3 ms·

As I understand it's not bonding chemically. "The process does not melt either metal, instead it plasticizes the surfaces of both metals, causing them to come

by elnate 13y ago

As I understand it's not bonding chemically. "The process does not melt either metal, instead it plasticizes the surfaces of both metals, causing them to come into intimate contact sufficient to create a weld. "

icegreentea 13y ago
You don't need to melt the metal for there to be "chemical bonding". And now for me to probably mangle what I remember from material science. When you're dealing with metals (and alloys) you're pretty much dealing with a lattice of atoms. It's not very useful to talk about individual atoms "chemically bonding" with each other, or at least not the way you would talk about say... bounding sodium and chlorine to get table salt, or nitrogen with oxygens to get nitrates. Instead, you have this lattice of metal atoms, within which you can incorporate other atoms. Say you had a lattice of iron. By incorporating other metals into the lattice (say magnesium, or nickle), or even other non-metallic elements (most famously carbon) into the lattice, then you get different alloys. But at no time are you actually bonding say an iron atom with a carbon atom making a "steel molecule". So when you weld dissimilar metals/alloys, what you're trying to do is to get two dissimilar lattices to "blend" into each other. When you have a crappy weld, one of the things that happen is you get a crappy boundary layer (potentially of stuff like metal oxides) sitting between the two lattices. Whereas if you have a "perfect" weld, the two lattices will seamlessly merge into each other - you'll get this diffusion zone. In the case of two similar metals, then the two lattices can actually perfectly merge - that is what Feynman is trying to get at there. So, back to the point. Creating a good weld is in many ways as close to "chemically bonding" as you can get with metals / lattice structured materials. Or at least, in a useful sense.

by elnate 13y ago

icegreentea 13y ago

You don't need to melt the metal for there to be "chemical bonding". And now for me to probably mangle what I remember from material science. When you're dealing with metals (and alloys) you're pretty much dealing with a lattice of atoms. It's not very useful to talk about individual atoms "chemically bonding" with each other, or at least not the way you would talk about say... bounding sodium and chlorine to get table salt, or nitrogen with oxygens to get nitrates. Instead, you have this lattice of metal atoms, within which you can incorporate other atoms. Say you had a lattice of iron. By incorporating other metals into the lattice (say magnesium, or nickle), or even other non-metallic elements (most famously carbon) into the lattice, then you get different alloys. But at no time are you actually bonding say an iron atom with a carbon atom making a "steel molecule". So when you weld dissimilar metals/alloys, what you're trying to do is to get two dissimilar lattices to "blend" into each other. When you have a crappy weld, one of the things that happen is you get a crappy boundary layer (potentially of stuff like metal oxides) sitting between the two lattices. Whereas if you have a "perfect" weld, the two lattices will seamlessly merge into each other - you'll get this diffusion zone. In the case of two similar metals, then the two lattices can actually perfectly merge - that is what Feynman is trying to get at there. So, back to the point. Creating a good weld is in many ways as close to "chemically bonding" as you can get with metals / lattice structured materials. Or at least, in a useful sense.