Scientific Literature Retrieval Final Report PDF

Scientific Literature
Retrieval Final Report

学院名称宇航学院
专业名称飞行器设计
学生姓名郭咦贝 Christovyo Sebastian Kusnadi
学生学号 LS1815205
指导教师刘丰睿
2018 年 12 月 13 日
Scientific Literature Retrieval Final Report
Study on Failure Prediction Method of
Composite Multi-Bolt Joints
Author: Christovyo Sebastian Kusnadi
Tutor: Liu Feng Rui 刘丰睿
Abstract
Applications of composite materials in aircraft structures has been increasing in the past few decades.
Especially with the design of composite mechanical joints, that has attracted an increasing attention in
the aircraft structure context. One of the frequently used for the mechanical joints of composite
structures are Bolted Composite Joints (can be multi-bolted or single-bolted joints), which have
advantages such as high load-carrying and load-transferring capacities, can be conveniently dismantled
and replaced, and are preferred connections in composite structures. However, due to the high stress
concentrations stemming from hole drilling and anisotropic multiphase composite components, both of
the multi-bolted and single-bolted joints are consistently potential weak spots in such composite
structures, and their failure may exert a critical influence on structural safety and integrity. Therefore,
failure behaviours of composite bolted joints have attracted much attention in the last decades. In this
study, we will focus on several different failure analysis or prediction method for composite multi-
bolted joints, such as bolt load transfer distribution analysis of multi-bolted joints by using finite
element method, failure prediction of single-bolted joints based on progressive damage method and the
effect of geometric dimension on the failure of composite bolted joints. In this study, several simulations
or tests using computational method by the help of ABAQUS software will be completed.
Keywords: Composites, Multi-bolt joints, Finite element analysis, Progressive damage analysis,
Failure mode map, Geometric dimension
Contents
1 Introduction .......................................................................................................................................... 1
1.1 Aerospace Engineering: Flight Vehicle Design ............................................................................ 1
1.2 Research topic ............................................................................................................................... 1
2 Databases ............................................................................................................................................. 3
2.1 Main databases used ..................................................................................................................... 3
2.2 Step by step process of the literature retrieval .............................................................................. 3
3 Information retrieval ............................................................................................................................ 5
3.1 English sources ............................................................................................................................. 5
4 Literature Review: Failure Prediction Method on Composite Multi-Bolt Joints ............................... 13
4.1 Composites in Aviation............................................................................................................... 13
4.2 Development of Failure Prediction Method for Composite Bolt Joints...................................... 14
4.2.1 Multi-Bolt Composite Joints ................................................................................................ 15
4.2.2 Single Bolt Composite Joints ............................................................................................... 20
4.2.3 Several Other Methods......................................................................................................... 22
5 Reference List .................................................................................................................................... 24
6 Supervisor evaluation......................................................................................................................... 26
1 Introduction
1.1 Aerospace Engineering: Flight Vehicle Design
Aerospace engineering is the primary field of engineering that is concerned primarily with the
development of aircraft and spacecraft. It has two major developing part and branches that
overlaps on each other which are the Aeronautical Engineering and Astronautical Engineering.
This sub branches which are critical to the production of aircraft and spacecraft, which includes
various technological and engineering disciplines such as aerodynamics, propulsion, avionics,
materials science, structural analysis and manufacturing. Because of the complexity and
number of disciplines involved, this field of aerospace engineering is usually carried out by
teams of engineers, designers and technicians that are each having their own specialized area
of expertise.
Developing an aircraft or a flight vehicle is a challenge by itself, so does the process in
designing an aircraft. The aircraft design or flight vehicle process is the engineering design
process by which aircraft are designed. These depend on many factors such as customer and
manufacturer demand, safety protocols, physical and economic constraints etc. Aircraft design
by itself is both art and science and it is a separate discipline of aerospace engineering which
is different from analytical discipline such as aerodynamics, structural analysis, flight control
and propulsion. It is an intellectual engineering process of creating on paper flying machine.
Aircraft design is by its nature an iterative process. This means that estimates and assumptions
have sometimes to be made with inadequate data such “guess estimates” must be check when
more accurate data on the aircraft is available. Requirements are set by prior design trade
studies. Concepts are developed to meet requirements. Design analysis frequently points
towards new concepts and technologies which can initiate a whole new design effort. All of
these activities are equally important in producing a good aircraft concept.
1.2 Research topic

Nowadays, most of structure designers especially in the field of aircraft design preferred to use
multi-bolt composite joints as the connections in composite structures because of the ease of
dismantling, convenient detection and high reliability in which having high load-carrying and
load-transferring capacity. Unfortunately, as this joint will act as the core component that will
do the transferring of high loads, the multi-bolt composite joint became a weak spot for the
composite structure which affect the problem of making it to be having a good structural
integrity and safety. In order to get to that goal of solving this problem, further analysis of the
multi-bolt composite joint’s strength is required.
The method of failure prediction for multi-bolt composite joints generally may include two
steps, in which the first one is to do an overall joint analysis to be able to determine the load
distribution between the bolts. The second one is to do a detailed local strength analysis around
a critical hole in order to ensure a safety margin that will cover for all failure modes. As
composites materials have a characteristic of anisotropy and brittleness, multi-bolt composite
joints will experience a minimal amount of plastic stress relief and mechanical softening as
compared to their metallic counterparts. This will lead into an uneven load distribution that
happen between bolts.
Each bolt hole in the joints will be subjected to differently combined bearing bypass loads.
Generally, due to the final catastrophic failure of the joints is determined by the combined
1
bearing bypass stresses at the critical hole, it is really important for conducting an accurate load
transfer analysis in multi-bolt composite joints in order to be able to determine the combined
stress status of the critical hole.
There are several factors that must be considered in doing the load transfer analysis of multi-
bolt joints, which includes joint materials, laminate lay-ups, laminate thickness, hole diameter,
fastener position (with ratios of width-to-hole diameter and edge-to-hole diameter), type of
fastener (countersunk or protruding head), bolt-hole clearance and the torque for bolt-
tightening.
With all of factors that needs to be considered as stated in the paragraph before, there are several
famous failure prediction or analysis method that is used to predict the failure behaviour of the
multi-bolted composite structure joints, which are analytical approaches, simple spring-based
methods, FE methods, Progressive damage methods, Characteristic length method and many
more. For this particular study on failure prediction method of composite multi-bolt joints, it
will be focusing on bolt load transfer distribution analysis of multi-bolted joints by using finite
element method, failure prediction of single-bolted joints based on progressive damage method
and the effect of geometric dimension on the failure of composite bolted joints. In this study,
several simulations or tests using computational method by the help of ABAQUS software will
be completed.
2
2 Databases
2.1 Main databases used
English sources: Elsevier ScienceDirect
2.2 Step by step process of the literature retrieval

