Composite load transfer

Example 3.17 Short carbon fibres with a diameter of 10 fim are to be used to reinforce nylon 66. If the design stress for the composite is 300 MN/m and the following data is available on the fibres and nylon, calculate the load transfer length for the fibres and also the critical fibre length. The volume fraction of the fibres is to be 0.3. 

Net-tension failures can be avoided or delayed by increased joint flexibility to spread the load transfer over several lines of bolts. Composite materials are generally more brittle than conventional metals, so loads are not easily redistributed around a stress concentration such as a bolt hole. Simultaneously, shear-lag effects caused by discontinuous fibers lead to difficult design problems around bolt holes. A possible solution is to put a relatively ductile composite material such as S-glass-epoxy in a strip of several times the bolt diameter in line with the bolt rows. This approach is called the softening-strip concept, and was addressed in Section 6.4. 

Optimum mechanical properties in composite materials are strongly related to the efficiency of load transfer. 

Proper reinforcement of rubber matrix using hllers can be achieved only if there exists adequate adhesion between the hller and the mbber. Rubber-mbber adhesion and rubber-hller adhesion both without and with adhesion promoters have been studied extensively [125-127]. Fiber-matrix adhesion in short fiber-rubber composites is always a field of extensive research. If the fibers are not bonded properly with the rubber matrix, fibers will shde past each other under tension deforming the matrix, thereby reducing the strength properties. In the case of short fiber-reinforced rubber composites, loads are not directly applied to the fibers, but are apphed to the matrix. To obtain a high-performance composite, the load must be effectively transferred to the fibers, which is possible only when the fiber-matrix interphase is sufficiently strong. In addition, the adhesion between the fiber and the matrix should be such that the failure occurs in the matrix rather than at the interphase [92]. 

In the macrocomposite model it is assumed that the load transfer between the rod and the matrix is brought about by shear stresses in the matrix-fibre interface [35]. When the interfacial shear stress exceeds a critical value r0, the rod debonds from the matrix and the composite fails under tension. The important parameters in this model are the aspect ratio of the rod, the ratio between the shear modulus of the matrix and the tensile modulus of the rod, the volume fraction of rods, and the critical shear stress. As the chains are assumed to have an infinite tensile strength, the tensile fracture of the fibres is not caused by the breaking of the chains, but only by exceeding a critical shear stress. Furthermore, it should be realised that the theory is approximate, because the stress transfer across the chain ends and the stress concentrations are neglected. These effects will be unimportant for an aspect ratio of the rod Lld> 10 [35]. 

Chamis, C.C. (1974). Mechanics of load transfer at the interface. In Interfaces of Polymer Matrix Composites, Composite Materials, Vol. 6, Ch. 2, (E.P. Plueddemann ed.). Academic Press, New York. 

Vautey, P. and Favre, J.P. (1990). Fiber/matrix load transfer in thermoset and thermoplastic composites-single fiber models and hole sensitivity of laminates. Composites Sci. Technol. 38, 271-288. 

Daabin, A., Gamble, A.J. and Sumner, N.D. (1992). The effect of the interphase and material properties on load transfer in fiber composites. Composites 23, 210-214. 

The methods quoted under a) above give bulk information only although they may be used in conjunction with composite models to test theories of the microstructure. The methods under b) are more closely related to structural elements. It is an interesting fact that even at the atomic level, displacement and therefore strain can be measured by several means (although average values are of course obtained) yet loads are only measurable in terms of the secondary effects they produce, for example elastic or inelastic displacements, strain-related optical effects or electronic transitions detectable by optical or infrared spectroscopy. The problem of load transference in a polymer is of great interest, yet very few methods exist by which it may be studied. 

A second method consists in observing the composite fracture surfaces. For example, in the case of MWNTs in a poly (hydroxy-aminoether) matrix, C. Bower et al. (70) observed a lot of pulled out nanotubes and concluded that the load transfer from polymer to nanotube was not sufficient to fracture the nanotubes. In the same time, a lot of kinked MWNTs were observed, which were believed to be plastically deformed. 

Fabrication methods have overwhelmingly focused on improving nanotube dispersion because better nanotube dispersion in polyurethane matrix has been found to improve the properties of the nanocomposites. The dispersion extent of CNTs in the polyurethane matrix plays an important role in the properties of the polymer nanocomposites. Similar to the case of nanotube/solvent suspensions, pristine nanotubes have not yet been shown to be soluble in polymers, illustrating the extreme difficulty of overcoming the inherent thermodynamic drive of nanotubes to bundle. Therefore, CNTs need to be surface modified before the composite fabrication process to improve the load transfer from the polyurethane matrix to the nanotubes. Usually, the polyurethane/CNT nanocomposites can be fabricated by using four techniques melt-mixing (15), solution casting (16-18), in-situ polymerization (19-21), and sol gel process (22). 

The gap between the predictions and experimental results arises from imperfect dispersion of carbon nanotubes and poor load transfer from the matrix to the nanotubes. Even modest nanotube agglomeration impacts the diameter and length distributions of the nanofillers and overall is likely to decrease the aspect ratio. In addition, nanotube agglomeration reduces the modulus of the nanofillers relative to that of isolated nanotubes because there are only weak dispersive forces between the nanotubes. Schadler et al. (71) and Ajayan et al. (72) concluded from Raman spectra that slippage occurs between the shells of MWNTs and within SWNT ropes and may limit stress transfer in nanotube/polymer composites. Thus, good dispersion of CNTs and strong interfacial interactions between CNTs and PU chains contribute to the dramatic improvement of the mechanical properties of the... 

A number of studies on CNT-polymer composites have focused on improving the dispersion and load transfer efficiency in other words the compatibility between the CNTs and polymer matrix through covalent chemical functionalization of CNT surface (12,40). Many of the studies reported above have used acid-functionalized CNTs to fabricate MWCNT-PMMA composites with improved mechanical properties using different processing methods (24,25,27,62). Yang et. al (68) modified the acid functionalized CNTs with octadecylam-ine (ODA) to obtain ODA-functionalized CNTs. These CNTs were reinforced in a copolymer P(MMA-co-EMA) to form composites with improved dispersion and mechanical properties. 

---