Stress induced propagating crack

Impact Properties. Chemical Nature of the Rubber. If the rubber is too compatible with the matrix, it will dissolve in the rigid material and disperse on a molecular scale. Little or no reinforcement will occur since the rubber particles become smaller than the radius of the tip of a stress-induced propagating crack. However if it is highly incompatible, good adhesion between rubber and matrix cannot be obtained. For example polybutadiene rubber adheres poorly to a styrene/acrylonitrile copolymer, but a nitrile rubber adheres well to the SAN copolymer. If grafting techniques are used however, compatibility is less of a problem since the rubber is chemically bonded to the matrix. 

Thermal shock resistance can also be improved by the surface treatments used to increase mechanical strength (see 17.3.3.2). The stresses induced by thermal shock must exceed the compressive surface stresses to cause a crack to propagate, resulting in failure during quenching. 

Crack-opening mechanisms have been proposed that simply relate to the effect of environment and local alloy composition on the atom-to-atom bond strength at the crack tip. Reduction in this bond strength has been attributed to stress-induced changes in alloy composition as just described and to adsorption of atoms from the environment. Since dislocation movement is not considered in the mechanism, breaking bonds in the plane of the crack propagation leads to a cleavage-type rupture (Ref 159). 

This type of attack does not require any specific environment to take place, since it can take place simply in neutral or acidic wet environments. Failure due to hydrogen is named hydrogen embrittlement since it leads to a brittle-Uke fracture surface. Indeed, the ductility of the bulk metal does not change, but the propagation of the crack is due to the mechanical stresses induced in the lattice by hydrogen accumulated near the crack tip. If hydrogen is present in the metal lattice before the application of loads a delayed fracture may occur, i. e. the steel does not fail when the load is applied, but after a certain time. 

Various efforts have been made to estimate the amount of corrosion that will cause spalling. It has been shown that cracking is induced by less than 0.1 mm of steel section loss, but in some cases far less that 0.1 mm have been needed. This is a function of the way that the oxide is distributed (i.e. how efficiently it stresses the concrete), the ability of the concrete to accommodate the stress (by creep, plastic or elastic deformation) and the geometry of rebar distribution that may encourage crack propagation by concentrating, stresses, etc., for example, in a closely spaced series of bars near the surface, or at a corner where there is less confinement of the concrete to restrain cracking. 

Rapid cooling of a ceramic or a glass is more likely to inflict thermal shock than heating, since the induced surface stresses are tensile. And, as you know, crack formation and propagation from surface flaws are more probable when the imposed stress is tensile. 

When a material obeys linear elastic fracture mechanics, its tendency to undergo crack initiation or propagation as a result of mechanical stress can be assessed in terms of fracture toughness parameters, such as (critical stress intensity factor) or Gj, (strain energy release rate). Analogous parameters can be used with thermally induced cracking. 

---