Madhu S. Madhukar

Fibre-matrix adhesion and its relationship to composite mechanical properties

Two major areas of enquiry exist in the field of fibre-matrix adhesion in composite materials. One is the fundamental role that fibre-matrix adhesion plays on composite mechanical properties. The other is what is the “best” method used to measure fibre-matrix adhesion in composite materials. Results of an attempt to provide an experimental foundation for both areas are reported here. A well-characterized experimental system consisting of an epoxy matrix and carbon fibres was selected in which only the fibre surface chemistry was altered to produce three different degrees of adhesion. Embedded single-fibre fragmentation tests were conducted to quantify the level of fibre-matrix adhesion. Observation of the events occurring at the fibre breaks led to the documentation of three distinct failure modes coincident with the three levels of adhesion. The lowest level produced a frictional debonding, the intermediate level produced interfacial crack growth and the highest level produced radial matrix fracture. High fibre volume fraction composites made from the same material were tested for on- and off-axis, as well as fracture, properties. Results indicate that composite results can be explained if both differences in adhesion and failure mode are considered. It will be further demonstrated that fibre-matrix adhesion is an “optimum” condition which has to be selected for the stress state that the interface will experience. The embedded single-fibre fragmentation test is both a valuable measurement tool for quantifying fibre-matrix adhesion as well as the one method which provides fundamental information about the failure mode necessary for understanding the role of adhesion on composite mechanical properties.

Fiber-Matrix Adhesion and Its Effect on Composite Mechanical Properties: II. Longitudinal (0°) and Transverse (90°) Tensile and Flexure Behavior of Graphite/Epoxy Composites:

An optimum level of interfacial bond strength between reinforcing fiber and a polymeric matrix in which it is placed is essential for acceptable composite mechanical properties and performance. The interfacial bond strength can be optimized only when the relationship between the level of fiber-matrix adhesion and the mechanical and fracture behavior of composites is clearly understood. This study establishes the relationship between the fiber-matrix interfacial shear strength and 0° and 90° tensile and flexure properties of graphite/epoxy composites. A well defined and characterized graphite fiber/epoxy system was chosen in which the level of adhesion between fiber and matrix was changed by using the same graphite fibers through the use of surface treatment and finish. The level of adhesion between the fiber and matrix associated with these changes resulted in an increase of fiber-matrix interfacial shear strength (ISS) by over a factor of two while the fiber and matrix properties remained unchanged. The experimental results demonstrated that the fiber surface modification did not have much effect on the tensile and flexural moduli and on the fiber dominated properties. However, the strengths and maximum strains that are governed by the matrix and interface properties were highly sensitive to the fiber surface modification. In addition, the major failure modes were also found to be affected by the fiber-matrix interfacial shear strength.

Fiber-Matrix Adhesion and Its Effect on Composite Mechanical Properties: IV. Mode I and Mode II Fracture Toughness of Graphite/Epoxy Composites

To optimize the level of fiber-matrix adhesion an understanding of the relationship between fiber-matrix interfacial bond strength and the mechanical and frac ture behavior of composites is essential. This study establishes the relationship between fiber-matrix interfacial shear strength (ISS) and interlaminar fracture toughness (both Mode I and Mode II) and failure modes for graphite/epoxy composites. A well defined and characterized graphite fiber/epoxy system was chosen in which the level of adhesion be tween fiber and matrix was changed by using the same graphite fibers with different sur face treatments. These surface treatments changed the level of adhesion between the fiber and matrix thus resulting in an increase of the fiber-matrix ISS by over a factor of two while the fiber and matrix properties remained unchanged. The Mode I and Mode II tests were conducted by the double cantilever beam (DCB) and end-notch flexure (ENF) tests methods, respectively. The Mode I fracture toughness (GIC ) of composites having low fiber-matrix ISS could not be determined from the DCB test because of extensive fiber bridging and crack meandering. For the composites having higher values of the ISS, the GIC increased with the ISS. The experimental results demonstrated that there is a strong dependency of Mode II fracture toughness (GIIC ) on fiber-matrix adhesion. Increased fiber-matrix adhesion in one group of composites significantly improved the GIIC , but the presence of brittle interphase around graphite fibers in another group of composites tended to cancel part of the improvement resulting from increased adhesion. Based on the major failure modes occurring during the Mode I and Mode II loading conditions, a causal link age between fiber-matrix adhesion and interlaminar fracture behavior of graphite/epoxy composites is established.

