In recent years, Fiber Reinforced Polymers (FRP) have emerged as a potential solution to the tribulations associated with materials technology. Lightweight composite materials have gained increasing interest for applications in aeronautical, transports and sports industries and in many other fields of engineering. The aerospace industry has capitalized the use of reinforced composites, known for their extensive outstanding material properties, for decades. The high strength to weight ratio, stiffness to weight ratio, corrosion and fracture resistance of FRP composites make them an attractive structural component for use in many applications. Bader (2002) reported that composite materials based glass fiber in thermosetting resins has been widely utilized in industry for more than 50 years. Woven GFRP composite is a classic example of a synthetic composite that combines the strength of small diameter glass fibers with the ductility of the polymeric matrix, the outcome of which is superior to either component alone. This supremely characteristic is due to the proper selection of fabrication process and this is also agreed by Avila and Morais (2005), which confirmed that performance is directly related to the manufacturing process used to produce the composite.
A good understanding of the mechanical performance of composite materials and the influence of production methods on their performance is vital in achieving the full potential of these materials in structural applications. However, a number of technical issues remain that must be addressed before the engineering practitioners can develop confidence in structural design with composite members. These issues include, but are not limited to, low stiffness, connection details, cost, confirmation of improved durability and toughness to fracture failure. However, delamination of laminated composite materials is a critical problem, whereby interlaminar failure during routine services becomes a major concern in structural composite design. The delamination in the composite panel will result in looses of stiffness and hence the design performances in structural application of fiber composite materials. The resistance or toughness to delamination and propagation of interlaminar defects or micro-cracks in fiber-reinforced composite is identified by the interlaminar separation subjected to no fiber breakage, pullout and fiber bridging. This defines a material property independent of test-specimen geometry and ply orientation that constitute the delaminating interface (Yeo and Dalimin, 2001).
The purpose of this study is to study on the mode-I interlaminar fracture toughness
by established standard (ASTM, 2001; BS ISO 15024, 2001) due to symmetrical
ply arrangement and various curing pressure of the reinforced woven glass fiber
material. In the interest of this matters, the discussion is limited to composite
based mainly on thermosetting (unsaturated polyester) resin with E-glass plain
weave woven fiber.
Many workers specializing on fracture mechanics approach to investigate the interlaminar failure of fiber composite materials can be found. As detailed by earlier workers (Faizal et al., 2006), commonly employed standard test geometry, ASTM D5528 and BS/ISO 15024, can be used for characterization of Mode-I interlaminar toughness. However, many practical applications usually involve only multidirectional laminates. Therefore, it is essential to study the interlaminar fracture toughness of multidirectional and woven composites (Morais et al., 2002). The Double Cantilever Beam (DCB) geometry has been widely used in this study for the characterization of delamination resistance or fracture energy of laminates (Blackman et al., 1991; Todo and Jar, 1997). This is also shown by Robinson et al. (1995) that, the conventional DCB specimen is most widely used in determining Mode-I critical interlaminar strain energy release rate (GIC) for unidirectional laminates. However, if this specimen is used for measuring GIC between plies of dissimilar fiber orientation then extensive fiber bridging and crack jumping can result, which prevents measurement of the critical energy release rate for pure delamination growth.
Toughness is a quantitative term that describes the energy consumed or work done in rupturing a specimen. Toughness is a material property independent of crack length, geometry or loading system. Therefore, fracture toughness is a generic term for measuring a resistance to extension of a crack (ASTM, 1996). The material is said to have a high toughness when it is hard to make a crack propagate into the system or body. Pinho et al. (2006) study on fracture toughness of the tensile and compressive fibre failure modes in laminated composites reported that lower value of GIC is associated with an initial planar fracture from the pre-crack plane without any significant fiber pullout and if this is the case, then the initiation value would be lay-up independent. Robinson et al. (1995) reported that laminated fiber reinforced plastic composites are particularly susceptible to failure by delamination initiation and growth owing to the relatively low tensile strength of the matrix. Schön et al. (2000) studied on the fracture mechanical properties of DCB specimens, reported that by comparing different types of damages, manufacturing or service induced, delamination growth occur as a consequence of interlaminar stresses. Perrin et al. (2003), has also reported that DCB specimen employed for studying interlaminar fracture toughness expressed, critical energy release rate, is very sensitive to fiber-matrix interface in composite laminates.
