UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Experimental 4-step braided glass fiber reinforced composite for dental CAD/CAM applications Lesniak, Robert Matthew 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_september_lesniak_robert.pdf [ 3.34MB ]
Metadata
JSON: 24-1.0368686.json
JSON-LD: 24-1.0368686-ld.json
RDF/XML (Pretty): 24-1.0368686-rdf.xml
RDF/JSON: 24-1.0368686-rdf.json
Turtle: 24-1.0368686-turtle.txt
N-Triples: 24-1.0368686-rdf-ntriples.txt
Original Record: 24-1.0368686-source.json
Full Text
24-1.0368686-fulltext.txt
Citation
24-1.0368686.ris

Full Text

  EXPERIMENTAL 4-STEP BRAIDED GLASS FIBER REINFORCED COMPOSITE FOR DENTAL CAD/CAM APPLICATIONS by  Robert Matthew Lesniak   B.Sc., The University of Alberta, 2007  D.D.S., The University of Alberta, 2012    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF    MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (CRANIOFACIAL SCIENCES)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)    June 2018    © Robert Matthew Lesniak, 2018      ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the thesis entitled:  Dr. Ricardo Carvalho, Dr. Dorin Ruse, Dr. Adriana Manso      submitted by   Dr. Robert Lesniak  in partial fulfillment of the requirements for  the degree of    Master of Science    in    Craniofacial Science     Examining Committee:    Dr. Ricardo Carvalho, Faculty of Dentistry  Supervisor   Dr. Adriana Manso, Faculty of Dentistry  Supervisory Committee Member    Dr. Dorin Ruse, Faculty of Dentistry Supervisory Committee Member   Dr. Vincent Lee, Faculty of Dentistry  Additional Examiner          iii  Abstract  Introduction: Computer-aided design/computer-aided manufacturing in dentistry has lead to rapid expansion of new dental materials.    A new product from Shofu (Shofu, Japan), TRINIA CAD/CAM blocks and pucks, offers an alternative to the existing resin composite CAD/CAM materials.  The claimed mechanical properties of TRINA provide merit to further research and development of 3D fiber reinforced composites for CAD/CAM dentistry.    Objective: To design and produce 3D braided fiber-reinforced composites for dental CAD/CAM applications.    Materials and Methods: Experimental groups were designed and produced in conjunction with the Department of Materials Engineering, UBC.  The proposed design parameters included: 50:50 volume fraction fibers, 45-degree braiding angle, and 14 mm*14 mm cross sectional area, the approximate size of currently available CAD/CAM blocks.  Continuous S-2 glass fiber roving was used as the reinforcing agent.  Three experimental 3D glass fiber preforms were produced with internal structural variations using a 4-step braiding technique.  A 50:50 UDMA:TEGDMA resin matrix blend with a thermal curing agent was used for infiltration via submersion under vacuum, followed by thermal curing.  An unreinforced resin blend and a unidirectional fiber-reinforced composite, with the same volume fraction and dimensions, served as control groups.  Samples were prepared to 4 mm*4 mm*45 mm beams for three-point-bend testing.  An Instron 5969 Dual Column Material Testing System was used to test the flexural properties of the samples.  Scanning electron microscopy (SEM) was used to evaluate the fractured samples.       iv   Results: Results showed that incorporation of anisotropic unidirectional fibers had the most significant effect in reinforcing the resin blend.  The unidirectional control group produced the greatest flexural strength and elastic modulus at 336.6 MPa and 37.3 GPa, respectively.  The 2-ply+axial group followed at 236.52 MPa and 20.75 GPa.  The 2-ply had values of 196.2 MPa and 11.83 GPa.  The 4-ply had values of 96.48 MPa and 4.906 GPa.  The 4-ply group failed to reinforce a resin control group, which had values of 102.65 MPa and 2.467 GPa.  Conclusion: Incorporation of anisotropic reinforcing fibers has most significant influence on the experimental materials flexural properties.                v  Lay Summary  This work aimed to produce and evaluate experimental 3D braided fiber reinforced composites for dental CAD/CAM applications.  Desirable properties of 3D braided fiber-reinforced composites compared to particulate composites include improved strength, elastic modulus, damage tolerance, fatigue life, and notch insensitivity.   Experimental groups were designed to have differing braid structures and were tested against control groups including an unreinforced resin, and a group with a 3D arrangement of straight (anisotropic) fibers.  A three-point bend test was used to test the samples.  The results demonstrated that a braided structure can reinforce a resin blend.  The most significant contributing factor to reinforcement was the inclusion of anisotropic fibers.    It is recommended that future research into 3D fiber reinforced composite consider the importance of incorporating anisotropic fibers in their design.           vi  Preface  This work was produced as a collaborative effort between Jasmine Zhang and myself, Robert Lesniak.  Jasmine Zhang is a visiting post-graduate research student from Donghua University, Shanghai, China.  Our respective supervisors, Dr. Frank Ko and Dr. Ricardo Carvalho, guided the research providing suggestions and aid as needed.  To the best of the research teams’ knowledge, the proposed experimental is a novel materials for CAD/CAM dentistry.                                  vii  Table of Contents     Abstract .................................................................................................................................iii  Lay Summary……………………………………………………………………………………………………………….………..v Preface………………………………………………………….………………………………………………………………………vi Table of Contents ..................................................................................................................vii List of Tables ...........................................................................................................................x List of Figures .........................................................................................................................xi List of Symbols .....................................................................................................................xiii List of Equations ...................................................................................................................xiv List of Abbreviations .............................................................................................................xv Acknowledgements ..............................................................................................................xvi Dedication ..........................................................................................................................xviii Chapter 1: Introduction ..........................................................................................................1      1.1 CAD/CAM restorative materials…………………………………………………………………………………………1 1.1.1 Ceramics/glass-ceramics………………………………………………………………………………………….2 1.1.2 Resin composites……………………………………………………………………………………………………..4 1.1.3 Alternative CAD/CAM materials……………………………………………………………………………….6 1.1.4 CAD/CAM fiber reinforced composites…………………………………………………………………….7 1.2 Dental fiber reinforced composites…………………………………………………………………………………..8 1.2.1 Production of glass fibers…………………………………………………………………………………………9 1.2.2 Resins for fiber reinforced composites……………………………………………………………..……12 1.2.3 Fiber reinforcement……………………………………………………………………………………………….15 viii  1.2.4 Fiber orientation and length………………………………………………………….……………………….18 1.3 CAD/CAM fiber reinforced composites…………………………………………………………………………...21 Chapter 2: Objectives, experimental plan, rational, hypothesis…………………………….…….……….23  2.1 Objective…………………………………………………………………………………………………………………………23 2.2 Experimental plan………………………………………………………………………………………….………………..23 2.3 Rational…………………………………………………………………………………………………………………………..24 2.4 Null hypothesis……………………………………………………………………………………………………….……….25 Chapter 3: Materials and methods ........................................................................................26 3.1 Material acquisition…………………………………………………………………………………………………………26 3.2 Preform fabrication…………………………………………………………………………………………………………26 3.3 4-Step braiding………………………………………………………………………………………………………………..28 3.4 Resin blend………………………………………………………………………………………………………………………34 3.5 PVS mold fabrication……………………………………………………………………………………………………….35 3.6 Impregnation and polymerization of experimental preforms………………………………………….36 3.7 Impregnation and polymerization of control groups……………………………………………………….37 3.8 Sample preparation…………………………………………………….…………………………………………………..39 3.9 Three-point bend testing………………………………………………………………………………………………...41 3.10 Scanning electron microscopy……………………………………………………………………………………….43 Chapter 4: Results. ................................................................................................................45 4.1 Total sample production………………………………………………………………………………………………….45 4.2 Experimental preform unit cell geometry………………………………………………………………………..45 4.3 Flexural strength……………………………………………………………………………………………………………..46 ix  4.4 Elastic modulus……………………………………………………………………………………………………………….49 4.5 SEM analysis of pre-test cross sections……………………………………………………………………………52 4.6 SEM analysis of tested samples……………………………………………………………………………………….55 Chapter 5: Discussion…………………………………………………………………………………………………………..59 5.1 Comparative analysis……………………………………………………………………………………………………….63 5.2 Future recommendations………………………………………………………………………………………………...64 Chapter 6: Conclusion ...........................................................................................................66  Bibliography .........................................................................................................................67                            x  List of Tables     Table 1.1 Types of glass fibers ……………........................................................................................11 Table 1.2 Composition of E, S, R glass fibers………………………………………………………………………….…12 Table 3.1 4-Step braiding parameters………………………………………………………………………………………27 Table 3.2 Quantity of glass rovings for experimental preforms………………………………………………..27 Table 3.3 Resin blend ratios…………………………….……………………………………………………………………….35 Table 4.1 Total number of samples produced……………………………………………………….…………………45 Table 4.2 Unit cell dimensions, braiding angles and unit cell number per sample……………………46 Table 4.3 Flexural strength statistics…………………………………………….…………………………………………47 Table 4.4 Fischer individual tests for difference of means……………………….………………………………47 Table 4.5 ANOVA groupings……………………………………………….……………………………………………………48 Table 4.6 Elastic modulus statistics……………………………………….…………………………………………………49 Table 4.7 Fischer individual tests for difference of means……………………………………………………….50 Table 4.8 ANOVA groupings…………………………………………………………………………………………………….50              xi  List of Figures    Figure 1.1 Methods to strengthen ceramics……………………………………………………………………….………8 Figure 1.2 Production of glass fibers…………………………………………………………………………………………10 Figure 1.3 Chemical Structure of common FRC monomers………………………………………………………14 Figure 1.4 Silanation of glass fibers………………………………………………………………………………………….16 Figure 1.5 Reinforcing effect of continuous fibers with differing orientations………………………….19 Figure 1.6 Influence of aspect ratio and orientation on flexural strength and elastic modulus…20 Figure 1.7 Relationship between aspect ratio and reinforcing efficiency………………………………….21 Figure 3.1 S-2 Glass roving for experimental groups………………………………………………………………..28 Figure 3.2 4-Step braider………………………………………………………………………………….………………………29 Figure 3.3 4-Step braiding set up……………………………………………………………………………………………..31 Figure 3.4: 4-Step braiding process………………………………………………………………………………………….32 Figure 3.5: Braiding front……………….………………………………………………………………………………………..33 Figure 3.6 Complete 3D 4-step braided preforms…………………………………………………………………….34 Figure 3.7 PVS mold fabrication……………………………………………………………………………………………….36 Figure 3.8 Impregnation and polymerization……………………………………………………………………………37 Figure 3.9 Unidirectional control group fabrication………………………………………………………………….38 Figure 3.10 Preforms unit cell width………………………………………………………………………………………..39 Figure 3.11 Unit cells in 4-step braided preforms…………………………………………………………………….40 Figure 3.12 Sample preparation……………………………………………………………………………………………….41 Figure 3.13 Pretest experimental group samples……………………………………………………………………..41 xii  Figure 3.14 Sample testing……………………………………………………………………………………………………….42 Figure 3.15 Displacement of testing sample from supporting strut………………………………………….43 Figure 3.16 Desk II and Quanta 650……………………………………………………………………………………….