(1) Go to the Beihang’s Library website (http://lib.buaa.edu.cn/):
(2) Choose either the English or Chinese database and search for keywords or topics by
the search bar:
(3) Scroll down to the bottom of the page and switch to English version of website:
3
(4) In the next step choose the web-based electronic sources:
(5) Then, a new webpage will be opened and we can find about 58 foreign databases in
the left column and 30 databases of Chinese sources in the right column:
(6) Then, we can continue to do literature search and retrieval by using any of the
databases and find any information, research papers or thesis that are related to the
research topic required.
4
3 Information retrieval
3.1 English sources
1. An analytical joint stiffness model for load transfer analysis in highly torqued multi-bolt
composite joints with clearances
Source: Elsevier ScienceDirect / Composite Structures 131 (2015) 625–636
Review: In this study, an analytical tri-linear joint stiffness model accounting for the effect of bolt-hole
clearances on the bearing chord stiffness is presented. A single-parameter iteration method is also
proposed to solve the nonlinear system of equations of the load transfer analysis, in which the recursive
formulae are verified to be monotonic. To verify the present joint stiffness model, a series of single-lap,
single-bolt joints with various bolt-hole clearances and bolt tightening torques were tested to
empirically confirm the influence coefficients of the joint stiffness model. The load distributions in
three single-lap, three-bolt joints with various clearances were also tested under finger-tight and highly
torqued conditions. The load distributions in multi-bolt joints predicted by the presented joint stiffness
model together with the single-parameter iteration method are in good agreement with the experimental
outcomes, verifying the accuracy of the presented joint stiffness model and the feasibility of the single-
parameter iteration method. In addition, the effects of the friction forces originated from the bolt-
tightening torque and bolt-hole clearances on the load distribution in multi-bolt joints are presented.
2. An interpretation of the load distributions in highly torqued single-lap composite bolted
joints with bolt-hole clearances
Source: Elsevier ScienceDirect / Composites Part B 138 (2018) 194–205
Review: This paper presented a detailed interpretation of load distribution in composite single-lap
bolted joints which is based on theoretical, experimental and numerical investigations. Modified
instrumented bolts (MIBs) were used to measure bolt shear loads consisting of the bolt bearing load and
the friction force between the laminate and nut. The bolt load carrying mechanism in single-bolt joints
and the load transfer mechanism in multi-bolt joints with various tightening torques and bolt-hole
clearances were revealed based on theoretical and numerical analyses and validated against
experimental results. The load transferred by each bolt in a multi-bolt joint was determined with greater
fidelity, and the load distributions obtained using a three-dimensional finite element (FE) model and a
spring-based method were compared. The load distributions obtained from the established 3D FE model
were found to be identical to those obtained using the spring-based method; the bolt-hole clearances
significantly impacted the load distributions, but the effects of the bolt tightening torques on the load
distributions could be ignored.
3. Investigation of bolt load redistribution and its effect on failure prediction in double-lap,
multi-bolt composite joints
Review: In this paper, a test method was presented to measure the bolt load distribution in the
progressive damage process of multi-bolt composite joints, which exhibits continuous changes with the
damage propagation in composite laminates. Based on the repeated loading-unloading cycles, bolt shear
loads of a double-lap, three-bolt joint were tested by alternately using modified instrumented bolt-nut
pairs and aircraft-grade fasteners, from which the bolt load redistribution history was achieved. In
addition, a progressive damage method was adopted to trace the load redistribution history of the joint.
The numerical predictions were in good agreement with the experimental outcomes, both of which show
that the load distribution tended toward equal proportions before the final failure of the joint. Using a
modified failure envelope method, the failures of two-, three- and four-bolt composite joints were
predicted based on the uniform load distribution and actual load distribution, and compared with that
based on the traditional load distribution calculated by a traditional stiffness method (TSM). It follows
5
that the effect of the actual load distribution on the failure of the multi-bolt joints can be more faithfully
reflected by adopting a uniform load distribution than by the traditional one.
4. A bi-material property-based FE modelling method for progressive damage analyses of
composite double-lap bolted joints
Source: Elsevier ScienceDirect / Results in Physics 11 (2018) 674–683
Review: This paper proposed a bi-material property method to describe mechanical properties of the
unidirectional composite lamina. It divided the lamina into the fiber layer and the resin layer, which
have different material properties and are responsible for the intra-laminar and inter-laminar properties,
respectively. Furthermore, a finite element modelling method with the composite lamina simulated by
the bi-material property method was presented. Combined with a Hashin-type failure criterion and a
micromechanics-based material degradation model, a novel progressive damage model (PDM) was
constructed for typical double-lap bolted joints. To validate the novel PDM, four double-lap, single-
bolt joints with various width-to-hole-diameter ratios were designed and quasi-static tensile tests of the
joints were performed. Numerical results of the novel PDM show better agreements with the
experimental outcomes than the traditional PDMs. Besides, the novel PDM can describe the inter-
laminar damage more precisely via a more reasonable simulation for the inter-laminar property of the
lamina, and can also obtain more accurate deformation of the hole along the thickness direction of the
laminate. Furthermore, influences of the width-to-hole-diameter ratio W/D on failure behaviours of the
double-lap, single-bolt joint were revealed, and an optimal width-to-hole-diameter ratio was
recommended for engineering practice. Combined with our previous study on effects of the end-
distance-to-hole-diameter ratio E/D, a failure mode map with varying W/D and E/D was determined for
the double-lap bolted joint. It is helpful for designers to seek safer designs of composite bolted joints
because the W/D and E/D are key geometrical parameters that significantly affect failure behaviours of
the joints.
5. Probabilistic bolt load distribution analysis of composite single-lap multi-bolt joints
considering random bolt-hole clearances and tightening torques
Review: This paper presents a stochastic bolt load distribution analysis method of composite single-lap
multi-bolt joints by using an improved three-stage spring-based method and Monte Carlo simulation.
In the proposed method, the random properties of tightening torque, bolt-hole clearance, lamina
properties and geometric parameters were involved. The allowable manufacturing tolerance band of
parameters and clearance fit were considered, and a combining distribution function consisting of
binomial distribution and normal distribution function was adopted for modeling the random tightening
torque relaxation. To validate the proposed method, a set of composite single-lap three-bolt joints were
designed and tested to achieve the stochastic bolt load distribution of the joint. Good agreements
between the numerical and experimental stochastic bolt load distribution validated the proposed method.
Furthermore, the probabilistic bolt load distributions of the three-bolt joints in four tightening torque
conditions, including finger-tight, wrench torque, standard tightening torque and relaxation tightening
torque, were investigated. It is found that tightening torque conditions and bolt-hole clearances have
significant influences on the variation of the bolt load distribution, while the effects of other random
parameters on the variation of the bolt load distribution are slight.
6. A micromechanical model for longitudinal compressive failure in unidirectional fiber
reinforced composite
Review: A micromechanics-based 2D numerical model is proposed to comprehensively trace the
longitudinal compressive failure of unidirectional carbon fiber reinforced polymers (UD-CFRP), in
which microscopic damage mechanisms (namely the fiber fracture, matrix plastic yield and fiber-matrix
interface crack) are considered. With this model, the predicted compressive modulus, strength and
failure pattern are in good agreements with the experimental results, which validate the effectiveness of
the proposed model. It concludes that the longitudinal compressive failure is governed by the matrix
6
damage rather than the fiber fracture and the interface crack triggers the matrix damage. In addition, the
influence of some microscopic parameters, including the fiber volume fraction, initial fiber
misalignment, initial fiber-matrix interface stiffness, interface strength and fracture energies, on the
longitudinal compressive failure is disclosed. The compressive modulus increases with the fiber volume
fraction while it is almost unaffected by the rest four parameters. The compressive strength increases
with the increment of the initial interface stiffness or the decrement of the initial fiber misalignment
while it is immune to the interface fracture energies. Moreover, a limited increment of the fiber volume
fraction or the interface strength is beneficial for enhancing the compressive strength.
7. A progressive damage analysis based characteristic length method for multi-bolt composite
joints
Review: This paper provides A progressive damage analysis based characteristic length method (PDA-
based CLM) with only unidirectional layer properties required is proposed to predict the failure of
composite multi-bolt joints. This method absorbs the convenience of the CLM on predicting failure of
multi-bolt joints and introduces the PDA to obtain the characteristic lengths, which avoids time-
consuming and expensive characteristic length tests. Meanwhile, only two simple and universal
structures, an open-hole laminate and a single-bolt joint, instead of complex multi-bolt joints, are
conducted by the PDA, which provides positive advantages for prediction accuracies and escapes from
unbearable time consumption due to large amount of numerical iterations. Two progressive damage
models (PDMs) which are composed by in situ strength, Hashin or modified Yamada–Sun criterion and
Camanho’s degradation rules are recommended. Static tensile tests of multi-bolt joints were also carried
out. Although the final failure load errors of multi-bolt joints predicted by the method proposed are
over-conservative, further schematically designed error transfer analysis discloses the errors are mainly
caused by the CLM itself. In fact, the errors induced by the numerical characteristic lengths are no more
than 4%, which suggests a good prospect of the proposed method in engineering.
8. A novel characteristic curve for failure prediction of multi-bolt composite joints
Review: In this paper, a novel characteristic curve determined by tensile, compressive and shear-out
characteristic length is presented to predict failure load and failure mode of mechanically fastened
composite joints. In contrast to the existing curves, the novel one introduces a shear-out characteristic
length and provides capability to predict shear-out failure. A mathematics formulation of the novel
curve is proposed based on the expression of Chang’s characteristic curve and shear-out characteristic
length. A series of tests were carried out and nonlinear finite element analyses were conducted to obtain
the tensile, compressive and shear-out characteristic lengths for the novel curve. Following tensile tests
of two-bolt and three-bolt joints were performed, and the failure loads and modes were numerically
predicted by the characteristic curve method with the novel curve and Chang’s. The numerical results
obtained from the curve presented show good agreements with the experimental outcomes and better
accuracy on both failure load and failure mode.
9. Secondary bending effects in progressively damaged single-lap, single-bolt composite joints
Review: In this paper, static tensile experiments and progressive failure simulations of single-bolt,
