-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses
By
Matthew Humphreys BEng Civil (Hons), MIEAust, CPEng, RPEQ
A thesis submitted to the School of Civil Engineering Queensland
University of Technology
in partial fulfilment of the requirements for the degree of
Doctor of Philosophy
December 2003
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses iii
Abstract
Fibre composite materials are gaining recognition in civil
engineering applications as
a viable alternative to traditional materials. Their migration
from customary
automotive, marine, aerospace and military industries into civil
engineering has
continued to gain momentum over the last three decades as new
civil engineering
applications develop. The use of fibre composite materials in
civil engineering has
now evolved from non-structural applications, such as handrails
and cladding, into
primary structural applications such as building frames, bridge
decks and concrete
reinforcement. However, there are issues which are slowing the
use of fibre
composite materials into civil engineering. Issues include high
costs, difficulties in
realising potential benefits, general lack of civil engineers’
familiarity with the
material and relatively little standardisation in the composites
industry. For
composites to truly offer a viable alternative to traditional
construction materials in
the civil engineering marketplace, it is essential that these
issues be addressed. It is
proposed that this situation could be improved by demonstrating
that potential
benefits offered by composites can be achieved with familiar
civil engineering forms.
These forms must be well suited to fibre composite materials and
be able to produce
safe and predictable civil engineering structures with existing
structural engineering
methods.
Of the numerous structural forms currently being investigated
for civil engineering
applications, the truss form appears particularly well suited to
fibre composites. The
truss is a familiar structural engineering form which possesses
certain characteristics
that make it well suited to fibre composite materials. In this
research a novel
monocoque fibre composite truss concept was developed into a
working structure
and investigated using analytical and experimental methods. To
the best of the
author’s knowledge the research presented in this thesis
represents the first doctoral
research into a structure of this type. This thesis therefore
presents the details of the
development of the monocoque fibre composite (MFC) truss concept
into a working
structure.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses iv
The developed MFC truss was used as the basis for a detailed
investigation of the
structural behaviour of the MFC truss elements and the truss as
a whole. The static
structural behaviour of the principal MFC truss elements
(tension members,
compression members and joints) was investigated experimentally
and analytically.
Physical testing required the design and fabrication of a number
of novel test rigs.
Well established engineering principles were used along with
complex finite element
models to predict the behaviour of the tested truss elements and
trusses. Results of
the theoretical analysis were compared with experimental results
to determine how
accurately their static structural behaviour could be
predicted.
It was found that the static structural behaviour of all three
principal truss elements
could be accurately predicted with existing engineering methods
and finite element
analysis. The knowledge gained from the investigation of the
principal truss elements
was then used in an investigation of the structural behaviour of
the MFC truss. Three
full-scale MFC trusses were fabricated in the form of
conventional Pratt, Howe and
Warren trusses and tested to destruction. The investigation
included detailed finite
element modelling of the full-scale trusses and the results were
compared to the full-
scale test results. Results of the investigation demonstrated
that the familiar Pratt,
Howe and Warren truss forms could be successfully manufactured
with locally
available fibre composite materials and existing manufacturing
technology. The
static structural behaviour of these fibre composite truss forms
was accurately
predicted with well established engineering principles and
finite element analysis.
A successful marriage between fibre composite materials and a
civil engineering
structure has been achieved. Monocoque fibre composite trusses
have been
developed in the familiar Pratt, Howe and Warren truss forms.
These structures
possess characteristics that make them well suited to
applications as primary load
bearing structures.
KEYWORDS: civil engineering, construction, structural
engineering, finite element
analysis, fibre composites, truss structures, Pratt truss, Howe
truss, Warren truss,
glass fibre, carbon fibre, epoxy resin, particulate filled
resin, fibre composites in civil
engineering.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses v
Publications
Humphreys, M.F., van Erp, G.M., and Tranberg, C.T. (1999),
“Monocoque Fibre
Composite Truss Joints”, ACUN 1 – Proceedings of the First
International
Composites Meeting, University of New South Wales, Australia, pp
247 - 251.
Humphreys, M.F., van Erp, G.M., and Tranberg, C.T. (1999), “An
Investigation into
the Structural Behaviour of Monocoque Fibre Composite Truss
Joints”, Proceedings
of ICCM12 – International Conference on Composite Materials,
International
Committee on Composite Materials, Paris, France, Paper 274.
Humphreys, M.F., van Erp, G.M., and Tranberg, C.T. (1999),
“Structural Behaviour
of Monocoque Fibre Composite Trusses”, Mechanics of Structures
and Materials,
Edited by Bradford M. A., Bridge, R. Q., Foster, S. J., Balkema,
Rotterdam, The
Netherlands, pp 501 - 506.
Humphreys, M.F., van Erp, G.M., and Tranberg, C.T. (1999), “The
Structural
Behaviour of Monocoque Fibre Composite Truss Joints”, Advanced
Composite
Letters, Vol 8 No. 4, Adcotec, London, UK, pp 173 - 180.
Humphreys, M.F., (2003)”, “Extending the Service Life of
Buildings and
Infrastructure With Fibre Composites”, PRRES9 - Proceedings of
the Ninth Pacific
Rim Real Estate Society Conference, Brisbane, Australia.
Humphreys, M.F., (2003), “The Use of Polymer Composite in
Construction”,
SASBE2003 – Proceedings of the Smart and Sustainable Built
Environment
conference, Brisbane, Australia, pp 585 - 593.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses vi
Contents
Statement of original authorship i
Acknowledgements ii
Abstract iii
Publications v
Contents vi
List of Figures xii
List of Tables xviii
Notation xxi
Chapter 1 – Introduction
1.1 Background 1.2 Aims 1.3 Scope 1.4 Thesis structure
1 - 1
1 - 2
1 - 6
1 - 7
1 - 8
Chapter 2 – Fibre Composite in Civil Engineering
2.1 Introduction 2.2 Fibre composite materials in construction
and civil
engineering
2.2.1 Rehabilitation and retrofit 2.2.2 Concrete structures
reinforced with fibre
composites
2.2.3 New fibre composite civil structures 2.3 Issues affecting
the use of fibre composites in civil
engineering applications
2.3.1 Cost 2.3.2 Structural performance 2.3.3 Durability 2.3.4
Familiarity and education 2.3.5 Specification and standardisation
2.3.6 Compatibility 2.3.7 Temperature and fire performance
2 - 1
2 - 1
2 - 2
2 - 3
2 - 4
2 - 7
2 - 8
2 - 9
2 - 15
2 - 24
2 - 29
2 - 31
2 - 32
2 - 33
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses vii
2.4 The need for a new approach 2.4.1 Lessons from history 2.4.2
Current approach
2.5 Summary
Chapter 3 –Monocoque Fibre Composite Trusses
3.1 Trusses 3.1.1 Truss definition 3.1.2 Brief history of
trusses 3.1.3 Characteristics of trusses suited to fibre
composite materials
2 - 36
2 - 36
2 - 38
2 - 40
3 - 1
3 - 2
3 - 2
3 - 3
3 - 6
3.2 FRP trusses 3.2.1 Concrete filled CFRP tubes 3.2.2
Experimental transmission towers with serrated
joints
3.2.3 CFRP roof truss 3.2.4 Expandable space trusses 3.2.5
Pultruded section pedestrian bridge 3.2.6 Areas for potential
improvement of existing
approaches to FRP trusses
3 - 8
3 - 8
3 - 9
3 - 11
3 - 11
3 - 12
3 - 13
3.3 The monocoque fibre composite (MFC) truss 3.3.1
Configuration 3.3.2 Form 3.3.3 Materials 3.3.4 Fabrication 3.3.5
Static load response
3.4 Adopted configuration of the MFC truss
Chapter 4 – Static Structural Behaviour of MFC Truss Tension
Elements
4.1 Introduction 4.2 Preliminary investigation of tension
elements
4.2.1 Material properties
3 - 15
3 - 17
3 - 31
3 - 38
3 - 51
3 - 56
3 - 63
4 - 1
4 - 1
4 - 3
4 - 4
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses viii
