% 20% 40% 60% 80% 100%

Norm

aliz

ed m

ate

rial pro

pert

ies

0.0

0.2

0.4

0.6

0.8

1.0 Ex

Ey

G

• Gap area percentage is varied with the gap width.

• The graphs are used in APDL codes to calculate properties of a defect element with any defect area

percentage.

y

x

Gap w

idth

PASINI LAB

Stress distribution

• The stress in y-direction for [+<26|45|26>] layer in Design (A) with gaps

41

Real gap distribution Stress in y-direction

0

0.257

0.386

0.515

0.644

0.772

0.901

1.030

1.158

0.129

Free Free

x

y

automated fiber placement composites for improved ...arrow.utias.utoronto.ca/ncss13/ncsa...

Documents