The formation of cracks perpendicular to the loading direction is the first stage of deformation. The periodicity of these cracks is a sign of the interface strength in addition to film thickness and fracture strength of the film. One can notice this simply by comparing two polyimide samples with and without the Cr layer: Cr serves as an adhesion promoter between polyimide and Al coating, and hence, crack periodicity decreases upon the addition of Cr.

The second phenomenon is buckling taking place after periodic cracking. It starts at one side of a cracked strip and arrests at the other side as described in Figure 2. Its initiation and extend are again functions of interface strength. By considering shear stresses along the interface and observing at what strain buckling takes place, we have calculated stress intensity factor to be for an interfacial crack at the onset of delamination. (pure Mode II) [1].

Fracture studies on thin films can also be carried out under biaxial stress state. In order to accomplish this, a diaphragm with a thin coating is subjected to pressure. The bulge setup is employed for this purpose, where the laser sensor is replaced with a microscope. Cracking of the thin coating in the middle of the diaphragm, where the stress state is equi-biaxial, is observed as a function of applied pressure. Resulting crack patterns look like the one given in Figure 3.

Using the dependence of the energy release rate on
1) stress at which steady state propagation of isolated cracks takes place,
2) elastic mismatch between the substrate and the film,
3) elastic properties of the film,
4) the thickness of the film, one can calculate the energy release rate of the film. Using polyimide as the substrate, critical mode I stress intensity factor for 150-nm-thick Al films has been calculated to be[2].

We also investigate patterns exhibited by desiccation cracks [3]. Approaching such crack patterns with an emphasis on their basic elements, junctions, provides a clear methodology. The study aims to address both experimental difficulties encountered in drying experiments and the resulting inconsistencies. This is carried out in three steps as summarized in Figure 4.

1) First of all, a complete isolation of the initiation phase of the desiccation cracks from the propagation phase is achieved by introducing stress-raisers into the drying medium. Propagation can be completely suppressed if the distance between stress-raisers is kept adequately small.
2) A controlled stress state is created within the drying medium to understand mechanics of kinking during propagation. This is achieved by a controlled drying experiment, where a certain solvent concentration distribution is imposed in a suspension and resulting crack patterns are recorded. It is concluded that kinks, diversions from a straight path frequently observed under ambient conditions, can be regarded as adjustments of crack paths to inhomogeneous stresses around them.
3) Finally, the connection between kinking and junction formation during propagation is established. A microscopic evidence for the nucleation of the third arm in a kink site is obtained. This observation can be used to explain the broad histogram for junction angles during propagation.

[1] B. E. Alaca, M. T. A. Saif, and H. Sehitoglu,
"On the interface debond at the edge of a thin film on a thick substrate",
Acta Materialia 50(5), 1197-1209 (2002).

[2] B. E. Alaca, J. C. Selby, M. T. A. Saif, and H. Sehitoglu,
"Biaxial testing of nano-scale films on compliant substrates - fatigue and fracture",
Review of Scientific Instruments 73(8), 2963-2970 (2002).

[3] K. B. Toga and B. E. Alaca,
"Junction formation during desiccation cracking",
Physical Review E 74, 021405 (2006).