…44 Figure 4.1 Unit cell geometry…………………………………………………………………………………………………..46 Figure 4.2 Flexural strength box plot……………………………………………………………………………………….48 Figure 4.3 Elastic modulus box plot……………….………………………………………………………………………..51 Figure 4.4 2-Ply+axial cross sectional SEM……………………………………………………………………………….52 Figure 4.5 2-Ply+axial cross sectional SEM.………………………………………………………………………………53 Figure 4.6 4-Ply cross-sectional SEM………………………………………………………………………………………..53 Figure 4.7 4-Ply cross-sectional SEM………………………………………………………………………………………..54 Figure 4.8 2-Ply cross-sectional SEM………………………………………………………………………………………..55 Figure 4.9 2-Ply cross-sectional SEM………………………………………………………………………………………..55 Figure 4.10 2-Ply+axial tested sample………………………………………………………………………………….…..56 Figure 4.11 4-Ply tested sample……………………………………………………………………………………………….56 Figure 4.12 2-Ply tested sample……………………………………………………………………………………………….57 Figure 4.13 Failure modes of short discontinuous fibers………………………………………………………….57 Figure 4.14 Tested resin control sample……………………………………………………………….………………….58 Figure 4.15 Tested unidirectional control sample…………………………………………………………………….58        xiii  List of Symbols     P = Maximum load L = Span between supports b = Specimen width (mm) t = Specimen thickness (mm) KIC – Fracture toughness  δP/δd – slope of the straight-line portion of the stress strain curve (MPa)  β – Internal braiding angle α – External braiding angle                        xiv  List of Equations     Equation 1 – Flexural strength (MPa)……………………………………………………………………………………….43 Equation 2 – Elastic modulus (GPa)………………………………………………………………………………………….43 Equation 3 – Internal braiding angle (degrees)…………………………………………….…………………………..46                                 xv  List of Abbreviations    ANOVA – Analysis of variance BisGMA – Bisphenol-A-glycidyl methacrylate CAD – Computer aided design CAM – Computer aided manufacturing FRC – Fiber reinforced composite GPa – Giga pascals MMA – Methyl methacrylate  MPa – Mega pascals  PMMA – Poly(methyl methacrylate) PRC – Particulate resin composite PVS – Polyvinylsiloxane  TEGDMA – Triethylene glycol dimethacrylate TERTBPB – Tert-butylperbenzoate Tex – Linear mass density (g/km) 3pb – Three point bend UDMA – Urethane dimethacrylate UHMWPE – Ultra high molecular weight polyethylene       xvi  Acknowledgements   This work has spanned three years and would have not been possible without the help and support from numerous individuals. Beginning with my supervisor Dr. Carvalho.  Your feedback and guidance throughout these past years was invaluable in progressing through this project.  By making yourself available to answer many questions, and review numerous documents and presentations kept this research moving forward.  When an obstacle presented itself, you were available to lend guidance and direct me to appropriate resources.  I thank you for supervising a rookie researcher on his first significant research endeavor.  Dr. Manso, your thoughtful comments and encouragement were instrumental in keeping me going.  Whether during a formal committee meeting, or a brief hallway run in, you were always supportive, understanding, and informative when challenges presented themselves. Dr. Ruse, your guidance and feedback in committee meetings and thesis review were greatly appreciated.  There is no doubt your skill as a researcher and educator left its mark on this work. Dr. Frank Ko and Jasmine Zhang, this work would have not been possible without the two of you.  Your expertise turned our initial idea into action.  Jasmine, I enjoyed the many hours we spent working together.  I thank you for your patience with me in answering my many questions about a discipline that was at the outset, foreign to me. To my family and Kristen.  You were a constant source of love, support and encouragement throughout this work.  I am truly thankful for the support you have given me xvii  throughout my academic pursuits.  Without the foundation you have provided me, all of this would not have been possible.                                      xviii  Dedication    This work is dedicated to Kristen.  These past three years would have not been possible without your love, support, and sacrifice.  I always looked forward to coming home to you after long days of clinical work and research.  Your constant encouragement and willingness to help on top of your busy schedule is greatly appreciated.  I know together, we can accomplish anything we put our minds to.  For all this and much more, I thank you.                 Chapter 1: Introduction  Computer-aided design and computer-aided manufacturing (CAD/CAM) has lead to increasing numbers of biomaterials for dental applications [1].  Applications are expanding and currently include restorative, prosthetic, maxillo-facial, and implant dentistry.  Fueling this material evolution is improving intra and extra oral scanning technology, coupled with improving milling and rapid prototyping technology, each of acceptable accuracy for clinical use [2].   Increasing demands from patients for white, esthetic, metal free dental restorations and prosthetics are propelling dental biomaterials research [3].  CAD/CAM technology opens opportunities to produce metal free, esthetic restorations of acceptable mechanical properties that would not be possible by traditional laboratory procedures [3].   To satisfy patients’ esthetic expectations and to comply with biosafety requirements, clinicians must be abreast of the latest CAD/CAM materials and applications.   This chapter will introduce concepts leading to development of an experimental three-dimensional (3D) fiber-reinforced composite (FRC) for dental CAD/CAM applications.  The chapter is divided into two sections: CAD/CAM restorative materials and glass fiber reinforcement in dentistry.   1.1 CAD/CAM restorative materials Two categories have been described encompassing available CAD/CAM material: ceramics/glass-ceramics and resin composites [4].  These two classes of materials encompass a diverse spectrum of varying mechanical properties and applications.   2  Advantages of ceramics/glass-ceramics over resin composites include greater flexural modulus, flexural strength and hardness, and better optical properties [4],  while brittleness, susceptibility to failure in the presence of flaws, and cost of machining constitute disadvantages [4].  Ceramics low fracture toughness present their most significant hinderance to dental applications.  This allows cracks to propagate readily with catastrophic failure as a consequence [5].  Figure 1.1 displays methods to strengthen ceramics, reducing the influence of this hinderance [6].  Figure 1.1 Methods to strengthen ceramics [6]   1.1.1 Ceramics/glass-ceramics In 1888, Charles Land published the Porcelain Dental Art, bringing feldspathic porcelain into the dental practice [7]. Improving mechanical properties of initial feldspathic porcelain resulted from dispersion strengthening [8].  Incorporating aluminum oxide crystals, and later leucite crystals, up to 35 % of the structural volume, improved mechanical properties of 3  feldspathic porcelain [9–11].  These materials are still in use today but are limited to scenarios of low expected flexural stress. The development of a lithium disilicate glass-ceramic further improved mechanical properties allowing cautious extension into posterior sextants [10].  Lithium disilicate occupies 70% of the volume of these glass-ceramics providing increased resistance to crack propagation [10].  The flexural strength of lithium disilicate is sufficient for anterior and posterior full coverage restoration, in addition to small span fixed dental prosthetics, should certain clinical requirements be satisfied [3]. Slip casting enables production of high volume fraction ceramic core that are infiltrated by glass [10].  Porous cores of spinell (MgAl2O4), alumina (Al2O3), or zirconia reinforced alumina are infiltrated with a molten lanthanum glass resulting in a highly filled glass ceramic.  Mechanical and physical properties vary depending on the core material [10].  Slip casting has been shown to produce materials with less porosity, fewer processing defects, increased strength and fracture toughness compared to feldspathic porcelains [11]. Final esthetics are tailored using feldspathic porcelain over slip casted cores [10]. Polycrystalline ceramics, specifically zirconia, have superior mechanical properties allowing restorations and prosthetics in situation of high mechanical stress.  Wide spread utilization of polycrystalline ceramics is possible due to CAD/CAM technology and industrial manufacturing of consistent, stable zirconia [12].  Clinical use of zirconia requires stabilization of tetragonal phase at room temperature [13].  Yttria is used as a doping agent to stabilize tetragonal zirconia at room temperature, producing yttria-stabilized tetragonal zirconia polycrystalline (Y-TZP).  Without yttria, phase transformation would occur upon cooling forming 4  monoclinic zirconia, which would be accompanied by bulk (3 to 5) % volume increase, which would result in material failure [13].  While yttria leads to the stabilization of the tetragonal structure at room temperature, trapped energy within the crystal structure still exists.  Trapped energy can convert a localized area to monoclinic zirconia, with accompanying local volume increase.  This local volume increase can then compress a progressing crack leading to transformation toughening [10,14].   Today, there is a spectrum of ceramics/glass-ceramic materials that are available for CAD/CAM dentistry.  Available ceramic/glass-ceramic materials allow multiple choices for varying clinical situation, depending on esthetic, strength, arch location, etc.  There is not one universal material that can be utilized in all clinical situations.   1.1.2 Resin composites Particulate resin composites (PRC) have been used in restorative dentistry for over 50 years [15].   They consist of three components: a polymer matrix, silanated reinforcing fillers, and chemicals for promotion and modulation of polymerization [15].  PRC have gone through multiple variations through alterations to particulate filler and matrix.   Filler alterations have involved incorporating smaller particles increasing surface area interacting with the resin matrix, improving wear resistance, polishing, and mechanical properties [16].  Increasing the proportion of filler also reduces the amount of polymerization shrinkage, independent of the matrix constituents [16].  Early PRC filler size ranged from 10-50 µm and were termed macrofill.  Filler size can now be as small as 5-100 nm [16].  Composites containing combinations of smaller and larger fillers are termed hybrid composites [16].   5  Mechanical properties are largely related to volume fraction filler content, with composites having the most filler being stronger, stiffer, and tougher [15].   Alterations to resin matrix continue to focus on methods to reduce polymerization shrinkage and improve adhesion between PRC and tooth structure [15].  Common resin monomers include bisphenol A glycidyl methacrylate (BisGMA), urethane dimethacrylate (UDMA), and triethylene glycol dimethacrylate (TEGDMA), among others [15].  They differ in properties, such as molecular weight, polarity, polymerization shrinkage, and viscosity [16].  In addition, monomers can influence mechanical properties and water sorption. Further research into restorative resin composites aims to address issues that have existed for  over 50 years: mechanical properties, polymerization shrinkage, and adhesion to tooth structure [15].  As awareness of potential harmful effects of PRC is increasing, improving biocompatibility and reducing elution of the chemical components are additional areas of research [16].  Traditional PRC are indicated for anterior and small to medium size posterior restorations [16].  However, their mechanical properties fall short of those needed for large restorations, cuspal replacement, or high stress situation, such as present in parafunctioning patients [15].  Bulk fracture is often cited as a common cause of PRC failure [17]. With CAD/CAM dentistry, industrial processing of pre-cured PRC eliminates concerns over polymerization shrinkage.  In addition, fabrication and polymerization of the composite can be controlled, producing consistent results without concern of contamination or voids that can result from direct PRC applications [18].  Mechanical properties of PRC can also benefit from industrial manufacturing CAD/CAM PRC indirect restorations.  High temperature and 6  pressure have been used to improve the mechanical properties of CAD/CAM PRC blocks [18]. An in vitro study evaluating the mechanical behavior of dental composites, demonstrated the maximum flexural strength was obtained with the hybrid composite Z250  at 160.8 MPa [19].  