Essential work of fracture (EWF) was used to determine the essential work of fracture of a fiber-reinforced dental composite.
The thickness of conventional composites covering the fiber-reinforced composite has no effect on the toughness of the system.
Increasing the thickness of fiber-reinforced composites will increase the toughness potential of the material.
Short fiber reinforced composites (SFRC) require a veneering layer of conventional composite when used as restorations in the oral environment. The current study investigates the toughening effects during the path of a preexisting crack propagating through the bilayer system as it confronts the interface, through the attempted alignment of fibers and matrix-fiber interactions in the SFRC, and the distance it travels in the SFRC.
Bilayer systems of SFRC and conventional composite were produced with aligned fibers perpendicular to load direction. Single-edge-notched bend (SENB) specimens (25 × 5 × 2.5 mm 3 ) with pre-crack length ( a ) to width ( W ) ratios ( a / W = 0.2−0.8) were produced and tested in 3-point bending configuration until complete fracture. The specific work of fracture (w e ) was deduced from calculating the area under the load-displacement curves. Fiber alignment was digitally evaluated from images taken from the top and side planes of the specimen.
The toughness of the bilayer system is optimal when maximum SFRC thickness is used. EWF methods showed toughness and increasing nonessential work of fracture scaling with ligament length. A longer distance is accompanied by a higher distribution of aligned fibers bridging behind the crack wake, reducing crack driving forces at the crack tip.
SFRC materials provide increasing toughening potential with increasing thickness, and have the ability to be more anisotropic than other composite materials. Clinically, the layer must have a conventional composite cover layer, but which thickness does not affect toughness potential. Therefore the thickness of the conventional composite can be dictated by wear behavior.
Fiber-reinforced dental composites are used as dentinal replacements and are expected to perform as reliable substitute materials under the masticatory loads experienced in the mouth. Conventionally, dental composite contains reinforcing particulates in nano-, micro-, and macro-scale lengths in a resin matrix [ ]. The fillers significantly influence mechanical properties, contribute to isotropy, and supply sufficient strength as an essential requirement for restorative materials [ , ]. Furthermore, dental materials must possess fracture toughness or the ability to resist fracture [ ]. It is here that short fiber-reinforced composites (SFRC) demonstrate completely different toughening mechanisms from their particulate reinforced (PRC) conventional counterparts. The interactions between the high aspect-ratio fibers (with potential anisotropy with alignment) with the resin matrix contribute to rising fracture toughness values with crack extension [ , ].
During crack propagation, there are notably four contributions to the toughening of SFRC; strain energy stored in the fiber over the debond length contributing to elastic bridging, residual strain energy over the debond length which has a negative contribution to toughening when the fiber breaks/degrades, the energy creating the debond fracture surface, and the frictional energy in fiber pull out [ ]. Additionally, in a tradeoff between strength and toughness, the interface between fiber and matrix also plays an essential role in how the propagating crack moves. If the interface is too strong, the composite itself acts as a rigid, brittle material, and the high loads needed to overcome the glass fibers will cause fiber fracture. On the contrary, if this interface is too weak, fibers pull out without much frictional resistance, and the composite will exhibit overall low strength and low toughness [ , ]. However, at the right balance, optimal interfaces can be achieved to maximize the frictional resistance, dissipating energy through shear interfacial forces and effectively diminishing the crack advancing forces felt at the crack tip [ ].
We can use fracture mechanics to evaluate the material response to an existing crack. Linear-elastic fracture mechanics (LEFM) is used in the case of brittle materials; it is given as the critical stress intensity factor K c or the critical energy release rate G c . LEFM is widely used to characterize the localized stress at the crack tip in materials displaying a linear-elastic fracture response. LEFM is then not applicable in characterizing materials with larger process zones as it fails to account for the energy lost in plastic deformation. Instead, one applies elastic-plastic fracture mechanics (EPFM), which is commonly given by the path-independent line integral – J integral , and the essential work of fracture (EWF). Both EPFM methods are derived experimentally by calculating the area under the load-displacement curves, or work done; however, the difference is in testing methods and energy partitioning. J-integral is calculated as the potential energy difference between specimens with different crack lengths, which requires detection of cracking onset followed by test interruption. At each loading cycle, each load-displacement curve is partitioned into elastic and plastic components adhered to under ASTM E1820 [ ], resulting in the calculation of the critical energy release rate under mode 1 loading conditions, J 1c from the J R – Δ a resistance curve (r-curve) [ ].
On the other hand, EWF originated from the idea of an inner autonomous region around the crack tip requiring increasing energy input to sustain crack growth, as the energy increasingly dissipates through the outer region [ ]. The inner region is energy partitioned as the essential work (W e ), and the outer region is coined as the non-essential or plastic work (W p ) component. The idea is that EWF partitions in the calculation of the specific essential work of fracture which is an inherent material property unaffected by specimen dimension or crack length, and simultaneously determines the plastic work as a function of ligament length (or uncracked portion of the specimen) [ , ]. Unlike the J-integral values requiring periodic test interruption for crack measurement, EWF values are derived experimentally by running a complete uninterrupted test of the specimen with different ligament lengths until failure. The total area under the load-displacement curve is the total work of fracture (W f ).