4.2.2 Specimen geometry 4.2.3 Prediction of static structural
behaviour 4.2.4 Fabrication of test specimens 4.2.5 Testing 4.2.6
Results 4.2.7 Discussion 4.2.8 Summary
4.3 Detailed investigation of tension elements 4.3.1 Proposed
approach 4.3.2 Specimen design 4.3.3 Prediction of stiffness and
strength 4.3.4 Fabrication 4.3.5 Test setup 4.3.6 Results 4.3.7
Comparison of prediction with test results 4.3.8 Discussion
4.4 Conclusions
Chapter 5 – Static Structural Behaviour of MFC Truss
Compression
Elements
5.1 Preliminary investigation of compression elements 5.1.1
Specimen details 5.1.2 Analysis and prediction of behaviour 5.1.3
Testing 5.1.4 Results and discussion
5.2 Detailed investigation of compression elements 5.2.1 Member
design 5.2.2 Analysis 5.2.3 Fabrication 5.2.4 Test setup 5.2.5
Results 5.2.6 Discussion
5.3 Summary and conclusions
4 - 4
4 - 5
4 - 7
4 - 8
4 - 9
4 - 11
4 - 16
4 - 17
4 - 17
4 - 18
4 - 20
4 - 35
4 - 36
4 - 37
4 - 41
4 - 45
4 - 50
5 - 1
5 - 1
5 - 2
5 - 3
5 - 7
5 - 8
5 - 11
5 - 12
5 - 16
5 - 22
5 - 26
5 - 29
5 - 34
5 - 39
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses ix
Chapter 6 – Monocoque Fibre Composite Truss Joints
6.1 Joint strength 6.1.1 Calculating joint strength 6.1.2
Preliminary experimental investigation 6.1.3 Joint testing
6.2 Joint rigidity 6.2.1 Secondary stresses 6.2.2 Truss
deflection
6.3 Simplified joint analysis 6.4 Conclusions
6 - 1
6 - 1
6 - 2
6 - 12
6 - 20
6 - 33
6 - 34
6 - 39
6 - 42
6 - 43
Chapter 7 – Static Structural Behaviour of MFC Trusses
7.1 Design of truss specimens 7.2 Analysis
7.2.1 Approach 7.2.2 Elements and mesh 7.2.3 Material properties
7.2.4 Loading and restraints 7.2.5 Results of finite element
analysis 7.2.6 Discussion of finite element analysis
7.3 Fabrication of truss specimens 7.4 Testing of MFC
trusses
7.4.1 Test description 7.4.2 Test results
7.5 Comparison of predictions with test results 7.5.1 Stiffness
7.5.2 Strength
7.6 Summary
7 - 1
7 - 1
7 – 6
7 - 6
7 - 7
7 - 9
7 - 29
7 - 30
7 - 36
7 - 37
7 - 44
7 - 44
7 - 46
7 - 54
7 - 54
7 - 56
7 - 57
Chapter 8 – Conclusions and Recommendations
8.1 Conclusions 8.1.1 Development of truss concept into
working
structure
8 - 1
8 - 1
8 - 2
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses x
8.1.2 Investigation of structural behaviour of principal truss
elements and demonstration of accuracy
with which structural behaviour can be predicted
8.1.3 Design and construction of prototype Pratt, Howe and
Warren trusses
8.1.4 Evaluation of the structural performance of the three
prototype trusses to characterise their
behaviour in terms of stiffness, strength, failure
mode, predictability and warning of failure
8.2 Recommendations for future research
8 - 4
8 - 9
8 - 9
8 - 10
Appendix A – Material Properties
A - 1
Appendix B – Tension Specimen Graphs
B - 1
Appendix C – Compression Specimen Graphs
C - 1
Appendix D – T-joint Graphs
D - 1
References
R - 1
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xi
List of Figures
Figure 1.1 – Applications of fibre composites in civil
engineering
structures
1 - 4
Figure 1.2 – Examples of fibre composite trusses 2 - 5
Figure 2.1 – Simply supported beam with UDL 2 - 18
Figure 2.2 –Behaviour of hybrid FRP 2 - 22
Figure 2.3 – a) Veirendeel load path b) “Long” truss form 2 -
24
Figure 3.1 – Truss terminology 3 - 2
Figure 3.2 – Early truss structures 3 - 3
Figure 3.3 – Nail-plate 3 - 6
Figure 3.4 – Reinforced concrete jointed CFRP tube bridge
(Karbhari,
1998)
3 - 9
Figure 3.5 – Serrated joints (Goldsworthy, 1998) 3 - 10
Figure 3.6 – CFRP roof truss (Agematzu, 1998) 3 - 11
Figure 3.7 – Telescopic space truss (NASA, 2002) 3 - 12
Figure 3.8 – Pultruded section pedestrian bridge (Milcovich,
2002) 3 - 13
Figure 3.9 – Conceptual configuration of monocoque truss 3 -
17
Figure 3.10 – Truss joint showing web member lapped onto bottom
chord
at panel point
3 - 17
Figure 3.11 – Lapping fibres from adjacent members 3 - 18
Figure 3.12 – Pratt truss fibre architecture with arbitrary fill
layers 3 - 19
Figure 3.13 – Pratt truss fibre architecture with aligned fill
layers 3 - 20
Figure 3.14 – Joint layup types 3 - 22
Figure 3.15 – Type 1 fibre architecture 3 - 23
Figure 3.16 – Type 2 fibre architecture 3 - 24
Figure 3.17 – Joint type 3 fibre architecture 3 - 25
Figure 3.18 – Typical MFC truss cross-sections 3 - 27
Figure 3.19 – Frame analysis truss configurations 3 - 33
Figure 3.20 – Typical frame analysis truss 3 - 34
Figure 3.21 – Typical displaced shape 3 - 35
Figure 3.22 – Typical combined stress 3 - 36
Figure 3.23 –Buckled shape of polystyrene foam core MFC truss
from 3 - 44
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xii
FEA
Figure 3.24 – Truss test configuration 3 - 44
Figure 3.25 – Foam and balsa truss core configuration 3 - 45
Figure 3.26 – Foam truss failure mode 3 - 45
Figure 3.27 – Load-displacement graph for polystyrene foam truss
member 3 - 46
Figure 3.28 – Initial balsa investigation 3 - 47
Figure 3.29 – Initial balsa investigation 3 - 48
Figure 3.30 – PFR 3 - 50
Figure 3.31 – Use of pre-fabricated sandwich panel to produce
MFC truss 3 - 53
Figure 3.32 – Fabrication of PFR core 3 - 55
Figure 4.1 – Typical MFC truss member 4 - 2
Figure 4.2 – Representative tension element 4 - 5
Figure 4.3 – Bi-linear load versus strain curve 4 - 7
Figure 4.4 – Representative tension element specimens 4 - 8
Figure 4.5 – Typical test configuration 4 - 9
Figure 4.6 – Typical load - displacement curves 4 - 9
Figure 4.7 – Type 1 specimen failure modes 4 - 12
Figure 4.8 – Transverse cracks in PFR specimens 4 - 14
Figure 4.9 – Transverse cracks in PFR specimens 4 - 15
Figure 4.10 – Tension specimen geometry and fibre architecture 4
- 19
Figure 4.11 – Two-dimensional FE model 4 - 24
Figure 4.12 – Principal stress distribution at maximum load 4 -
26
Figure 4.13 – Deformed FE model of RVE 4 - 27
Figure 4.14 – Two-stage prediction of load vs strain behaviour
(PFR core) 4 - 27
Figure 4.15 – Incremental prediction of load vs strain behaviour
(PFR core) 4 - 29
Figure 4.16 – Predicted load vs strain curve for plaster core
specimens 4 - 30
Figure 4.17 – Predicted load vs strain curve for foam core
specimens 4 - 31
Figure 4.18 – Predicted load vs strain curve for neat resin core
specimens 4 - 32
Figure 4.19 – σ11 stress distribution of RVE 4 - 34
Figure 4.20 – Typical tension specimens 4 - 36
Figure 4.21 –Test set-up 4 - 37
Figure 4.22 – Typical tension specimen failure modes 4 - 37
Figure 4.23 – Predicted and Experimental load vs strain graphs
for tension 4 - 43
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xiii
elements
Figure 4.24 – SEM of typical transverse crack 4 - 45
Figure 5.1 – Location of cut line 5 - 2
Figure 5.2 – Typical preliminary test specimens 5 - 3
Figure 5.3 – Typical MFC truss compression member section
geometry 5 - 5
Figure 5.4 - Test set-up 5 - 8
Figure 5.5 – Typical compression load vs axial shortening curves
5 - 8
Figure 5.6 – Typical failure modes 5 - 9
Figure 5.7 – Compression member failure zones 5 - 12
Figure 5.8 – Proportions of stocky specimens with a non-compact
cross
section
5 - 14
Figure 5.9 – Stocky / non-compact member section 5 - 15
Figure 5.10 – Typical slender member cross section 5 - 16
Figure 5.11 – First critical buckling mode (λ = 77.9, applied
load = 4 kN) 5 - 19
Figure 5.12 – Slender member (symmetric about x-x neutral axis)
cross
section
5 - 21
Figure 5.13 – Locations of extracted stocky compression
specimens with
compact cross section
5 - 22
Figure 5.14 – 5 mm thick PFR core 5 - 23
Figure 5.15 – Stocky specimen with non-compact cross section 5 -
24
Figure 5.16 – Surface profile measurement of stocky specimen
with a non-
compact cross section
5 - 24
Figure 5.17 – Production of slender specimen cores 5 - 25
Figure 5.18 – Out-of-straightness measurement of slender
specimens 5 - 26
Figure 5.19 – Finished slender specimens 5 - 26
Figure 5.20 – Test rig for stocky member with a compact cross
section 5 - 27
Figure 5.21 – Test setup for stocky member with non-compact
cross
section
5 - 28
Figure 5.22 – Test setup for slender specimens 5 - 28
Figure 5.23 – Typical load vs strain graph for Pratt and Warren
specimens 5 - 30
Figure 5.24 – Typical Pratt specimen failure modes 5 - 31
Figure 5.25 – Typical Warren specimen failure modes 5 - 31
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xiv
Figure 5.26 – Strain gauge and dial gauge readings 5 - 32
Figure 5.27 – Typical slender specimen load vs strain
measurements 5 - 33
Figure 5.28 – Typical flexural test specimen 5 - 37
Figure 5.29 – Revised local buckling model (applied load = 4000
N, λ =
161.