Experimental PRC manufactured under high temperature, high pressure had flexural strength approaching 200 MPa [18],  an improvement over traditional composite, but a value too low for consideration for large posterior onlays, crowns, and fixed dental prosthesis.  While resin composite material for CAD/CAM applications fall short of the ceramics/glass-ceramics in terms of mechanical properties, they do have advantages such as  more cost-effective machining and potential for direct intra oral repair [4].  Their direct use for small to medium cavities or anterior restorations is advocated and they have shown acceptable clinical results over 5 years [20].  Extending direct PRC into high stress situations results in unacceptable incidence of failure [20].     1.1.3 Alternative CAD/CAM materials A challenge faced in developing new indirect restorative materials is merging a balance between machinability and cost effectiveness of resin composite material, while obtaining similar or superior mechanical properties to ceramics/glass-ceramics.  Fracture toughness becomes of paramount importance in situation of increased mechanical stress, such as posterior regions, fixed dental prosthetics, and implant-supported frameworks [15]. Two novel indirect resin composite CAD/CAM materials include Enamic and Trinia.  Enamic is described as a polymer infiltrated ceramic network [21], while  Trinia is described as a multidirectional interlacing of fiberglass and resin in several layers [22]. 7  Structurally, Enamic consists of a porous feldspar ceramic network infiltrated by a resin matrix via capillary action.  In a mechanical characterization of resin ceramic CAD/CAM materials, Enamic failed to impress with a flexural strength 137 MPa [23].  The elastic modulus was stated to be 22.1 GPa [23].   Limited information is available on Trinia, aside from manufacturer provided data. Trinia is described as a 3D fiber reinforced composite (FRC) manufactured for CAD/CAM applications.  The manufacturer reports mechanical properties as follows: flexural strength 393 MPa, elastic modulus 18.8 GPa, and fracture toughness (KIC) of 9.7 MPa·m1/2 [19].   The flexural strength of 393 MPa brings Trinia into the realm of ceramics/glass-ceramics, possibly expanding its application into posterior regions of the mouth, with potential for use in fixed dental prosthesis and implant-supported frameworks [19].  Independent product characterization and long term clinical trials are required to ascertain its suitability in restorative and prosthetic dentistry.   1.1.4 CAD/CAM fiber reinforced composites An area of current interest lies in indirect machinable FRCs.  Indirect manufacturing can control all aspects of fiber production, orientation, impregnation, and polymerization, theoretically improving mechanical properties of the resultant FRC.  Although lacking external evidence, Shofu’s claimed mechanical properties of Trinia are sufficient for broad application in CAD/CAM restorative and prosthetic dentistry.  CAD/CAM FRCs may bridge an existing discrepancy between mechanical properties of CAD/CAM ceramics/glass-ceramics and resin composites.   8  This introduction of a machinable FRCs also merges fields of textile fabrics for structural composites with machinable CAD/CAM dental materials for clinical applications.  Textile fabrics for structural composites have been used in multiple high impact, high stress situations, including aircrafts manufacturing, sporting equipment, and military applications [25,26].   The following section will review attributes influencing the mechanical properties of dental FRCs.   1.2 Dental fiber reinforced composites  FRCs consist of a polymer matrix and a high aspect ratio fiber reinforcing agent [27].  Their use in dentistry predates the development of the BisGMA monomer by Bowen in 1956 [28,29].  Initial uses involved reinforcing poly(methyl methacrylate) (PMMA) dentures with glass fibers for repair or strengthening [30].  Progressing from initial applications in removable prosthodontics, fiber reinforcement was incorporated into multiple dental disciplines including; restorative dentistry, prosthodontics, implant reconstructions, intra-canal posts, orthodontic retainers, splints (periodontal, traumatic), and facial/cranial prosthesis [30,31]. Multiple fiber reinforcing materials are available for dental application. Carbon, polyaramide, ultra-high molecular weight polyethylene (UHMWPE), and glass are examples of such reinforcing fibers [27,31].  Compared to glass fibers, alternatives have clear disadvantages:  carbon fibers are not esthetic;  polyaramide fibers are challenging to handle and difficult to cut and polish;  UHMWPE fibers suffer from poor matrix adhesion, preventing effective load transfer between the matrix and fibers [31].   9  Due to the fact that glass fibers can bond to matrix through silanization, form composites that have adequate transparency and acceptable flexural properties, they encompass most of reinforcement in dentistry today [27,32].  From this point onward, all reference to fiber reinforcement will imply glass fiber reinforcement, unless otherwise stated.       1.2.1 Production of glass fibers Production of fiber glass is largely accomplished through a direct draw process [33,34].  Components of glass are ground, blended, and mixed into a batch.  Batched components are delivered to a furnace for melting and homogenization, producing a melt.  Following refining, the melt is transferred to fiber forming stations where conversion to fibers is accomplished via  attenuation and rapid cooling upon extrusion from a heated bushing [34].  Fiber diameter is a function of attenuation rate, melt viscosity, bushing temperature, and extrusion pressure [34].  Direct draw glass fiber production can produce glass fibers of diameters between 3-20 µm [33,34].     Extruded fibers are cooled, rinsed and applied with sizing [35].  Sizing is a coating applied to a fiber to serve specific applications [27].  Primary components of a sizing include a film former and coupling agent [35].  Film formers lubricate fibers, protecting from fiber to fiber abrasion and facilitate fiber handling and wetting.  Coupling agents, often silane compounds for glass fibers, serve to improve matrix adhesion and protect the fibers [36]. Continuously sized fibers are gathered into rovings, or yarns, prior to being wound for packaging and distribution.  Chopping of continuous fibers is also possible after sizing, should short span glass fibers be required.  Rovings consist of several fibers gathered without twisting.  10  Yarns consist of several fibers or rovings that are twisted before being wound [33].  Continuous yarns and rovings can be used to produce a variety of preform structures for various applications.  The process of glass fiber production is diagramed in Figure 1.2 [37].   Figure 1.2 Production of glass fibers [37]    Glass fiber development can be tailored by altering composition of the glass to fulfill requirements determined by the purpose of the fibers [36,38].  Various types of glass fibers are available for differing applications.  These fibers are listed in Table 1.1 [39].  11   Table 1.1 Types of glass fibers [38]    Of the glass fibers available, those considered to be alkali free are most suitable for dental application [27].    Glasses of high alkali metal oxide content area susceptible to strength reduction in the presence of water, thus unsuitable for dental application.  E-glass and S-glass fit this requirement and are the most commonly used fibers in dental applications [27].  These glass fibers have less than 1 % alkali metal oxides, Na2O and K2O, and have relative resistance to the presence of water.  R-glass is also suitable for dental applications; however, its use is not common.  Composition of E, R, and S-glass fibers is listed in table 1.2 [27].         12  Component Type of glass  E R S SiO2 53-55 60 62-65 Al2O3 14-16 25 20-25 CaO 20-14 6-9 - MgO 20-14 6-9 10-15 B2O3 6-9 - 0-1.2 K2O <1 0.1 - Na2O <1 0.4 0-1.1 Fe2O3 <1 0.3 0.2 Table 1.2 Composition of E, S, R glass fibers [27]   E-glass composition is calcium-aluminum-borosilicate [38].  Standard E-glass is chemically resistant to water, has good dielectric properties, sufficient strength, and low thermal coefficient of linear expansion [38].  Aside from dental applications, E-glass is used in approximately 90 % of all fiber glass applications.    S-glass composition is magnesium-alumino-silicate glass [38].  S-glass strength and stiffness is 1.3 times and over 3 times that of E-glass, respectively [33].  S-glass has been incorporated into a variety of mechanically demanding situations, including military applications, automotive, and sporting equipment, among others.  High production cost, due to the high temperature requirements, limits the broad use of S-glass [36,38].  The tensile strength of E-glass is listed in the 3100-3800 MPa range and that of S-glass at 4700 MPa [40].     1.2.2 Resins for fiber reinforced composites 13  The general function of the resin matrix in a FRC is to hold the fibers together, transfer stress, and protect the fibers from the outside environment [39].  Two general categories of dental polymers are commonly used:  linear polymers and cross-linked polymers [27]. Linear polymers are also known as thermoplastic polymers.  They involve a monomer with only one reactive vinyl group for chain growth.  Individual linear polymer chains are not covalently bound together, they are held together by weak van der Waals forces [27].  This group of dental resins is generally made up of methyl methacrylate (MMA), which is the dominant monomer in denture bases [27].  Polymerization of MMA is activated by heat or chemically via a tertiary amine.  In either method, benzoyl peroxide breaks down to form free radicals that initiate the polymerization [27].  Typical usage of MMA involves a mixture of pre-polymerized PMMA beads with free monomer.  This formulation has high viscosity, complicating direct impregnation of fibers in a FRC [41]. Cross linked, or thermoset polymers, originate from monomers that have more than one unsaturated bond per molecule and are generally dimethacrylates.  These monomers can form a covalently bonded, 3D cross-linked structure [27].  Dimethacrylates have a variety of uses in dentistry, including adhesives, sealants, cements, restorative materials, and others [42].   Thermoset polymers are generally based on acrylic polyester resins [27].  The presence of ester links renders these polymers susceptible to degradation via hydrolysis. Polymerization is activated by light, heat, or chemically.  Light activation relies on the photosensitizer camphorquinone to produce free radicals to initiate polymerization.  Heat and chemical activation were discussed above.  The mechanical properties of the resultant polymer are dependent on the degree of conversion and the network structure [42].   14  Typical resins systems used in dental FRC include dimethacrylates based on Bis-GMA, TEGDMA, and UDMA [27].   Of the three resin systems, the maximal flexural strength of a homopolymer is found with UDMA at 133.8 MPa [42].  The molecular structures of Bis-GMA, TEGDMA, and UDMA are shown in Figure 1.2 [42].   Figure 1.3 Chemical Structure of common FRC monomers  [42]   Homopolymers are generally not used in dentistry, as mixtures of various resins are tailored for specific applications.  Bis-GMA is often used as a base monomer, combined with others to customize properties.  By itself, bis-GMA has high intrinsic reactivity and resistant to degradation by water [27,42,43].   However, the aromatic ring and hydroxyl groups hinder the mobility of the monomers and increase the viscosity of the resin [44].  This results in undesirable properties, including a decrease in degree of conversion, high viscosity, and poor handling [27,42].  Combining bis-GMA with lower viscosity monomers, such as TEGDMA and 15  UDMA, improve the handling properties of the resin system [42,44].  This can influence the achievable volume fraction filler and impregnation of reinforcing fibers [27,42].     1.2.3 Fiber reinforcement The interaction between the two components, high aspect ratio fibers and resin matrix, produces effective load transfer between the resin matrix and reinforcing fibers.  Integral to this load transfer is both impregnation of the fibers by the resin matrix and adhesion between the two components.  Impregnation is dependent on the space between fibers, viscosity of the resin, and wetting properties of the fibers [27].  Adhesion can be accomplished by a sizing procedure that applies a silane coupling agent to the fibers during production [27,31].   Space between fibers and viscosity of resin interplay to influence the quality of fiber impregnation.  