SFRC materials, although composed of two constituents that would qualify individually for LEFM, are structured in a way that large amounts of energy are dissipated through fiber interactions as a pre-existing crack is loaded. Fracture mechanics of SFRC are better suited with EPFM. Previously, we measured the r-curve behavior of fiber-reinforced composites [ ], determined that different r-curves are created when fibers are aligned or randomly oriented [ ] and that the fiber-matrix interface is unaffected by water hydration [ ]. Fibers are known to be harmful to the lungs if inhaled due to their length and aspect ratio [ ]; therefore, it is precautionary to cover SFRC with a PRC material and prevent exposure to the oral environment. In practice, SFRC is confined to a finite space within a tooth cavity. Therefore it is of significance to understand the fracture mechanics in terms of thickness of SFRC and the thickness of the PRC covering of which is mandated for its use. As such, sufficient insight into the cover thicknesses and interface mechanisms to advancing cracks has not been effectively realized. Here we use the EWF method, as the conventional term ‘ligament’ translates to the clinical term ‘thickness’ of the SFRC and will hereafter be referred to as thickness. Therefore, this study investigates the thickness influence of the PRC material when veneered on the aligned SFRC using the EWF method. Moreover, we provide a discussion on the interface between the two composites, how it relates to clinical practice, and the feasibility and validity of our attempt to create an anisotropic structure with aligning fibers.
The materials used in this study are commonly used in combination with bulk-fill restorations ( Fig. 1 a). Short-fiber reinforced composite (SFRC; everX Flow™, GC Corporation, Japan) with E-glass fibers (200−300 μm, ⌀ = 7, ∼20 vol%) and particulates (barium glass ∼26 vol%), in a resin matrix (Bis-MEPP:UDMA:TEGDMA. 4:1:1 ratio) (E modulus = 10.0 ± 1.6 GPa [ ]); was used with a conventional composite (PRC; Essentia® HiFlo, filler fraction 50 vol%, strontium glass 200 nm, silica particles 16 nm, GC Corporation, Japan)(E modulus = 8.8 ± 0.4 GPa [ ]). Bilayered single-edge notch beam (SENB) specimens (2.5 mm × 5.0 mm × 25.0 mm) with different thickness ratios of SFRC:PRC (where SFRC thickness was 1.0 mm, 1.5. mm, 2.0 mm, 2.5 mm, 3.0 mm, 3.5 mm, and 4.0 mm) were constructed by carefully measuring and cutting Mylar strips of length 5.0 mm and heights corresponding to SFRC thickness (n = 3 for each group). The Mylar strips were placed in a custom tungsten carbide split mold (for specimens 2.5 mm × 5.0 mm × 25.0 mm) and SFRC was carefully extruded to the Mylar strip height. Care was taken to extrude in the long axis direction through the provided nozzle (∼1 mm radius opening), to align the fibers. Side surfaces (where depth of cure was 2.5 mm for all specimens) were photo-activated using a halogen light-curing unit 750 mW/cm² Elipar Trilight, 3M Oral Care, USA) over five overlapping points for 20 s each on both sides (according to ISO 4049). After removing the Mylar strips, the remaining space was filled with PRC to complete the specimen and further photo-activated again with the same protocol. PRC thickness was minimum 1 mm (up until 4 mm to correspond with SFRC thickness) ( Fig. 1 a) for ease of processing. It provided a minimum 0.5 mm distance for the crack to propagate through the PRC (vertical wear rates of flowable dental composites are reported to be <100 μm/equivalent year [ , ]. Monolayered SFRC and PRC specimens (n = 2 for each material) were also produced with the same mold and photo-activation protocol.
A notch (∼0.5 mm) in the PRC was created in all specimens using a 0.1 mm-thick diamond disk to length 0.35 mm, and the remaining 0.15 mm was produced with a sharp razor blade with fine alumina polishing paste (1 μm). The notch and SFRC height was checked and measured under a light microscope (SteREO Discovery.V8, Zeiss, Germany), coupled with a camera (AxioCam, Zeiss) and measuring software (Zen 2, Zeiss).
Alignment and fiber orientation
In order to confirm the fiber alignment, a quantitative orientation analysis was performed. Specimens were finely polished (P4000 silicon carbide wet/dry), and surface images were captured using light microscopy (Leica Microsystems GmbH, Germany) with attached camera attachment (AxioCam, Zeiss, Germany). Images were processed using Adobe Photoshop CS6 (Adobe Inc. USA), removing color, reducing noise, and increasing the contrast. Fiber characterization was performed in ImageJ software (NIH, Bethesda, Maryland, USA) with the OrientationJ plugin [ ]. Visual directional analysis, distribution of orientations, and dominant direction with coherency (where 0 = isotropic, 1 = highly orientated) were performed on representative images (n = 4) taken from the side and top planes of the specimens. Each image had a minimum of 100 fibers.
Fig. 2 is a representative color-mapped image from the light microscope, illustrating the side planes were aligned in the area of crack propagation. The area showed fibers generally aligned perpendicular to load direction. The dominant direction (where 0 degrees indicates absolute perpendicular) and related coherency (where 1 > C > 0, C = 1 when the local image features have one dominant direction, and C = 0 when the image is isotropic), in the images from the side plane, was −3.210° (0.424 coherency). Images taken from the ends of the beam were less aligned, showing the difficulty of alignment as the working space becomes increasingly challenging.
Specimens were tested in a universal testing machine (Z2.5, Zwick/Roell, Germany) in a 3-point bending (3-PB) configuration at a crosshead speed of 0.01 mm/min. Axial load-line displacement was recorded using an attached laser extensometer (LaserXtens, Zwick/Roell, Germany) with laser beam illumination following the digital speckle correlation method. The energy partition of the EWF method is schematically illustrated in Fig. 1 b, where the inner zone represents the inherent material property of essential work, and the outer zone is the size-dependent plastic zone. The area under the curve is the total work of fracture W f ( Fig. 1 b).
The total work of fracture is given by Eq. (1) :