5 - 38
Figure 6. 1 – Sensitivity of lap strength to lap length 6 -
5
Figure 6. 2 – Failure loads of lap specimens 6 - 6
Figure 6.3 – Simplified Goland and Reissner model 6 - 6
Figure 6.4 – Stress distribution of Goland and Reissner model 6
- 7
Figure 6.5 – Stress distribution of Goland and Reissner model 6
- 8
Figure 6.6 – Section through truss joint 6 - 9
Figure 6.7 – Section A-A 6 - 9
Figure 6.8 – FE model of joint section 6 - 10
Figure 6.9 – Stress distribution along bond plane of model 6 -
11
Figure 6.10 – Modified lap-shear specimen 6 - 13
Figure 6.11 – Modified lap-shear test configuration 6 - 13
Figure 6.12 – Average tensile failure stress of modified lap
specimens 6 - 14
Figure 6.13 – Test specimen configuration 6 - 15
Figure 6.14 – “T” joint specimens 6 - 16
Figure 6.15 – Test fixtures 6 - 17
Figure 6.16 – “T” joint test set-up 6 - 17
Figure 6.17 – Typical failure mode of Type 1 (lap joints) 6 -
18
Figure 6.18 – Typical load vs displacement curve for Type 1 (lap
joints) 6 - 18
Figure 6.19 – Typical failure mode of Type 2 (loop joints) 6 -
19
Figure 6.20 – Typical load vs displacement curve for Type 2
(loop joints) 6 - 19
Figure 6.21 – Extracted Pratt joint 6 - 21
Figure 6.22 – Joint geometry (solid areas hatched) 6 - 21
Figure 6.23 – Full casting mould 6 - 23
Figure 6.24 – Finished PFR core 6 - 24
Figure 6.25 – Finished Type 1 joint 6 - 24
Figure 6.26 – Finished Type 2 joint 6 - 25
Figure 6.27 – Actual PFR core section 6 - 25
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xv
Figure 6.28 – Finished Type 4 joint 6 - 26
Figure 6.29 – Schematic test rig 6 - 27
Figure 6.30 – Test set-up 6 - 27
Figure 6.31 – Typical failure mode of Type 1 joint 6 - 28
Figure 6.32 – Typical failure mode of Type 2 joint 6 - 29
Figure 6.33 – Typical failure mode of Balsa core 6 - 29
Figure 6.34 – Failed PFR core 6 – 30
Figure 6.35 - First noise location 6 – 32
Figure 6.36 – Finite element mesh, loading and restraints 6 –
34
Figure 6.37 – Finite element model section geometry and
material
properties
6 - 35
Figure 6.38 – Bending moment distribution in simply supported,
four-
panel, rigid jointed Pratt truss
6 - 36
Figure 6.39 – Member stress distribution with secondary stresses
6 - 37
Figure 6.40 – Effect of stress reversal on member strain in
primary tensile
member
6 - 38
Figure 6.41 – Relaxation of primary stress 6 - 39
Figure 6.42 – Deflection of fixed-end beam 6 - 40
Figure 6.43 – Deflection of truss panel 6 - 40
Figure 6.44 – Four-panel MFC Pratt truss frame model with UDL 6
- 41
Figure 6.45 – Truss deflection 6 - 41
Figure 6.46 – Typical 2D Plate element Pratt joint model 6 -
42
Figure 7.1 – Truss geometries and dimensions 7 - 2
Figure 7.2 – Typical joint details 7 - 3
Figure 7.3 – Nominal truss core cross section dimensions 7 -
3
Figure 7.4 – Void extents 7 - 4
Figure 7.5 – Fibre architecture 7 - 5
Figure 7.6 – Typical truss member cross section 7 - 5
Figure 7.7 – Strain measurement locations 7 - 7
Figure 7.8 – Strand 7 plate element types used in finite element
analysis 7 - 8
Figure 7.9 – Final finite element meshes for Pratt, Howe and
Warren
trusses
7 - 8
Figure 7.10 – Number of layers per face in MFC truss tension
members 7 - 11
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xvi
Figure 7.11 – Typical MFC truss member RVE 7 - 12
Figure 7.12 – Symmetric FEA model of RVE 7 - 12
Figure 7.13 – Deformed RVE, εeq = 6.3E-5mm/mm 7 - 13
Figure 7.14 – Typical material properties for tension members
(Pratt truss
shown)
7 - 16
Figure 7.15 – Number of layers per face in MFC truss
compression
members
7 - 18
Figure 7.16 – Typical material properties for compression
members (Pratt
truss shown)
7 - 19
Figure 7.17 – Material properties and orientations in a typical
Pratt truss
joint
7 - 21
Figure 7.18 – Typical detailed and simplified truss joints 7 -
23
Figure 7.19 – Typical detailed and simplified truss joints 7 -
24
Figure 7.20 - ε11 Max for different joint forms 7 - 25
Figure 7.21 – Typical displacement points (Pratt truss shown) 7
- 27
Figure 7.22 – Typical isotropic material properties used in
truss joints 7 - 29
Figure 7.23 – Typical external loading and constraints (Pratt
truss shown) 7 - 30
Figure 7.24 – Truss midspan top chord deflections 7 - 31
Figure 7.25 – Pratt truss FE model – fibre direction strain
distribution 7 - 33
Figure 7.26 – Pratt truss local compressive strain concentration
7 - 34
Figure 7.27 – Howe truss FE model – fibre direction strain
distribution 7 - 34
Figure 7.28 – Howe truss local tensile strain concentration 7 -
35
Figure 7.29 – Warren truss FE model – fibre direction strain
distribution 7 - 35
Figure 7.30 – Warren truss local tensile strain concentration 7
- 36
Figure 7.31 – Typical mould configuration 7 - 38
Figure 7.32 – Typical void former restraint 7 - 39
Figure 7.33 – Typical finished MFC trusses 7 - 40
Figure 7.34 – Typical MFC member cross section 7 - 42
Figure 7.35 – Typical MFC truss joints 7 - 42
Figure 7.36 – Peel ply at strain gauge locations 7 - 45
Figure 7.37 – Typical strain gauge locations 7 - 45
Figure 7.38 – Typical test setup (Pratt truss shown) 7 - 46
Figure 7.39 – MFC truss load versus displacement graphs 7 -
47
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xvii
Figure 7.40 – Strain versus load curves for Pratt, Howe and
Warren trusses 7 - 49
Figure 7.41 – Failure zone of Pratt truss 7 - 51
Figure 7.42 – Location of first noise in Pratt truss 7 - 51
Figure 7.43 – Failure zone of Howe truss 7 - 52
Figure 7.44 – Location of first noise in Howe truss 7 - 52
Figure 7.45 – Failure zone of Warren truss 7 - 53
Figure 7.46 – Location of first noise in Warren truss 7 - 53
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xviii
List of Tables Table 2.1 – Material cost comparison between
traditional materials and
fibre composites
2 - 10
Table 2.2 – Indicative Strength/dollar ratio for different
materials 2 - 11
Table 2.3 – Comparison of specific strength and specific elastic
modulus
between traditional materials and fibre composite materials
2 - 16
Table 2.4 – Issues of emerging materials 2 - 36
Table 3.1 – Common characteristics of mechanical and adhesive
joints 3 - 14
Table 3.2 – Member length and number of joints 3 - 34
Table 3.3 – Truss midspan deflection 3 - 35
Table 3.4 –Maximum combined stress in each truss 3 - 36
Table 3.5 – Percent increase in stresses due to joint rotation
in each truss 3 - 37
Table 3.6 – Mechanical properties of ADR246 / West Systems 105 3
- 40
Table 3.7 – Mechanical properties of LB100 polystyrene foam
typical 3 - 43
Table 3.8 – Mechanical properties of end-grain balsa (Diab,
2002) 3 - 47
Table 3.9 – PFR (44% SL150 filler by vol) typical mechanical
properties 3 - 49
Table 3.10 – Summary of materials adopted 3 - 51
Table 3.11 – Contribution to member compressive stiffness by
core
materials
3 - 57
Table 4.1 – Material properties for preliminary tension element
tests 4 - 4
Table 4.2 – Tension test results 4 - 10
Table 4.3 – Additional tension test observations for Type 1
specimens 4 - 11
Table 4.4 – Core and hardpoint materials 4 - 19
Table 4.5 – Four core / laminated face systems 4 - 20
Table 4.6 – Material properties for prediction of PFR tension
member pre-
cracking behaviour
4 – 23
Table 4.7 – Material properties of RVE used in FEA 4 - 25
Table 4.8 – Incremental equivalent elastic modulus 4 – 28
Table 4.9 – Material properties for prediction of plaster core
tension
member behaviour
4 - 30
Table 4.10 – Material properties for foam core tension member
behaviour 4 - 31
Table 4.11 – Material properties for neat resin core tension
member 4 - 32
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xix
behaviour
Table 4.12a – PFR core specimen results 4 - 39
Table 4.12b – Additional PFR core specimen results 4 - 40
Table 4.13 – Plaster core specimen results 4 – 40
Table 4.14 – Foam core specimen results 4 - 41
Table 4.15 – Neat resin core specimen results 4 - 41
Table 4.16(a) – Comparison of predicted results with
