Interfacial cracking

Sakai M., Takeuchi S., Fischbach D.B. and Bradt R.C. (1988). Delamination toughening from interfacial cracking in ceramics and ceramic composites. In Proc. Ceramic Microsiruclures 86 (J.A. Pask and A.G. Evans, ed.), Plenum Press, New York, pp. 869-876. 

After polishing, the specimens were stored in a desiccator until testing. Once mounted in the specimen holder, the specimen was scanned for fibers that were spaced a minimum of 2 /rm and a maximum of one half of their diameter from the nearest neighbor. Matrix conditions near the fiber of interest were also considered. Fibers near voids, cracks, or surface damage or those fibers that showed a pre-existing interfacial crack were rejected. Once a suitable fiber end was located, its diameter and distance to the nearest neighbor were entered for data reduction purposes. 

The fiber modulus and matrix shear modulus are also required for the analysis. The fiber s coordinates are recorded directly from the stage controllers to the computer. The operator begins the test from the keyboard. The x and y stages move the fiber end to a position directly under the debonder tip the z stage then moves the sample surface to within 4 yum of the tip. The z-stage approach is slowed down to 0.04 jan/step at a rate of 6 steps/s. The balance readout is monitored, at a load of 2 g the loading is stopped, and the fiber end returned to the field of view of the camera. The location of the indent is noted and corrections are made, if necessary, to center the point of contact. Loading is then continued from 4 g in approximately 1 g increments. Debond is determined to have occurred when an interfacial crack is visible for 90-120° on the fiber perimeter. The load at which this occurs is used to calculate the interfacial shear stress at debond. 

IFSS of the sized fibers vs. the bare. These results indicate that the application of the sizing and the consequent formation of the interphase have resulted in an increase in the level of fiber-matrix adhesion. It could reasonably be expected that under the multiaxial state of stress at the fiber-matrix interphase, stronger fiber-matrix adhesion would reduce the tendency to grow an interfacial crack, thereby placing more energy into driving the matrix crack as observed for the sized fiber. 

Kendall, K. (1975), The effects of shrinkage on interfacial cracking in a bonded laminate , Journal of Physics D Applied Physics, 8, 1722-1732. 

Kolhe, R., Wi, C.Y.I., Ustandag, E. and Sass, S.L., Residual thermal stresses and calculation of the critical metal particle size for interfacial crack extension in metal-ceramic matrix composites , Acta Mater, 1996 44(1) 279-287. 

J. Schmittbuhl et al Roughness of interfacial crack fronts Stress-weighted percolation in the damage zone. Phys. Rev. Lett. 90, 045505 (2003)... 

The results on different PECVD films, showed that the failure of the systems was different, depending on the associated materials and on the interfacial behavior when the debonding occurred with buckling, or by interfacial crack deviation, the transverse crack saturation was reached at lower strains. No... 

The simulations in this paper give failure modes sequences very similar to the actual ones observed in the experiments. The model predicts the formation of shear-dominated inter-layer (or interfacial) cracks that initiate first and that such cracks grow very dynamically, their speeds and shear nature being enhanced by the large wave mismatch between the core and the face sheet. The triggering of the complex mechanism of the intra-layer failure of the core structure is also well reproduced. 

The transition from simple chain scission to crazing is also confirmed by TEM observations of interfacial cracks between PS and PVP in thin films [28] (as... 

After buckling, at a critical compressive stress in the scale, an interfacial crack starts from the periphery of the buckled area which leads to an increase of the delaminated area. The spalling of the scale occurs by the deflection of the interface crack toward... 

It is clear that if the criterion (la) or (3a) holds for interfacial separation before fibril drawing (whether on account of AG being small or Oy being large) then there will not be a large dissipation of energy,, per unit advance of the macroscopic separation front, and there will be very little work for the applied force to do. So the force required to open up an interfacial crack, in brittle separation, will be small ... 

Two representative probe test curves for the detachment of an SIS adhesive from steel and from EP surfaces are shown in Pig. 22.17 while the initial portion of the curve is identical, the force drops rapidly to zero for the EP surface, and never forms the characteristic fibrillar plateau observed on steel surfaces. How does this happen As qualitatively described by Creton et al. [55] for a detachment from a polydimethylsiloxane layer, when the resistance to crack propagation is low, cavities are nucleated (around the peak stress) and then propagate as interfacial cracks at the interface between the probe and the adhesive, and eventually coalesce. This process of crack propagation and coalescence is responsible for the sharp drop in force observed in Fig. 22.17 for the EP surface and occurs at rather low values of nominal strain. In this case no formation of the characteristic foam stracture responsible for the high debonding energy is observed. 

This balance between interfacial propagation and bulk deformation has been described for linear elastic materials [56] and results from the competition between two mechanisms the velocity of propagation of an interfacial crack, which is controlled by the critical energy release rate Gc, and the bulk deformation, which is controlled by the cavitation stress and hence essentially by the elastic modulus E or G. In the hnear elastic model, the key parameter is the ratio GJE, which represents the distance over which an elastic layer needs to be deformed before being fuUy detached from the hard surface. This model has been verified experimentally for elastic gels [57]. 

However, for PSA layers we need to introduce two modifications which complicate the analysis the adhesives are both viscoelastic and strained in the nonlinear elastic regime. In other words the term Gc will include a dissipative term and the term E should be replaced with a high-strain equivalent controlled by the nonlinear elastic properties as shown in Fig. 22.14 and 22.15. As a result, the crack front will not have the same shape as the classical interfacial crack and the exact nature of the stress distribution at the crack tip will be unknown. 

The value of the dissipative factor tan S=G IG for the four adhesive blends is shown in Fig. 22.21. Clearly the addition of diblock in the blend has no effect on the dissipative properties of the adhesives at high frequencies but it has a significant effect at low frequencies. A relaxation experiment such as that described in Fig. 22.20a and b involves a very slow growth of interfacial cracks, precisely in the regime where linear viscoelastic properties differ. We can therefore propose, at least qualitatively, that the spectacular improvement in adhesive properties observed for the high diblock adhesives on EP surfaces is due to their more dissipative character, which slows crack propagation considerably at the interface, therefore avoiding early coalescence between separate cavities, and favors the formation of a fibrillar stracture with the cavity walls. 

Failure in simple shear is still more complex. An approximate treatment for an interfacial crack, starting at one edge, yields a relation analogous to Eq. (10.5) ... 

T. Akfaolu, M. Tokyay and T. elik Effect of Coarse Aggregate Size on Interfacial Cracking under Uniaxial Compression, Materials Letters Vol. 57, No. 4 (2002), p. 828-833. 

---