single- and double-lap joints were carried out to comparatively investigate secondary bending effects,
which present significant eccentric-loading phenomena in single-lap joints but are almost non-existent
in symmetric double-lap joints. Progressive damage models of single-lap and double-lap joints were
established, from which the numerical predictions were found to be in good agreement with the
experimental outcomes. Experimental macro-scope failure patterns and seven numerical micro-scope
failure modes obtained from the progressive damage analyses were presented for the two types of joints.
The effects of secondary bending on the mechanical degradation and failure mechanism of single-lap
joints were revealed. Some characterizations of secondary bending in the joints, such as a characterized
parameter on the AGARD points, joint deformations and contact states, were quantitatively traced
7
during the total progressive damage process. All these characterizations increased the understanding of
the effects of secondary bending on the failure process of a single-bolt, single-lap joint.
10. A probabilistic model for strength analysis of composite double-lap single-bolt joints
Review: This study presents a probabilistic model involving the randomness of basic design parameters
to predict the probabilistic strength of a typical composite double-lap single-bolt joint. In the
probabilistic model, statistical models of the random parameters including geometrical dimensions,
fiber orientation angles and material properties were constructed. In addition, a parametric failure
prediction model employing a modified characteristic curve method was established. Static tensile tests
of fifteen composite double-lap single-bolt joints made of T800 carbon/epoxy composites were carried
out. The probabilistic failure load of the joint obtained from the proposed model is in good consistency
with that from the experiments, which verifies the effectiveness and accuracy of the proposed model.
Meanwhile, it shows that the proposed model is not sensitive to the probability distribution type of
random variables, although the probability distribution type of random variables slightly influences the
statistical parameters of the probabilistic failure load. Furthermore, a relationship between the failure
probability and the tensile load is determined, and longitudinal compressive strength XC, ply thickness
ply and longitudinal elastic modulus E11 of T800 unidirectional lamina are the key factors affecting the
probabilistic failure load and reliability of the joint remarkably.
11. A progressive fatigue damage model for composite structures in hygrothermal environments
Source: Elsevier ScienceDirect / International Journal of Fatigue 111 (2018) 299–307
Review: This paper proposed an improved progressive fatigue damage model (PFDM) involving
hygrothermal effects to predict fatigue failure of composite structures in hygrothermal environments.
In the improved PFDM, a unified model was established to evaluate the variation of composite material
properties caused by hygrothermal environments. The hygrothermal-induced material properties were
utilized in the stress analysis model, material degradation models and fatigue failure criterion. The
hygrothermal strains were introduced into the constitution equation to account for the hygrothermal
effects. A residual-strain-based gradual material degradation model and a micromechanics-based
sudden material degradation model were enhanced to describe the damage of composite materials in
hygrothermal environments. Besides, fatigue tests and progressive fatigue damage analyses of an open-
hole laminate were performed in hygrothermal environments. The good consistency between the
numerical and experimental results validated the improved PFDM. Both the experimental and
numerical outcomes indicate that the effective moisture equilibrium considerably reduces the
compressive strength and compressive fatigue life of the open-hole laminate made of T800
carbon/epoxy composites.
12. Design and analysis of a novel bolted composite π joint under bending load
Source: Elsevier ScienceDirect / Materials and Design 98 (2016) 201–208
Review: This study proposed a novel design of out-of-plane composite π joints to elevate the load
carrying capability. This design adopts an in-plane π-shaped section scheme along the loading direction
to eliminate the delamination damage, which commonly occurs in traditional composite π joints, and
uses bolts to connect the joint and other structural components. Static experiments of the novel bolted
composite π joint and an aluminum counterpart were conducted. The bolted composite π joint exhibited
a larger load carrying capability with a massive weight reduction. Compared with existing static
experimental results from a traditional design, the novel bolted composite π joint also showed a greater
load carrying capability. A progressive damage analysis of the bolted composite π joint was further
performed and validated by experimental data. Based on numerical simulation, the progressive damage
process and failure mechanism of the bolted composite π joint were discussed. It follows that the novel
design not only eliminated potential weaknesses in deltoid fillers but also brought positive factors of
composite into play and substantially strengthened the composite π joint.
13. A modified failure envelope method for failure prediction of multi-bolt composite joints
Source: Elsevier ScienceDirect / Composites Science and Technology 83 (2013) 54–63
8
Review: This paper presents a modified failure envelope to predict final failure mode and strength of
multi-bolt composite joints based on the conventional failure envelope method, which was presented
by Hart-Smith from abundant strength tests of open hole laminates, double-lap single and multi-bolt
joints with quasi-isotropic or near-quasi-isotropic lay-ups. In contrast to the bearing ‘‘cut-off’’ of
conventional failure envelope method, the modified one takes into account the effect of bypass load on
the bearing failure and a new polyline consisted of two oblique lines is proposed. It is able to be
established by an additional compressive strength test of laminates and semi-analytical or numerical
analysis of fastener-hole laminates and open-hole laminates, combined with the point stress criterion.
A flowchart of integration structure failure analysis process with the modified failure envelope method
is provided. Series of tests were carried out to provide basic parameters and finite element analysis was
conducted for the modified failure envelope. The method proposed was applied to two-bolt and four-
bolt carbon-to-carbon double-lap joints with near-quasi-isotropic lay-ups. The predictions of the method
proposed and the conventional one were compared with the tests data. The results indicate that the
method proposed can obtain effective prediction of failure modes, and more accurate ultimate failure
loads in double-lap composite bolted joints.
14. A progressive failure analysis of all-C/SiC composite multi-bolt joints
Review: A progressive damage model (PDM) was proposed to analyse the failure of plain-weave
carbon fibre-reinforced silicon carbide (C/SiC) composite joints with C/SiC bolts, which were
manufactured by the chemical vapour infiltration (CVI) technique. A bilinear constitutive model was
established to describe the pseudo-plasticity of the C/SiC composites. Several failure criteria, including
the maximum stress failure criterion, Tsai-Wu failure criterion, and Hashin-type criterion, and several
material degradation models were compared. Moreover, the deposition models accounting for the
deposition effects of CVI in the all-C/SiC composite joints were discussed. A series of comparative
failure analyses of a C/SiC open-hole composite laminate with available experimental data in the
literature was conducted to determine the most suitable combination of the Hashin-type criterion and
the combined material degradation model was filtered for C/SiC composite structures. Based on the
PDM, the failure strength and failure pattern of an all-C/SiC composite single-lap, three-bolt joint were
predicted. Static tension experiments testing the all-C/SiC composite single-lap, three-bolt joints
fabricated by the CVI technique were also conducted. Good consistency between the numerical failure
predictions and the experimental outcomes was obtained, indicating the effectiveness of the proposed
PDM for failure analyses of all-C/SiC composite multi-bolt joints.
15. Failure load prediction of composite bolted joint with clamping force
Review: In this paper, failure load prediction for composite joints with clamping force was conducted
using a characteristic length method combined with Tsai-Wu failure criteria. Tensile and bearing tests
and stress analyses were conducted to determine tensile and compressive characteristic lengths. A
characteristic curve was used to perform failure load prediction for single lap-shear bolted joints. This
prediction was accompanied by experiments and stress analysis. The predicted failure load was
compared with the actual failure obtained from experiments, and the results were found to be in good
agreement.
16. Effect of damage on failure mode of multi-bolt composite joints using failure envelope
method
Review: In this paper, tensile experiments of two-bolt and three-bolt joints were conducted to obtain
the failure modes and other properties. Three dimensional (3D) finite element models involving damage
were constructed, and the results were consistent with experimental data. Then, the models were used
to study the effect of damage on failure mode of multi-bolt composite joints. The results show that
failure mode prediction with failure envelope method should involve the effect of composite damage.
Compared with net-section damage, bearing damage has greater influence on failure mode. Based on
conventional failure envelope method, a new evaluation procedure involving damage was proposed.
9
17. Global sensitivity analysis of load distribution and displacement in multi-bolt composite
joints
Review: This paper presents the development of a novel methodology for global sensitivity analysis of
multi-bolt composite joints. Two sources of uncertainty were considered, i.e., the mechanical properties
of composite plies and the bolt-hole clearance which comes from assembly phase. The uncertainty
propagation in the composite plate was derived through the classical laminate theory, while the load
distribution within bolts and the joint displacement were determined by an analytical spring-based
method. A kind of Monte Carlo simulation was used to calculate the variance-based sensitivity indices.
Furthermore, the proposed methodology was applied to a generic single-lap composite joint, which can
also be extended to represent double-lap joints, to identify the significant input random variables with
respect to structural responses of the joint. Useful conclusions for the design of multi-bolt composite
joints were obtained from the computational results and discussions.
18. An improved spring method for calculating the load distribution in multi-bolt composite
joints
Review: In this paper, an improved spring method considering the hole clearance and friction effects is
developed for efficient modelling of the composite bolted joint structures. This method is based on
spring method and finite element theory. A seven-bolt double-lap joint model is built as an example to
study the effects of plate width, plate thickness, bolt pitch and clearances on the load distribution. The
improved spring method is also adopted in the Monte Caro simulation with clearances considered as
uncertain-but-bounded parameters. The results show that the proposed method is time efficient for the
load distribution prediction of the bolted joint composite structures.
19. An analytical model for strength prediction in multi-bolt composite
joints at various loading rates
Review: This paper presents an enhanced analytical model to determine the complete load displacement
curve of single- and multi-fastener composite joints. It is an extension of a previous spring-based
method, which included the effects of bolt pre-load and clearance, but not local bearing damage. The
model has advanced beyond the state-of-the-art as loading rate effects, in addition to bearing damage,
are accounted for through a novel conic damage approximation function. Using the developed model,
an accurate prediction of the load displacement response of single- and multi-fastener joints to complete
failure can be obtained in a matter of seconds. The method is validated against experimental data and
excellent correlation was observed. Further studies carried out using the model suggest that slight
variations in the energy absorption characteristics at each fastener hole in a multi-fastener joint can
