strength
Strength is the stress a material can carry before it breaks, and for ceramics it is both the reason they are wanted and the reason they are feared. A ceramic can be extraordinarily strong: some silicon nitrides survive over 1000 MPa, more than many steels. Yet quoting a single strength for a ceramic is misleading, because unlike a metal, a ceramic does not have one dependable strength. It has a strength that is set by its worst flaw, so it scatters from part to part, differs between tension and compression, and shrinks as the part gets bigger. Strength is a distribution, not a number.
Because ceramics are so hard to grip and pull without breaking at the grips, their strength is usually measured in bending rather than in pure tension. A bar is loaded on three or four points until it snaps, and the peak surface stress at fracture is reported as the flexural strength or modulus of rupture. Fracture mechanics ties strength directly to toughness and flaw size: sigma_f = K_IC / (Y times sqrt(pi times c)), where K_IC is the fracture toughness, Y a geometry factor near one, and c the size of the worst flaw. This is why strength depends on the flaw, not the bond. The theoretical strength set by the bonds is about E/10, but real ceramics reach only around one part in a hundred of that, because a pre-existing flaw always fails first.
The tension-compression asymmetry is central to using ceramics well. In tension, flaws are pulled open into running cracks, so tensile strength is low; in compression, flaws are squeezed shut, so compressive strength is typically ten to fifteen times higher. That is why ceramics live in compression: brick arches, prestressed structures, and dies. Because strength is flaw-controlled and scattered, designers never trust a single measured value; they characterise the whole distribution with Weibull statistics, apply a size correction, and often proof-test critical parts to guarantee a minimum. Always ask of any quoted ceramic strength: measured how, at what volume, in tension or bending?
The same alumina grade may be quoted as 300 MPa in tension but 350 to 400 MPa in a four-point bend test, and over 3000 MPa in compression. None of these is the strength; each is what that flaw population yields under that particular loading and stressed volume.
Strong in compression, weak in tension, and higher in bending than in pure tension: one ceramic, many strengths.
There is no single strength for a ceramic. It depends on the worst flaw, the stressed volume, the surface finish, and how it is loaded, so bend strength always reads higher than tensile strength for the same material. Treat any quoted number as one point on a distribution.