experimental results 4 - 42
Table 4.16 (b) – Comparison of predicted results with observed
continued 4 - 42
Table 4.17 – Difference in average ultimate properties of PFR
and plaster /
foam core specimens
4 - 49
Table 5.1 – Mechanical properties of 720gsm UD Colan E-glass and
BASS
Pacific 420gsm heatset UD E-glass
5 - 5
Table 5.2 – Adopted material properties 5 - 6
Table 5.3 – Predicted capacity of compression specimens 5 -
7
Table 5.4 – Comparison of test results with predictions 5 -
10
Table 5.5 – Slenderness ratio of test specimens 5 - 13
Table 5.6 – Calculation of predicted failure load for specimen
W1 5 - 17
Table 5.7 – Specimen predicted failure loads 5 - 17
Table 5.8 – Material properties used in FEA of stocky member
with non-
compact cross section
5 - 18
Table 5.9 – Test results for stocky specimens with a compact
cross section 5 - 29
Table 5.10 – Slender specimen results summary 5 - 33
Table 5.11 – Comparison of predicted and test values 5 - 34
Table 5.12 – Flexural elastic modulus 5 - 37
Table 6.1 – Estimated joint and member strengths 6 - 3
Table 6.2 – FE model properties (refer Figure 6.7) 6 - 11
Table 6.3 – Stress / capacity ratio 6 - 12
Table 6.4 – Lap lengths 6 - 13
Table 6.5 – Failure loads of Type 1 (lap joints) 6 - 18
Table 6.6 – Failure loads of Type 2 (loop joint) 6 - 19
Table 6.7 – Test joint materials 6 - 22
Table 6.8 – Joint fibre architecture 6 - 22
Table 6.9 – Joint ultimate capacity 6 - 28
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xx
Table 7.1 – RVE FEA model properties 7 - 13
Table 7.2 – Gross tension member tensile elastic moduli 7 -
14
Table 7.3 – Apparent tension member core tensile elastic moduli
7 - 15
Table 7.4 – RVE FEA model properties 7 - 18
Table 7.5 – Comparison of actual joint strains with simplified
joint strains 7 - 26
Table 7.6 – Comparison of IP deflection 7 - 28
Table 7.7 –Truss deflections and strain gauge readings at 100kN
load 7 - 32
Table 7.8 – Truss stiffness results from FE models 7 - 32
Table 7.9– Derived load carrying capacity 7 - 36
Table 7.10 – MFC truss weight, first noise load and crack
spacing 7 - 48
Table 7.11 – Truss deflections at intermediate loads and
ultimate load,
stiffness at failure and span-to-deflection ratio at failure
7 - 48
Table 7.12 – Summary of member strains at ultimate load 7 -
50
Table 7.13 – Truss stiffness comparison 7 - 54
Table 7.14 – Comparison of deflection and strain for Pratt
Truss1 7 - 54
Table 7.15 – Comparison of deflection and strain for Howe Truss1
7 - 55
Table 7.16 – Comparison of deflection and strain for Warren
Truss1 7 - 55
Table 7.17 – Strength / failure mode comparison summary (Pratt)
7 - 56
Table 7.18 – Strength / failure mode comparison summary (Howe) 7
- 56
Table 7.19 – Strength / failure mode comparison summary (Warren)
7 - 57
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xxi
Notation
Abbreviations
AFRP = aramid fibre reinforced plastic (polymers)
CERF = Civil Engineering Research Fund
CFRP = carbon fibre reinforced plastic (polymers)
COV = coefficient of variation
CTE = coefficient of thermal expansion
DB = double bias fibres
FCDD = Fibre Composites Design and Development
FEA = finite element analysis
FRP = fibre reinforced plastic (polymers)
GFRP = glass fibre reinforced plastic (polymers)
gsm = grams per square metre
HFS = high failure strain
HSHF = high strength / high failure strain
HSLF = high strength / low failure strain
LFS = low failure strain
LSHF = low strength / high failure strain
LSLF = low strength / low failure strain
MFC = monocoque fibre composite
NASA = National Aeronautics and Space Administration
PFR = particulate filled resin
QUT = Queensland University of Technology
RVE = representative volume element
UD = unidirectional fibres
UDL = uniformly distributed load
USQ = University of Southern Queensland
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xxii
Symbols
k = stiffness
σ = stress
ε = strain
δ = deflection
φ = ultimate limit state capacity factor
ν = Poisson’s ratio
σ11.c.ult = ultimate compressive stress in the fibre
direction
ε11.c.ult = ultimate compressive strain in the fibre
direction
ε11.t.ult = ultimate tensile strain in the fibre direction
σ11.t.ult = ultimate tensile stress in the fibre direction
ι12 = shear stress
σ22.c.ult = ultimate compressive stress perpendicular to the
fibre direction
ε22.c.ult = ultimate compressive strain perpendicular to the
fibre direction
σ22.t.ult = ultimate tensile stress perpendicular to the fibre
direction
ε22.t.ult = ultimate tensile strain perpendicular to the fibre
direction
εfirst noise = strain at which first noise occurs
σu = ultimate limit state stress
A = area
Acore = area of core
Afaces = area of faces
Ag = gross area
b = cross sectional width
d = cross sectional depth
E = elastic modulus
Ixx = second moment of area about the X-X axis
E11.c = compressive elastic modulus in the fibre direction
E11.t = tensile elastic modulus in the fibre direction
E11.t.core = tensile elastic modulus of core in the fibre
direction
E11.t.eq = equivalent tensile elastic modulus in the fibre
direction
E11.t.faces = tensile elastic modulus of faces (nominally 20 GPa
for E-glass)
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xxiii
E22.c = compressive elastic modulus perpendicular to the fibre
direction
E22.t = tensile elastic modulus perpendicular to the fibre
direction
EAF1 = elastic modulus of face material in area 1
EAF2 = elastic modulus of face material in area 2
EBP = elastic modulus of “BASS Pacific” laminate
Eeq = equivalent elastic modulus based on transformed cross
section
Ef = final elastic modulus
Ei = initial elastic modulus
EL = elastic modulus in the member’s longitudinal direction
EPFR = elastic modulus of PFR
G12 = shear modulus
h = depth of truss (between panel points)
I = second moment of area
Imin = second moment of area about the minor principal axis
Ixx.AF1 = second moment of area of face area 1 about the X-X
axis
Ixx.AF2 = second moment of area of face area 2 about the X-X
axis
Ixx.BP
= second moment of area of “BASS Pacific” laminate about the
X-X
axis
Ixx.PFR = second moment of area of PFR about the X-X axis
Iyy = second moment of area about the Y-Y axis
L = length
Lav = average debond length
Nc = nominal member capacity in compression
Ns = nominal section capacity of compression member
P = point load
PE = Euler critical buckling load
Sav = average crack spacing
T = tension force in truss member
t = thickness
w = uniformly distributed load
δ = deflection
ε = strain
ε11.c.ult = ultimate compressive strain in the fibre
direction
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses xxiv
ε11.t.ult = ultimate tensile strain in the fibre direction
ε22.c.ult = ultimate compressive strain perpendicular to the
fibre direction
ε22.t.ult = ultimate tensile strain perpendicular to the fibre
direction
εfirst noise = strain at which first noise occurs
φ = ultimate limit state capacity factor
ι12 = shear stress
ν = Poisson’s ratio
σ = stress
σ11.c.ult = ultimate compressive stress in the fibre
direction
σ11.t.ult = ultimate tensile stress in the fibre direction
σ22.c.ult = ultimate compressive stress perpendicular to the
fibre direction
σ22.t.ult = ultimate tensile stress perpendicular to the fibre
direction
σu = ultimate limit state stress
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 1
Chapter 1 - Introduction
Fibre composite materials are gaining recognition in civil
engineering applications as
a viable alternative to traditional materials. Their migration
from customary
automotive, marine, aerospace and military industries into civil
engineering has
continued to gain momentum over the last three decades as new
civil engineering
applications develop. The use of fibre composite materials in
civil engineering has
now evolved from non-structural applications, such as handrails
and cladding, into
primary structural applications such as building frames, bridge
decks and concrete
reinforcement.
