K UNGLIGA T EKNISKA HÖGSKOLAN
M
ASTERT
HESISImpact Resistance of CFRP Products
Author:
Noel CHACKO
Supervisor:
Guillaume MOREAU
A thesis submitted in fulfilment of the requirements for the degree of Master of Science
in
Aerospace Engineering
Aeronautical and Vehicle Engineering Department
October 1, 2018
iii
KUNGLIGA TEKNISKA HÖGSKOLAN
Abstract
School of Engineering Sciences
Aeronautical and Vehicle Engineering Department
Master of Science
Impact Resistance of CFRP Products
by Noel CHACKO
This thesis investigated the impact performance of CFRP products within the sports industry. The primary aim of this thesis was to evaluate different configurations, matrix system, and technologies to find the best performing solutions for impact.
During this work, an extensive literature study was conducted and various solutions were reviewed. Further on, several tubes were manufactured, impacted and put through a 2 point bending test to find out the residual strength.
It was found that TeXtreme R fabrics positively affected the impact performance when compared to conventional fabrics and UD depending on the placement lo- cation. Thin plies proved to be better than conventional plies. Newer technologies such as CNT stitching requires further investigation before it can be qualitatively assessed.
Keywords:Impact performance, TeXtreme R, Composite, Thin plies, Sports
v
Acknowledgements
I would like to acknowledge Oxeon AB for giving me the opportunity and support to perform this thesis. To Guillaume Moreau, my supervisor at Oxeon, I am thank- ful for your continued support and guidance throughout this project. To Fredrik Ohlsson at Oxeon, I appreciate your input and discussions during various points in this project. To Anders Sjögren at AD Manus, thank you for your help which was vital to complete this project on time. To Antoine Battin, at SHD composites, thank you for taking the time out to answer all my questions. To Robin Olsson at Swerea, thank you for the very fruitful discussion. I would also like to thank my examiner at KTH, Stefan Hallström, I truly value your input and help. And lastly, I would like to thank the team at Oxeon for making my time there a memorable and educative experience.
vii
Contents
Abstract iii
Acknowledgements v
1 Introduction 1
1.1 Contextual background . . . 1
1.2 Scope of work . . . 1
1.3 Literature review . . . 2
1.3.1 Resin system . . . 2
1.3.2 Layer interface and stacking sequence . . . 3
1.3.3 Fibre . . . 4
1.3.4 Thin ply effect . . . 4
2 Materials 5 2.1 Comparison of conventional HS fabrics with TeXtreme R . . . 6
2.2 Effect of different types of fibres. . . 6
2.3 Positioning of TeXtreme R fabrics in the lay-up . . . 7
2.4 Effect of ply scattering . . . 7
2.5 New developments for impact . . . 8
2.6 Effect of ply thickness. . . 10
3 Methodology 11 3.1 Tube manufacturing . . . 11
3.2 Drop mass test. . . 17
3.3 2 point bending test . . . 18
4 Results and Discussion 21 4.1 Comparison of conventional HS fabrics with TeXtreme R . . . 21
4.2 Effect of different types of fibres. . . 23
4.3 Positioning of TeXtreme R fabrics in the lay-up . . . 25
4.4 Effect of ply scattering . . . 26
4.5 New developments for impact . . . 28
4.6 Effect of ply thickness. . . 31
4.7 Variation in stiffness . . . 32
5 Conclusion 35 5.1 Future work . . . 35
A Test results from AD Manus 37
B Specimen measurements 59
C Fibre volume fraction of specimens 61
viii
D Estimated absorbed energy 63
Bibliography 65
ix
List of Figures
2.1 Configurations with TeXtreme at different locations . . . 7
2.2 Thick and thin variations. . . 8
2.3 Configuration of tube with N12 . . . 9
2.4 Configuration of tube with Xantu.Layr. . . 9
2.5 Tubes F1 and F2 . . . 10
3.1 Tube geometry. . . 11
3.2 Mold with heating elements and thermocouple embedded . . . 11
3.3 The open mold . . . 12
3.4 Latex bladder . . . 12
3.5 Coupling for air inlet to bladder. . . 12
3.6 Tube with a fold on the inside . . . 13
3.7 Tube with a smooth inner surface . . . 13
3.8 Mandrel wrapped around with release film . . . 14
3.9 Formed stack on the mandrel . . . 14
3.10 Dry spots on the tube . . . 15
3.11 Wrinkles when forming the first layer . . . 15
3.12 Schema for forming pre-stacks . . . 15
3.13 Curing cycle . . . 16
3.14 Perforated stack of 45/-45 UD . . . 16
3.15 Partial vacuum applied on the mold . . . 16
3.16 Three specimens cut from a tube . . . 17
3.17 Drop mass set-up . . . 17
3.18 2 point bending test . . . 19
3.19 Tube fixture . . . 19
3.20 Cylindrical fixture . . . 20
4.1 Comparison of impact performance of tubes in category A . . . 21
4.2 Specimen A1.1 . . . 22
4.3 Percentage reduction in strength of tubes in category A . . . 22
4.4 Specimen B2.1 . . . 23
4.5 Comparison of impact performance of tubes in category B . . . 23
4.6 Percentage reduction in strength of tubes in category B . . . 24
4.7 Load-displacement curves for B1 . . . 25
4.8 Comparison of impact performance of tubes in category C . . . 25
4.9 Percentage reduction in strength of tubes in category C . . . 26
4.10 Comparison of impact performance of tubes in category D . . . 27
4.11 Comparison of impact performance of tubes in category D . . . 27
4.12 Percentage reduction in strength of tubes in category D . . . 28
4.13 Percentage reduction in strength of tubes in category D . . . 28
4.14 Load-displacement curve for impacted specimen in category D . . . . 29
4.15 Comparison of impact performance of tubes in category E . . . 29
x
4.16 Comparison of impact performance of tubes in category E . . . 30
4.17 Specimen E3.1 and E3.2. . . 30
4.18 Percentage reduction in strength of tubes in category E . . . 31
4.19 Percentage reduction in strength of tubes in category E . . . 31
4.20 Comparison of impact performance of tubes in category F . . . 32
4.21 Percentage reduction in strength of tubes in category F . . . 32
4.22 Load - displacement curves for tubes with different TeXtreme fabrics on the outside (Undamaged) . . . 33
4.23 Load - displacement curves for tubes with TeXtreme at different posi- tions in the lay-up (Undamaged) . . . 33
4.24 Load - displacement curves for tubes with new technology (Undam- aged) . . . 34
A.1 Test result for tube A1 . . . 37
A.2 Test result for tube A2 . . . 38
A.3 Test result for tube A3 . . . 39
A.4 Test result for tube B1 . . . 40
A.5 Test result for tube B2 . . . 41
A.6 Test result for tube B3 . . . 42
A.7 Test result for tube B4 . . . 43
A.8 Test result for tube C2. . . 44
A.9 Test result for tube C3. . . 45
A.10 Test result for tube D1 . . . 46
A.11 Test result for tube D2 . . . 47
A.12 Test result for tube D3 . . . 48
A.13 Test result for tube D4 . . . 49
A.14 Test result for tube D5 . . . 50
A.15 Test result for tube E1 . . . 51
A.16 Test result for tube E2 . . . 52
A.17 Test result for tube E3 . . . 53
A.18 Test result for tube F1 . . . 54
A.19 Test result for tube F2 . . . 55
A.20 Test result for tube R1. . . 56
A.21 Test result for tube R2. . . 57
xi
List of Tables
2.1 Material used in the test programme . . . 6
2.2 Tubes in category A . . . 6
2.3 Tubes in category B . . . 7
2.4 Ply scattering . . . 8
2.5 New developments for impact . . . 8
xiii
List of Abbreviations
CFRP Carbon Fibre Reinforced Plastics CNT Carbon Nano-Tube
FAW Fibre Areal Weight
GSM Grams (per) Square Metre HM High Modulus
HS High Strength
IM Intermediate Modulus RC Resin Content (% weight) UD Uni-Directional
VACNT Vertically Aligned Carbon Nano-Tube
xv
List of Symbols
h Height m
m Mass kg
g Acceleration due to gravity m s−2
v Velocity m s−1
Vf Fibre volume fraction
Wf Weight of fibres g
Wm Weight of matrix g
ρf Fibre density g cm−3
ρm Matrix density g cm−3
1
Chapter 1
Introduction
1.1 Contextual background
Composite products have gained popularity due to the various advantages they of- fer. The ability to tailor the composition of composites to meet specific requirements make them an attractive prospect. The weight savings offered by composite prod- ucts have found them being used extensively where this translates to fuel savings in industries such as aerospace, automobile etc.