Focus on the present work is to undertake a detailed investigation of the interlaminar fracture toughness of woven E-glass fiber/Unsaturated polyester composite material using the fracture mechanics approach. By using DCB specimen the critical mode-I energy release rate, GIC, can be measured and this information can then be used to increase the understanding of interlaminar toughness in composite materials. In particularly, to investigate the effects of curing pressure to symmetrical ply arrangement, which induces matrix variation and the relationship with the interlaminar fracture toughness energy value, GC.
MATERIALS AND METHODS
The experimental method describes in detail the materials and its fundamental constituents, the specimen preparation for the fabrication of GFRP composite panels and the experimental test method that has been carried out.
Material: Plain-weave type woven glass fiber, commercial code: NISER (EWR-600B) had been used as the GFRP material for the purpose of this study. The basis of EWR-600B was the glass fabric of E-glass, based according to the 0.5% content of Na2 + K2O. The E-glass plain weave had been produced by interlacing warp thread and weft thread and was mainly used in hand lay-up processes. The EWR-600 had a round cross sectional shape approximately 9 μm and chemically contains less than 1% alkali. The basic physical properties of EWR-600B were described in Table 1.
The package of the woven roving was stored horizontally as recommended by reference
(BS 3396-1, 1991; BS 3396-2, 1987; BS 3396-3, 1987) in order to avoid distortion
of the fabric weaves and the physical outlook of the woven roving EWR-600B as
shown in Fig. 1.
properties of EWR-600B|
Before cutting the fabrics, it was ensured
that uniformly woven, selvedges were well made, straight and even, free yarn
defects and free from streaks, stains and also other contaminations and permanent
Laminate composite panel: Woven laminates had been prepared by hand
lay-up at room temperature. Four experimental composite panels were fabricated
with symmetry arrangement of E-glass fiber cloth/ply. A plain weave woven roving
with fibers at 0° and 90°
direction, were symmetrically layered up about a central plane as shown in the
An average mass of 71g woven roving ply was positioned manually in a rectangular open mold. The polyester resin was then poured into the mold and the ply flattens using brush or flat stick. The plain weave woven E-glass fabric was stacked into ten layers symmetrically as mentioned earlier. A panel of 10 ply laminates had been fabricated for different curing pressure of 35.8 kg m2, 70.1 kg m2, 104 kg m2 and 138.2 kg m2. The stacked plies for each group of laminate were then allowed to cure for 24 h at room temperature at 23°C/297.15 K.
DCB sample: The geometry of the DCB specimen consists of a rectangular uniform thickness as illustrated in the Fig. 3. The DCB specimen contains a thin, non-adhesive starter film embedded at the mid-plane, which was used to simulate an initial delamination (BS ISO 15024, 2001). From the Fig. 3, b was the specimen width, h the specimen thickness, ao the initial delamination length, a the total delamination length, A the insert length, l the specimen length, l1 the distance from center of loading pin to mid-plane of specimen, l2 the distance from center of loading pin to edge of end block, l3 the end block length and H the end block thickness.
In this study, dimensions of the specimen had been selected in accordance with the standard test method BS ISO 15024 (2001) specifications and as well as other publications were referred. The Fig. 3 shows the summary of DCB specimen design according to established test standards and other researchers work. Perrin et al. (2003) reported that improper selection of thickness of the DCB specimens could cause the need for large deflection correction, which was due to large displacement of end block rotation.
Test method: The testing was undertaken in accordance with the ASTM D5528-01 (2001) and British test method BS ISO 15024 (2001). All tests were conducted at room temperature 23°C/297.15 K, with dry laminates using an axial loading servo-hydraulic test machine, INSTRON Series 8801, under displacement control, as shown in the Fig. 4. Prior to testing, a thin layer of white correction fluid is applied along the edge of the DCB specimen for clear identification of crack position or intervals during the testing. The specimen edge was marked with initial 1mm interval up to 5mm from the starter crack end and thereafter at 5mm intervals. All specimens were not pre-cracked before the testing in order to obtain an initiation value free of fiber bridging. All tests were conducted under displacement control at constant crosshead speed of 3 mm min1.
