It is known that brittle materials tend to fail under normal stresses, that is, tensile or compressive stresses. Brittle materials often are bound together into crystals by ionic bonds between atoms, with alternating positive and negative charged atoms.
I remember also that brittle materials fail when ionic bonds between layers are broken and the layers shift, causing like-charges to move close and therefore repel each other. The above picture I found illustrates this.
But wouldn't this situation happen more likely in shear, instead of compression or tension? If we had a cube of brittle material, wouldn't a shearing stress effectively shift those atomic layers in opposite directions, bringing like charges together, just like in the above picture? But apparently brittle materials fail more likely in tension or compression, and that's why maximum principal stress failure theory is used for brittle materials. This seems contradicting to me. Wouldn't a compressive stress, for example, simply squeeze the sample of material together instead of moving its layers?