The sports sector has adapted composite technology since its development. Using this technology, manufacturers are able to specifically optimise their products for better performance and lighter weight. Composites are used in the production of high end hockey sticks, bicycle frames, tennis racquets etc.
Although composites provide numerous advantages, they do have drawbacks. They are an expensive investment, which could be offset by long term benefits in terms of fuel savings. One of the critical issues with composite structures is their dam- age response. They are inherently brittle in nature, therefore they do not deform significantly before fracture compared to conventional materials such as steel.
The lack of reinforcement in the out of plane direction makes composite products susceptible to forces in that direction. This makes them weak in impact and restricts their use in impact critical structures. In recent years, there has been a considerable amount of focus on improving this particular characteristic of composite materials.
This work was carried out at Oxeon AB, a company that produces Spread Tow Car- bon Reinforcements under the brand name TeXtreme R. Oxeon’s unique spreading and weaving technologies were developed by Dr.Nandan Khokar of Chalmers Uni- versity of Technology in Gothenburg, Sweden. Oxeon AB was founded in 2003 at the university’s School of Entrepreneurship. Presently Oxeon is head-quartered in Borås, Sweden.
1.2 Scope of work
The presented work tried to outline the factors influencing the impact performance of composite products. This was done within the scope of the sports industry. In this work, different configurations, matrix systems, and technologies were investigated by manufacturing tubes and conducting drop mass tests on them. The primary aim
2 Chapter 1. Introduction
of the work was to build an initial database based on the conclusions of which, fur- ther detailed investigations could be carried on.
The aim of this work was to:
• Conduct an extensive literature study to gather information about impact per- formance of composite products
• Choose the material configurations and concepts to be tested
• Streamline the manufacturing process in order to produce acceptable test spec- imens
• Manufacture test specimens and conduct tests to obtain the residual strength after impact
• Analyse the results and draw conclusions based on which further tests can be carried out
1.3 Literature review
The literature review was focused on three main areas in composite design:
• Resin system
• Layer interface and stacking sequence
• Fibre
Apart from these aspects, the thin ply effect is also explored here briefly as it is relevant to this study.
1.3.1 Resin system
The matrix of the composite is an important factor when it comes to damage inhi- bition as any kind of failure begins with the matrix cracking. These cracks in the matrix propagate through the composite and leads to further failures often charac- terised as delamination and fibre breakage. Therefore, any form of improvement in the ability of the matrix to resist damage, will improve the impact resistance of the composite product significantly.
There are various forms of tougheners available in the industry that can be used to improve impact performance. However, Graphene is a relatively new development in this field and has emerged as a popular option to enhance the properties of the resin. The high mechanical strength and large specific area of graphene increases the toughness of the resin. The ability to easily disperse graphene in the resin, makes it an attractive option.
The use of carbon nanotubes (CNTs) has been prevalent. Adding CNTs to the resin improves its impact performance as they bring in additional modes of energy ab- sorption such as nanotube fracture, nanotube pullout and micro-crack bridging [1].
Their incorporation also allows production of multifunctional composites due to
1.3. Literature review 3
their excellent electrical and thermal conductivities. Although CNTs provide nu- merous advantages, their usage is limited by the difficulties in efficiently dispersing them in the resin [2].
Core shell particles (CSPs), which consist of a shell and a core, are used to modify the resin to achieve better mechanical properties. In recent years, studies have shifted their focus from pure polymer based CSPs due to their inability to effectively im- prove the impact resistance of epoxy thermosets. A recent approach has been to in- tegrate inorganic nanoparticles to form core shell nanoparticles. They seem to show much better properties due to the effect of mechanical performance from inorganic nanoparticles and the functionality arising from organic polymers [3].
1.3.2 Layer interface and stacking sequence
Delamination is the most common form of damage in composite parts subjected to impact. This is due to cracks that propagate and branch out. Therefore, strengthen- ing these regions will improve the impact resistance of the product.
A common way of increasing the delamination resistance is through the introduction of interleaves. They increase the energy absorption by deforming plastically. This inhibits crack propagation. Thermoplastic interleaves were found to have better energy absorbing ability when compared to thermoset interleaves [4], [5]. The com- patibility of the interleaf with the resin is an important criterion for selection. When the crack encounters the filaments of the veil while growing between the plies, it has to spend energy to go across the filament. If the energy required to do so is small, there will not be a significant improvement in the impact resistance.
Out of plane stitching as a method to introduce fibres in the z direction has been evaluated, but is not a popular method to improve the impact resistance. This is due to the fact that stitching disturbs the material architecture in the regions sur- rounding the stitches [6]. It introduces defects such as fibre misalignment and fibre breakage that could act as stress concentrations. A recent development in this area involves stitching together the plies using CNTs. They do not behave like conven- tional stitches as being on a nano-scale, they are unobtrusive, thereby maintaining the in-plane properties while improving the impact resistance [7].