In recent years, researchers around the world have been seeking
to develop new and
innovative structural forms that can successfully exploit the
benefits offered by these
materials (Karbhari & Zhao, 2000; Gowripalan, 1999; van Erp,
1999d; Hooks et. al.,
1997 and Creative Pultrusions, 2002a). One structural form that
appears to have
significant potential is the truss. The framed nature of trusses
can be used to
minimise the shortfalls of fibre composite materials and
maximise their benefits. To
date, much of the development of fibre reinforced polymer (FRP)
trusses has
focussed on connection of continuous profile members using
mechanical joints or
secondary bonding (Creative Pultrusions, 2002b; NASA, 2002;
Goldsworthy 1995;
Morsi & Larralde, 1994a and Strongwell, 2002). Mechanical
joints can introduce
performance compromises into fibre composite trusses while the
use of continuous
profile members tends to reduce the freedom of the designer to
use materials
efficiently and effectively.
This dissertation presents an investigation into a novel fibre
composite truss
proposed by The University of Southern Queensland Fibre
Composite Design and
Development. The truss uses monocoque construction to eliminate
the use of
mechanical fasteners or secondary bonding to connect members,
whilst allowing the
designer freedom to locate material where it will perform most
efficiently and have
the greatest effect. It will be shown that structures of this
type can be developed with
predictable structural behaviour and a level of strength and
stiffness suitable for civil
engineering applications.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 2
1.1 Background
Composite materials combine and maintain two or more distinct
phases to produce a
material that has properties far superior than either of the
base materials. Fibre
composites are two-phase materials in which one phase reinforces
the other. High
strength fibres are used as the primary means of carrying load
and a polymer resin
binds the fibres into a cohesive structural unit. The
combination of fibres and resin
produces a bulk material with strength and stiffness governed
primarily by the fibres
and chemical resistance provided by the resin.
Evidence has been found to suggest that fibre composite
materials have been used in
construction for thousands of years. Straw has been used to
reinforce bricks for over
3000 years and this method is still used today. Chinese bridge
builders used timber as
early as 1600 BC and Greek builders were apparently the first to
reinforce masonry
with metal around 1000 BC. By comparison, the development of
modern synthetic
fibre composites is relatively new, beginning in the early 20th
century.
In the first half of the 1900’s the development of synthetic
resins which could cure at
room temperature, combined with the serendipitous discovery of a
method to
manufacture fine glass fibres, led to an increased use of
synthetic fibre composites.
The unique properties offered by glass fibre reinforced
composites made them
particularly desirable for marine, aerospace, military and
automotive applications.
Material characteristics such as low weight, tailorable
durability, good fatigue
resistance and manufacturing versatility were some of the
characteristics initially
recognised as potential advantages over existing materials. Over
the last sixty years
fibre composites have enjoyed widespread use in these industries
and have evolved
significantly. Developments include production of advanced
reinforcing fibres, such
as carbon and aramid, which offer improved strength and elastic
modulus and better
impact performance over glass, and the evolution of modern
polyester, vinylester and
epoxy resin formulations which offer improved performance over
the original high
temperature curing phenolic resin in areas such as mechanical
properties, chemical
resistance and bonding.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 3
As these materials find new applications their significance to
engineering fields with
which they were previously unfamiliar is becoming more
pronounced. In the last
twenty years fibre composite materials have established
themselves as a viable and
competitive option for rehabilitation and retrofit of existing
civil structures. By
externally bonding fibres, pre-fabricated strips and jackets to
deteriorated or obsolete
structures, strength and stiffness can be re-introduced or
improved to increase service
life. Fibre composites are also used to replace steel as a
reinforcing and stressing
material in concrete for some specialised applications. To a
lesser extent new civil
structures have been created almost entirely from fibre
composite materials by
joining standard structural sections or modular components to
produce complete
structures.
Civil and structural engineering is seeing an increased push for
the use of these
materials in mainstream structures. This push is being driven by
both composites and
civil engineering groups. Civil engineers are driven primarily
by the desire to realise
the potential performance benefits offered by these materials.
Potential benefits
include high strength, low weight and environmental durability.
For the civil /
structural engineer high strength and light weight may translate
into reduced
construction times and costs through the use of products which
can be brought to
near-complete state in an off-site factory and then easily
transported to site for rapid
deployment. The advantages of a material with superior
durability include potential
reductions in maintenance over the life of a structure and hence
lower cost for the
owner.
The composites industry, on the other hand, is driven by a
desire to participate in
what is arguably the world’s largest industry. The global civil
engineering and
construction market turnover has been estimated at US$800
billion per annum
(Marsh, 2000). With the global composites market in the year
2000 being valued at
only around US$8 billion (Marsh, 2000), even a 1% stake in the
construction market
would double its current size.
Regardless of the reasons, the international pressure to use
composites in mainstream
civil engineering has been growing over the past decade. During
this time a large
number of experimental and commercial structures have been
constructed around the
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 4
world to demonstrate the potential of composites in major civil
engineering
structures. Figure 1.1 provides examples of these
structures.
While these applications help to demonstrate that fibre
composite materials are
structurally capable, there are issues which are slowing the
ingress of fibre composite
materials into the civil engineering industry. One such issue is
that there is currently
a significant cost premium associated with their use.
b) Road Bridge - AUS (Source: FCDD, 2002)
a) Lattice Towers - USA
(Source: Strongwell, 2002)
c) Dome roof structures – Libya
(Source: Hollaway, 2002)
Figure 1.1 – Applications of fibre composites in civil
engineering structures
Other issues exist such as difficulties in realising potential
benefits, general lack of
civil engineers’ familiarity with the material and relatively
little standardisation in
the composites industry. For composites to truly offer a viable
alternative to
traditional construction materials in the civil engineering
marketplace, it is essential
that these issues are addressed. It is proposed that this
situation could be improved by
demonstrating that potential benefits offered by composites can
be achieved with
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 5
familiar civil engineering forms that are well suited to fibre
composite materials and
whose structural behaviour is consistent with safe civil
engineering structures and
predictable with existing structural engineering analysis and
design methods.
One structural form that would appear to offer significant
potential for composites in
civil engineering is the truss. Trusses have been accepted as
efficient structural
elements for centuries and offer a number of advantages over
solid web members.
Typically they use much less material than solid web members and
their framed
nature allows material to be located where it has the greatest
effect. Given the
significantly higher costs of fibre composite laminates in
comparison to traditional
structural materials such as steel and concrete, this low
material usage may address
some of the cost disparities between traditional and composite
structures. In addition
to this, the loading within truss elements is largely axial. It
is thought that this would
enable the easy tailoring of reinforcement along the load paths,
resulting in a high
level of efficiency in material usage.
The concept of a fibre composites truss is not entirely new.
Several examples of this
type of structure have been constructed around the world using a
variety of joint
configurations (see Figure 1.2). However, these trusses are very
much a reflection of
traditional technology developed for timber or metal members. As
a result they fall
short of fully exploiting the potential offered by
composites.
a) FRP truss bridge with pultruded
members (Source: Berenberg, 1997)
b) Bolted connection of pultruded
members (Source: Berenberg, 1997)
Figure 1.2 – Examples of fibre composite trusses
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 6
c) Inserts used to connect FRP truss
members (Source: Hollaway, 2002)
d) Serrated FRP truss joint
(Source: Goldsworthy 1995)
Figure 1.2 – Examples of fibre composite trusses
In 1998 the USQ proposed a new type of fibre composite truss.
Unlike previous
composite trusses, the new concept was based on monocoque design
and avoided the
need for secondary joints. Initial investigations into this
truss were extremely
promising and the concept was identified as one with significant
development
potential. However, there is a need to fully understand the
structural behaviour of this
new monocoque fibre composite (MFC) truss before it can be used
confidently.
1.2 Aims
The primary aim of this thesis is to develop and improve the
fundamental
understanding of the structural behaviour of monocoque fibre
composite trusses and
to advance the civil engineering community’s knowledge in the
use of fibre
composite materials in civil engineering structures.
In fulfilling the broad aim above, the study will:
develop the MFC truss concept into a working structure
investigate the structural behaviour of the principal truss
elements (tension
and compression members and joints)
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
hallaThis figure is not available online. Please consult the
hardcopy thesis available from the QUT Library
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 7
demonstrate that the structural behaviour of principal truss
elements can be
predicted using well established structural engineering analysis
and design
methods
design a prototype truss in Pratt, Howe and Warren
configurations using
understanding gained from truss element and joint
investigations
evaluate the structural performance of the three prototype
trusses through
physical testing and computer analysis
characterise the behaviour of the MFC trusses in terms of
stiffness, strength,
failure mode, predictability and warning of failure
conclude on the ability of the MFC trusses to provide a safe,
predictable and
adequate structural solution
1.3 Scope
In order to maintain focus on the primary aims of this
investigation, adhere to
budgetary constraints of this research project and work within
limitations of available
fabrication and testing equipment, the following restrictions
were imposed on the
scope of this project:
Truss tests were undertaken for a single load / support system
with
defined restraint conditions and loading points
A single prototype of each full-scale MFC truss design was
fabricated
and tested (in this thesis the term “full-scale” refers to
trusses that are
the same size as the finite element model)
Structural behaviour investigations were limited to
short-term,
pseudo-static loading at ambient temperature in a
non-aggressive
environment
Long-term durability was not considered
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 8
1.4 Thesis structure
Chapter 2 – Fibre Composites In Civil Engineering: presents a
more in-depth
discussion of the background to this project. It examines the
history of composites
and the potential benefits that make them attractive to civil
engineers. Three basic
areas of fibre composite use in civil engineering type
applications are discussed,
followed by an examination of the issues that affect the use of
fibre composite
materials in civil engineering.