Fiber-Matrix Adhesion and Its Effect on Composite Mechanical Properties: I. Inplane and Interlaminar Shear Behavior of Graphite/Epoxy Composites

An experimental investigation was performed to establish the relationships between fiber-matrix adhesion as determined by single fiber interfacial shear strength tests with the inplane and interlaminar shear properties of graphite/epoxy composites. ±45°-tension, Iosipescu, and short beam shear tests were conducted on three identical sets of composites differing only in their fiber-matrix interfacial shear strength. The fiber-matrix interphase and consequently the interfacial shear strength was varied by using the same graphite fiber was different surface modifications, namely untreated, surface treated, and surface treated and coated with a thin layer of epoxy. The surface modification changed the interfacial shear strength by more than a factor of two, while the properties of fibers remained unchanged. The experimental results showed that both inplane and interlaminar shear strengths increased approximately in the same ratio as the interfacial shear strength, however, the inplane shear modulus was relatively insensitive to the fiber surface modifications. The fracture surface analysis revealed that when the fiber matrix interfacial shear strength was increased from low to intermediate to high values, the major failure modes changed from primarily interfacial failure to a combination of interfacial and matrix failure to primarily matrix failure, respectively, in a manner identical to that observed with the single fiber fragmentation tests.

A New Method to Reduce Cure-Induced Stresses in Thermoset Polymer Composites, Part III: Correlating Stress History to Viscosity, Degree of Cure, and Cure Shrinkage

Non-thermoelastic effects such as cure shrinkage of a polymer can play a role in residual stresses in composite parts. Studies have shown that cure shrinkage can place significant stresses on fibers. Therefore, the cure cycle of 3501-6 epoxy resins was modified to change its cure shrinkage characteristics to minimize the stresses. New cure strategies were developed using volumetric dilatometry, differential scanning calorimetry, dielectric cure monitoring, and a unique single fiber stress test method. Cure cycles were modified to balance the resins thermal expansion with its cure shrinkage. In some cases, a region of constant volume was achieved for a short time. However, the cure shrinkage eventually dominated over thermal expansion in all cycles as the polymer gelled. Changing the cure cycle affected the degree of cure at the point where the fiber/matrix interface developed as well as the amount of cure shrinkage occurring afterwards. A higher degree of cure at this point leads to longer stress relaxation time. Furthermore, less cure shrinkage at this point leads to less stress on the fibers. Also, slow heating rates allow more time for the polymer to relax and relieve stresses caused by cure shrinkage. Finally, a cure cycle that minimizes stresses due to cure shrinkage has been demonstrated.

A New Method to Reduce Cure-Induced Stresses in Thermoset Polymer Composites, Part I: Test Method

The inhomogeneous structure of polymeric composites causes internal stresses to develop due to matrix volume changes during processing. The volume changes occur during cure and during cooldown after the cure is completed. Most of the previous studies on residual stresses concentrated on stress development during cooldown. In this study, a new test method was used to monitor fiber stresses that develop during cure in single fiber model composites. The method was used to study the effect of changing the cure cycle on curing induced fiber stresses. It was seen that changing the cure cycle changes the resulting stress significantly. Also, it was shown that the cure-induced stresses and their contribution to final residual stresses vary for different resins. A cure cycle with almost zero cure-induced stresses is demonstrated. The new cycle was found to satisfy the cure requirements such as glass transition temperature and cure cycle duration.