3D woven composites have proven to be a relatively new area of research when it comes to impact study. 3D weaving tries to address the limitation of composite parts in the out of plane direction by introducing fibres in that direction. Studies have shown that they have better energy dissipating capabilities due to additional mechanisms brought in [8].
Some studies have shown that grouping of plies can have a negative effect on the impact performance of the composite part [9]. Grouping together of plies can lead to higher interlaminar stresses at the interface due to the increase in the difference of bending stiffness between the grouped plies. It also implies that there are fewer surfaces for delamination and hence the delaminations that occur tend to be larger in size.
Quasi-isotropic laminates have demonstrated better ability to withstand impact dam- age [10]. This could be attributed to the lower bending stiffness mismatch between the layers since the difference in angles are smaller. They also provide more sites for the onset of delamination, when compared to grouped plies.
4 Chapter 1. Introduction
1.3.3 Fibre
Although fibre may not be the most important factor for impact, it still is quite sig- nificant as the properties of the combination of fibre-matrix is what affects the me- chanical properties of the composite part.
There are generally three kinds of fibres available in the market. They are High Strength (HS), High Modulus (HM) and Intermediate Modulus (IM) fibres. HM fibres have the highest stiffness among the three and hence they are used in appli- cations where the stiffness is of paramount importance. IM fibres have the highest tensile strength and therefore they are used in applications where strength matters the most. HS fibres have properties that are a compromise between the two. Gener- ally, IM and HM fibres are much more expensive when compared to HS fibres.
Hybrid materials are gaining interest in the composite industry due to the unique advantages they bring. Carbon fibre composites on their own are brittle in nature and tend to break apart. But when combined with other fibres, the brittle behaviour can be avoided making it safer for the user of any composite product.
Dyneema is a fibre produced by gel spinning of ultra high molecular weight polyethy- lene. They have a low density, making them an attractive option for lightweight structures. They have high tensile strength but low compressive strength. Dyneema hybrids are being employed in bullet proof vests due to their impact resistance and low weight. Some initial studies have shown that Dyneema can be of particular in- terest in applications where the limiting load case is impact resistance [11]. For the purpose of this study, sourcing Dyneema during the limited time frame proved to be difficult.
CNT yarns are a new advancement in the composite industry. Presently, they are used extensively in applications that require electrical or thermal conductivity. For example, they have become popular in the aerospace industry as a replacement for copper wires. Their extremely low weight means that replacing all the wires in an aircraft with CNT yarns produces considerable weight savings. CNT yarns have not found extensive use in structural applications although they have five times the tensile strength of carbon fibre [12].
1.3.4 Thin ply effect
Thin plies are produced by spreading large tows of carbon fibre. This way, ply thick- ness can be reduced to as much as 1/6 of the conventional 0.12 mm thickness [13], [14].
Thin ply laminates offer several advantages over conventional plies. Apart from being able to produce thinner and lighter laminates, for the same laminate thickness, more plies can be accommodated when using thin plies. This expands the design space considerably.
The failure mechanism of thin ply vary from conventional plies. Thin plies delay the onset of delamination and matrix cracking and increases failure loads [15]. Failure modes for regular thick plies are usually large delaminations. This is not true for thin plies, for which fibre failure usually happens before free edge delamination.
5
Chapter 2
Materials
The tubes were manufactured using prepregs supplied by SHD composites. The base resin selected for the test was MTC 510 from SHD. This is a standard epoxy based resin used in the sports industry.
The entire test matrix was set-up to conduct tests in six different categories (A) Comparison of conventional HS fabrics with TeXtreme R
(B) Effect of different types of fabrics
(C) Positioning of TeXtreme R fabrics in the lay-up (D) Effect of ply scattering
(E) New developments for impact (F) Effect of ply thickness
The tests were divided into categories for easier comparison of results. Although most of the results presented in this report sticks to comparison within the cate- gories, some comparisons are made across the categories as well.
The tubes manufactured are named based on the alphabet of the category they be- long to. For example, the first three tubes belonging to category A will be called A1, A2 and A3 respectively. Further on, since each tube is cut into three specimens, the specimens will be referred to by a decimal point followed by a number. Therefore, the three specimens of tube A1 will be referred to as A1.1, A1.2 and A1.3. The speci- mens that are impacted will always be numbered 1 and 2. This system makes it easy to clearly identify and refer to results. This test programme uses two tubes as the reference for comparison of results. They will be called R1 and R2.
The base lay-up for comparison was chosen as 45/-45/0/0/45/-45/0/0/45/-45.
This is a standard lay-up used in the sports industry. All the TeXtreme R material used in this test were biaxial woven fabrics. Table2.1 shows all the prepreg used for this test. The resin content was decided based on the limitation of the supplier and global fibre volume fraction of the tube. Whenever the outer plies of the base layup are swapped with other fabrics from table2.1, it is done so by ensuring that the global thickness and quantity of fibre in the tube remains similar i.e, the number of plies on the outer layer can differ based on the FAW. In table2.1, TeXtreme 1000*
is the only material with a different resin, hence the usage of ’*’ is used to distinguish it from the other standard material. Unless explicitly specified, "TeXtreme TR50" in this report refers to the 100 GSM fabric (TeXtreme 1000 in table2.1).
6 Chapter 2. Materials
Prepreg Fibre FAW (GSM) Resin System RC (%)
UD T700 100
MTC 510
38
NCF 45/-45 T700 200 42
3K Twill weave T300 200 42
TeXtreme 1000 TR50 100 38
TeXtreme 1171 MR70 42 60
TeXtreme 1162 Innegra-TR50 Hybrid 130 38
TeXtreme 1009 M30S 76 45
TeXtreme 1025 TR50 200 38
TeXtreme 1051 UTS50 64 48
TeXtreme 1000* TR50* 100 MTC 9800 40
TABLE2.1: Material used in the test programme
2.1 Comparison of conventional HS fabrics with TeXtreme
RThe aim of this category is to provide a comparison between using conventional fabrics on the outer layers and using TeXtreme R on the outer layers. The inner layers have standard UD. The conventional fabrics used in this test were NCF and 3K twill weave. It has to be noted that 3K twill fabrics used in this category consist of T300 fibres which does differ in properties from the other fibres used. The reference used in this category was a tube manufactured completely using UD prepreg, as this will provide the advantages/disadvantages of replacing the outer layers with other fabrics. The tube configurations used are shown in table2.2.
Tube Designation Inner Layer Outer Layer
Lay-up (Inner to Outer) Material Lay-up Material A1 45/-45/0/0/45/-45/0/0 UD (45/-45)2 TeXtreme TR50
A2 45/-45/0/0/45/-45/0/0 UD 45/-45 3K Twill
A3 45/-45/0/0/45/-45/0/0 UD 45/-45 NCF
R1 45/-45/0/0/45/-45/0/0 UD 45/-45 UD
TABLE2.2: Tubes in category A
2.2 Effect of different types of fibres
In this category, the affect of having different types of fibres on the outer layer was in- vestigated. M30S and MR70 are IM fibres. IM fibres are a popular option to enhance the impact resistance of composite products due to their high tensile strength. MR70 in particular is of specific interest as it is the fibre with the highest tensile strength available in the market. Tubes B2 and B3 have M30S and MR70 respectively on the outer layers.