significantly alter the bolt-load distribution in the joint.
20. A three dimensional implicit finite element damage model and its application to single-lap
multi-bolt composite joints with variable clearance
Review: In this paper, a three-dimensional, implicit, finite element composites damage model is
developed, and applied to a problem involving highly complex, three-dimensional loading, i.e. a single-
lap, multi-bolt, composite joint, having variable clearances. The model extends a previous model,
adding Puck’s criteria to allow delamination-like, transverse cracking-like or mixed mode failures to be
simulated. Plasticity is also added for non-reversible strains in the undamaged parts of the matrix, and
a mesh independence strategy is incorporated. A back-up strategy for when the Newton–Raphson
method fails to converge, adds significantly to model robustness. The model is implemented in a
commercial, implicit solver and demonstrates excellent robustness, being capable of following damage
progression from onset, through bearing failure at the holes, to catastrophic, net-tension failure. The
predicted failure modes and loads are in good agreement with experiment. The effect of clearance on
secondary bending, and consequently on damage progression, is demonstrated. The current model
10
shows better capability than our previous model for predicting matrix damage near the hole, which will
be important for future fatigue modelling studies.
21. A study of intra-laminar damage in double-lap, multi-bolt, composite joints with variable
clearance using continuum damage mechanics
Review: This paper proposes a damage modelling approach, based on a continuum damage model
(CDM) formulation, which is applied to the problem of double-lap, multi-bolt, fibre-reinforced
composite joints with variable clearances, subjected to quasi-static tensile loading. A new method of
dealing with fibre failure is included in the CDM model, which is implemented in a commercial implicit
finite element analysis code. The model is validated at element and structural levels by comparing with
experimental data. It has been found that, for the joints examined in this paper, our formulation is
capable of modelling development of damage from bearing failure onset all the way to ultimate
catastrophic net-tension failure without numerical problems, which is an advance over previous work.
The predictions from the CDM model of net-tension failure modes and ultimate loads are in good
agreement with those from the experiments. Furthermore the model is capable of explaining some non-
intuitive experimental findings, such as the larger energy absorption obtained in joints with higher
clearances.
22. An analytical model for the prediction of load distribution in multi-bolt composite joints
including hole-location errors
Review: In this paper, an analytical model is developed to evaluate load distribution in an aluminium-
composite double-lap joint, in the presence of clearance and hole-location errors. The model also
includes the nonlinear behaviour of the bolt implied by bearing degradation. The analytical model is
validated by comparing results obtained by finite element analysis and experiments. Finally, an
industrial use of the analytical model is presented.
23. Numerical failure assessment of multi-bolt FRP composite joints with varying sizes and
preloads of bolts
Review: In this paper, a three dimensional (3D) progressive damage model (PDM) is developed for the
failure assessment of fibre reinforced plastic (FRP) composite bolted joints subjected to bolt preloads.
Material constitutive equations and material damage model are defined in a user-subroutine UMAT
which is integrated with the finite element (FE) software ABAQUS. Fibre and matrix failure in tension,
compression or in shear along with delamination failure criteria are incorporated in the present
numerical model. Load-displacement behavior, laminate surface strains, propagation of damage and
failure of a double-lap multi-bolt composite joint have been studied and present results are validated
with the results available in literature. Failure of double-lap multi-bolt FRP joints with different bolt
diameters and bolt tightening torques have been investigated using the proposed progressive damage
model.
24. A numerical failure analysis of multi-bolted joints in FRP laminates based on basalt fibers
Source: Elsevier ScienceDirect / Procedia Engineering 109 (2015) 492 – 506
Review: This paper aims to model the progressive damage of multi-bolted joints connecting structural
elements made up of FRP (fiber-reinforced polymers) composite laminates and comprising different
fiber materials (namely, based on basalt, carbon and glass), as well as different stacking sequences.
Differences in failure mode and ultimate-load values are numerically investigated. A numerical home-
made finite element model has been conceived, implemented, and validated by means of available
experimental data. The numerical model is based on an incremental displacement-based approach and
on a plane-stress bi-dimensional formulation. The stress analysis has been performed by accounting for
micro-structural stress-strain localization mechanisms, and describing the progressive damage process
by implementing a failure criterion operating at the constituents’ scale (namely, the Huang’s criterion).
11
Proposed results have highlighted that bolted joints based on basalt-FRP laminates and defined by a
double-bolted configuration exhibited bearing failure loads comparable to those computed for glass-
FRP and carbon-FRP laminates. In the case of single-bolted joints, the use of carbon-FRP laminates
allowed to obtain the best mechanical properties, although joints based on basalt-FRP laminates
numerically-experienced mechanical response and strength features always comparable with those of
glass-FRP.
25. A global bolted joint model for finite element analysis of load distributions in multi-bolt
composite joints
Source: Elsevier ScienceDirect / Composites: Part B 41 (2010) 317–325
Review: This paper presents the development and validation of a global bolted joint model (GBJM), a
highly efficient modelling strategy for bolted composite joints. Shell elements are used to model the
composite laminates and the bolt is represented by a combination of beam elements coupled to rigid
contact surfaces. The GBJM can capture effects such as bolt–hole clearance, bolt-torque, friction
between laminates, secondary and tertiary bending in the laminates as well as the load distribution in
multi-bolt joints. The GBJM is validated using both three-dimensional finite element models and
experiments on both single and multi-bolt joints. The GBJM was found to be robust, accurate and highly
efficient, with time savings of up to 97% realised over full three-dimensional finite element models.
12
4 Literature Review: Failure Prediction Method on
Composite Multi-Bolt Joints
4.1 Composites in Aviation
In the designing process and manufacturing of an aircraft, the main problems that engineers and
designers faced are how to make the aircraft as light and as efficient as possible with the best
performance and giving a better customer experience. Previously, most of them turn to aluminium, steel
and titanium as the main material for an aircraft. With the recent development and findings in the world
of material science, composite materials have shown a new hope for the future of aviation industry. If
the composite materials are designed well, we will be able to get the best qualities out of the constituents
inside the composite materials and often some new qualities that neither constituent possesses. This
include high stiffness-to-weight ratios, high strength-to-weight ratios, better fatigue performance, non-
corrosive characteristics, and less susceptibility to crack growth.
With the advantages of composites as shown above, nowadays most of engineers and designers are able
to make their vehicles lighter, more fuel-efficient and more comfortable for passengers. Common
composite materials used on aircrafts include fiberglass, carbon-fiber, and fiber-reinforced matrix
systems or any combination of any of these. Increasingly, the use of this materials had been
implemented on aircrafts parts from small, not crucial parts to a large, crucial aircraft parts. We can see
some examples in the recent aircrafts that Boeing and Airbus has manufactured, which are the Boeing
787 and Airbus A350 which nearly over 50% of their parts are made of composite materials. With the
usage of this composite materials, the aircraft produced will have a more efficient fuel usage, reduced
weight, and a better customer experience compared to previously metal aircrafts. Besides that, with a
property of non-corrosive characteristics, composite materials don't corrode as easy as other types of
structures made from metals or other materials. Structures made from composite materials also don't
crack from metal fatigue and they hold up well in structural flexing environments. Composite designs
also last longer than aluminium, which means fewer maintenance and repair costs to take care the
aircraft.
As the introduction of composite materials in aviation industries brings major changes and development
that turns out to help solve problems faced by the engineers and designers before, on the downside,
composites have their own distinct problems. The damage modes experienced by composite materials
are far different than those of metals. Common damage modes for composite materials are fiber
breakage, fiber pulling out of matrix, matrix cracks and delaminations, in which they can occur
internally to the structure with no visible damage on the surface of the material. This poses a problem
for doing structural inspection, a solution which has been implemented by manufacturers to solve for
this problem is by embedding sensors into the structure in order to detect internal damage. Besides that,
the manufacturing process needed to make composites are much different, with a necessity of having
different and more advance equipment and tooling.
A final issue for composite materials revolves around the joining of composite components to metal
structures or other composite structures. In which, as we know that the composite is stiffer than the
metal, so from an engineering standpoint, the composite will carry most of the loads. In order to
compensate for this additional stress, manufacturers must build up the joint with more material, a
process that adds more weight to the aircraft. Moreover, as the joint is experiencing load, the metal will
be able to expand and contract much more than the composites do. If this happen, a joint failure might
happen in the reason of the imbalance that is happening in the joint. One simple solution is to use more
composites, unfortunately joining composite to composite may eliminate this issue but also brings up
another. In an example, the joining of composites is often done with an adhesive layer that under certain
types of load is prone to delamination. Other method to join composites is by using fasteners or bolted
joints. Although this might solve the delamination problem, it will also present other difficulties due to
13
stress concentrations from drilled holes and thermal expansion mismatch between the composites and
the fasteners. So, for ideal situation, most designers and engineers are looking to create more integral
composite structures that do not require joining.
As there are so many technical variables which can impact the use of composites, the use of computer
simulation to design, test and analyse the composite structures are needed before constructing the actual
aircraft to make sure the strength and safety of the structure. However, as composite is a material which
are made with different layers of fibers that have a directionality that can influence the response of the
structure, the simulation becomes more complex and consumes more time to analyse. Fortunately, as
technologies develop, more advance simulation software has been enabling engineers to account for
these technical variables in a more reliable and systematic fashion. Several simulation software for
composite material design, analysis and simulation that are reliable and used around the world are
ABAQUS, ANSYS, CATIA, MSC Patran & Nastran, Solidworks, etc.
In conclusion, we can say that composites are likely to continue being developed into assuming an ever
more crucial position in the aviation industry. As composites started to be used in aircraft for small parts
that are not crucial and now has went on to play a larger role, the industry is able to expect a lift that
might takes it and its customers to new levels of safety, comfort and efficiency on an aircraft.
4.2 Development of Failure Prediction Method for Composite Bolt Joints