In applications using monomers with PMMA beads or particles, the viscosity and size of the fillers impose limitations on fiber spacing [27].   The space between fibers cannot be smaller than the beads/particles mixed into the resins [27].  In addition, the higher viscosity imparted on the resin due to the particles also reduces ease of impregnation.  Pre-impregnated fibers for direct and indirect use have also been developed to overcome these challenges of impregnation.   Pre-impregnated fibers allow near complete impregnation of fibers and predictable bonding to a veneering composite [31].   Using dimethacrylate monomers, the problem of powder beads or fillers of specific dimensions is negated.  Less space is required between fibers, allowing more complex fiber arrangements at higher volume fractions [27].  The resin matrix can be tailored to a viscosity 16  suitable for impregnation and clinical handling by varying composition of the resin monomers [27]. Critical to load transfer between the reinforcing fibers and the matrix is adhesion between the components [27].  Adhesion can be accomplished through a silane coupling agent applied during production of the fibers [31,33].  Silane coupling agents are synthetic bifunctional compounds capable of interacting with the glass surface via a silanol group and with the methacrylic groups of the resin monomers via a methacrylic group. [27]. These interactions produce reliable covalent adhesion between fibers and matrix.  In addition to covalent adhesion between the matrix and reinforcing agent, silanation improves wettability of the matrix, facilitating impregnation of the fibers [27].  The chemical reaction of glass fiber silanation is displayed in Figure 1.3 [27].   Figure 1.4 Silanation of glass fibers [27]   17  The fiber volume fraction can influence the ability to effectively impregnate and adhere the two components.  By decreasing fiber diameter, the number of fibers that can be included in a FRC can be increased [27].  At fiber diameters of 13-15um, a common diameter used in dental FRC, a 68% volume fraction produces maximal flexural properties of a FRC specimen [27].  Beyond a 68% volume fraction, incorporating additional fibers results in deficient adhesion and impregnation of the fibers, negatively affecting the flexural properties [31].  Recent in vitro research is focused on reducing fiber dimensions to nanometer size to further increase surface area without sacrificing appropriate impregnation and adhesion.  Initial results show promise [27,45].   Areas of incomplete impregnation and adhesion result in voids within the FRC, which are areas of internal weakness and are susceptible to water sorption and discoloration [31].  These voids also house oxygen, which can locally inhibit polymerization of the resin matrix.  Should adhesion be inadequate, effective load transfer from the matrix to reinforcing fibers cannot occur.  In either scenario, or in combination, a significant reduction in the theoretical flexural properties can be expected.  Should impregnation and adhesion be acceptable, flexural properties of a FRC should lie between that of the resin, and reinforcing fibers [27].  Moisture and water sorption have potential to affect each of the above-mentioned components and these effects are influenced by the quality of impregnation and adhesion.  Moisture can induce corrosive/dissolutive effects to FRC through action on the surface and body via diffusion into the FRC structure [30,31].  Moisture alters glass composition through hydrolysis of alkali earth metals, leaching ions from fibers, degrading matrix adhesion and fiber structural integrity.  It also affects the resin matrix by inducing a plasticizing effect through 18  hydrolysis of the ester groups in the polyester monomers [31].  Composition and sizing of fibers along with matrix composition determine susceptibility to moisture [31].  Compositions consisting of E or S-glass, and those based on a Bis-GMA monomer system are more resistant to the effects of moisture for reasons reviewed above. The quality of impregnation and adhesion are also important when discussing the effects of moisture on the FRC.  Inadequate impregnation and adhesion results in voids and tracks for moisture diffusion, increasing potential degradative effects of water sorption.  Good adhesion via silane coupling agent increases hydrolytic stability; however, there is potential for rehydrolysis of the silane coupling agent [31].  Negative effects of moisture on polymers are well known and will continue to be a limitation of resin composite materials [30,31].     1.2.4 Fiber orientation and length Unlike particulate composites that are purely isotropic, high aspect ratio fillers, in the form of continuous or discontinuous fibers, can impart anisotropic properties resulting in significant improvement in flexural properties [46].   Krenchel’s factor predicts the influence of continuous fiber orientation on the strength of a FRC [27,46].  Keeping volume fraction the same in all instances discussed, fibers aligned parallel to tensile stress will produce maximal reinforcing effect.  Altering this alignment to a configuration perpendicular to tensile stress produces no reinforcement.  Incorporating a bidirectional weave both perpendicular and parallel to tensile stress results in a reinforcing effect in two directions at 50 % of a completely ansiotropic specimen.  An oblique orientation of a bidirectional weave is expected to reduce reinforcement to 25 % of that of a completely 19  anisotropic orientation.  Finally, a random orientation of fibers produces to lowest expected reinforcement as an isotropic specimen at 20 %.  Figure 1.3 displays Krenchel’s factor [27].   Figure 1.5 Reinforcing effect of continuous fibers with differing orientations [27]   Short discontinuous fibers can also be aligned to impart anisotropic flexural properties to a FRC.  Their reduced aspect ratio results in reduced flexural properties compared to an anisotropic continuous FRC [46].  Continuous fibers have a greater surface area of interaction and adhesion to the matrix, accounting for their superior flexural properties [27].  Altering the orientation to random produces an isotropic FRC with a further reduction in flexural properties.  Figure 1.6 displays the influence of aspect ratio between anisotropic continuous vs short discontinuous, and random short discontinuous fibers [46]. 20    Figure 1.6 Influence of aspect ratio and orientation on flexural strength and elastic modulus [46]   Short discontinuous fibers can be considered high aspect ratio fillers, should they exceed a critical fiber length [46].    This length represents a minimum length that will cause a fiber to fail at its midpoint in a FRC, as opposed to interfacial fracture between the fiber and matrix [27,46].  It is suggested that the critical fiber length is approximately 50 times the fiber diameter.  As fiber length increases, the properties are expected to approach those of continuous fibers [27,47].  Failure to exceed this critical length results in isotropic reinforcement, regardless of orientation. In addition to expected differences in flexural properties, continuous and discontinuous FRC differ in their modes of failure.  Continuous FRC have four modes of failure: axial tensile failure, transverse tensile failure, shear failure, and buckling failure.  Discontinuous FRC have three modes of failure: crack propagation through the matrix, debonding of the fibers, and fiber fracture [27].  The type of failure can be influenced by the fibers length, aspect ratio, volume 21  fraction, and interfacial adhesion between the fibers and matrix.  The relationship between aspect ratio and reinforcing efficiency is presented in Figure 1.6 [27].   Figure 1.7 Relationship between aspect ratio and reinforcing efficiency [27]   The use of discontinuous fibers is required when there is no space to incorporate higher aspect ratio continuous fibers, as in the case of restorations and core build ups.  Continuous fibers, either unidirectional or of various textile architecture, are more suited to clinical situations, including intra canal posts, periodontal splinting, denture repair and fixed dental prosthesis [27].     1.3 CAD/CAM fiber reinforced composites CAD/CAM dentistry has expanded the spectrum of available dental materials.  Just as the spectrum of ceramics/glass-ceramics and resin composites have taken advantage of this technology, FRC materials can be optimized through controlled manufacturing and CAD/CAM dentistry. 22  This area of CAD/CAM FRC is emerging with the addition of TRINIA to the commercial market.  At the time of this study, only the manufacturer’s claimed data of this commercially available product was available.  The flexural properties of Trinia were discussed previously.  The manufacturer claims benefits that include: lightweight, durable and resilient, no firing requirements, high flexural and compressive strength, biocompatibility, and adjustability [19].  Its potential uses include posterior and anterior crowns and bridges, substructure, and telescopic restorations [19].  With these flexural properties and potential applications, 3D FRC has potential to be a viable material for clinical dentistry. One additional experimental 3D FRC has been recently reported in 2016 [48].  The test specimen was a 3D woven FRC.  Flexural testing yielded a flexural strength of 575.7 MPa, fracture toughness of 11.26 MPa·m1/2, and an elastic modulus of 24.6 GPa [48].  In terms of flexural properties, this experimental 3D FRC surpasses Trinia and many of the currently available ceramic materials. 3D FRC appear to be a potential CAD/CAM material that can bridge or eliminate a gap between the mechanical properties of CAD/CAM glass-ceramics and resin-based materials.  Further research and development, including long term clinical trials are required to determine the feasibility and durability of these CAD/CAM FRC dental materials.     23  Chapter 2: Objective, Experimental Plan, Rational, Hypothesis   2.1 Objective Working in collaboration with the Department of Materials Engineering, specifically with Dr. Frank Ko, and Jasmine Zhang, the following experimental objective was devised:  to engineer and produce a 3D 4-step braided FRC with superior properties to similar materials available for CAD/CAM dentistry, specifically Trinia.   2.2 Experimental plan As stated above, the experimental plan was to design and produce 3D braided FRC by 4-step braiding and evaluate the resultant flexural properties, specifically flexural strength and elastic modulus.  Test specimens were to approximate the dimensions of available CAD/CAM restorative blocks and be tested in accordance with ISO 10477.  Experimental groups were designed by the Materials Engineering team and fabricated as a collaborative effort.  Three experimental groups were proposed by Materials Engineering, with the following parameters: 50 % volume fraction fibers, 14mm*14mm cross section, and 45° braiding angle.  The experimental groups were:  1. Double ply of glass fibers rovings with unidirectional glass fibers (2-ply+axial) 2. Double ply of glass fibers rovings (2-ply) 3. Quadruple ply of glass fibers rovings (4-ply)  24  Evaluation of the experimental FRC was to discern effects of unit cell size and unidirectional glass fiber inclusion on the resultant flexural properties.   The resin blend selected was a 50:50 blend of UDMA and TEGDMA, with a thermal polymerization system.  This blend was deemed by Materials Engineering to be of suitable viscosity to adequately impregnate the designed preforms at a 50 % volume fraction.  It is consistent with dental resins currently available. Two control groups were developed: a 50:50 TEGDMA:UDMA resin blend with no reinforcing fibers and resin reinforced with a 3D arrangement of unidirectional glass fibers at a 50 % volume fraction.   2.3 Rationale  3D braided composites hold key structural advantages over two dimensional preforms.  These include delamination resistance, through thickness reinforcement and improved damage tolerance [49,50].  3D structural composites offer several advantages over materials such as metals or ceramics including lower densities, availability, and low-cost fabrication [26,49].  Advantageous aspects of these fabrics can be extrapolated from engineering and military avenues and applied to dental biomaterials in multiple areas of dentistry [26].   The use of 3D braided preform as fiber reinforcement offers a unique approach to a fiber reinforced composite for dental CAD/CAM use.  The braid can distribute load throughout the structure, has a mechanism to arrest micro cracking, offers an improved strength to weight ratio over woven structures, and reduces potential for delamination [49–51].   