Innegra is a low density polypropylene fibre that is usually used along with carbon fibre to provide weight reduction, impact resistance and durability to a composite product. Addition of Innegra prevents the composite part from sudden fracture as the Innegra fibres hold together the composite part. Tube B1 is manufactured using an Innegra-TR50 hybrid fabric on the outer layer.
2.3. Positioning of TeXtreme R fabrics in the lay-up 7
Tube Designation Lay-up (Inner to Outer)
Inner Lay-up Material Outer Lay-up Material
B1 45/-45/0/0/45/-45/0/0 UD (45/-45)2 Innegra-TR50
B2 45/-45/0/0/45/-45/0/0 UD (45/-45)2 TeXtreme M30S
B3 45/-45/0/0/45/-45/0/0 UD (45/-45)4 TeXtreme MR70
B4 45/-45/0/0/45/-45/0/0 UD (45/-45)3 TeXtreme UTS50
R1 45/-45/0/0/45/-45/0/0 UD 45/-45 UD
TABLE2.3: Tubes in category B
UTS50 is a fibre which is very similar to TR50 and T700 in terms of performance.
The reason for its inclusion in the test matrix is because UTS50 fabrics fall into the category of ultra thin plies. As explained in section 1.7, thin plies bring various advantages. Tube B4 is manufactured using UTS50 on the outside to observe the effect of thin plies. Table2.3shows a complete list of the tubes in this category.
2.3 Positioning of TeXtreme
Rfabrics in the lay-up
In this category, plies of UD in tube R1 were replaced by TeXtreme TR50 fabric at different positions as shown in figure2.1.
FIGURE2.1: Configurations with TeXtreme at different locations
The results of the three tubes are compared with reference tube R1 to understand how placing TeXtreme R woven fabrics at different points in the lay-up affects the impact resistance.
2.4 Effect of ply scattering
In this group different lay-ups were tested to understand how scattering the plies will effect the impact properties of the tube. Table2.4shows the three tubes with dif- ferent lay-ups. All the tubes in this category are manufactured using only TeXtreme R material. All the tubes in this group are compared to R2 for reference. R2 is a tube with the standard lay-up and complete TeXtreme R material.
8 Chapter 2. Materials
Tube Designation Lay-up (Outer to Inner) Material
D1 [(0/90)/(45/-45)]5 TeXtreme TR50
D2 [(45/-45)/(0/90)]5 TeXtreme TR50
D3 [(45/-45)/(0/90)/(45/-45)/(0/90)/(45/-45)]s TeXtreme TR50 R2 (45/-45)2/(0/90)2/(45/-45)2/(0/90)2/(45/-45)2 TeXtreme TR50
TABLE2.4: Ply scattering
Apart from the tubes in table2.4, two more tubes are included in this category. They are shown in figure2.2. These tubes are variations of different fabrics of differing thickness. This is done so to understand how the mixing of different thickness affects the performance.
FIGURE2.2: Thick and thin variations
2.5 New developments for impact
In this group, some recently developed technologies were tested to find out how it compares to the standard in the industry. Tube E1 (table2.5) is manufactured using a resin in which graphene has been dispersed. The resin MTC 9800 was supplied by SHD composites. MTC 9800 has a very similar viscosity curve compared MTC 510.
Hence, the same curing cycle is used.
Tube Designation Lay-up Material
E1 (45/-45)2/(0/90)2/(45/-45)2/(0/90)2/(45/-45)2 TeXtreme TR50*
E2 figure2.3 figure2.3
E3 figure2.4 figure2.4
R1 45/-45/0/0/45/-45/0/0/45/-45 UD
R2 (45/-45)2/(0/90)2/(45/-45)2/(0/90)2/(45/-45)2 TeXtreme TR50 TABLE2.5: New developments for impact
Tube E2 (figure2.3) is completely made up of TeXtreme TR50 fabrics with outermost layers having VACNTs transferred onto the surface. As explained in section 1.3.2, the CNTs on curing stitch together the plies delaying delamination. The CNT transfer was provided by N12 Technologies.
2.5. New developments for impact 9
FIGURE2.3: Configuration of tube with N12
Tube E3 has UD on the inside with TeXtreme R on the outermost four plies. The outermost four interfaces has Xantu.Layr R in them. Xantu.Layr R is a thermoplas- tic nano-veil. It has kilometers of nano-fibres in every square metre of the material.
These nano-veils are manufactured by using an electro-spinning technology. These veils improve the impact resistance of the composite part. They were applied di- rectly onto the prepreg manually.
FIGURE2.4: Configuration of tube with Xantu.Layr Tubes E3 is compared with R1, while E1 and E2 are compared with R2.
10 Chapter 2. Materials
2.6 Effect of ply thickness
The aim of this category is to understand the effect of using thicker plies on impact performance. As shown in figure2.5, tube F1 has 200 GSM plies and F2 has 200 GSM on the inside with the last two plies of F1 replaced with four plies of 100 GSM fabric.
FIGURE2.5: Tubes F1 and F2
11
Chapter 3
Methodology
3.1 Tube manufacturing
The tubes are manufactured through a bladder molding technique. This process is used to manufacture hollow composite parts by using prepregs.
The mold used in this thesis was designed to manufacture tubes according to the geometry indicated in figure3.1.
FIGURE3.1: Tube geometry
The mold (figure 3.2) was manufactured from aluminium to allow for a lighter weight and consequently easier handling. The mold has heating elements and ther- mocouples embedded within for precision temperature control (figure3.3).
FIGURE3.2: Mold with heating elements and thermocouple embed- ded
A high temperature latex bladder (figure3.4) was used which was connected to a coupling (figure3.5) that allows for air supply to go into the bladder. A latex blad- der was chosen over a nylon or silicon due to its advantages. Nylon bladders would
12 Chapter 3. Methodology
FIGURE3.3: The open mold
produce wrinkles on the inner surface while silicon would prove tough and expen- sive to obtain. Latex bladders perform better than nylon and are not as expensive as silicon. Each bladder was checked for any leaks every time before using it for the curing cycle. The latex bladder and mold were prepared for each curing cycle by applying a water based release agent (Zyvax WATERSHIELD [16]). Three coats of the release agent was applied on the mold after every two cycles of curing. The release agent (3 coats) was re-applied on the bladder after each use.