As it has been stated in the previous section, most of structure designers in the aviation industry has
implemented the use of composite materials in their design and most of them nowadays preferred to
use composite bolted joints as the connections in composite structures. The bolted joints that can be use
in the joint between two composite materials can be divided into single-bolt and multi-bolt joints and
with single or double lap (more preferred) of the composite materials that are going to be joined together.
Composite bolted joint has been preferred by most of the engineers in the world because of the ease of
dismantling, convenient detection and high reliability in which having high load-carrying and load-
transferring capacity. Unfortunately, as this joint will act as the core component that will do the
transferring of high loads, the composite bolted joint became a weak spot for the composite structure
which affect the problem of making it to be having a good structural integrity and safety. In order to get
to that goal of solving this problem, further analysis of the composite bolted joint’s strength is required.
The method of failure prediction for composite bolted joints generally may include two steps, in which
the first one is to do an overall joint analysis to be able to determine the load distribution between the
bolts. The second one is to do a detailed local strength analysis around a critical hole in order to ensure
a safety margin that will cover for all failure modes. As composites materials have a characteristic of
anisotropy and brittleness, composite bolted joints will experience a minimal amount of plastic stress
relief and mechanical softening as compared to their metallic counterparts. This will lead into an uneven
load distribution that happen between bolts.
Each bolt hole in the joints will be subjected to differently combined bearing bypass loads. Generally,
due to the final catastrophic failure of the joints is determined by the combined bearing bypass stresses
at the critical hole, it is really important for conducting an accurate load transfer analysis in multi-bolt
composite joints in order to be able to determine the combined stress status of the critical hole.
There are several factors that must be considered in doing the load transfer analysis of multi-bolt joints,
which includes joint materials, laminate lay-ups, laminate thickness, hole diameter, fastener position
(with ratios of width-to-hole diameter and edge-to-hole diameter), type of fastener (countersunk or
protruding head), bolt-hole clearance and the torque for bolt-tightening.
With all of factors that needs to be considered as stated in the paragraph before, there are several famous
failure prediction or analysis method that is used to predict the failure behaviour of the multi-bolted
14
composite structure joints, which are analytical approaches, simple spring-based methods, FE methods,
Progressive damage methods, Characteristic length method and many more.
As composite multi-bolted joints with double lap are more preferred by the engineers in the world, in
this paper we will mostly present the development of the failure prediction method for the composite
multi-bolt joints with some several parts on single bolt joints also.
4.2.1 Multi-Bolt Composite Joints