Chapter 3 – Monocoque Fibre Composites Trusses: introduces the
MFC truss
concept and the methodology used to develop the concept into a
working structure.
The Chapter begins with a brief discussion of the history of
trusses and some of the
characteristics that saw them become the structure of choice for
bridge construction
during mid 1800’s. Focus is then shifted to existing FRP
trusses, a selection of which
are critically examined to identify potential shortfalls of the
current approaches. The
philosophy behind the MFC truss is then presented and compared
with existing FRP
truss construction, in particular how the MFC truss has the
potential to overcome
some of the shortfalls of existing FRP truss technology. Chapter
3 then examines and
evaluates a number of materials and structural form options for
the MFC truss,
finally presenting the adopted MFC truss configuration that will
form the basis of the
study.
Chapter 4 - Static Structural Behaviour of MFC Truss Tension
Elements:
presents the results of an investigation into the behaviour of
axial MFC truss tension
elements constructed using the prototype cross-section developed
in Chapter 3. MFC
truss tension members with a variety of core materials are
investigated
experimentally and analytically to determine their load
response, failure mode,
predictability and warning of failure. An incremental method
based on established
structural engineering principles is developed to predict the
tension member
behaviour. These predictions are compared to results obtained
from testing to
determine their accuracy.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 1 - 9
Chapter 5 – Static Structural Behaviour of MFC Truss Compression
Elements:
investigates the structural behaviour of typical MFC truss
compression members.
Three classes of compression member are investigated, namely;
stocky members
with a compact cross-section, stocky member with a non-compact
cross section and
slender members. Established closed form solutions as well as
finite element analysis
are used to predict their structural behaviour. These
predictions are then compared to
results obtained through testing to determine the accuracy of
the predictions.
Chapter 6 - Monocoque Fibre Composite Truss Joints: shifts the
focus of the
investigation to the MFC truss joints. In this Chapter the work
is primarily concerned
with the ability of the joint to provide adequate connection to
members. Joint
strength is studied experimentally and analytically and the
effect of the rigid nature
of the joint on connected members and truss deflection is
investigated. Finally, the
suitability of a proposed simplified approach to joint analysis
is discussed.
Chapter 7 - Static Structural Behaviour of MFC Trusses: applies
the findings of
Chapters 4, 5 and 6 in the examination of a series of prototype
truss designs. These
designs are analysed using finite element analysis techniques
and predicted
behaviour is evaluated through fabrication and testing of
several sample trusses.
Observations including strength, failure mode, stiffness and
provision of warning of
imminent failure are made.
Chapter 8 – Conclusions and Recommendations: draws together key
findings on
the structural behaviour of MFC trusses and methods for
predicting such behaviour.
Finally, several recommendations for future work are identified
and discussed.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 1
Chapter 2 - Fibre Composites in Civil Engineering
Synthetic fibre composite materials have existed for over a
century and have been
used widely in military, aerospace, sporting and automotive
applications. In the last
three decades they have been increasingly considered for civil
applications mainly in
retrofit of existing structures, reinforcement for concrete and
to a lesser extent
complete fibre composite structures. Their use in civil
engineering applications has
been the focus of worldwide research by both the composites
industry and the civil
engineering industry. This research has highlighted a number of
potential benefits as
well as some concerns with the use of this material in civil
engineering applications.
To date the bona fide use of fibre composite materials as load
bearing structural
elements in civil applications remains rare and it is unclear
why this is the case. This
chapter examines key fibre composites issues in relation to
civil engineering, draws
conclusions on their viability as civil engineering materials
and recommends a way
forward.
2.1 Introduction
Modern fibre composites originated in the late 19th century when
the first man-made
polymer, phenol-formaldehyde, was reinforced with linen fibre to
make Bakelite. In
1936, DuPont patented the first room temperature curing resin,
unsaturated polyester.
The first epoxy resin system was produced in 1938 and Ciba
introduced the widely
recognised Araldite epoxy resin system in 1942. At the same time
reinforcing fibres
were undergoing rapid development and in 1941 Owens-Corning
began production
of the world’s first woven glass fabric.
The defence and marine industries were amongst the first to
exploit some of the
potential advantages of reinforced polymer composites such as
relatively high
strength-to-weight ratio, good durability and fatigue
performance and radiowave
transparency. The automotive industry was able to capitalise on
characteristics such
as efficient production techniques leading to inexpensive
tooling, rapid turn-around
production and high quality surface finishes. The Cold War
prompted significant
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 2
effort by the military in the development of fibre composite
materials. This led to
improvements in composites processing technology and mechanical
properties of
laminates. The reader is directed to Gibson (1994), Herakovich
(1997), Swanson
(1997), Kaw (1997), Vinson & Chou (1995), Mallick (1997),
Johnson (1994), Lubin
(1982), Ayers (2002) and Owens Corning (2002) for more detailed
information.
The use of fibre composite materials in civil structures can be
traced back to the
1960’s when glass reinforced plastic rods were used to reinforce
concrete. In the
1970’s civil FRP structures included roofs, pedestrian bridges,
pipes, in-ground tanks
and phone boxes (see Yoosefinejad and Hogg, 1997; Liao et al,
1998; Holloway,
2002 and Sharjah Airport Corporation, 2002). However, the end of
the Cold War in
the late 1980’s contributed to a glut of fibre composite
resources and the composites
industry began a concerted effort to migrate fibre composites
technology to
infrastructure applications.
The 1990’s saw significant development in the application of
fibre composite
technology to civil infrastructure led by rehabilitation and
retrofit projects.
Developments were also made in the use of fibre composite
materials as concrete
reinforcement and to a lesser extent civil structures comprised
primarily of fibre
composite materials. These developments will be discussed in
more detail in the next
section.
In the last few years the use of fibre composites in civil
engineering applications has
made some progress. However, these materials have not enjoyed
the level of
widespread acceptance predicted by its proponents over the last
decade. This slow
ingress appears to be due to a number of issues, which are
discussed later in this
chapter.
2.2 Fibre composite materials in construction and civil
engineering
Non-structural fibre composites have enjoyed widespread use in
the construction
industry for many years in non-critical applications such as
baths and vanities,
cladding, decoration and finishing. However, the use of
structural fibre composites in
critical load-bearing applications remains rare mainly
consisting of rehabilitation and
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 3
retrofit of existing infrastructure, reinforcement of new
concrete structures and new
civil structures constructed predominantly of fibre composite
materials. These are
discussed next.
2.2.1 Rehabilitation and retrofit
Rehabilitation and retrofit currently represents the largest
structural use of fibre
composite materials in civil engineering applications. The
widespread deterioration
of infrastructure in Canada, the USA and Europe is well
documented (Head, 1994;
Karbhari, 1997, 1998, 2000; Rizkalla, 1999 and Green, 2000). The
estimated cost to
rehabilitate and retrofit existing infrastructure worldwide is
around CAD$900B
(ISIS, 1998). In Australia it is estimated that $500M per annum
is required to repair
and upgrade concrete structures (Oehlers, 2000). In addition to
this a large amount of
infrastructure is reaching the end of its design life as
revisions in structural codes and
loading codes combined with increased traffic demands are
raising load limits on
existing infrastructure. Earthquakes in Loma Prieta (1989),
Northridge (1994) and
Kobe (1995) have demonstrated the vulnerability of many of the
existing concrete
structures to the effects of earthquake (Karbhari, 1998;
Rizkalla, 1999). In many
cases demolition and rebuilding of these structures is difficult
to justify in
economical terms so engineers seek inexpensive and effective
methods to strengthen
them.
Traditional rehabilitation and retrofit methods use concrete or
external steel sheets to
re-introduce or improve structural properties such as strength
and ductility. The
ability of concrete to form complex shapes and its suitability
to submerged
installation has seen it used for encapsulation of elements such
as bridge piers
(Carse, 1997). However, concrete’s relatively low stiffness,
high density, frequent
requirement of complex formwork and difficulties in achieving
sufficient bonding to
the substrate and sufficient compaction to properly protect
reinforcing steel are seen
as drawbacks in general retrofit applications. Steel, on the
other hand, can be bonded
or bolted to deteriorated concrete structures to provide
strength and stiffness
improvements with relatively little additional weight. However,
steel plates can be
difficult to use on complex shapes and protective coatings
required by steel plates
can be compromised during installation. Both the concrete and
steel systems tend to
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 4
inherently provide additional stiffness to the structure which
can attract additional
load. This can be a disadvantage in some cases, particularly if
foundation capacity is
limited.