FIGURE3.4: Latex bladder
FIGURE3.5: Coupling for air inlet to bladder
The release agent was applied using a lint free cloth. A lint free cloth does lot leave behind any residue compared to a normal cloth. The surface was first wiped with a lint free cloth soaked with the release agent and then wiped dry after waiting for 1-2 minutes. A waiting time of 15 minutes was observed between the application of each coat as recommended by the supplier of the release agent. The release agent
3.1. Tube manufacturing 13
was allowed to cure for at least 30 minutes after the final coat before starting the cure cycle for the tube.
The prepreg patches were cut into rectangular dimensions of 185 mm by 950 mm.
The rectangular patches were cut in the required angles for all material except for the UD prepreg. In the very beginning, the patches were directly formed on the bladder. This was done by pressurising the bladder slightly in order to provide support. Tubes manufactured in this manner had a fold on the inside (figure3.6) as the final diameter of the formed stack was higher than the diameter for the geometry in figure3.1.
FIGURE3.6: Tube with a fold on the inside
Following this, a mandrel with a much smaller diameter was used to form the plies which solved this problem (figure3.7). The mandrel was wrapped around in a re- lease film held together by release tape as shown in figure3.8. The film makes it easy to remove the formed plies from the mandrel.
FIGURE3.7: Tube with a smooth inner surface
The first layer of the prepreg was formed onto the mandrel carefully with the help of a heat gun. Using a heat gun makes the prepreg more tacky and hence it will stick to the mandrel better. Subsequent layers were easier to form as prepreg to prepreg adhesion is better than prepreg to film. As each patch of prepreg has the same width, the amount of overlap will be lower on the outside than the inside. The stack was turned by a quarter after each patch has been formed. This ensures that the overlap is spread through out in the tube. Since the method of forming is a constant for all
14 Chapter 3. Methodology
FIGURE3.8: Mandrel wrapped around with release film
the tubes, this should not influence the results. Figure3.9shows the formed stack on the mandrel.
After forming all the plies of prepreg, the stack was removed from the mandrel.
When doing this, the release film sticks to the inside of the formed plies. There- fore, the release film was scraped off from the inside. The bladder was then pulled through the stack with the help of a cable tied to the end of the coupling.
FIGURE3.9: Formed stack on the mandrel
The formed plies were then placed in the mold and the mold was closed and tight- ened using screws. A heat blanket was used to cover the mold to prevent heat loss during the curing cycle. After the complete cycle, the mold was cooled down in two hours with the help of a fan. The mold was cooled down this way to increase the pace of the manufacturing. The first trials using this method of curing produced dry spots on the surface of the tube (figure3.10).
It was initially thought that the issue was due to the uneven compaction of fibres.
This was due to the difficulty in forming the first layer on the mandrel. Because of this difficulty, the first layer would be very much wrinkled as seen in figure3.11.
Since the issue with wrinkles were only with UD, it was decided to form pre-stacks of 45/-45. To reduce wastage of material, a schema was followed to form the pre- stack. This is shown in figure3.12. This had two advantages, it made it easier to form the first layer and this way it was possible to have 45/-45 patches which were 950 mm in length as the prepreg roll of UD came at a width of 600 mm.
3.1. Tube manufacturing 15
FIGURE3.10: Dry spots on the tube
FIGURE3.11: Wrinkles when forming the first layer
FIGURE3.12: Schema for forming pre-stacks
Although, in this manner it was possible to form a complete stack without any major wrinkles, the issue with dry spots remained. To make sure it was not an issue with the release agent, a solvent based release agent was also tried out but this did not help either. Up until this point, the curing cycle used had a dwell at 80◦C as recom- mended by the supplier. It was decided to increase this temperature to 100◦C as this was the point of lowest viscosity of the resin. This seemed to improve the surface finish of most tubes with TeXtreme R on the outer layer but tubes with UD on the outside still had issues. The curing cycle used in this process is depicted in figure 3.13
It was hypothesised that these spots were due to entrapped air and in order to re- move them, 45/-45 stacks of the UD were vacuum bagged for 30 minutes under 1 bar of pressure. Before forming, these stacks were then perforated to further facili- tate the movement of air (figure3.14).
The perforation seemed to improve the surface finish but it was not yet perfect.
16 Chapter 3. Methodology
FIGURE3.13: Curing cycle
FIGURE3.14: Perforated stack of 45/-45 UD
FIGURE3.15: Partial vacuum applied on the mold
3.2. Drop mass test 17
Therefore, a partial vacuum was applied on the mold to help pull out air during the cure cycle. The mold was designed to include this feature (figure3.15). Although it would have been possible and better to run a complete vacuum, this would slow down the manufacturing cycle. The partial vacuum was run through out the cure cycle. This method seemed to produce tubes with an acceptable level of surface finish.
3.2 Drop mass test
The manufactured tube was cut into three parts of 300 mm length each (figure3.16).
Two of these specimens, the ones closest to the air inlet, were impacted. The last specimen was left undamaged. This ensured that all three specimens had the same manufacturing cycle and history.
FIGURE3.16: Three specimens cut from a tube
The set up to conduct the drop mass test on the tubes is shown in figure3.17. The set-up includes an anti-rebound system to prevent multiple impacts. A slow motion camera capable of recording at 120 fps and 1080p resolution was used to capture the rebound height. This can be used to approximate the absorbed energy.
FIGURE3.17: Drop mass set-up
The drop mass and height was selected as 1280.24g and 0.82m. This was chosen in order to achieve an impact velocity of 4m/s and energy as 10.3 J. This energy level will cause visible damage on the tube. It is similar to a situation in which a bike might encounter a crash. The velocity and energy was calculated from the conservation of energy principle
18 Chapter 3. Methodology
mgh= 1
2mv2 (3.1)
It was assumed that the energy loss from friction is negligible.
Before impacting the specimens, their weight, length, and thickness were recorded in order to back calculate the fibre volume fraction. In order to do so, the resin content is first estimated from
Wm =Specimen weight− (FAW·specimen length·patch width·Number o f plies) (3.2)
RC=Wm/Specimen weight (3.3)
Wf = (1−RC) ·Specimen weight (3.4) From this, knowing the fibre density and resin density, it is possible to calculate the fibre volume fraction by
Vf = Wf/ρf
Wf/ρf +Wm/ρm (3.5)
3.3 2 point bending test
All the specimens manufactured were tested for their flexural strength through a 2 point bending test. The 2 point bending tests were conducted at AD Manus materi- alteknik AB. The set-up for conducting the test is shown in figures3.18and3.19.
The part of the tube between the two fixed points is supported from the inside by a cylindrical fixture shown in figure3.20. This is to prevent the tube from collapsing inwards during the test.
It was ensured that the part of the specimen that had been impacted was facing upward when the end was pulled up. The specimens always fail due to compression on the top. The machine was set to have a displacement rate of 2 mm/min. All tests were halted after a drop of 10% in the measured load.