For multi-bolted composite joints, several different methods have been used to do failure
prediction of the joints.
Firstly, we will discuss a method that presents an analytical joint stiffness model that has been
done by several researchers around the world. Sharos et al. (2014) [19] has presented an analytical
model that has been enhanced as an extension of a previous spring-based method that are used to
determine the complete load displacement curve of a single- and multi-bolted composited joints.
The method presented proved could be a very useful approximation of the damage response given
by composite bolted joints as a preliminary design tool. They were able to prove it by assuming
a non-linear monotonically increasing stage followed by a quasi-linear softening stage, that are
both governed by the energy absorption observed during failure of an equivalent single-bolt joint,
which were then applied to a more complicated problem of a single column, three-bolt joints. In
result, they were able to get good approximations of the load displacement curved for all
considered loading rates with a reasonable degree of accuracy in a matter of seconds. This result
has been validated against experimental data and excellent correlation was obtained. Lecomte et
al. (2014) [22] were able to present an analytical model which are useful to represent load transfer
in multi-bolt double-lap joint. The model that was presented can simultaneously take into account
the loss of stiffness due to bearing damage, the location of adherent holes and bolt-hole clearance.
The performance of the analytical model was then validated and confirmed by a numerical model
and experimental results. This proves that it enables us to add value to bolted joint design in an
industrial context. Thus, the errors due to drilling and the mechanical performance of the joint
become possible to rely on directly. Then, by evaluation different flow-process grids and their
induced location tolerances we are able to make a global choice, integrating cost and
dimensioning criteria. Liu et al. (2015) [1] has presented a novel tri-linear joint stiffness model for
highly torqued multi-bolt composite joints with clearances. To conveniently solve the nonlinear
systems of equations of the load transfer analysis, in which the recursive formulae they were
using verified to be monotonic, they proposed a single-parameter iteration method. The joint
stiffness model that they developed has been compared and verified to the experimental load
distribution measurements in three single-lap, three-bolt joints. The result that they obtained by
using the numerical predictions on the presented joint stiffness model and the single-parameter
iteration method have shown significantly small errors of no more than 6% which demonstrates
the accuracy of the presented joint stiffness model and the feasibility of the proposed single-
parameter iteration method. Additionally, they also presented the effects of bolt-hole clearance
and bolt-tightening torque on the joint stiffness and the load transfer history of the bolt that are
determined based on the experimental measurements and numerical investigation. They found
that the initial stiffness remains constant in the joint stiffness model, as it is not affected by the
bolt-hole clearance or bolt-tightening torques. With the transition load is directly proportional to
the bolt tightening torque and the bearing chord stiffness is linearly dependent on the clearance
and independent on the bolt-tightening torque, it can then be said for highly torqued multi-bolt
joints, the friction forces between the laminates as well as the bolt and laminates reacted to the
joint load first.
Some other research that had been conducted recently has focused on how to determine and
predict the load distribution on the bolts from the composite bolted joints. According to Xiang et
15
al. (2016) [18], “Fast and reliable prediction of the load distribution is necessary in the initial design
of the composite bolted joint structures. Traditional spring method is time consuming in the
modelling of a composite
bolted joint structure”. In answer to this problem, they have presented an improved spring method
that considers the hole clearance and friction effects that is developed for efficient modelling of
composite bolted joint structures. The method they presented is based on spring method and finite
element theory. They were able to obtained both the modelling efficiency and the calculating
efficiency, which then proved that this improved spring method is more suitable to the initial
design stage in comparison to the traditional spring methods when the bolted joint model is
modified frequently. They found that for double-lap joints there is an optimum skin plate
thickness that can be used for load distribution. The result is that as bolt pitch increases and plate
width decreases, an uneven load distribution will be resulted. This improved spring method then
also was adopted in the Monte Caro simulation with clearances that are considered as uncertain-
but-bounded parameters. In conclusion, from the results they got, we may see that the proposed
method is time efficient for load distribution prediction of bolted joint composite structures.
Regarding method to present the measurement of bolt load distribution, Liu et al. (2018) [3]
presented a test method that is able to measure the bolt load distribution in the progressive damage
process of multi-bolt composite joints. This method exhibits continuous changes with the damage
propagation in composite laminates. In their research they tested a three-bolt joint as an example.
They found that both the numerical and experimental results were in good agreement that both of
them show the load distribution in the three-bolt joint varies throughout the loading process and
eventually tends toward uniformity before the final failure of the joint. Then, by using a modified
failure envelope method, they were able to predict the failures of two-, three- and four-bolt
composite joints with a base of prediction on the uniform and actual load distribution. The
prediction obtained then is compared with the traditional load distribution that was calculated by
traditional stiffness method (TSM). It was then found that the uniform load distribution can reflect
the effect of the actual load distribution on the failure of the multi-bolt joints more faithfully than
the traditional load distributions. Another bolt load distribution analysis method also has been
presented by Liu et al. (2018) [5], which is a stochastic or probabilistic bolt load distribution
analysis method for single-lap three-bolt joints that was developed based on the spring-based
method and Monte Carlo simulation. In this proposed method, they were able to use an improved
three-stage spring-based method to analyse the bolt load distribution efficiently. For random input
variables of the method, basic design parameters such as, material properties, stacking angles,
geometric parameters, bolt-hole clearances and tightening torques were involved. Numerous
composite single-lap three-bolt joints were designed and tested by using the modified
instrumented bolts to achieve the stochastic bolt load distribution of the joint in order to validate
the proposed method. The result shows that good agreements were able to be achieved between
the numerical and experimental stochastic bolt load distribution which then validated the
proposed method. Then, with the proposed method they investigated the probabilistic bolt load
distributions of the three-bolt joints in four tightening torque conditions, including finger-tight,
wrench torque, standard tightening torque and relaxation tightening torque. They found that a
high level of tightening torque reduces the variation of the bolt load distribution, while the
relaxation of the tightening torque increases the variation of the bolt load distribution. In
conclusion, after doing a further sensitivity analysis, they concluded that the bolt-hole clearances
have significant influences on the variation of the bolt load distribution, while for other random
parameters the effects on the variation of the bolt load distribution are slight. Besides that, Liu et
al. (2018) [2] also present detailed interpretation method of load distribution in composite single-
lap bolted joints which is based on theoretical, experimental and numerical investigations. In
order to elucidate the load distributions in highly torqued single-lap composite bolted joints with
bolt-hole clearances, they had performed free-body diagram analysis, 3D FE analysis and static
16
load distributions experiments for the cases of both single- and three-bolt joints. By using a
spring-based method, they were able to calculate the load distributions in three-bolt joints and
compare the results to the 3D FE method results. In their experiments they were able to extract
the bolt shear load of each bolt and validate their suitability as substitutes for the real fasteners,
by utilizing a modified instrumented bolts (MIBs). They were able to also found that the
corresponding experimental results were in good agreement with the numerical curves of the
loads vs. the hole deformation for a single-bolt and of the bolt shear loads vs. the external load
for three-bolt joints, which then validates the effectiveness of the 3D FE models and the
theoretical analysis. They were then able to determine with greater fidelity the load transferred
by each bolt in a multi-bolt joint. For the load distributions of the multi-bolt joint, were compared
between a 3D FE model and a spring-based method. As a result, it was obtained that the load
distributions from the established 3D FE model are identical to those from the spring-based
method; the bolt-hole clearances significantly impacted the load distributions, while the effects
of the bolt tightening torque on the load distributions can be ignored.
The next method to be discussed is the progressive damage or failure analysis for multi-bolt
composite joints. Zhang et al. (2014) [7] proposed a method using numerical progressive damage
method in order to obtain characteristics lengths by basing on the characteristic lengths method
to predict failure of the multi-bolt joints. By using the proposed method, failure prediction of
multi-bolt composite joint can not only be carried out in a small-scale computation but also, we
are able to avoid time-consuming and expensive tests. Other than that, this method also makes
the failure analysis conducted become simple and more efficient as we will only need
unidirectional layer properties. Then with the accompaniment of simple and convenient
characteristic point method (CPM), the characteristic lengths that are obtained from progressive
damage models (PDMs) and conventional characteristic length tests, they were able to predict the
failure of multi-bolt joints. After doing tests and comparisons with the experimental outcomes,
they were able to then recommend two progressive damage models which are composed by in
situ strength as Hashin or modified Yamada-Sun Criterion and Camanho’s degradation rules. In
result of their research, after comparison with the experiment outcomes of multi-bolt joints, they
were able to obtained overconservative predictions with the PDA-based CPM. Although in there
are considerable part of the total error that comes from the CPM itself, the PDA-based
characteristic length calculations give a small error with no more than 4%, which means that the
characteristic length tests replaced by PDA is a feasible option for accurate and reliable failure
predictions for composite multi-bolt joints. Aside from a new variation of PDM based on
characteristic length method, Zhao et al. (2018) [14] has proposed an analysis of the failure
behaviours of plain-weave carbon fibre-reinforced silicon carbide (C/SiC) composite joints with
C/SiC bolts by using a progressive damage model. This analysis is proposed in order for them to
explore the potential of using C/SiC composites in the thermal protection systems of hypersonic
aircrafts. This analysis was conducted with a basis on mechanical property tests as they developed
a bilinear constitutive model that will characterize the non-linearity of the C/SiC material. They
employed the Hashin-type criterion and the combined degradation model in order to establish a
model of progressive failure for the C/SiC composite structures with a basis on a series of
progressive damage analysis of a C/SiC composite open-hole laminate and comparisons with
experimental results. In results, they achieved to get a close-agreements between the numerical
predictions and experimental outcomes that validated the effectiveness of this proposed model to
predicting the failure of all-C/SiC composite bolted joints.
For the next method in failure prediction, a failure envelope method will be discussed. Liu et al.
(2013) [13] presented a modified failure envelope method that consists of two oblique lines
corresponding to tensile envelope and bearing envelope respectively, in order to predict final
failure mode and strength of multi-bolt composite joints. This method was able to be established
by using an additional compressive strength test of laminates and semi-analytical or numerical
17
analysis of fastener-hole laminates and open-hole laminates, which are combined with the point
stress criterion. They were able to also illustrated a flowchart of integration structure analysis
process with the modified failure envelope method. They also did a comparison between the
experimental results and the failure prediction data of two-bolted and four-bolted carbon-to-
carbon double-lap joints by using both the modified and conventional envelope method. The
results of this comparisons indicate that the method proposed is able to be used to make effective
predictions of failure modes and more accurate ultimate failure loads for the double-lap multi-
bolted composite joints. Another method that also utilizing the failure envelope method is the one
that is been brought by Cheng et al. (2017) [16] who had stated that failure envelope method is a
method that is “based on the assumption that bolt load distribution proportion is constant till
failure. However, bearing damage and net-section damage must have some influence on the load
distribution and then change structure failure mode.” In order to overcome the downside of this
method, they did a study of the effect of damage on failure mode of composite multi-bolt joint by
using the failure envelope method and conducted tensile experiments of two-bolt and three-bolt
joints to obtain failure modes and other properties of the joint. They also constructed three-
dimensional finite element models containing composite damage, calculated the bolt load
distribution, bearing stress and bypass stress, predicted the failure mode of the composite joints,
and analysed the effect of bearing damage and net-section damage. As the results of the tests were
consistent with the experimental data, they were able to conclude that the effect of composite
damage should be involved within the failure mode prediction that utilize the failure envelope
method. Another addition to their findings is that the bearing damage they calculated in the tests
has a greater influence on the failure mode compared to the net-section damage.
The next section will discuss another method of failure analysis which is the numerical failure
analysis method. Nerilli et al. (2015) [24] have modelled the progressive damage method of multi-
bolted joints connecting structural elements that were made up of FRP (fibre-reinforced polymers)
composite laminates which comprised of different fibre materials and different stacking
sequences. They have numerically investigated difference failure mode with varying ultimate-
load values. This numerical homemade finite element model which has been conceived and
validated with the available experimental data in a good agreement, is made based on an
incremental displacement-based approach and an on a plane-stress bi-dimensional formulation.
The proposed numerical results that they got have shown that the use of bolted joints based on
basalt-FRP laminates lead to failure loads that are fully comparable to those numerically-
experienced in the case of glass-FRP and carbon-FRP laminates. Although the joints based on
basalt-FRP laminates numerically-experienced a good mechanical response and strength features
that are always comparable with those of glass-FRP, but in the case of single-bolted joints, the
use of carbon-FRP laminates show the best mechanical properties. In results with this promising
numerical evidence, together with the good physico-chemical qualities and low production cost,
encourages scientist and engineers to do further studies on basalt fibres as reinforcement
constituents to be employed in effective connections among composite structural elements.
Another proposed research that has been done on numerical failure analysis method is the one
that was done recently by Mandal and Chakrabakti (2018) [23] which developed a three-
dimensional progressive damage model (PDM) for the failure assessment of fibre-reinforced
plastic (FRP) composite bolted joints subjected to bolt preloads. By using a finite element (FE)
software, ABAQUS, they have integrated material constitutive equations and material damage
model which are defined in a user-subroutine UMAT. Several different and new interesting
results were generated by using the developed model in consideration of different torque levels
and variation of the diameter of bolts. They also have studied and presented results that are
validated with good agreement to the available literature on load-displacement behaviour,
laminate surface strains, propagation of damage and failure of a double-lap multi-bolt composite
joint. Beside that, they also have investigated failure behaviour of double-lap multi-bolt FRP
18
joints with different sizes of bolt diameters and different values of bolt-torques. After several
different conditions of the test, Mandal and Chakrabakti (2018) [23] concluded that “At higher
torque level the trend is found to be merging to the case when torque is varied equally for all three
bolts.”
Several other methods that are also worth to be mentioned as a development for failure prediction
method of composite multi-bolt joints will be discussed in this particular section. Gray and
McCarthy (2010) [25] has presented the development and validation of a global bolted joint model
(GBJM) which is a highly efficient modelling strategy that can be used for finite element
simulation of load distributions in multi-bolt composite joints. In this model, they used shell
elements as the composite laminates and a combination of beam elements coupled to rigid contact
surfaces for representing the bolt. This GBJM is able to capture the physical parameters of a
bolted joint with a high degree of accuracy. These physical parameters that are captured may be
describes such as bolt–hole clearance, bolt-torque, friction between laminates, secondary and
tertiary bending in the laminates as well as the load distribution in multi-bolt joints. This GBJM
modelling strategy has been proved and validated by using both three-dimensional finite element
models and experiments on both single and multi-bolt joints. The final result shown that GBJM
is both accurate and robust, and a highly efficient tool for predicting the load distributions in
multi-bolt joints with time savings that may reach up to 97% realised over full three-dimensional
finite element models.
Another failure prediction method is proposed by Zhang et al. (2014) [8] is by presenting a novel
characteristic curve which is determined by tensile, compressive and shear-out characteristic
length in order to predict failure load and failure mode of mechanically fastened multi-bolt
composite joints. This proposed curve introduces a shear-out characteristic length and able to
predict shear-out failure in contrast to existing curve. They proposed a mathematic formulation
of the novel curve by superimposing the expression of Chang’s characteristic curve and with an
additional term that is related with the shear-out length. They also have carried out a series of
tests and conducted nonlinear finite element analyses in order to obtain the tensile, compressive
and shear-out characteristic lengths for the novel curve. In the end result, the numerical
predictions and experimental outcomes show good agreements between both of them which can
be concluded that the characteristic curve method with the novel characteristic curve is able to do
a correct prediction for failure modes and an accurate failure loads for multi-bolt composite joints.
Next in line, we also have Zhou et al. (2014) [21] who propose a damage modelling approach by
taking a basis on the continuum damage model (CDM) formulation, which is then applied to the
problem of double-lap, multi-bolt, fibre-reinforced composite joints with variable clearances and
subjected to quasi-static tensile loading. They also have developed a detailed three-dimensional
FE model of double-lap, multi-bolt composite joints and introduced a new method to deal with
fibre failure that is included in the CDM model and implemented in a commercial implicit finite
element analysis code. By comparing the model to the existing experimental data, they were able
to validate the model at an element and structural levels. For the joints they have tested, the
formulation that they proposed has shown an advancement of previous work, for it is capable of
modelling the development of damage from bearing failure onset all the way to ultimate
catastrophic net-tension failure without numerical problems in a stable and sensible results.
Additionally, the predictions from the CDM model of net-tension failure modes and ultimate
loads have good agreement with those from the experiments. In results, they suggest that this
method could be used to predict the effects of many other joint parameters and showed that it is
a useful design tool which can also give a significant additional insight into the detailed
progression of damage.
19
For the next method, we have Zhou et al. (2015) [20] who have developed a 3D implicit finite
element composite damage model that has been validated and tested on complex problem of
three-dimensional loading, i.e. a single-lap, multi-bolt composite joint that has variable clearances.
The model that they proposed is a development which extends a previous model by adding Puck’s
criteria to give more rational approach to final cut-off of the material stiffnesses, allowing for
different delamination-like, transverse cracking-like or mixed mode failures to be simulated. For
non-reversible strains in the undamaged parts of the matrix, this model has incorporated plasticity.
This model also has a mesh independence strategy that has been incorporated to it. They found
that this model has showed an exceptional level of robustness, in which it is able to follow the
damage development in the joints all the way to the final net-tension failure. Besides that, they
also found that the predicted failure modes and loads are in good agreement with the experiment,
in which it is able to predict many subtle effects that happen in single-lap, multi-bolt joints, such
as fibre damage through thickness, effect of secondary bending and re-distribution of load when
bearing failure occurred at a particular hole. In conclusion Zhou et al. (2015) [20] stated that “The
current model shows better capability than our previous model for predicting matrix damage near
the hole, which will be important for future fatigue modelling studies.”
The final failure prediction method for multi-bolt composite joint that will be discussed here is
presented by Li et al. (2017) [17] who developed a methodology for global sensitivity analysis of
multi-bolt composite joints. In this study, they have considered two sources of uncertainty which
were the mechanical properties of composite plies and the assembly uncertainty. By using the
classical laminate theory, they were able to derived the uncertainty propagation in the composite
plate. For assembly uncertainty, it is represented by the uncertainty in bolt-hole clearance which
is determined by a spring-based method. For the simulation that was used to calculate the
variance-based sensitivity indices, they used a simulation similar to Monte Carlo simulation. In
their study, this proposed methodology was applied to a generic single-lap composite joints that
can be extended to represent double-lap joints, with a goal to identify the significant input random
variables with respect to the structural responses of the joint. So then, based on the computational
results and discussions they have done, they found several useful conclusions in which two of
them are able to be applied to other cases. The first conclusion they found is that the proposed
method is useful and very efficient (only requires one or two minutes to generate a rational
estimate) for preliminary design tool from uncertainty propagation and quantification point of
view. The second one they found is that the uncertainties in the bolt-hole clearances were
identified to be the dominant input random variables compared to the uncertainties of the
mechanical properties.
4.2.2 Single Bolt Composite Joints