Fibre composites are often used as a surface layer that either
protects and/or
improves the response of the encapsulated element. In these
cases the materials are
usually bonded externally to the structure in the form of tows
(fibre bundles), fabrics,
plates, strips or jackets. The strength of circular or near
circular concrete members
can be improved through confinement provided by tangentially
oriented reinforcing
fibres without the introduction of significant stiffness.
Strength or stiffness
improvement in bending members, such as beams and slabs, can be
achieved by
bonding laminated strips to the underside of the member.
The advantages offered by composites in these forms include
their ability to bond
well to many substrate materials and to follow complex shapes.
Composites also
offer a potential benefit over isotropic retrofit materials,
such as steel, by allowing
enhancement of strength without increasing stiffness and vice
versa. This can be an
advantage for strength enhancement of bridge piers where
increasing stiffness could
attract unwanted extra load.
An important consideration in the decision to use traditional or
FRP rehabilitation
methods is cost. FRP rehabilitation methods are available, but
they are currently
carried out by specialist subcontractors and tend to be
expensive. However, Carse
(2003) provides an example of an application in which FRP
rehabilitation was
competitive on cost and provided a more desirable solution in
terms of aesthetics and
protection compared to traditional methods. In this case bridge
piers were
rehabilitated with concrete below the waterline and FRP above
the waterline.
2.2.2 Concrete structures reinforced with fibre composites
Concrete reinforced with FRP materials has been under
investigation for decades.
Unstressed FRP reinforcement has been developed in a number of
forms including
ribbed FRP rod similar in appearance to deformed steel
reinforcing bar, undeformed
E-glass and carbon fibre bar bound with polyester, vinylester or
epoxy resin, E-glass
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 5
mesh made from flat FRP bars and prefabricated reinforcing cages
using flat bars and
box sections (Shapira et al, 1997; Ko, 1997; Harris, 1998 and
Gowripalan, 1999).
Stressed FRP reinforcement is also available, usually consisting
of bundles of rods or
strands of fibre reinforced polymer running parallel to the axis
of the tendon. These
are used in a similar fashion to conventional steel tendons (El
Kady et al, 1999 and
Gowripalan, 2000).
The durability performance of FRP reinforcements is considered
by Ko (1997),
Harris (1998) and Gowripalan (1999) to offer a possible solution
to the problem of
corrosion of steel reinforcement, a primary factor in reduced
durability of concrete
structures. Other reported advantages of FRP rebar include
enhanced erection and
handling speeds (Karbhari, 1999) and suitability to applications
which are sensitive
to materials which impede radiowave propagation and disturb
electromagnetic fields.
However, in many cases corrosion of reinforcing steel can be
traced back to
deterioration of concrete resulting from poor design, materials
or workmanship.
Well-designed steel-reinforced concrete can produce extremely
durable structures
and examples exist of concrete elements found to be in excellent
condition despite
continuous tidal zone wetting and drying in a saline marine
environment for over 70
years (Carse, 1997). Good design and construction is likely to
be a more
economically feasible approach than use of expensive FRP rebar
which can be up to
eight times (Tilco, 2002) as expensive as uncoated reinforcing
steel and around one
and a half times as expensive as stainless steel reinforcement
(Arminox, 2003).
In terms of erection and handling speeds, FRP rebar can
sometimes prove more
difficult to work on site than traditional reinforcing steel. An
example of this is the
common requirement for reinforcing steel to be bent on site, a
characteristic not
possessed by thermoset FRP rebars. In relation to radiowaves and
electromagnetic
fields, carbon fibre rebar would offer little benefit over
traditional steel
reinforcement as it is not transparent to radio waves or
electromagnetic radiation. In
these cases E-glass bars are often used as they are able to
combine good radiowave
and electromagnetic transparency with adequate structural
performance. However,
GFRP rebar has disadvantages such as relatively low strength,
which could require
up to seven times the area of steel to satisfy deemed-to-comply
requirements of
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 6
AS3600 for minimum strength (SA, 2001) and susceptibility to
stress rupture which
can require a large amount of reinforcing bar to keep stresses
low. In addition to
this, the relatively low stiffness of GFRP rebar can require
deeper sections and
greater reinforcing areas to achieve serviceability limits.
Applications that are not sensitive to radiowave or
electromagnetic transparency may
use stronger and stiffer carbon fibre rebar. CFRP rebar can
offer up to three times the
stiffness and three times the strength of GFRP rebar. However,
these bars are usually
more expensive than GFRP rebars and in most cases will never be
able to develop
their claimed high strength due to stringent serviceability
limits which restrict the
maximum strain developed at serviceability failure. For example,
members
supporting masonry commonly adopt a deflection limit of around
span/500 (SA,
2001). Based on this deflection limit, a concrete member with a
typical span-to-depth
ratio of around 12 will develop approximately 0.1% strain in the
rebar. Carbon fibre
typically has an ultimate strain of around 1% and as CFRP rebar
is predominantly
unidirectional, the material will be used to approximately 1/10
of its capacity at the
serviceability limit state. The result is inefficient use of FRP
material and difficulties
in producing a serviceable concrete member. It is likely that
unstressed FRP rebar
will be limited to applications not governed by serviceability
and will need to
consider the material cost fibre composites compared to other
available traditional
materials. This is discussed further in Section 2.3.1.
On the other hand, stressed FRP rebar can allow more efficient
materials usage at
serviceability limit states. The action of pre-stressing can
take up some of the excess
strain capacity of FRP materials allowing fuller development of
the material’s
characteristic high strength and production of a large area of
concrete in compression
resulting in a stiff and strong concrete member. However,
tendons are required to
carry sustained load for long periods of time and issues such as
stress rupture and
creep must be addressed. Materials such as CFRP and AFRP are
generally favoured
over GFRP as they are not susceptible to stress rupture, except
at higher stress levels,
and their tendency to creep can be accommodated at the design
stage by over
stressing. A significant difference between FRP stressing
tendons and steel tendons
is that unidirectional FRP tendons are predominantly brittle and
do not cope well
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 7
with stress concentrations at crack locations, unlike ductile
steel. It is therefore
advisable for FRP tendons not to be bonded to the concrete.
The use of stressed FRP reinforcement can provide a more
economical use of
material than unstressed FRP and therefore will probably be
favoured in applications
requiring FRP reinforced concrete for reasons of cost.
2.2.3 New fibre composite civil structures
A small number of new load bearing civil engineering structures
have been made
predominantly from FRP materials over the last three decades.
These include
compound curved roofs (Hollaway, 2002 and Sharjah Airport
Corporation, 2002)
pedestrian and vehicle bridges and bridge decks (Hazen and
Bassett, 1998; Karbhari,
2000; Kollar, 1998 and FHWA, 2002), energy absorbing roadside
guardrails (Bank
and Gentry, 2000), building systems (Barbero and GangaRao,
1991), access
platforms for industrial, chemical and offshore (Hale, 1997),
electricity transmission
towers, power poles, power pole cross-arms and light poles
(Goldsworthy, 1998;
ISIS, 1998 and Weaver, 1999), modular rooftop cooling towers
(Barbero and
GangaRao, 1991) and marine structures such as seawalls and
fenders (Weaver,
1999). The benefits most often claimed to be offered by fibre
composites include
high specific strength and specific stiffness, tailorable
durability, good fatigue
performance and the potential to reduce long-term costs.
However, in many cases
these claims are difficult to substantiate and are often based
on sparse and irrelevant
data. Currently many civil engineers are sceptical of the
material’s ability to provide
a viable alternative to traditional materials and bona-fide
applications are scarce.
Many of the existing applications are experimental in nature and
are aimed at
demonstrating the ability of fibre composite materials to
perform in certain
applications. To this end they are often successful in terms of
structural performance,
but offer little by way of meaningful financial performance
data. In most cases
groups of interested parties combine to design, fabricate and
install the structure at a
reduced cost. Examples of this cooperation are evident in
projects such as the
Bennett’s Creek crossing on New York State route 248 (Allampali
et al, 2000) and
the Tech 21 road bridge (Farhey, 2000).
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 8
Allampali et al (2000) provides some financial data for the
Bennett’s Creek crossing.
These figures indicate a cost of around US$400,000 to produce
the temporary bridge,
based on reduced rates and omitting some items such as deck
wearing surface and
some engineering costs. Based on average costs to produce a
similar pre-stressed
concrete plank bridge in Australia, at around $1000/m2, it is
likely that a complete
permanent bridge could have been built for half the cost of the
FRP option.
However, the extra cost of an FRP option could be amortised over
the expected
service life of a project. This would offer little justification
in the case of a temporary
bridge but, as will be discussed later, may be a significant
consideration in structures
with a long service life, or a temporary structure which can be
re-used. This would
only be an advantage if potential benefits such as high
durability could be realized to
keep maintenance costs down and provided that the material and
structure were
durable enough to allow multiple re-use.