DATA REDUCTION SCHEME
The Linear Elastic Fracture Mechanics (LEFM) analysis for analyzing the data obtained in the load-displacement traces can be tackled by different approaches. However, the various methods available for calculating mode-I strain energy release rate from DCB geometries could result in different values (Ashcroft et al., 2001). In this study however, a load/displacement plot provided the data for each crack length used for the calculation of GIC by adopting the experimental modified beam theory method (Stevanovic et al., 2000).
Modified Beam Theory, MBT: The experimental beam theory expression for GI assumes perfectly built-in (clamped at the delamination front) of DCB specimen that can be derived as
where, P is the load, δ the load point or opening displacement, b the specimen width and a the delamination length. However, this equation could over estimate the GI by imperfect built-in assumption, which could be affected by the rotation at the delamination front. Correction of this rotation is to treat the DCB as if it contains a slightly longer delamination and this led to the introduction of Modified Beam Theory (MBT) method. For this data reduction strategy, then the critical energy release rate is resolved as
where, Δ is determined experimentally by generating a least square plot of the cube root of compliance, C1/3, as a function of delamination length as shown in the Fig. 5. The parameter-C is the compliance or also the ratio of the load point displacement to the applied load, δ/P and N is the loading block correction.
Parameter Δ, is the crack length correction factor which, in a real specimen,
accounts for the effects of rotation at the crack tip.
||Typical plant-weave E-glass woven roving, EWR-600B
||Symmetry arrangment of E-glaas woven ply
||Specimen geometry and parameters
hydraulic testing machine, INSTRON
of the Δ-correction
The correction to the crack length ensures that the opening displacement δ
can be related to the corrected crack length (a+|Δ|), by the
simple beam theory expression for a cantilever (Robinson et al., 2004),
If the Δ is negative, ESIS recommends that it should be set to zero whereas ASTM D5528-01 recommends that the modulus should be used. As reported by Robinson et al. (2004), neither of these recommendations is followed since a negative value of Δ is physically appropriate due to the existence of the bridging stress.
As a result, the critical energy release rate GIC is determined by multiplying the F/N factor to the GI in Eq. 2, such that;
where F is the large displacement effect and N is the loading block correction.
RESULTS AND DISCUSSION
A successful curing process of composite fabrication will be determined by its outstanding properties. The variation of curing cycle and curing pressure could contribute to the variation in the material properties. As a result, the characteristic standard curing cycle and pressure variation, toughness and fractographic analysis of the fabricated composite would reviewed any such distinction and correlation.
Physical characteristic: Micrographs of through thickness cross-section of the woven has been obtained to determine its fiber and matrix characteristic via the JEOL JSM 5610LV scanning electron microscope (SEM), as shown in Fig. 6. Generally, voids are found in small quantities about 1 to 4% across the thickness section that appears as black dots in the micrographs. The voids formed are found to cause local rearrangement of tows, which reflects the non-uniform stress distribution through the thickness. However, the fiber weight fraction (%) has shown a variation increasing from 50 to 56% and remains about constants with respect to increasing curing pressure. As a result, the quality of these composites produced can be classified as of a mediocre quality (Berthelot et al., 1999).
Toughness characteristic: A typical load-displacement curve for the symmetrical lay-up system has a distinct decreasing load shift with increasing curing pressure, as shown in the Fig. 7. By increasing the curing pressure, resin depletion characteristic at the interlaminar fracture interface has been identified. This depletion of resin at the interlaminar interface reduces the localized volume of plastic strain absorption, thus resulting in a lower load necessary to fracture the specimen.