This section will specifically discuss the current trends or development for the single bolt
composite joints.
For the first method, here we have Zhao et al. (2016) [9] who have carried out a combination of
experimental and numerical methods to investigate the secondary bending effects on the
mechanical behaviour of single-bolt, single-lap joint in comparison to a double-lap joint. They
implemented the progressive damage analysis of both joints type to investigate the differences in
load-bearing mechanism and the effects of secondary bending during the progress toward total
failure. The effectiveness of the proposed progressive damage models was proved as good
agreements was achieved between the experimental and numerical results. Based on the result,
they were able to get that the single-bolt, single-lap joint had a nonlinearity onset strength which
was lower by half than that of the double-lap joint. They also presented experimental macro-
scope failure patterns and seven numerical micro-scope failure modes which obtained from the
progressive damage analysis for both type of the joints. As a result, they were able to show that
the secondary bending that was analysed on the mechanical degradation and failure mechanism
20
of single-lap joints has shown effects more significant than the double-lap joint. During the total
progressive damage process, they were able to also quantitatively trace the characterized
parameter on the AGARD points, joint deformations and contact states, which in the end may
increased the understanding of the effects of secondary bending on the failure process of a single-
bolt, single-lap joint.
As a next development in failure prediction method for single-bolt composite joint, Zhao et al.
(2017) [10] have presented a probabilistic model accounting for the randomness of geometrical
dimensions, fibre orientation angles and material properties in order to predict the probabilistic
failure load of a composite double-lap single-bolt joint. They have established a parametric failure
prediction that employs a 3D parametric FEM and a modified characteristic curve method for
predicting the failure load of the joint. Besides that, they were able to also carry out static tensile
tests of fifteen composite double-lap single-bolt joints that were made of T800 carbon/epoxy
composites. After comparing the probabilistic failure load obtained from the proposed model with
the result from the experiments, good consistency and agreement were achieved, which validates
the effectiveness and accuracy of the proposed model and shows that the model is not sensitive
to the probability distribution type of random variables. In contrast to the insensitivity of the
proposed model, they were able to get that the probability distribution type of random variables
slightly influences the statistical parameters of the probabilistic failure load after several
comparisons between the fitted distribution parameters have been conducted. Then, with the
determination of the relationship between the failure probability and the tensile load, they
concluded that longitudinal compressive strength Xc, thickness of the ply and the longitudinal
elastic modulus of E11 of the T800 unidirectional lamina are the top three factors that significantly
affect the probabilistic failure load and reliability of the joint remarkably.
For the next method, we have Choi et al. (2018) [15] who have conducted a failure load prediction
method for composite joints with clamping force by using a characteristic length method that is
combined with Tsai-Wu failure criteria. They were able to carried out tensile and bearing tests
with stress analysis in order to determine tensile and compressive characteristic lengths. They
also used different characteristic lengths and characteristic curves for different geometrical bolt
shapes to perform failure load prediction for single lap-shear bolted joints. After comparison, this
prediction of the failure load was found to be in good agreement with the actual failure obtained
from experiments and it was found that clearly the failure loads of mechanically fastened joints
subjected to clamping forces could be predicted to within 10% via the suggested failure prediction
method.
As the methods of failure prediction for single-bolted composite joint are not as much different
with the methods for the multi-bolt composite joint, the next method will conclude the
development of the failure prediction method for single-bolted composite joint up to recent
development. In which, Shan et al. (2018) [4] have proposed a bi-material property method to
describe mechanical properties of the unidirectional composite lamina. This method divided the
lamina into the fibre layer and the resin layer, which have different material properties and are
responsible for the intra-laminar and inter-laminar properties, respectively. In this study, they
presented a finite element modelling method which simulated the composite lamina by the bi-
material property method. A novel progressive damage model (PDM) was also constructed for
typical double-lap bolted joints with a combination with a Hashin-type failure criterion and a
micromechanics-based material degradation model. Four double-lap, single-bolt joints with
various width-to-hole-diameter ratios were designed and quasi-static tensile tests of the joints
were performed in order to validate the novel PDM. The final numerical results show a better
agreement happened between the novel PDM with the experimental outcomes than the traditional
PDMs compared to the experimental outcomes. Besides that, it was found that the novel PDM
proposed can describe the inter-laminar damage more precisely via a more reasonable simulation
for the inter-laminar property of the lamina, and it can also obtain a more accurate deformation
21
of the hole along the thickness direction of the laminate. Moreover, they also revealed influences
of the width-to-hole-diameter ratio W/D on failure behaviours of the double-lap, single-bolt joint
and recommended a good ratio to be used for engineering practice which is an optimal width-to-
hole-diameter ratio. Then as it is combined with their previous study on the effects of the end-
distance-to-hole-diameter ratio E/D, they were able to determine a failure mode map with varying
W/D and E/D for the double lap bolted joint. In the end, they suggested that designers should
seek safer designs of composite bolted joints as W/D and E/D are key geometrical parameters
that affect failure behaviours of the joints significantly.
4.2.3 Several Other Methods