New civil structures may also benefit from the ability to
produce large modular
components allowing rapid deployment of an FRP structure,
although the benefit of
this would be most significant where the cost of public
inconvenience and traffic
management is high. Other potential benefits such as high
specific stiffness and high
specific strength may exist, but at a cost which makes them
uncompetitive with
traditional materials. These issues suggest that fibre
composites would mainly suit
non-standard civil applications or those which can balance extra
cost against some
unique composites property.
2.3 Issues affecting the use of fibre composites in civil
engineering applications
To date fibre composite materials have not enjoyed widespread
use in civil projects.
To understand the reasons for this it is necessary to examine
the key issues affecting
the use of fibre composite materials in structures in a civil
engineering context. A
number of issues affecting the use of fibre composite materials
in civil engineering
applications have been highlighted by research undertaken over
the last decade
(Ballinger, 1990; Meier, 1991; Gangarao, 1993; Morsi and
Larralde, 1994; Karbhari,
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 9
1996; Scalzi et al, 1999; van Erp et al, 2000). The following
key issues will be
examined in the context of application to civil engineering:
1. Cost
2. Structural performance
3. Durability
4. Familiarity and education
5. Specification and standardisation
6. Compatibility
7. Temperature and fire performance
2.3.1 Cost
In most civil engineering structures, good design requires
provision of a solution
which can satisfy design requirements for the lowest cost. Cost
can be considered in
terms of short-term costs, such as design, construction and
installation, and long-term
costs such as maintenance, modification, deconstruction and
disposal. These can be
further grouped into direct costs, such as materials and
production, and indirect costs,
such as interruptions to traffic, depreciation, resale value and
impact on the
environment.
Short term costs of fibre composites
Currently, fibre composite materials are expensive when compared
to conventional
construction materials on an initial cost basis. This is
demonstrated in Tables 2.1 and
2.2 which compare the cost per stiffness and cost per strength
of FRP materials with
traditional construction materials. In engineering terms
stiffness can be expressed as
LEAk =
Where E is the Modulus of Elasticity of the material (N/mm2), A
the area (mm2) and
L the length (mm).
Strictly speaking this equation applies to axial loading only
but it can also be used to
compare materials loaded in bending assuming the dimensions of
the cross section
are fixed.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 10
Determining how much “stiffness” can be bought for a dollar is a
realistic way to
compare the stiffness capabilities of different materials. The
cost of a material can
be expressed as
( ) ( ) densitylengthareakgAcost ×××= /$
Assuming a length of 1 m, the stiffness in N/mm per dollar is
given by
( ) densitykgAEk
×=
/$1000 .
Table 2.1 – Material cost comparison between traditional
materials and fibre
composites
Material E
(N/mm2)
Cost
(A$/kg)
Density
(kg/m3)
Stiffness/dollar
(N/mm per A$)
Pultruded glass composites 30,000 7.00 1,800 2,381
Carbon composites 90,000 30.00 1,400 2,142
Standard Construction Steel 200,000 2.50 7,850 10,190
Steel Rebar 200,000 1.20 7,850 21,230
Stainless Steel Rebar 200,000 6.00 7,850 4,246
50 MPa concrete 30,000 0.10 2,500 120,0001
Hard wood timber 16,000 2.50 650 9,846
New polymer concrete 11,000 0.75 1,900 7,719
1. Compression only
A similar table can be assembled for strength. For a 1 m long
bar, the load carrying
capacity per dollar can be expressed as:
( ) densitykgANu
××
=/$
106φσ
where φ is an indicative ultimate limit state capacity factor
and uσ is the ultimate
limit state stress (N/mm2).
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 11
Table 2.2 – Indicative strength/dollar ratio for different
materials
strength/dollar
(N/A$)
Material uφσ
(N/mm2)
Cost
(A$/kg)
Density
(kg/m3)
tension compression
Pultruded glass
composites
0.6x600(T*)
0.6x480(C)
7.00 1,800 17,143
13,714
Carbon composites 0.7x900(T)
0.6x720(C)
30.00 1,400 15,000
12,000
Standard Construction
Steel
0.9x300 2.50 7,850 13,758 13,758
Steel Rebar 0.9x500 1.20 7,850 47,770 47,770
Stainless Steel Rebar 0.8x500 6.00 7,850 8,492 8,492
50 MPa concrete 0 (T)
0.8x50(C)
0.10 2,500 0
160,000
Hard wood timber 0.6x40(T)
0.6x50(C)
2.50 650 14,770
18,462
New polymer concrete 0.6x10(T)
0.6x50(C)
0.75 1,900 4,210
21,053
(*T is tension, C is compression)
These results clearly show that fibre composites struggle to
compete financially with
traditional construction materials both in terms of stiffness
and strength. Although
these performance criteria are rarely considered alone, they are
often fundamental in
“material-evaluation”.
There are a number of factors contributing to the high cost of
composite materials
including; high cost of raw materials and processing, the use of
imported materials,
the general acceptance of high prices in markets such as marine
and aerospace and
occasional low availability of material (Goldstein, 1996). The
production of
materials locally is likely to reduce material cost, however
with America and Europe
making up 35% and 27% of the worldwide market respectively
(Weaver, 1999),
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 12
there appears to be little incentive for material manufacturers
to provide production
facilities in Australia.
In line with the evolution of other composites uses, such as
sporting equipment,
many researchers and composites commentators believe it is
likely that production
volume increases resulting from the use of fibre composites in
civil engineering
applications will lead to decreased cost of materials (Karbhari,
1998a; Weaver, 1999;
Hastak and Halpin, 2000). However, in the Author’s opinion, this
should be viewed
as an optimistic outlook. The majority of fibre composites
materials used in Australia
are imported and are therefore subject to a range of
international economic variables.
For example, fluctuations in the local price of imported
materials would be affected
by overseas production costs, transport and import costs and
fluctuations in the
exchange rate between Australia and countries such as Europe,
United States and
Japan, which supply us with carbon, aramid, E-glass fibres and
many resins.
When this is considered in conjunction with the tendency of
suppliers to provide
price reductions for purchase of large quantities of some
materials, accurate costing
of an FRP civil project can seem difficult. The uncertainty that
exists in our ability to
accurately cost FRP civil project suggests that it could be some
time before
anticipated price drops could significantly influence project
cost.
Fabrication cost
In addition to relatively high material costs, the short-term
cost of FRP materials is
dependant on fabrication. Most fibre composite manufacturing
techniques were
originally developed for the aircraft, marine and/or car
industries. The civil
engineering industry is vastly different to these as civil and
structural engineers tend
to be concerned with the design and construction of rather
large-scale structures.
Most of these structures have to meet different design
specifications and therefore
very little duplication of design solutions occurs. As a result
most civil engineering
projects tend to be ‘one-off’ jobs. This situation is in
contrast with the manufacturing
industries, where mass production of one design solution is
common. As a result,
design and manufacturing methods which are highly successful in
the manufacturing
industry are often not viable in civil engineering.
-
Development and Structural Investigation of Monocoque Fibre
Composite Trusses 2 - 13
Most civil engineering applications of fibre composites are
currently based around
the pultrusion process (similar to extrusion of aluminium). The
pultrusion process is
ideal for the continuous production of elements of constant
cross-sectional geometry
and moderate complexity. The advantages are relatively low
labour cost, minimal
material wastage, consistent quality and high production rates.
However, the
pultrusion process has some serious disadvantages such as high
initial costs of setting
up for a production run and relatively few producers.
Consequently, pultruders offer
a limited range of standard structural sections similar to those
available in steel. A
non-standard section would require a significant volume to
justify the high setup
cost. In many civil engineering projects, this high volume
simply does not exist and
for these situations, a cheaper, more flexible manufacturing
approach is required.
One of the most flexible methods of FRP fabrication is
hand-layup. This method
allows individual placement of fibres and resin onto surfaces of
virtually any
topography. However, this method tends to be inefficient and
laborious and is
usually avoided as much as possible where large volumes are
required. The most
likely way forward is through use of modular components
fabricated using a
combination of manual and automated manufacturing procedures
such as embedment
of standard pultruded sections in cast components, filament
winding, automated tape
laying and resin transfer moulding. These procedures can be
computer controlled to
produce accurate components with lower labour costs. Integration
of these methods
could be particularly suitable for civil applications, which
often require highly
aligned fibres and fabrication versatility.
Some short-term costs, such as transport and erection, may
benefit from production
of large, lightweight, modular components. Lower weight can
translate into reduced
transport and cranage costs, while the use of fewer large
modular components can
reduce erection time. The implications on indirect short-term
costs such as consumer
inconvenience and traffic management can be substantial. Meiers
(2000) points out
that although it is difficult to quantify indirect savings, they
have a cost that is
present. He believes that savings can be accrued at the systems
level due to faster
construction thereby causing less distress and disruption to the
community, lower
dead weight requiring smaller and l