25  Its potential for high strength, high modulus, high damage tolerance, and fatigue resistance can be of value for dental applications including: prosthesis copings, implant prosthetic frameworks, intra canal posts, among others.  To the best of the team’s knowledge, a 3D braided FRC is a novel approach to a FRC for CAD/CAM dentistry.    2.4 Null Hypotheses Four null hypotheses were tested:  1. The addition of 3D braided glass fibers to the resin blend will have no effect on the flexural properties of the resultant resin blend. 2. At equal volume fractions, 3D braided FRC will not differ in flexural properties from a completely anisotropic FRC. 3. Altering unit cell size of the 3D braided glass fibers performs will have no effect on the flexural properties of the resulting FRC. 4. Inclusion of unidirectional fibers to 2-ply 3D braided preforms will have no effect on the flexural properties of the resulting braided FRC.           26  Chapter 3: Materials and Methods   3.1 Material Acquisition The experimental design required acquisition of resin monomers with a thermal curing agent, and suitable glass fibers for structural composites.   Resin monomers, UDMA and TEGDMA, were provided by Bisco (Schaumburg, Il USA).    The thermal curing agent provided was tert-butylperbenzoate (TERTBPB). AGY America (Aiken, SC USA) generously donated 14 kilograms of 449 S-2 Glass® Roving.  The product included numerous 9-micron glass filaments, pre-treated with an amino-silane compatible with both epoxy and selected urethane resins.  AGY describes the 449 S-2 Glass® Roving as having a combination of strength, impact resistance, stiffness, radar transparency, and fatigue resistance.  Examples of product applications include helicopter blades, pressure vessels, aircraft fuel tanks, golf shafts, and aircraft flooring [52].  3.2 Preform Fabrication The department of Materials Engineering designed three experimental preforms, anticipating to obtaining flexural properties superior to those of Trinia. The initial step of design involved determining the quantity of S-2 glass rovings required for each preform.  This was determined by Material Engineering based on established experimental parameters, S-2 glass roving characteristics, and machine bed dimensions.  These parameters are summarized in Table 3.1.    27  Variable Value Volume Fraction 50% Cross Section 14mm*14mm = 196mm2  Braiding Angle 45o Tex 203 g/km Density 2.46 g/cm3 Machine Bed Columns 14 Machine Bed Rows 14 Table 3.1 4-Step braiding parameters   The number of required rovings for each experimental group is listed in Table 3.2.  It should be noted that the 2-ply experimental group will significantly deviate from the proposed cross-sectional area due to limitations in the dimensions of the machine bed.  Figure 3.1 demonstrates roving bundling for the experimental groups.  Experimental Groups Number of Rovings 1. 2-ply braiding + 2-ply axial 224 braiding + 196 bundled axial rovings 2. 4-ply 224 bundled braiding rovings 3. 2-ply 224 bundled braiding  Table 3.2 Quantity of glass rovings for experimental preforms    28   Figure 3.1 S-2 Glass roving for experimental groups   The number of rovings for the unidirectional control group was determined from the above parameter to produce 50 % volume fraction at a 14 mm*14 mm cross sectional area.  1400 rovings were required for the unidirectional control group.   3.3 4-Step braiding  4-Step braiding requires extending fibers from a beam extended horizontally, centered above a machine bed.  The machine bed is organized in columns and rows of mobile braiding carriers and immobile fasteners for axial fibers running along the columns.   The horizontal beam is located at a variable distance above, centered over the machine bed, based on the desired length of the preform and limited by the construction of the apparatus itself.   The braiding front will originate at the horizontal beam and progress towards the machine bed as the braiding cycles are completed.  Figure 3.2 displays the braiding apparatus to produce the three experimental preforms.  29   Figure 3.2 4-Step braider   Rovings were cut approximating 1.25 times the distance from the machine bed to the braiding front.  The rovings were then quadrupled forming 112 4-plied (4-ply experimental group) or doubled into 112 2-plied (2-ply and 2-ply+axial experimental groups) bundles, as needed for each of the experimental groups.    An elastic Spandex loop, approximating 0.50 times the distance to braiding front unstretched, was knotted to each end of the plied rovings.  The resultant rovings were looped through the horizontal beam, allowing each plied roving to serve as two total plied rovings.  By looping each bundled roving through the horizontal beam resulted in the 112-plied roving to produce 224 braiding rovings.  Spandex loops were fasted to braiding columns and rows.  In the 2-ply+axial group, 98 2-plied rovings were also included, looped through the horizontal beam and fastened to 196 axial carriers. 30  Fastening the rovings to the carriers via Spandex was accomplished to not intertwine the rovings from the horizontal beam down to the machine bed.  To accomplish this, each half of the rovings was fastened to its respective half on the machine bed, at their respective equivalent braiding front or axial attachment.  The process was repeated until all braiding fronts were used, and the axial fasteners in the case of the 2-ply+axial group.    The Spandex held the rovings in tension during braiding.  This set up protocol was used for the three experimental groups.                31    Figure 3.3 4-Step braiding set up   4-Step braiding relies on 4 sequential steps to produce the basic unit cell.  The general procedure is outlined in multiple publications referenced in this work [26,50,51].  The first step involves moving carries in alternate rows along the X-axis one position.  The second step involves movement of alternate carries along the Y axis.   Steps 3 and 4 are the reverse motions of steps 1 and 2.    Following a four step cycles, the braid carriers should have ended in their 32  original pattern on the machine head.  Each 4-step cycle of braiding produces single unit cell that will elongate along the Z axis from the braiding front.  This is depicted in figure 3.4 [53], and an in production braiding front in figure 3.5.   Figure 3.4: 4-Step braiding process [53]   33   Figure 3.5: Braiding front    Following each 4-step cycle, a jamming procedure was carried out to pack the rovings at the resulting braiding front.  A ¼ inch metal dowel was used for the jamming step at 3-4 random points distributed along the X-axis or Y-axis.   Jamming at either the X or Y-axis was alternated after each 4-step process.  The jamming dowel was inserted in a manner to not intertwine the rovings, but directly through two planes of glass fibers.  While subject to some variation, consolidation of the unit cells was attempted to be done with equal amounts of pressure at each jamming procedure.  Jamming pressure has implications in the preforms unit cell width, density, and braiding angle [49].  To lubricate the glass fibers, to prevent glass to glass abrasion, a light misting of water was sprayed at the braid front every 3-5 4-step cycles. Braiding and jamming were conducted until the lengths of usable rovings had been exhausted.  It is noted that structural variations do exist due to the manual fabrication process, 34  alterations in jamming force, and increasing tension applied by Spandex attachments as the preform was developed.  To secure and prevent braid relaxation, the braided masking tape was used at both the braid front and at the origin at the horizontal beam.  The preform was then removed from the 4-Step braiding machine for impregnation, sample preparation and three-point bend testing.  The resultant preforms are displayed in Figure 3.6.   Figure 3.6 Complete 3D 4-step braided preforms   3.4 Resin blend  The experimental resin matrix, 50:50 wt% blend of UDMA and TEGDMA, was mixed with 3 % TERTBPB as the thermal curing agent (Table 3.3).  This ratio was suggested by the supplier, Bisco.  Measurements were made using an analytical balance.  Mixing was carried out at room temperature manually and then using magnetic stirring overnight.  The blend was degassed under vacuum for 24 h.  The resulting resin blend was stored in the fridge.  Prior to use, the 35  blend was warmed to room temperature, hand mixed and degassed for 1-2 h.  Multiple 200 mL batches of the resin blend were produced as required throughout the experiment.   PROPORTION COMPONENTS WEIGHT (%) 50:50 UDMA 48.5% TEGDMA 48.5% TERTBPB 3% Table 3.3 Resin blend ratios   3.5 PVS mold fabrication  Control group molds were fabricated using polyvinyl siloxane (PVS) putty wash technique with Aquasil (DENTSPLY Caulk, Millford, DE) putty and light body impression material.  The positive replica for the mold was 3D printed using the FLASHFORG NEW Creator Pro Dual Extrusion 3D Printer (FLASHFORGE, CA, USA).  The dimensions of the positive replica were 14 mm*14 mm*170 mm.  Molds were flattened against the counter top to produce a level surface. The same protocol and materials were used to fabricate molds for the experimental preforms.  Dimensions of each preform were measured and the positive 3D printed replicas were designed to be 2mm larger than the preform, in all dimensions.  This increased dimension was to allow complete submersion into the resin blend during the impregnation process discussed below.  Examples of the PVS molds are displayed in Figure 3.7.  36   Figure 3.7 PVS mold fabrication   3.6 Impregnation and polymerization of experimental preforms Continuous sections of the experimental preforms were bound with heat resistant Kapton® Tape (Milton ON, Canada) at approximate lengths of 16 cm.  The three experimental preforms were submerged in the prepared resin blend, at room temperature, in a clean Tuperware® container of size sufficient to accommodate without stressing the preforms.  The container was then placed under vacuum for 48 h.  At two instances, the preforms were removed from the vacuum and rotated 90° around their long axis to further remove entrapped air from the preforms.  Following the 48-h period, the impregnated preforms were transferred to their respective PVS molds.  Additional degassed resin was added to completely submerge the preforms.  The molds were then placed under vacuum for 24 h.  Without removing from the vacuum chamber, the temperature of the vacuum chamber was increased to 120 °C  for 15 min.  The experimental fiber reinforced composites were removed from the vacuum oven, allowed to 37  cool to room temperature, removed from the mold, and stored dry.  Figure 3.8 displays the material at different stages of impregnation and polymerization.    Figure 3.8 Impregnation and polymerization   3.7 Impregnation and polymerization of control groups A cluster of 1400 roving yarns, 15 cm in length, were cut from the roving spool.  This cluster was maintained manually in a unidirectional orientation and was submerged completely in the resin blend at room temperature and placed under vacuum for 48 h.  Gentle manipulation of the glass fiber was carried out periodically to remove entrapped air inclusions.  Care was taken to maintain the unidirectional orientation.  Following 48 h, the glass fibers were transferred to the prepared unidirectional mold and resin was further added to submerge the glass fibers.  An additional 24 h under vacuum was carried out.  Due to the lack of structural organization and dimensional maintenance, the unidirectional fibers were compressed into the mould using a ¼ inch glass slab and binding clips.  This step confined the unidirectional fibers to 38  the prepared mold, a necessary step to maintain the 50 % volume fraction calculated to the molds dimensions.  The apparatus was then placed in the vacuum chamber and heated at 120 °C for 25 min.  The additional 10 min heating time was added arbitrarily due to the ¼ inch glass covering of the specimen.  It was sufficient for thermal curing of the sample.  The sample was removed from the vacuum chamber, allowed to cool, removed from the mold, then stored dry at room temperature.  Figure 3.9 displays the unidirectional control group at various stages of production.     Figure 3.9 Unidirectional control group fabrication   Using the prepared PVS mold for the resin group,  resin blend was added to the mold at room temperature.  The mold was placed under vacuum and degassed for 48 h.  Without removal from the vacuum, the oven was heated to 120 °C for 15 min.  The resin beam was removed from the oven, allowed to cool, and stored dry at room temperature.    39   3.8 Sample preparation  Samples from the experimental groups and control groups were prepared in the same fashion.  Prior to sample preparation, a decision was made to deviate from ISO 10477 standard for three-point bend testing (TPBT).   