The Effects of Fiber Surface Modification and Thermal Aging on Composite Toughness and Its Measurement

A detailed experimental study was conducted to establish the structure-property relationships between elevated temperature aging and (1) fiber-matrix bonding, (2) Mode II interlaminar fracture toughness, and (3) failure modes of carbon fiber/PMR-15 composites. The fiber-matrix adhesion was varied by using carbon fibers with different surface treatments. Short beam shear tests were used to quantify the interfacial shear strength afforded by the use of the different fiber surface treatments. The results of the short beam shear tests definitely showed that, for aging times up to 1000 hr, the aging process caused no observable changes in the bulk of the three composite materials that would degrade the shear properties of the material. Comparisons between the interlaminar shear strength (ILSS) measured by the short beam shear tests and the G,,, test results, as measured by the ENF test, indicated that the differences in the surface treatments significantly affected the fracture properties while the effect of the aging process was probably limited to changes at the starter crack tip. The fracture properties changed due to a shift in the fracture from an interfacial failure to a failure within the matrix when the fiber was changed from AU-4 to AS-4 or AS-4G. There appears to be an effect of the fiber/matrix bonding on the thermo-oxidative stability of the composites that were tested. The low bonding afforded by the AU-4 fiber resulted in weight losses about twice those experienced by the AS-4 reinforced composites, the ones with the best TOS. The results are in agreement with those of previous work completed by the authors.

Adhesion to carbon fiber surfaces: surface chemical and energetic effects

Abstract IM6 carbon fibers have been surface treated in steps to 20, 200 and 600% as well as the nominal 100% surface treatment. Quantifiable changes in the surface oxygen and nitrogen composition as well as in the surface energetics have been measured and compared against the untreated IM6 fiber. Single fiber fragmentation testing of these fibers in an epoxy matrix has shown that there is a direct effect of the surface chemical changes on fiber-matrix adhesion. Composites were fabricated from these same materials, and interlaminar shear, transverse flexural and Mode II fracture toughnesses were measured. Results showed that the composite shear and transverse flexural strengths were significantly increased after only the 20% surface treatment level, with further increases in the surface treatment level having only a minor effect. However, Mode II fracture toughness continued to increase with increasing surface treatment.

A New Method to Reduce Cure-Induced Stresses in Thermoset Polymer Composites, Part II: Closed Loop Feedback Control System

A closed loop feedback control system has been developed to obtain cure cycles which reduce the cure-induced stresses in single fiber/matrix model composites. In this feedback system, thermal expansion and stress relaxation are used to counteract the stresses resulting from the chemical shrinkage. The feedback system is applied to two different fibers (carbon and glass) in four different resins [3501-6, 934, and 977-3 epoxies and 5250-4 bismaleimide (BMI)]. The completeness of the cure is verified by comparing glass transition temperature (Tg) for polymers cured using the standard cure cycles (cycles recommended by the prepreg manufactures) and the feedback cure cycles. The durations of the feedback cure cycles are within that of the standard cycles. Test results indicate that feedback cure cycles reduce the curvature of unsymmetrical laminates as a measure of residual stresses.

Thermo-Oxidative Stability and Fiber Surface Modification Effects on the Inplane Shear Properties of Graphite/PMR-15 Composites

Experiments were conducted to study the effects of thermo-oxidative stability (weight loss) and fiber surface modification on the inplane shear properties of graphite/PMR-15 unidirectional composites. The isothermal aging was conducted at 316°C and up to 1000 hours of aging times. The role of fiber surface treatment on the composite degradation during the thermo-oxidative aging was investigated by using A-4 graphite fibers having three different surface modifications, namely untreated (AU-4), surface treated (AS-4), and surface treated and sized with epoxy-compatible sizing (AS-4G). Weight loss of matrix, fibers, and composites was determined during the aging. The effect of thermal aging was seen in all the fiber samples in terms of their weight loss and reduction in fiber diameter. Calculated values of weight loss fluxes for different surfaces of rectangular unidirectional composite plates showed that the largest weight loss occurs at those cut surfaces where fibers are perpendicular to the surface. Consequently, the largest amount of damage was also noted on these cut surfaces. Optical observation of neat matrix and composites plates subjected to the different aging times revealed that the degradation (such as matrix microcracking, void growth, etc.) occurred within a thin surface layer near specimen edges. The inplane shear modulus of the composites was unaffected by the fiber surface treatment and the thermal aging. The shear strength of the composites having the untreated fibers was the lowest and it decreased with aging. Fracture surface examination of the composites having untreated fibers suggests that the weak interface allows the oxidation reaction to proceed along the interface and thus expose the inner material to further oxidation. The results indicate that fiber-matrix interface affects the composite degradation process during its thermal aging and that the weak interface accelerates the composite degradation.