This section will mainly present several other methods that has correlation for the failure
prediction of composite bolted joints that has been developed just recently. Firstly, we have Zhou
et al. (2016) [12] who have proposed a novel design of out-of-plane composite π joints that are
able to significantly improve the load carrying capability. By taking the advantages of the in-
plane properties of composites this method successfully eliminates the delamination damage
which commonly occurs in traditional π joints, with the uses of bolts to connect the joint and
other structural components. For this design, they also designed, fabricated and tested a new
bolted composite π joint which is used in static experiments of the novel bolted composite π joint
and an aluminium counterpart. The experimental results show that the bolted composite π joints
exhibited a larger load carrying capacity with little deformation and a massive weight reduction
if compared to its aluminium counterpart. Besides that, in comparison with the traditional design
of composite π joint, the new joint proposed showed a greater load carrying capability while
maintaining a similar structural weight. They also performed a progressive damage analysis of
the bolted composite π joint which is validated by experimental data, in order to deeply
understand the failure process and failure mechanism of the novel bolted composite π joint. In
result, they were able to conclude that inside the novel bolted composite π joint proposed
catastrophic failure such as delamination and cracking propagation was inhibited and changed to
non-catastrophic bearing failure around the bolt holes, which shows that this is a beneficial
development that maybe able for the safety of integrated composite structures.
For the next method, we have Shan et al. (2018) [11] who proposed a novel progressive fatigue
damage model that involes hygrothermal effects in order to predict fatigue failure of composite
structures in hygrothermal environments. In their proposed novel PFDM, to evaluate the
degradation of the material properties that are caused by hygrothermal environments, they
established a unified model. They also utilized the hygrothermal-induced material properties that
were obtained from the unified model in the stress analysis model, material degradation models
and fatigue failure criterion. By improving a residual-strain-based gradual material degradation
model and a micromechanics-based sudden material degradation model, they were able to
describe the damage of the composite materials in hygrothermal environments. They also
performed fatigue tests and progressive fatigue damage analysis of an open-hole laminate in
hygrothermal environments. The result validated the improved PFDM by showing a good
consistency between the numerical and experimental results. In which, both the experimental and
numerical outcomes give an indication that the effective moisture equilibrium considerably
reduces the compressive strength and compressive fatigue life of the open-hole laminate made of
T800 carbon/epoxy composites.
For the concluding method in this section we have, Zhou et al. (2018) [6] who proposed a 2D
micromechanical model to capture the longitudinal compressive failure of unidirectional carbon
fibre reinforced polymers (UD-CFRP), in which they considered the microscopic damage
mechanisms (such as fibre fracture, matrix plastic yield and fibre-matrix interface crack). The
proposed model performs well on predicting the compressive modulus and strength as well as the
kink band phenomenon and as it is compared with the experimental results, it shows good
agreements which means a validation of effectiveness of the proposed model. From this model
22
can be concluded that the longitudinal compressive failure is controlled by the matrix damage
rather than the fibre fracture and the interface crack triggers the matrix damage. Additionally, on
the longitudinal compressive failure, the influence of some microscopic parameters (such as fibre
volume fraction, initial fibre misalignment, initial fibre-matrix interface stiffness, interface
strength and fracture energies) is disclosed. Besides that, they also summarize the influence of
key parameters on the compressive failure that “The compressive modulus increases with the
fibre volume fraction while it is almost unaffected by the rest four parameters. The compressive
strength increases with the increment of the initial interface stiffness or the decrement of the initial
fibre misalignment while it is immune to the interface fracture energies”. Furthermore, as the
strength increases with the fibre volume fraction and the interface strength, and then decrease
after certain limitation, to it is considered reasonable to improve the longitudinal compressive
performance by conducting an enhancement on the fibre-matrix interface and avoiding the initial
imperfection through an improved manufacturing strength.
In conclusion, we can get that there are several good methods in predicting failure modes of composite
bolt joints such as progressive damage method, analytical joint stiffness method, probabilistic load
distribution analysis, finite element method, characteristic length method, envelop method and many
more. Each of them has their own uniqueness and advantages that mostly has been improved from
previous method. With this level of development, there are still further studies and researches that can
be conducted in the future to accurately and efficiently predict the failure for multi-bolt composite joints.
Some of the method that had me taken interest in are the progressive damage failure method and also
research into the load distribution analysis on composite bolts. With this literature review research,
recent development on failure prediction for composite multi-bolt joint has been discussed, it leaves us
with still many new developments await to be found.
23
5 Reference List
[1]. Liu, F.R., Zhang, J.Y., Zhao, L.B., Xin, A., Zhou, L.W., 2015. An analytical joint stiffness model
for load transfer analysis in highly torqued multi-bolt composite joints with clearances. Composite
Structures, 131, 625-636.
[2]. Liu, F.R., Lu, X.H., Zhao, L.B., Zhang, J.Y., Hu, N., Xu, J.F., 2018. An interpretation of the load
distributions in highly torqued single-lap composite bolted joints with bolt-hole clearances. Composites
Part B, 138, 194–205.
[3]. Liu, F.R., Lu, X.H., Zhao, L.B., Zhang, J.Y., Xu, J.F., Hu, N., 2018. Investigation of bolt load
redistribution and its effect on failure prediction in double-lap, multi-bolt composite joints. Composite
Structures, 202, 397–405.
[4]. Shan, M.J., Liu, F.R., Fang, Z., Zhao, L.B., Zhang, J.Y., 2018. A bi-material property-based FE modelling
method for progressive damage analyses of composite double-lap bolted joints. Result in Physics, 11,
647-683.
[5]. Liu, F.R., Shan, M.J., Zhao, L.B., Zhang, J.Y., 2018. Probabilistic bolt load distribution analysis of
composite single-lap multi-bolt joints considering random bolt-hole clearances and tightening
torques. Composite Structures, 194, 12-20.
[6]. Zhou, L.W., Zhao, L.B., Liu, F.R., Zhang, J.Y., 2018. A micromechanical model for longitudinal
compressive failure in unidirectional fiber reinforced composite. Result in Physics, 10, 841-848.
[7]. Zhang, J.Y., Liu, F.R., Zhao, L.B., Chen, Y.L., Fei, B.J., 2014. A progressive damage analysis based
characteristic length method for multi-bolt composite joints. Composite Structures, 108, 915–923.
[8]. Zhang, J.Y., Liu, F.R., Zhao, L.B., Fei, B.J., 2014. A novel characteristic curve for failure prediction of
multi-bolt composite joints. Composite Structures, 108, 129–136.
[9]. Zhao, L.B., Xin, A., Liu, F.R., Zhang, J.Y., Hu, N., 2016. Secondary bending effects in progressively
damaged single-lap, single-bolt composite joints. Results in Physics, 6, 704–711.
[10]. Zhao, L.B., Shan, M.J., Liu, F.R., Zhang, J.Y., 2017. A probabilistic model for strength analysis of
composite double-lap single-bolt joints. Composite Structures, 161, 419–427.
[11]. Shan, M.J., Zhao, L.B., Hong, H.M., Liu, F.R., Zhang, J.Y., 2018. A progressive fatigue damage model
for composite structures in hygrothermal environments. International Journal of Fatigue, 111, 299–307.
[12]. Zhou, L.W., Zhang, D.H., Zhao, L.B., Zhou, X.H., Zhang, J.Y., Liu, F.R., 2016. Design and analysis
of a novel bolted composite π joint under bending load. Materials and Design, 98, 201–208.
[13]. Liu, F.R., Zhao, L.B., Mehmood, S., Zhang, J.Y., Fei, B.J., 2013. A modified failure envelope method
for failure prediction of multi-bolt composite joints. Composites Science and Technology, 83, 54–63.
[14]. Zhao, L.B., Wen, Y.G., Cao, T.C., Li, H.B., Liu, B.R., Zhang, C., Zhang, J.Y., 2018. A progressive
failure analysis of all-C/SiC composite multi-bolt joints. Composite Structures, 202, 1059–1068.
[15]. Choi, J.-I., Hasheminia, S.M., Chun, H.-J., Park, J.-C., Chang, H.S., 2018. Failure load prediction of
composite bolted joint with clamping force. Composite Structures, 189, 247–255.
[16]. Cheng, X.Q., Wang, S.W., Zhang, J., Huang, W.J., Cheng, Y.J., Zhang, J.K., 2017. Effect of damage
on failure mode of multi-bolt composite joints using failure envelope method. Composite Structures, 160,
8–15.
24
[17]. Li, H.-S., Gu, R.-J., Zhao, X., 2017. Global sensitivity analysis of load distribution and displacement
in multi-bolt composite joints. Composites Part B, 116, 200-210.
[18]. Xiang, J.W., Zhao, S.W., Li, D.C., Wu, Y.N., 2017. An improved spring method for calculating the
load distribution in multi-bolt composite joints. Composites Part B, 117, 1-8.
[19]. Sharos, P.A., Egan, B., McCarthy, C.T., 2014. An analytical model for strength prediction in multi-
bolt composite joints at various loading rates. Composite Structures, 116, 300–310.
[20]. Zhou, Y.H., Nezhad, H.Y., Hou, C., Wan, X.P., McCarthy, C.T., McCarthy, M.A., 2015. A three
dimensional implicit finite element damage model and its application to single-lap multi-bolt composite
joints with variable clearance. Composite Structures, 131, 1060–1072.
[21]. Zhou, Y.H., Nezhad, H.Y., McCarthy, M.A., Wan, X.P., McCarthy, C., 2014. A study of intra-laminar
damage in double-lap, multi-bolt, composite joints with variable clearance using continuum damage
mechanics. Composite Structures, 116, 441–452.
[22]. Lecomte, J., Bois, C., Wargnier, H., Wahl, J.-C., 2014. An analytical model for the prediction of load
distribution in multi-bolt composite joints including hole-location errors. Composite Structures, 117, 354–
361.
[23]. Mandal, B., Chakrabarti, A., 2018. Numerical failure assessment of multi-bolt FRP composite joints
with varying sizes and preloads of bolts. Composite Structures, 187, 169–178.
[24]. Nerilli, F., Marino, M., Vairo, G., 2015. A numerical failure analysis of multi-bolted joints in FRP
laminates based on basalt fibers. Procedia Engineering, 109, 492 – 506.
[25]. Gray, P.J., McCarthy, C.T., 2010. A global bolted joint model for finite element analysis of load
distributions in multi-bolt composite joints. Composites: Part B, 41, 317–325.
25
6 Supervisor evaluation
Evaluation idea: “Failure prediction of composite bolted joints has attracted much attention in the
last decades. In view of the native English and lucid logic in the information retrieval and literature
review, it is an excellent Scientific Literature Retrieval Report.”
刘丰睿老师联系方式：
邮箱: frliu@buaa.edu.cn
电话：+86 134-2601-7408
Signatures（签名）
Student（学生签名）________________________ Date（日期） ___________________
Tutor（导师签名） _________________________ Date（日期）____________________
26

Scientific Literature Retrieval Final Report PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Scientific Literature Retrieval Final Report PDF

Uploaded by

Copyright:

Available Formats

Scientific Literature

Retrieval Final Report

Tutor: Liu Feng Rui 刘丰睿

1.2 Research topic

2.2 Step by step process of the literature retrieval

3.1 English sources

4.2 Development of Failure Prediction Method for Composite Bolt Joints

4.2.1 Multi-Bolt Composite Joints

4.2.2 Single Bolt Composite Joints

4.2.3 Several Other Methods

Student（学生签名）______ Date（日期） _

Tutor（导师签名） _____ Date（日期）

You might also like