ISO 10477 standard call for samples to be prepared to 2 mm*2 mm*20 mm for TPBT, the key ratio between thickness and span being 1:10. Based on the size of the plied rovings and experimental groups, the ISO recommended size of the testing beams was deemed too small to account for alterations in unit cell sizes between the experimental groups.  It was recommended to incorporate at least 4-unit cells per specimen cross sectional area.  Figure 3.10 displays a potential 2 mm*2 mm*20 mm resin specimen aligned to the experimental preforms.   The beam size was altered to 4 mm*4 mm*45 mm, maintaining the 1:10 ratio.  This increase is sufficient to attain a minimum of 4 cross-sectional unit cells based on the measured width of the experimental preform unit cell.  An additional 5 mm was added to the length of the samples to facilitate TPBT.  Figure 3.10 Preforms unit cell width  40  The samples were cut and prepared using a low speed precision cutting machine (SYJ-150, Low Speed Diamond Saw, MTI Corporation, USA).  3D braided FRC have three different unit cell geometries: corner, surface, and interior [54].  These are displayed in Figure 3.11.  Based on the size of the sample specimen, these unit cell variations have potential to influence the flexural properties of the testing samples [54].  To prepare uniform samples of the braided experimental groups and to negate the potential influence on flexural properties, the surface and corner unit cells of the structural composites were cut up to 1 mm within the structural composite.  The result also produced a square cross section of the FRC that facilitated the preparation of 4 mm*4 mm*45 mm beams for testing.  While the removal of surface and corner unit cells does not apply to the control specimens, the procedure was still followed.  The number of beams per group were limited by the amount of available FRC produced.  All samples were polished using fine grit sandpaper and stored dry at room temperature.  Figure 3.12 illustrates this sample preparation procedure.  Figure 3.13 displays the experimental group samples produced.   Figure 3.11 Unit cells in 4-step braided preforms [54]   41   Figure 3.12 Sample preparation    Figure 3.13 Pretest experimental group samples   3.9 Three-point bend testing A Instron 5969 dual column, Material Testing System (Instron USA, Norwood MA, USA) with a 2 kN load cell was used to test all specimens.  Prior to testing each sample, they were measured exactly in the center using digital calipers and the width and thickness were input into the testing software.  In addition, the samples were marked for directional orientation to 42  discern the compression side from the tension, should this not be readily discernable following three-point bend testing.  The 45 mm samples were placed on manufacturer provided supporting struts with the center points set to 40 mm apart, Figure 3.14.   The samples were centered under the 10 mm upper anvil pin with the marked compression side facing up.  A smaller upper anvil was not available.  The central force applied via the upper anvil moved at a set rate of 1 mm/min until fracture occurred or until the test parameters were exceeded.  If sample fracture did not occur, testing was terminated by the operator.  Instances requiring operator termination included: time exceeding 15 minutes, excessive vertical displacement of the specimen from the supporting struts, and machine error.  Due to machine errors, the total sample number produced did not equal the total sample number evaluated.  Displacement of the sample from the supporting strut is shown in Figure 3.15.   Figure 3.14 Sample testing  43     Figure 3.15 Displacement of testing sample from supporting strut   Bluehill 2 Software (Instron USA, Norwood, MA, USA) was used to calculate the results for flexural strength and elastic modulus based on the following equations.  Equation 1: Flexural strength    Equation 2: Elastic Modulus   3.10 Scanning electron microscopy Cross sections of each group and representative samples near the means of each group tested were examined by scanning electron microscopy (SEM).  Cross sections were prepared 44  using a low speed precision cutting machine.  Cross sections were then polished using fine grit sand paper, followed by a polishing cloth.  Specimens were mounted on a carrier plate, gold coated using Denton Vacuum Desk II (Denton Vacuums, Moorestown, NJ USA) and examined using a Quanta 650 (ThermoFisher Scientific, Hillsboro, OR USA) SEM.  SEM was used to evaluated cross sections for quality of infiltration, fiber wetting, and inclusion of voids or impurities.  The tested samples were evaluated to discern potential modes of failure.   Figure 3.16 Desk II and Quanta 650                45  Chapter 4: Results   4.1 Total sample production The total number of samples prepared for the three experimental groups and two control groups are listed in Table 4.1.  The total samples produced does not equal the total samples tested (Table 4.3) for reasons described in Chapter 3.  Group Total Sample Number Resin control 13 Unidirectional control 9 Experimental group 1: 2ply + axial 10 Experimental group 2: 2ply 9 Experimental group 3: 4ply 12 Total samples produced 53 Table 4.1 Total number of samples produced   4.2 Experimental preform unit cell geometry Experimental preform unit cell dimensions and braiding angles were measured on each preform at multiple locations.  Average unit cell dimensions are displayed in Figure 3.10 above. The number of unit cells per sample cross sectional area of a 4 mm*4 mm=16 mm was determined.  Minor deviations from 4 mm*4 mm existed for each individual sample, resulting from sample preparation and polishing.  These variations were assumed to be consistent across all groups.  The external and internal braiding angles of the experimental preforms are also listed.  The external angle was measured at multiple points on each preform and averaged for each experimental group.  The internal braiding angle was calculated based on the equation 46  listed below, derived by Materials Engineering in accordance with the proposed unit cell geometry displayed in Figure 4.1.    This data is listed in Table 4.2.  ( )arctan 2 tanβ α=  Equation 3: Internal braiding angle     Figure 4.1 Unit cell geometry    Preform Unit Cell Width (mm) Unit Cell Area (mm2) Unit Cell Length (mm) Unit Cell Volume (mm3) External Braiding Angle Internal braiding Angle (Unit cells) / (sample cross section) 2Ply + Axial 1.80 3.24 3.00 9.72 30o 39.2o 4.94 4Ply 1.82 3.31 2.63 8.71 34.7o 44.4o 4.83 2Ply 1.33 1.73 2.50 4.42 28o 37.0o 6.4 Table 4.2 Unit cell dimensions, braiding angles and unit cell number per sample   4.3 Flexural strength  47  The mean, standard deviation, and ANOVA groupings for flexural strength are listed in Tables 4.3 through 4.5.  Figure 4.2 displays this information graphically.  All statistics for both flexural strength and elastic modulus were calculated using data analysis software Minitab (Minitab Inc., State College PA).  Group N Mean (MPa) St. Deviation Unidirectional 9 336.6 54.7 2Ply + Axial 9 236.52 27.1 2Ply  9 196.2 36.1 Resin 10 102.65 23.67 4Ply 9 96.48 14.93 Total samples tested 46   Table 4.3 Flexural strength statistics   Difference of Levels Difference of Means SE of Difference 95% CI T-Value Adjusted P-Value C2 - C1 234.0 15.6 (202.6, 265.4) 15.03 0.000 E1 - C1 133.9 15.6 (102.4, 165.3) 8.60 0.000 E2 - C1 93.5 15.6 (62.1, 125.0) 6.01 0.000 E3 - C1 -6.2 15.6 (-37.6, 25.3) -0.40 0.694 E1 - C2 -100.1 16.0 (-132.4, -67.9) -6.27 0.000 E2 - C2 -140.4 16.0 (-172.7, -108.2) -8.80 0.000 E3 - C2 -240.2 16.0 (-272.4, -207.9) -15.04 0.000 E2 - E1 -40.3 16.0 (-72.6, -8.1) -2.53 0.016 E3 - E1 -140.0 16.0 (-172.3, -107.8) -8.77 0.000 E3 - E2 -99.7 16.0 (-132.0, -67.5) -6.24 0.000 C1 – Resin, C2 – Unidirectional, E1 – 2Ply + Axial, E2 – 2Ply, E3 – 4Ply Table 4.4 Fischer individual tests for difference of means   48  Group N Mean (MPa) Grouping Unidirectional 9 336.6 A 2Ply + Axial 9 236.52 B 2Ply  9 196.2 C Resin 10 102.65 D 4Ply 9 96.48 D Table 4.5 ANOVA groupings    C1 – Resin, C2 – Unidirectional, E1 – 2Ply + Axial, E2 – 2Ply, E3 – 4Ply Figure 4.2 Flexural strength box plot  From the descriptive statistics and ANOVA analysis of flexural strength, the following conclusions were drawn:  • Unidirectional reinforcement produced the greatest flexural strength with a significant difference from all other groups. • Incorporating axial fibers into a braided structure (2-ply+axial) produced a significant difference in flexural strength compared to the two other experimental groups, and resin control group. 49  • The 2-ply group had a flexural strength significantly superior to the 4-ply group, and resin control group. • The 4-ply experimental group failed to produce a significant reinforcing effect on the resin control group.  It had the lowest flexural strength of all groups tested, not significantly different from the resin group without fibres.   4.4 Elastic modulus The mean, standard deviation, and ANOVA groupings for elastic modulus are listed in Tables 4.6 through 4.8.  Figure 4.3 displays this information graphically.    Group N Mean (GPa) Std. Deviation Unidirectional 9 37.30 0.612 2Ply + Axial 9 20.75 6.14 2Ply  9 11.83 3.37 Resin 9 2.467 4.01 4Ply 10 4.906 1.269 Table 4.6 Elastic modulus statistics                50  Difference of Levels Difference of Means SE of Difference 95% CI T-Value Adjusted P-Value C2 - C1 34.84 1.66 (31.47, 38.20) 20.93 0.000 E1 - C1 18.28 1.66 (14.92, 21.64) 10.99 0.000 E2 - C1 9.36 1.66 (6.00, 12.72) 5.63 0.000 E3 - C1 2.44 1.66 (-0.92, 5.80) 1.47 0.150 E1 - C2 -16.55 1.71 (-20.00, -13.10) -9.69 0.000 E2 - C2 -25.47 1.71 (-28.92, -22.02) -14.92 0.000 E3 - C2 -32.40 1.71 (-35.85, -28.95) -18.97 0.000 E2 - E1 -8.92 1.71 (-12.37, -5.47) -5.22 0.000 E3 - E1 -15.84 1.71 (-19.29, -12.40) -9.28 0.000 E3 - E2 -6.93 1.71 (-10.37, -3.48) -4.06 0.000 C1 – Resin, C2 – Unidirectional, E1 – 2Ply + Axial, E2 – 2Ply, E3 – 4Ply Table 4.7 Fischer individual tests for difference of means  Group N Mean (GPa) Grouping Unidirectional 9 37.30 A 2Ply + Axial 9 20.75 B 2Ply  9 11.83 C Resin 9 2.467 D 4Ply 10 4.906 D Table 4.8 ANOVA groupings  51   C1 – Resin, C2 – Unidirectional, E1 – 2Ply + Axial, E2 – 2Ply, E3 – 4Ply Figure 4.3 Elastic modulus box blot  From the descriptive statistics and ANOVA analysis of elastic modulus, the following conclusions were drawn:  • Unidirectional reinforcement produced the greatest elastic modulus and was significantly different from all other groups. • Incorporating axial fibers into a braided structure (2-ply+axial) produced a significant difference in elastic compared to the two other experimental groups, and resin control group. • The 2-ply experimental group resulted in a significant increase in elastic modulus compared to the resin control, and 4-ply experimental group. • The 4-ply experimental group failed to produce a significant increase on the resin control group elastic modulus.  It had the lowest elastic modulus of all groups tested, not significantly different from the resin group without fibers.   52   4.5 SEM analysis of pre-test cross sections Pre-test cross sections of the braided preforms were evaluated by SEM to characterize the quality of the 4-step braiding procedure.  The resin control group was not analyzed by SEM in cross section.  2-Ply+axial cross-section is shown in Figures 4.4 and 4.5 and it clearly displays axial and braiding rovings.  Axial rovings appear circular or triangular.  Braiding rovings have a varied elliptical appearance.  Multiple pre-test fracture lines were noted sectioning through the braiding rovings.  Scattered voids and inclusions can be appreciated in cross section.  Adequate wetting of the fibers can be appreciated in Figure 4.5.  No space can be visualized between the fiber and the matrix at this magnification.   Figure 4.4 2-Ply+axial cross sectional SEM   53   Figure 4.5 2-Ply+axial cross sectional SEM    4-Ply cross section is shown in Figure 4.6 and 4.7.  The general display is an unorganized, irregular arrangement of rovings that can be described with geometric shapes of elliptical or hexagonal.  Scattered voids and inclusions can be appreciated in cross section.   Adequate wetting of the fibers can be appreciated in Figure 4.7.  Areas of poor impregnation are also noted.   Figure 4.6 4-Ply cross-sectional SEM  54    Figure 4.7 4-Ply cross-sectional SEM    2Ply cross-section is shown in Figures 4.8 and 4.9 and it appears to have the most consistent cross-sectional arrangement of braided rovings.  The general geometry of the rovings can be considered hexagonal and some evidence of repeating units exist.  Some voids and areas of deficient impregnation are still present.  Adequate wetting of the fibers can be appreciated in Figure 4.9.  