The Fig. 8 shows the failure mechanisms that occur during
initial elastic loading that develops into initiation and damage growth of the
DCB specimen. This initial microscopic fracture was observed before the characteristic
macroscopic fracture or delamination during the crack propagation.
||SEM micrograph of through thickness cross-section of laminate
||Load-displacement characteristics for symmetrical ply arrangement
||Failure mechanism about the crack tip of woven GFRP/unsaturated
delamination-resistance curve (R-curve) for the Mode-I fracture
fracture toughness versus curing pressure, using MBT|
Stable crack propagation was obtained when the crack propagated continuously
under the external loading.
The delamination-resistance characteristic curves or the R-curves had been characterized as shown in the Fig. 9 for the different curing pressure in the symmetrical of lay-up arrangements. The visually observed initiation GIC values were labeled VIS and the propagation values labeled as PROP on the graph. The characteristic R-curves for different curing pressures all indicated trends of approaching to a plateau. Fracture deformation at initiation values was very similar in all cases, while during propagation the separation of plies leads to formation of fiber bridging phenomenon. These mechanisms had contributed to the slow and stable crack propagation and extensive half-arm fibers bridged that were observed in specimens with lower curing pressure. As a result, higher GIC propagation values had been obtained as compared to increasing curing pressure. This phenomenon was the main contributor to the rising R-curve shown in the Fig. 9. The increasing R-curves with small displacements are generally created from the testing on thicker specimens. While the stable propagation value of GIC was a significance of fiber bridges breaking at roughly the same rate as they were created.
mode and mechanism of E-glass woven GFRP laminate|
The resultant interlaminar toughness with respect to the curing pressure had been characterized with respect to the initiation and propagation toughness values as in Fig. 10. The fracture toughness values of woven GFRP were computed using the analysis of MBT method. Comparatively, the GIC propagation values were significantly higher than the GIC initiation values. Since the mechanism of initiation toughness involves only a layer of resin-rich and fiber-free crack-front, which depletes as the curing pressure increases. The depletion also means that the volume of localized plastic zone size had been reduced, leading to the characteristic decreasing toughness with increasing curing pressure.
Fractographic analysis: A fractographic investigation was undertaken during and after the fracture of DCB specimens for each of the curing pressure of 35.8, 70.1, 104 and 138.2 kg m2. Thickness cross-section fractured along crack propagation distance of 15 mm was systematically observed and captured as shown in Fig. 11. The failure by separation of the two double-cantilever arm illustrates the mode of fracture mechanisms involving tow cracking, splitting and separation, fiber bridging and fiber breakage. The fracture profile obtained from the test shows that fiber bridging occurred as early as in the first 5 mm of the crack propagation. The phenomenon then changed with fiber breakage as the crack propagated reaches up to 15 mm. This kind of behavior has also been reported by Shivakumar et al. (2006).
Under the Scanning Electron Microscopic analysis, Fig. 12,
fiber bundle fracture and pullout on warp and weft direction of the fiber yarn
can be clearly reviewed. However, fractured surfaces remain predominantly brittle
matrix failure. Broken fiber ends indicating brittle failure mode and interfacial
failure between the fibers and resin was common on the fracture surface.
failure and fiber bunch pullout of woven GFRP laminate|
brittle fragmentation of fibers by cleavage load is indicated by the flat broken
fiber-ends. Similar evident characteristic is found with the resin-fractured regions.
Fiber bundle pullout was also scarcely observed, which indicates fiber-matrix
interfacial bond strength was exceeded before the tensile failure strength of
the composite was attained.
Studies of influence of different curing pressure in mode-I interlaminar fracture toughness of symmetrical ply arrangement of woven GFRP composite was conducted. Quasi-static mode-I critical fracture toughness of advanced woven E-glass fiber reinforced composite material was an important controlling parameter to the interlaminar failure. The investigation found that both initiation and propagation GIC values were consistently affected by the different curing pressure applied. Although the fiber weight fraction was not particularly significant. Fractured morphology has also indicated fiber bundle brittle fracture and pullout for the symmetrical ply arrangements. The study had provided a useful insight to different types of curing pressure as an important and useful controlling parameter to interlaminar failure of woven GFRP composites.
The authors wish to express their sincere appreciation to the Center of Materials and Minerals (CMM), Universiti Malaysia Sabah, for their support in this project.