3.3. 2 point bending test 19
FIGURE3.18: 2 point bending test
FIGURE3.19: Tube fixture
20 Chapter 3. Methodology
FIGURE3.20: Cylindrical fixture
21
Chapter 4
Results and Discussion
4.1 Comparison of conventional HS fabrics with TeXtreme
RAll the tubes in this category have a fibre volume fraction between 45% and 46%.
The residual and flexural strengths of the tubes are shown in figure4.1. The percent- age reduction in their strengths are presented in figure4.3. Of all the specimens in this group, specimen A1.1 failed in a different manner compared to the rest of the specimens.
FIGURE4.1: Comparison of impact performance of tubes in category A
From figure4.2, it can be seen that the fracture of the specimen after the 2-point bending test proceeds in a diagonal manner. This is probably the reason why the value for this specimen seem to be lower than A1.2. It could be due to a defect in manufacturing or could be a result of the point of impact during testing. When woven fabrics are impacted on the interlacing points, the damage and failure mech- anism vary considerably. This variation depends on the length scale of the weave. It has to be noted that A2 does not show this problem which has a twill weave. This needs further investigation. Due to this reason, A1.1 is neglected from the analysis of the results.
22 Chapter 4. Results and Discussion
FIGURE4.2: Specimen A1.1
FIGURE4.3: Percentage reduction in strength of tubes in category A
A1, A2 and A3 have higher flexural strength than the reference with A3 being the highest. The impact performance of A1 seem to better than both A2 and A3. A1 has a residual strength which is slightly higher than the reference. Both A2 and A3 are not as effective in providing damage resistance to the tube when compared to the reference and A1.
Part of the reason why A2 and A3 are performing poorly when impacted might be due to the fact that they both have 200 GSM fabrics on the outermost layers. A1 and R1 both have two plies of 100 GSM on the outer layer which are thinner. Due to this, there is probably a thin ply effect acting which delays the crack initiation and propagation. Other factors that influence these results could be the presence of crimp in 3K and stitching threads in NCF.
In terms of percentage reduction in strength, A1 performs the best. A1 and R1 are very similar in this factor. A1 is around 60% better when compared to A2 and 30%
better when compared to A3.
4.2. Effect of different types of fibres 23
All the damage free specimens in this test programme had localised failure, i.e they failed at the support. Due to this, it stands to reason that flexural strengths of these specimens could be underestimated and hence percentage reduction plots in this report cannot be considered to be entirely accurate.
4.2 Effect of different types of fibres
All the tubes in this category have a fibre volume fraction between 43% and 46%
except for B1 which is 50%. B1 has Innegra-TR50 fabrics on the outer layer which is similar in aerial weight to the others. Since Innegra also has a lower density, it means that more fibres are required to have the same weight as others. The residual strength and percentage reduction in strength are shown in figures4.5and4.6.
From figure4.4, it can be seen that the fracture of B2.1 is similar to A1.1. Therefore, due to the same reasons B2.1 is neglected from the analysis.
FIGURE4.4: Specimen B2.1
FIGURE4.5: Comparison of impact performance of tubes in category B
24 Chapter 4. Results and Discussion
From figure4.5, it can be seen that B2 has the best performance compared to all other tubes. This should be due to the IM fibres on the outer layer. On the other hand, B3 has low impact resistance although it has the best available IM fibres in the market on the outer layer.
If the resin movement between the plies are restricted, it is possible to have a vary- ing local fibre volume fraction. Both B2 and B3 have the same specimen weight and inner lay-up. This means that any excess resin must have been squeezed out and they must have very similar quantity of resin. Hence the only difference between the two tubes is that B3 has slightly more amount of fibres due to the variation in the FAW. This small variation should not bring such a drastic change. It is therefore hypothesised that MR70 fibres and MTC510 resin system do not probably integrate well. However, it has to be noted that there were only two specimens with MR70 impacted here. More tests would have to be conducted to reach a definitive conclu- sion.
FIGURE4.6: Percentage reduction in strength of tubes in category B
Figure4.7shows the force displacement curves for B1. It can be seen that B1.1 breaks suddenly while B1.2 is still capable of taking load after the first major failure. B1.2 is more characteristic of Innegra though there is still a major load drop before the curve levels out. This could be due to the adhesion between the matrix and Innegra.
Innegra fibres usually have adhesion issues which is why the matrix should be care- fully selected. Although MTC 510 is a standard epoxy in the industry, it could be the poor adhesion between Innegra and the resin system that causes this behaviour.
Another factor influencing the behaviour of B1 could be that B1 uses commingled fibres of Innegra and TR50 which are spread. This separates the individual fibres of Innegra from each other with fibres of carbon in between them. Therefore, on impact, the carbon fibre could fracture and cut through the Innegra negating the influence of Innegra.
B4 also has a strong performance. This could be attributed to the thin ply effect as the plies used here are 64 GSM each. It has to be noted that UTS50 fabrics have
4.3. Positioning of TeXtreme R fabrics in the lay-up 25
FIGURE4.7: Load-displacement curves for B1
larger squares than the other material. But this should not influence the impact performance.
4.3 Positioning of TeXtreme
Rfabrics in the lay-up
All the tubes in this category have a fibre volume fraction around 46%. The residual strength and percentage reduction in strength are shown in figures4.8and4.9.
FIGURE4.8: Comparison of impact performance of tubes in category C
From figure 4.8 it can be seen that adding TeXtreme R improves the performance in all cases except for C2. When TeXtreme R is added on the extreme (inner and outermost) layers (C2), the performance is affected in a negative manner. The reason for this could be a defect in the tube and therefore it requires more investigation.
26 Chapter 4. Results and Discussion
FIGURE4.9: Percentage reduction in strength of tubes in category C
Overall, the performance of C3 is the best in this group. It could be that TeXtreme R performs better when experiencing lower energies. Since in C3, TeXtreme R is on the innermost layer, most of energy dissipates on the outer layers, exposing the inner layer to lower loads and energies. It should however be kept in mind that all of these tests would need to be repeated multiple times to have a credible data set. It also has to be noted that adding thin plies positively reinforces the tensile properties while it does not necessarily affect the compressive properties. Therefore, adding thin plies on the inner-most layers which are on tension in a 2 point bending test would provide better results. This could explain the behaviour of C3.
4.4 Effect of ply scattering
The tubes in this group have a fibre volume fraction around 47%. The residual strengths are shown in figures4.10and4.11. The percentage reduction in strengths are shown in figures4.12and4.13.
From figure4.10, it can be seen that scattering of plies does affect the flexural strength of the undamaged specimen but there is no real difference in the residual strength of impacted specimens.
The reason why D1 has a lower flexural strength when compared to D2 could be attributed to the fact that it has the 90 degree fibres closer to the surface of the spec- imen. These 90 degree fibres does not contribute to the flexural strength. During a 2 point bending test, the stress will be the highest on the surface and decreasing towards the inner regions.