55   Figure 4.8 2-Ply cross-sectional SEM    Figure 4.9 2-Ply cross-sectional SEM   4.6 SEM Analysis of tested samples Select samples consistent with the means of each group were examined by SEM.  The objective of this was to evaluate the fractured samples and discern potential modes of failure.  56  Similar findings were observed for all experimental groups, as presented in Figures 4.10 to 4.12.   Consistent finding between all experimental groups include minimal visible resin adhering to the fractured fibers and generalized fiber pull out.   General failure mechanism includes debonding between the matrix and fibers, and fiber fracture.  Figure 4.13 diagrams failure mechanisms of short discontinuous fibers [27].   Figure 4.10 2-Ply+axial tested sample    Figure 4.11 4-Ply tested sample   57   Figure 4.12 2-Ply tested sample    Figure 4.13 Failure modes of short discontinuous fibers [27]   Tested resin control group sample SEM images are shown in Figure 4.14.  The resin sample displays a clear compression curl.  Other features of brittle material fracture (mirror, mist, hackle lines) were not clear in the sample from this view.  The unidirectional control group sample in Figure 4.14 appears to suffer the same problem as the experimental group, i.e. minimal resin adhering to the fibers.  Failure modes of continuous fibers are also diagrammed in Figure 4.14 [27].  In addition, features of shear delamination,  and buckling failure are present.  These failure modes are consistent with unidirectional reinforcement.  No outright fracture of any unidirectional control groups occurred.   58    Figure 4.14 Tested resin control sample     Figure 4.15 Tested unidirectional control sample [27]         59  Chapter 5: Discussion  The primary research objective, producing a 3D 4-step braided FRC with superior flexural properties to Trinia, was not attained.  Multiple factors could have impeded the experimental groups ability to meet our objective.   One of them was the manual 4-step braiding procedure, relying on operator-controlled braiding movements and jamming steps.   With each preform requiring several hundred 4-step procedures and subsequent jamming steps, it is expected that significant intragroup variation existed in each experimental group.  This variability within each preform and testing specimen is expected to reside in braiding angle, unit cell size and structural consistency  [53].  This can influence the flexural properties of the samples [49,55].  The influence of this manual process is evidenced by the cross-sectional SEM images.  4-Step braiding is a process that is automated and readily available for small dimension preforms [51,56].  The manual production of these experimental preforms is a dated and subject to multiple instances of operator error; however, it was readily available, and its use was recommended by Materials Engineering. In the order of preform production, 2-ply+axial, 4-ply, and 2-ply, improvement in organization and consistency was noted, resulting in a preform of progressively higher quality.  This is exemplified in the cross-sectional SEM Figures 4.4 to 4.9.  Based on visual evaluation of the cross sectional-SEM images, roving geometry, distribution, and inter-roving alignment is more consistent with increasing experience with manual 4-step braiding.  In the case of the unidirectional control group, fibers were manually cut and adapted into molds for production.  It is inevitable that a certain percentage of rovings did not remain anisotropic, resulting in some 60  internal structural variation in fiber orientation.  The influence of these variations and those of the experimental group were not evaluated.  Structural consistency within and between braided preforms is cited as a challenge when producing 4-step braided structural composites [51,57].   Preform production often required several weeks of periodic braiding sessions.  Throughout the entire braiding process, individual rovings were left exposed to external conditions of an industrial laboratory.  Significant amounts of particulate matter and other contaminants are likely to have been incorporated into the braided preforms.  More concerning than particulate inclusion is the potential effect of environmental exposure on impregnation and adhesion between the resin matrix and silanated glass fibers.  Silanation is a covalent interaction and not likely to be degraded by environmental contaminants; however, their presence may prevent appropriate wetting, contact, and subsequent adhesion.  Failure in adhesion between the component fibers and matrix can result in the fibers weakening the resultant material by diminishing effective load transfer [27,58].  SEM images of fractured samples displayed little resin adhering to the fractured fibers and fiber pull out.    Impregnation of the preforms appeared reasonable in pretest cross-sectional SEM images, with minimal voids and apparent direct contact between the fibers and matrix.  It needs to be noted that quality of adhesion and impregnation was not evaluated as part of this work.  A pretest evaluation of the compatibility between the resin matrix and silanated glass fibers was not conducted.  It is important to have the silane be compatible with the fibers and matrix selected for the FRC [58].  The provided S-2 glass fibers were silanated with an amino-silane, stated to be compatible with both epoxy and select urethane resins.  The exact silane 61  coupling agent used was not provided by the manufacture.  A urethane-based resin was used in this experiment; thus, assuming the functionalized fibers were compatible with the resin; however, the quality of adhesion to the silanated glass fibers was not evaluated.    For a prospective CAD/CAM dental biomaterial, this crude manufacturing in suboptimal conditions fails to compare with the production quality of the currently available CAD/CAM material Trinia, reviewed above, which is obtained through optimized manufacturing, with minimal defects, inclusions, and voids.   While the above problems in manufacturing are noted, they are expected to be present in all groups in approximately equal proportions.  As all samples were tested under the same conditions, the results of the TPBT can still be analyzed with acknowledgment of these known limitations.   With exception of the unidirectional control group and 2-ply+axial experimental group, the process of sample preparation for 2-ply and 4-ply groups produced testing samples consisting of short discontinuous fibers [49].  Their arrangements could be described as quasi-isotropic short discontinuous fibers [49].  The 2-ply+axial testing samples are better described as a mixture of anisotropic continuous fibers and quasi-isotropic short discontinuous fibers.  The initial removal of edge fiber continuity in itself is expected to result in significant reduction in flexural properties [59].  With the notion that two of the experimental groups once prepared contained no significant continuous fibers, the experimental design would have benefitted from an additional control group, an isotropic discontinuous short fiber group.  This group would have served to evaluate the effect of some degree of anisotropic organization (quasi-isotropic) of short discontinuous fibers, compared to an isotropic (random) short discontinuous fiber control group. 62   The process of FRC sample preparation discloses a limitation in attempting to produce complex 3D continuous FRC for CAD/CAM dentistry.  Milling prefabricated discs and blocks of a continuous FRC material will inevitably produce discontinuous short fibers of various orientations during production of crowns, frameworks, prosthesis, etc.  Even in the case of incorporating continuous anisotropic fibers, there are limited applications in dentistry where these will remain intact following CAD/CAM machining and be oriented parallel to a tensile load clinically.  Unfortunately, prefabrication of high quality FRC for CAD/CAM application is unlikely to be able to take advantage of pure anisotropic benefits of continuous or discontinuous short fiber FRC.  Fiber alignment within CAD/CAM blocks or disks will require customization to a clinical scenario to maximize anisotropic benefits of fiber reinforcement.  A resulting product is suspected to reside in a middle ground, between anisotropic and isotropic material.  In vitro results, based on beam testing, need to be interpreted with caution.   While they appear promising, their relevance to a clinical scenario requiring curvilinear geometric shapes can negate much of structural benefit provided in a uniform 3D FRC sample.  Further research on both Trinia and other experimental 3D FRC for CAD/CAM applications is required to ascertain the validity of this assertion.  Based on Krenchel’s factor described above, it is not surprising that the completely anisotropic, unidirectional, control group produced the highest flexural properties.  The 2-ply+axial group, also containing unidirectional fibers, resulted in the second highest results.  Considering the 2-ply and 4-ply groups as short discontinuous quasi-isotropic groups, the 2-ply group had a significant advantage over the 4-ply group in flexural properties.  Among potential factors that can account for this is the internal braiding angle of 37.0 o of the 2-ply group 63  compared to 44.4 o of the 4-ply group.  The lower braiding angle increases anisotropic alignment improving flexural properties [51].  Differences in braiding angle can be a function of jamming, as well as of roving diameter [50].  As the roving diameter increases, the braiding angles also increase [50].  This was consistent with the experimental groups as the 4-ply preform had the greatest external and internal braiding angle.  The unit cell dimensions between the two groups could also contribute to this difference.  The 2-ply preform had 1.33 times the unit cells per sample cross-sectional dimension as the 4-ply group.  This may have an influence on arresting crack propagation through the samples.    We are unable to discern the significance of the internal braiding angle, or unit cell quantity on the flexural properties due to the intra-structural variations.  Again, these structural variations are a common disadvantage of a 4-step braided FRC making them challenging to evaluate experimentally and incorporate into industry predictably [49,57,60].  In the FRC samples in this work, extensive failure in adhesion is suspected to be the overwhelming failure mechanisms.     The 4-ply group failed to provide significant reinforcement over the resin control group.  This can be a function of poor adhesion, internal braiding angle of 44.4o providing minimal anisotropic reinforcement, and a reduced number of unit cells.       5.1 Comparative analysis  Limited studies were available to compare this work to.  The study referenced above by Petersen et al., displayed a significant advantage over the experimental groups flexural properties in this work [48].  Multiple differences exist between the two studies. These include preform structure, fabrication methodology, sample dimensions and FRC materials, each with potential to contribute to the differing flexural strengths and elastic moduli.  A key difference 64  warranting acknowledgement is differences in preform fabrication, impregnation and polymerization.  Petersen et al., involved third parties in preform production, impregnation, and polymerization of their FRC.  This is anticipated to produce a FRC of superior quality to those produced in their entirety via a laboratory/bench top setting.  Based on Krenchel’s factor, the bidirectional weave used by Petersen et al. is also expected to have a superior reinforcing effect compared to non-orthogonally oriented fibers of the braided structure used in this work.  With these two factors, it is not surprising that experimental 3D FRC produced by Petersen et al. had superior flexural properties.  Direct comparison to Trinia cannot be made as the glass fiber reinforcement architecture, production, and testing conditions are not described by Shofu; moreover, Trinia was not made available for external flexural testing.  If comparing the manufacturer’s proclaimed data to the experimental groups of this work, only the 2-ply+axial group proved to have a superior elastic modulus to Trinia.  This is suspected to be a result of inclusion of unidirectional anisotropic fibers aligned parallel to the applied tensile stress.  This is a speculative assertion.    5.2 Future recommendations   Both Trinia and the experimental 3D FRC produced by Petersen et al. are FRC CAD/CAM materials with in vitro tested or proclaimed flexural properties of acceptable values for multiple dental applications.  