Hacker News Reader

Crack kills.

Metals are able to deform plastically. A metallic bond is not so directional, and you typically thing of all the molecules in a metal as a big blob with a cloud of electrons that can freely move about (also the reason for good electrical conductivity).

There are "elastic" and "plastic" deformations. Think about bending a paperclip. Use it gently, and you can open it up and get it to squeeze a few papers together. Pull it off and it returns to its original shape. That's elastic. Now open it wider or bend it into another shape, and it will no longer return to the original shape. That's a plastic deformation.

Metals can deform plastically which means layers of atoms are sliding past one another in waves. Sometimes these waves are stopped by various irregularities in the crystal lattice. The waves start to pile up, preventing further plastic deformation. Keep bending that paperclip back and forth; it will eventually snap. But in the meantime, it absorbed a fair amount of energy. That's "toughness".

In a ceramic, the electrons are not free to move about. If a crack develops, it doesn't take much for there to be terrific stress at the tip of the crack (much like a lever and fulcrum). The crack propagates through the material, and you have catastrophic failure.

There are various ways to increase the toughness of ceramics. I'm not all that familiar with them, but a common one is Yttria-stabilized zirconia. That's zirconium oxide with added yttrium. Under stress from a crack, the material can undergo a phase transformation (from one crystal structure to another; not solid to liquid or anything like that). A phase transformation absorbs energy (just like boiling water into steam does) which may be enough to stop the crack from propagating. My ceramics professor had a YSZ hammer that was gifted to him by Coors ceramics (same family as the beer company). YSZ is not transparent however.

http://en.wikipedia.org/wiki/Yttria-stabilized_zirconia

> Metals are able to deform plastically. A metallic bond is not so directional, and you typically thing of all the molecules in a metal as a big blob with a cloud of electrons that can freely move about (also the reason for good electrical conductivity).

Also the reason why metals absorb electromagnetic radiation on a very broad spectrum.

Which means it's very difficult, if not impossible, to make materials that behave mechanically like metals (hard, yet not fragile) but are transparent.

> My ceramics professor had a YSZ hammer that was gifted to him by Coors ceramics (same family as the beer company).

Slightly wandering off topic, but I thought it's interesting to note that Coors was also given the contract to construct the ceramic fuel elements for the nuclear ramjet in the Pluto nuclear-powered cruise missile. And now the name is pretty much just known for crappy beer.

What you're taling about is the field of fracture mechanics. Engineers have spent countless hours trying to figure out how cracks propagate in different materials. It turns out different types of cracks behave differently (shocker). It's a field applied not only to ceramics, but also to metals other materials.

But defects are also a big part of how metals deform. If you calculate the theoretical strength of a perfect crystal of, say, aluminum you get a strength of several times what it's actual strength is. This is because real crystals have defects at the atomic level - misalignments in the crystal structure that allow for the metal to deform more easily. These are called "dislocations" and are similar to cracks, except they're not - as I said these are atomic-level "shifts" in crystal structure. The misalignment allows groups of atoms to slide more easily by other groups.

Interestingly, adding dislocations to a metal will decrease it's strength- to a point. But when you add too many dislocations, they interfere with each other and actually increase the strength of the metal. This is how materials like copper work harden (When you deform them, you create dislocations. It's also why when you heat them up, they soften again - the heat allows the crystal structure to resettle in a more organized structure, decreasing dislocations.

The article actually hinted at it, at the second to last "section" when they start talking about failure and failure mode of glasses (and gorilla glasses) as well as tantalizing hint when they talk about glass strengthening (whether heat-based or chemical).

> I am not entirely sure why, but what's happening is that in most ceramics, a scratch creates a place where the stresses inside the material get amplified in a way that doesn't exist with metals.

Wouldn't that be because the strength of strengthened glasses come from these internal stresses in the first place? Chips and scratches mean stresses concentrate on the weak point which ends up blowing.

Also, you're incorrect that this does not affect metals, it does (the Silver Bridge failed due to stress corrosion cracking) but it's usually less flagrant or quick than in glasses or ceramics. Stress fractures in metal also tend not to yield explosive releases, so they're less impressive.

> It's more obvious to see why wood doesn't behave this way -- it's made of lots of long strands glued together, so when you crack some of those strands, the crack "runs into" the glue and doesn't propagate. We've actually already stolen this idea from wood -- as fiberglass. Plastics have these really long molecules which also line up like the wood fibers and create a very irregular environment for cracks to propagate, so that may be why plastics resist fracture.

One of the most impressive materials on that front is teeth which are highly resistant to crack growth due to material structure (enamel is essentially full of cracks which tends to block and hinder propagation of new cracks) and layering

Ceramic materials like glass have high bulk modulus and hardness. The forces at the tip of a crack are much higher in these high-hardness materials so despite being harder to bend they have poor toughness [1]. That is cracks propagate much more easily. The likelihood of your glass breaking when you drop it is almost entirely governed by the small scratches existing before you drop it - if they are bigger then there will be more stress concentrated and one or more will propagate catastrophically across the glass. New glass is substantially tougher than old glass.

In comparison, metals can absorb the energy in cracks through several mechanisms ranging from the movement and formation of defects in the atomic arrangement in the metal crystals to large scale plastic flow.

[1] http://en.wikipedia.org/wiki/Fracture_toughness

> I am not entirely sure why, but what's happening is that in most ceramics, a scratch creates a place where the stresses inside the material get amplified in a way that doesn't exist with metals. So in metals those cracks also don't like to propagate much; it's really in glass where you see cracks propagate through the whole material.

Covalent versus metallic bonds between atoms.

You can model glass like a bunch of tiny spheres (atoms) joined by sticks (covalent bonds). You break a stick by moving two balls, it tends to stay broken most of the time.

You can model metals like a bunch of tiny spheres (atoms) submerged in a glue-y fluid (cloud of shared electrons that makes the metallic bonds). You can move the balls around within that sticky cloud, to some extent.

The comparison is very, very metaphorical, but it gives an idea why metals are more resilient to cracking.