From figure4.11, it can be seen that D4.3 has one of the highest flexural strength.
This specimen has a 100 GSM fabric and 200 GSM fabric on the last two outermost layers. The 200 GSM fabrics were found to have a lower resin content compared to the 100 GSM fabric. If there is no movement of resin between the plies, this could translate to a high local fibre volume fraction on the surface. On the other hand, D5
4.4. Effect of ply scattering 27
FIGURE 4.10: Comparison of impact performance of tubes in cate- gory D
FIGURE 4.11: Comparison of impact performance of tubes in cate- gory D
has the 64 GSM and 200 GSM fabrics on the last two outermost layers. Since 64 GSM fabrics have a comparatively higher resin content, it would have a lower local fibre volume fraction on the surface.
From figure4.14, it can be seen that the failure occurs at similar loads for all speci- mens but the post failure behaviour differs. When there is ply scattering involved, there is less of load drop and the failure is more gradual.
28 Chapter 4. Results and Discussion
FIGURE4.12: Percentage reduction in strength of tubes in category D
FIGURE4.13: Percentage reduction in strength of tubes in category D
4.5 New developments for impact
Tubes R2 and E2 have around 48% fibre volume fraction while E1 has around 45%.
This is due to the higher resin content on the plies for E1. E3 has 46% while R1 is at 45%. Figures4.15and4.16show the impact performance of the tubes by comparing the residual strengths. Figures4.18and4.19provide a comparison of the percentage drop in the residual strengths.
From figure4.15, it can be observed that CNT stitching and graphene loaded resin does not provide any improvement in terms of impact performance. However, they do show improvements in flexural strength of the undamaged specimen. The per- formance of E2 could be improved if more interfaces with CNTs are added as in this
4.5. New developments for impact 29
FIGURE4.14: Load-displacement curve for impacted specimen in cat- egory D
FIGURE 4.15: Comparison of impact performance of tubes in cate- gory E
test only the two outermost layers had CNT stitching.
From4.16, it is obvious that adding Xantu.Layr R decreases the flexural strength of the undamaged specimen. This could be a result of the thermoplastic fibres in the specimen. E3.1 and E3.2 were however the only specimens with no visible damage after they had been impacted (figure4.17).
30 Chapter 4. Results and Discussion
FIGURE 4.16: Comparison of impact performance of tubes in cate- gory E
FIGURE4.17: Specimen E3.1 and E3.2
From the results, it can be inferred that there were cracks on the inside, but it is still interesting that adding Xantu.Layr R seemed to make the tubes absorb the impact in a more elastic manner, i.e the rebound height of the impactor was higher than the rest. This was also evident in the manner in which these specimens bounced out of the fixture on impact. Although the undamaged specimen had a low strength, the drop in the strength in the damaged specimens were not significant. In this test, Xantu.Layr R was added to the last four interfaces. The results could be improved by reducing this to the last two or by using a version with a lower areal weight.
4.6. Effect of ply thickness 31
FIGURE4.18: Percentage reduction in strength of tubes in category E
FIGURE4.19: Percentage reduction in strength of tubes in category E
4.6 Effect of ply thickness
R2 has a fibre volume fraction of 48%, while F1 and F2 have 51% and 49% respec- tively. The higher fibre volume fraction of F1 stems from using only 200 GSM fabrics as these fabrics have the lowest resin content. Figures4.20and4.21show the resid- ual strength and the percentage reduction in the strength of the tubes.
From figure4.20, both F1 and F2 have higher flexural strength for the undamaged
32 Chapter 4. Results and Discussion
FIGURE 4.20: Comparison of impact performance of tubes in cate- gory F
specimen. This is probably due to the higher fibre volume fractions. F1 has higher strength values for both the damaged and undamaged specimens. This could be attributed to the effect of thinner plies.
FIGURE4.21: Percentage reduction in strength of tubes in category F
4.7 Variation in stiffness
Figures4.22, 4.23and4.24shows the comparison of the load-displacement curves between different tubes. Although, the stiffness cannot be accurately estimated from this curve, the slope of the curve gives an idea about the stiffness.
From figures4.22,4.23and4.24, the common theme is that all specimens with TR50 on the outer most layers (A1.3, C2.3 and R2.3) have a lower stiffness compared to the rest. All these tubes have the same behaviour. The slope of these curves seem
4.7. Variation in stiffness 33
FIGURE 4.22: Load - displacement curves for tubes with different TeXtreme fabrics on the outside (Undamaged)
to change around 2000 N. This implies that there must be some delamination on the outer layers at this load.
FIGURE4.23: Load - displacement curves for tubes with TeXtreme at different positions in the lay-up (Undamaged)
From figure4.24, it can be seen that the tubes with N12 and graphene loaded resin seem to inhibit this behaviour, i.e they suppress the delamination that occurs in the other tubes. This change in stiffness/early failure could be attributed to the interac- tion between the fibre and the resin. This behaviour is not as evident in B4.3 which has UTS50 on the outer layers which are very similar in properties to TR50.
34 Chapter 4. Results and Discussion
FIGURE4.24: Load - displacement curves for tubes with new tech- nology (Undamaged)
35
Chapter 5
Conclusion
From the results obtained, it seems that TeXtreme R provides some improvement over other conventional choices. This improvement does depend on the position at which TeXtreme R is added as explained in section 4.3. However, it is important to consider that there is no statistical evidence to support this as the sample size was small. Also, it was not possible to obtain prepregs that had the same fibre type in order to eliminate the influence of different fibres.
Thin plies seem to provide better performance but requires a broader sample size to confirm its advantages. Most notably, the tubes with UTS50 64 GSM plies on the outer layer seemed to consistently perform well. The thinnest plies used in this test programme were MR70 fabrics. The performance of these fabrics turned out to be poor. This result is contradictory and as explained in section 4.2, it could be due to the poor compatibility between the matrix and fibres. This is an area that requires further investigation.
Scattering of plies did not show any difference in impact performance in terms of strength. But the post failure behaviour of the damaged tubes was different when there was ply scattering. On the other hand, scattering of plies showed better per- formance for undamaged specimens.
New technologies such as N12, graphene based resin MTC9800 and Xantu.Layr R showed mixed results. While on certain parameters the performance seemed to be better, it was the same or worse in other cases. These cases require further investi- gation and as already mentioned in section 4.5, tweaking certain parameters could bring about a significant change in the performance. Therefore, these technologies require further investigation before they can be completely disregarded.
5.1 Future work
This test programme was able to confirm some of the advantages of using TeXtreme R fabrics. But there is still more work that could be done to expand this. While in this work, positioning of TeXtreme R was tested for three different configurations, there are more that could be of interest. For example, it would be beneficial to under- stand how replacing all the 45/-45 plies in a UD with TeXtreme R would affect the performance. It would also be beneficial to conduct a test programme in which the same tube is manufactured and tested several times to have statistical evidence to the findings.