The experimental 3D 4-step braided FRC produced in this work failed to reach these flexural values.  This failure can be attributed to the multiple factors reviewed above.  This experiment also highlighted some potential limitations of 3D CAD/CAM FRC dental biomaterials.  Despite these limitations, the flexural properties of both Trinia and the 3D FRC 65  produced by Petersen et al., suggest that 3D FRC CAD/CAM materials should continue to be researched and evaluated as an alternative to the existing classes of CAD/CAM materials: ceramics, and glass-ceramics. Based on the results of this study, several recommendations can be made.  The first is to improve the quality and consistency of the experimental preforms by addressing preform fabrication, impregnation, adhesion of the components.  This should be considered regardless of preform design and complexity.  The work by Petersen et al. can be used as an example as these critical steps were delegated to appropriate third parties focused on their respective steps.  Other authors also acknowledged the challenges of impregnating complex preforms and recommended delegating this task to a party with expertise in preform fabrication [27].    The second is to recommend that future experimental 3D FRC incorporate fibers aligned perpendicular to the applied load in a flexural test.  Should a 4-step braided approach be employed in the future, it is recommended to reduce the braiding angle to further improve alignment of the inevitable quasi-isotropic short discontinuous fibers to an anticipated tensile load.  Should improved adhesion be possible, an additional recommendation is to reduce the roving bundle size to incorporate more unit cells per cross sectional area.  This has potential to increase resistance to crack propagation [51].   For future work to successfully satisfy the objective stated in this study, a considerable number of additional factors will need to be researched and evaluated.  Some of these factors include, but are not limited to, the effects of water storage, compatibility with veneering composites, resistance to cyclic loading, resistance to off axis loading.   66  Chapter 6: Conclusion  Based on the results presented above, the following can be stated about the null hypotheses tested in this study:  1 The first null hypothesis was partially rejected, as all groups, except the 4-ply experimental group, were able to significantly reinforce the resin matrix. 2 The second null hypothesis was rejected, as a 3D braided FRC differed from an anisotropic FRC fibers with significant reductions in flexural properties.  3 The third null hypothesis was rejected, as altering the unit cell size from infinite (unidirectional control), 2-ply, to 4-ply resulted in significant reductions in flexural properties.   4 The forth null hypothesis was rejected, as the inclusion of anisotropic fibers into the 2-ply 3D braided FRC resulted in significant improvement in flexural properties.  Based on the many limitations of this study, the only inference to be extrapolated into future development of 3D FRC is the importance of incorporating anisotropic fibres.             67  Bibliography  [1] Davidowitz G, Kotick PG. The Use of CAD/CAM in Dentistry. Dent Clin North Am 2011;55:559–70. doi:10.1016/j.cden.2011.02.011. [2] Assembly FG. CAD/CAM Dentistry. Int Dent J 2018;68:18–9. doi:10.1111/idj.12373. [3] Li RWK, Chow TW, Matinlinna JP. Ceramic dental biomaterials and CAD/CAM technology: State of the art. J Prosthodont Res 2014;58:208–16. doi:10.1016/j.jpor.2014.07.003. [4] Ruse ND, Sadoun MJ. Resin-composite blocks for dental CAD/CAM applications. J Dent Res 2014;93:1232–4. doi:10.1177/0022034514553976. [5] Al-Sanabani F a, Madfa A a, Al-qudaimi NH. Alumina ceramic for dental applications : A review article. Am J Mater Res 2014;1:26–34. [6] Babu PJ, Alla RK, Alluri VR, Datla SR, Konakanchi A. Dental Ceramics: Part I – An Overview of Composition, Structure and Properties. Am J Mater Eng Technol 2015;3:13–8. doi:10.12691/MATERIALS-3-1-3. [7] 1847-1922 C. H. (Charles Henry), L. Porcelain dental art : the new process of restoring decayed and defective teeth to their original appearance in shape, size and color. Detroit : O.S. Gulley, Bornman; 1888. [8] Miyazaki T, Hotta Y. CAD/CAM systems available for the fabrication of crown and bridge restorations. Aust Dent J 2011;56:97–106. doi:10.1111/j.1834-7819.2010.01300.x. [9] Conrad HJ, Seong W-J, Pesun IJ. Current ceramic materials and systems with clinical recommendations: a systematic review. J Prosthet Dent 2007;98:389–404. doi:10.1016/S0022-3913(07)60124-3. [10] Shenoy A, Shenoy N. Dental ceramics: An update. J Conserv Dent 2010;13:195–203. doi:10.4103/0972-0707.73379. [11] Probster L, J. Geis-Gerstirfer, E. Kirchner, P. Kanjantra. In vitro evaluation of a glass – ceramic restorative material. J Oral Rehabil 1997:636–45. [12] Von Steyern PV, Ebbesson S, Holmgren J, Haag P, Nilner K. Fracture strength of two oxide ceramic crown systems after cyclic pre-loading and thermocycling. J Oral Rehabil 2006;33:682–9. doi:10.1111/j.1365-2842.2005.01604.x. [13] Bona A Della, Pecho OE, Alessandretti R. Zirconia as a dental biomaterial. Materials (Basel) 2015;8:4978–91. doi:10.3390/ma8084978. [14] Piconi C, Maccauro G. Zirconia as a ceramic biomaterial. Biomaterials 1999;20:1–25. doi:10.1016/S0142-9612(98)00010-6. [15] Ferracane JL. Resin composite - State of the art. Dent Mater 2011;27:29–38. doi:10.1016/j.dental.2010.10.020. [16] Klapdohr S, Moszner N. New inorganic components for dental filling composites. Monatshefte Fur Chemie 2005;136:21–45. doi:10.1007/s00706-004-0254-y. [17] da Rosa Rodolpho PA, Cenci MS, Donassollo TA, Loguércio AD, Demarco FF. A clinical evaluation of posterior composite restorations: 17-year findings. J Dent 2006;34:427–35. doi:10.1016/j.jdent.2005.09.006. [18] Nguyen JF, Migonney V, Ruse ND, Sadoun M. Resin composite blocks via high-pressure high-temperature polymerization. Dent Mater 2012;28:529–34. doi:10.1016/j.dental.2011.12.003. [19] Ilie N, Hickel R. Investigations on mechanical behaviour of dental composites. Clin Oral 68  Investig 2009;13:427–38. doi:10.1007/s00784-009-0258-4. [20] Beck F, Lettner S, Graf A, Bitriol B, Dumitrescu N, Bauer P, et al. Survival of direct resin restorations in posterior teeth within a 19-year period (1996-2015): A meta-analysis of prospective studies. Dent Mater 2015;31:958–85. doi:10.1016/j.dental.2015.05.004. [21] VITA ENAMIC® - redefines load capacity https://www.vita-zahnfabrik.com/en/CAD/CAM/Single-tooth-restoration/VITA-ENAMIC-24970,27568.html (accessed April 21, 2018). [22] TRINIA | The Next Generation CAD/CAM Material http://www.trinia.com/ (accessed April 21, 2018). [23] Awada A, Nathanson D. Mechanical properties of resin-ceramic CAD/CAM restorative materials. J Prosthet Dent 2015;114:587–93. doi:10.1016/j.prosdent.2015.04.016. [24] CAD / CAM Material WHY TRINIA ? http://www.bicon.com/downloads/pdf/TRINIA_Brochure.pdf (accessed May 1, 2018) [25] Bilisik K. Three-dimensional braiding for composites: A review. Text Res J 2013;83:1414–36. doi:10.1177/0040517512450766. [26] Kadir Bilisik NSK and NEB. 3D Fabrics for Technical Textile Applications  [27] Matinlinna J. 2 – Types of FRCs used in dentistry. Elsevier Ltd; 2017. doi:10.1016/B978-0-08-100607-8.00002-2. [28] Hervás-García A, Martínez-Lozano MA, Cabanes-Vila J, Barjau-Escribano A, Fos-Galve P. Composite resins. A review of the materials and clinical indications. Med Oral, Patol Oral y Cirugía Bucal 2006;11:215–20. [29] Peutzfeldt A. Resin composites in dentistry: the monomer systems. Eur J Oral Sci 1997;105:97–116. doi:10.1111/j.1600-0722.1997.tb00188.x. [30] Kruijt A, Turin M. A review on fibre reinforced composite resin. Lang Soc 2017;46:257–69. doi:10.1017/S0047404517000161. [31] Khan AS, Azam MT, Khan M, Mian SA, Rehman IU. An update on glass fiber dental restorative composites: A systematic review. Mater Sci Eng C 2015;47:26–39. doi:10.1016/j.msec.2014.11.015. [32] Vallittu PK. High-aspect ratio fillers: Fiber-reinforced composites and their anisotropic properties. Dent Mater 2015;31:1–7. doi:10.1016/j.dental.2014.07.009. [33] Wallenberger FT, Watson JC, Hong L. Glass Fibers. ASM Hanb 2001;21:27–34. doi:10.1361/asmhba0003353. [34] Loewenstein KL. Manufacture of Continuous Glass Fibres. Platin Met Rev 1975;19:82–7. [35] Glass Fibre Manufacturing and Sizing Stability https://www.azom.com/article.aspx?ArticleID=7561 P (accessed April 21, 2018) [36] Li H, Charpentier T, Du J, Vennam S. Composite reinforcement: Recent development of continuous glass fibers. Int J Appl Glas Sci 2017;8:23–36. doi:10.1111/ijag.12261. [37] Weblet Importer http://nptel.ac.in/courses/116102006/9 (accessed April 21, 2018). [38] Kolesov YI, Kudryavtsev MY, Mikhailenko NY. Types and Compositions of Glass for Production of Continuous Glass Fiber (Review). Glas Ceram 2001;58:197–202. doi:10.1023/a:1012386814248. [39] Zhang M, Matinlinna JP. E-Glass Fiber Reinforced Composites in Dental Applications. Silicon 2012;4:73–8. doi:10.1007/s12633-011-9075-x. [40] Azom.com. S - Glass Fibre 2001:1–3. 69  [41] Vallittu PK. Some aspects of the tensile strength of unidirectional glass fibre-polymethyl methacrylate composite used in dentures. J Oral Rehabil 1998;25:100–5. doi:10.1046/j.1365-2842.1998.00235.x. [42] Gajewski VES, Pfeifer CS, Fróes-Salgado NRG, Boaro LCC, Braga RR. Monomers used in resin composites: Degree of conversion, mechanical properties and water sorption/solubility. Braz Dent J 2012;23:508–14. doi:10.1590/S0103-64402012000500007. [43] Lovell LG, Stansbury JW, Syrpes DC, Bowman CN. Effects of composition and reactivity on the reaction kinetics of dimethacrylate dimethacrylate copolymerizations. Macromolecules 1999;32:3913–21. doi:10.1021/ma990258d. [44] Chen M-H. Update on Dental Nanocomposites. J Dent Res 2010;89:549–60. doi:10.1177/0022034510363765. [45] Zucchelli A, Focarete ML, Gualandi C, Ramakrishna S. Electrospun nanofibers for enhancing structural performance of composite materials. Polym Adv Technol 2011;22:339–49. doi:10.1002/pat.1837. [46] Vallittu PK. High-aspect ratio fillers: Fiber-reinforced composites and their anisotropic properties. Dent Mater 2015;31:1–7. doi:10.1016/j.dental.2014.07.009. [47] Behr M, Rosentritt M, Lang R, Handel G. Flexural properties of fiber reinforced composite using a vacuum/pressure or a manual adaptation manufacturing process. J Dent 2000;28:509–14. doi:10.1016/S0300-5712(00)00031-2. [48] Petersen R, Perng-Ru L. 3D-Woven fiber-reinforced composite for CAD/CAM dental applications 2016;8:583–92. Sampe J doi:10.1002/aur.1474.Replication. [49] Mouritz AP, Bannister MK, Falzon PJ, Leong KH. Review of applications for advanced three-dimensional fibre textile composites. Compos Part A Appl Sci Manuf 1999;30:1445–61. doi:10.1016/S1359-835X(99)00034-2. [50] Du GW, Ko FK. Unit cell geometry of 3-D braided structures. J Reinf Plast Compos 1993;12:752–68. doi:10.1177/073168449301200702. [51] Bilisik K. Three-dimensional braiding for composites: A review. Text Res J 2013;83:1414–36. doi:10.1177/0040517512450766. [52] 449 S-2 Glass ® Roving High-Strength Solutions for Your Toughest Reinforcement Challenges https://www.agy.com/wp-content/uploads/2014/03/449_S-2-Aerospace.pdf (accessed April 21, 2018) [53] Byun JH, Chou TW. Process-microstructure relationships of 2-step and 4-step braided composites. Compos Sci Technol 1996;56:235–51. doi:10.1016/0266-3538(95)00112-3. [54] Zeng T, Wu L zhi, Guo L cheng. A finite element model for failure analysis of 3D braided composites. Mater Sci Eng A 2004;366:144–51. doi:10.1016/j.msea.2003.09.054. [55] Dauda B, Oyadiji SO, Potluri P. Characterising mechanical properties of braided and woven textile composite beams. Appl Compos Mater 2009;16:15–31. doi:10.1007/s10443-008-9073-3. [56] Bilisik K. Three-dimensional axial braided preforms: Experimental determination of effects of structure-process parameters on unit cell. Text Res J 2011;81:2095–116. doi:10.1177/0040517511414978. [57] Fan Z, Zhenguo L, Zhe W, Guoquan T. A New Scheme and Microstructural Model for 3D Full 5-directional Braided Composites. Chinese J Aeronaut 2010;23:61–7. 70  doi:10.1016/S1000-9361(09)60188-6. [58] Zakir M, Ashraf U, Tian T, Han A, Qiao W, Jin X, et al. The Role of Silane Coupling Agents and Universal Primers in Durable Adhesion to Dental Restorative Materials - a Review. Curr Oral Heal Reports 2016;3:244–53. doi:10.1007/s40496-016-0108-9. [59] Crane RM, Camponeschi ET. Experimental and analytical characterization of multidimensionally braided graphite/epoxy composites. Exp Mech 1986;26:259–66. doi:10.1007/BF02320051. [60] Sun X, Sun C. Mechanical properties of three-dimensional braided composites. Compos Struct 2004;65:485–92. doi:10.1016/j.compstruct.2003.12.009.        .      

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0368686/manifest

Comment

Related Items