36 Chapter 5. Conclusion
Although a slow motion camera was used to record the rebound height in order to calculate the energy absorbed, the data was not used as it seemed inaccurate. It would be better to conduct these tests with load and speed sensors embedded in the drop test set up to accurately calculate the energy absorbed.
It could also be interesting to see how TeXtreme R behaves under smaller energies of impact. Impact scenarios such as stones hitting the frame of a mountain bike are at much lower energy levels. A test programme could be designed in which the tube is impacted multiple times with a low energy. The tube with TeXtreme R on the inside fared much better than the rest, which could be directly related to better performance of TeXtreme R architecture at lower energies. It would also be of interest to investigate the performance of TeXtreme R at higher energies.
It was mentioned earlier that in woven fabrics, the point of impact could be a major influence. The impact damage could be especially critical at the interlacing points which would result in cracks due to shear. Therefore, it could be important to un- derstand this more. It could also be possible to suppress damage due to impact by coating the surface with a thin layer of resin although this could increase the weight slightly. This could also be investigated further.
Although thin plies are better in terms of performance, it is not entirely practical to use thin plies everywhere. This is mainly due to the higher processing times which translates to a higher cost. This can be countered using thicker plies to build up thickness of the composite part and using thinner plies where performance matters the most. While there were some combinations of thick and thin plies in this test programme, there are a lot more possibilities for combining thick and thin plies. An entire test programme could be designed just to investigate this phenomena, which would be beneficial from an industrial standpoint.
Along with combinations of different thickness, it would be beneficial to see how combinations of IM, HM and HS fibres perform under impact. It could be possible to find an optimal configuration that performs sufficiently well not just with impact, but also in other mechanical characteristics.
37
Appendix A
Test results from AD Manus
FIGUREA.1: Test result for tube A1
38 Appendix A. Test results from AD Manus
FIGUREA.2: Test result for tube A2
Appendix A. Test results from AD Manus 39
FIGUREA.3: Test result for tube A3
40 Appendix A. Test results from AD Manus
FIGUREA.4: Test result for tube B1
Appendix A. Test results from AD Manus 41
FIGUREA.5: Test result for tube B2
42 Appendix A. Test results from AD Manus
FIGUREA.6: Test result for tube B3
Appendix A. Test results from AD Manus 43
FIGUREA.7: Test result for tube B4
44 Appendix A. Test results from AD Manus
FIGUREA.8: Test result for tube C2
Appendix A. Test results from AD Manus 45
FIGUREA.9: Test result for tube C3
46 Appendix A. Test results from AD Manus
FIGUREA.10: Test result for tube D1
Appendix A. Test results from AD Manus 47
FIGUREA.11: Test result for tube D2
48 Appendix A. Test results from AD Manus
FIGUREA.12: Test result for tube D3
Appendix A. Test results from AD Manus 49
FIGUREA.13: Test result for tube D4
50 Appendix A. Test results from AD Manus
FIGUREA.14: Test result for tube D5
Appendix A. Test results from AD Manus 51
FIGUREA.15: Test result for tube E1
52 Appendix A. Test results from AD Manus
FIGUREA.16: Test result for tube E2
Appendix A. Test results from AD Manus 53
FIGUREA.17: Test result for tube E3
54 Appendix A. Test results from AD Manus
FIGUREA.18: Test result for tube F1
Appendix A. Test results from AD Manus 55
FIGUREA.19: Test result for tube F2
56 Appendix A. Test results from AD Manus
FIGUREA.20: Test result for tube R1
Appendix A. Test results from AD Manus 57
FIGUREA.21: Test result for tube R2
59
Appendix B
Specimen measurements
Tube Designation Weight (g) Thickness (mm) Length (mm)
.1 .2 .3 .1 .2 .3 .1 .2 .3
A1 101.43 96.92 98.29 1.595 1.543 1.552 302 295 298 A2 98.18 100.99 97.74 1.557 1.595 1.491 297 296 299 A3 100.65 99.46 98.68 1.584 1.611 1.589 298 298 299 B1 101.02 101.55 104.67 1.649 1.691 1.714 296 299 298 B2 94.48 98.24 99.28 1.534 1.578 1.557 296 296 298
B3 98.5 99.73 98.35 1.567 1.55 1.546 298 297 296
B4 99 100.77 98.98 1.571 1.531 1.526 298 298 298
C2 97.56 99.68 98.46 1.519 1.533 1.645 297 296 299 C3 100.53 96.01 98.97 1.511 1.524 1.558 303 295 293
D1 94.48 93.69 94.99 1.523 1.49 1.449 299 297 297
D2 96.85 94.63 95.12 1.498 1.463 1.464 300 297 296
D3 95.03 94.27 95.92 1.521 1.53 1.52 295 296 298
D4 101.5 101.9 102.24 1.618 1.641 1.591 297 299 298 D5 100.55 101.77 102.8 1.609 1.591 1.604 296 298 299
E1 97.7 97.97 98.54 1.552 1.53 1.543 295 299 296
E2 93.22 93.32 93.16 1.41 1.416 1.46 296 297 298
E3 104.77 106.12 103.15 1.586 1.582 1.493 299 298 296 F1 90.88 90.74 90.96 1.515 1.469 1.485 297 298 297 F2 93.89 92.93 92.56 1.442 1.573 1.518 299 297 296 R1 98.48 99.44 100.12 1.483 1.569 1.591 301 299 296 R2 96.45 93.32 95.29 1.497 1.462 1.475 299 296 299
61
Appendix C
Fibre volume fraction of specimens
Tube Designation Fibre Volume Fraction (%)
.1 .2 .3
A1 45 46 46
A2 46 44 47
A3 45 45 46
B1 51 51 49
B2 45 43 43
B3 44 43 44
B4 45 44 45
C2 46 45 46
C3 46 47 45
D1 48 48 47
D2 47 48 47
D3 47 48 47
D4 47 47 47
D5 47 47 47
E1 45 46 45
E2 48 49 49
E3 44 44 45
F1 50 51 50
F2 49 49 49
R1 46 46 45
R2 47 48 48
63
Appendix D
Estimated absorbed energy
This data was not considered for the analysis of results due to its inconsistency.
Tube Designation Energy Absorbed (J)
.1 .2
A1 5.8 5.5
A2 6.4 6.5
A3 6.3 6.4
B1 5.4 6.6
B2 5.5 5.3
B3 6.9 6.5
B4 6 6.3
C2 5.5 5.5
C3 6.5 6.4
D1 8.9 9.4
D2 6.7 9.4
D3 9.1 6.5
D4 9.7 6.9
D5 8.2 8.6
E1 8.7 8.7
E2 9.2 8.2
E3 5.3 6.7
F1 7.1 6.9
F2 6.9 7
R1 7 7.2
R2 6.9 9.2
