Cold welding is the process of joining metals using pristine surfaces without contaminents(oxide layers, oil, dust...) using pressure through atomic diffusion between the metal joints. Why does this process not work with carbon alloys such as stainless steel? I've only seen 1 article mention it worked with low carbon steel but it cannot be done with higher carbon steels but no explanation as to why.

  • 1
    $\begingroup$ I guess the metal has to be at least somewhat soft and pliable, which ordinary steel just isn't. $\endgroup$ Jan 25 at 8:21
  • $\begingroup$ While it wouldnt exactly be cold welding anymore, would a modified process which heats the higher carbon steel to be more ductile work with pressure? Or is the carbon the real problem? $\endgroup$
    – Andi Iacob
    Jan 25 at 8:28
  • $\begingroup$ The key properties of many steels are based on careful creation of the domain structure of the bulk material (eg grain size and composition) and surface effects (eg strong oxide layers on the surface to protect against corrosion). They are often products of the specific production method which involves the heating history. Cold welding will alter the key properties in ways that don't preserve the key features of the original alloy. This matters a lot less in more uniform alloys, and might explain the differences. $\endgroup$
    – matt_black
    Jan 25 at 12:30
  • $\begingroup$ Its good for welding aluminum in the 7xxxx series though, which isnt exactly uniform. it just seems like carbon diffusion is the problem, probably due to ductility being decreased. But it would be great to have a research paper dedicated as to why atomic diffusion doesn't take place and in what circumstances it does. $\endgroup$
    – Andi Iacob
    Jan 25 at 16:51
  • 1
    $\begingroup$ The carbon inhibits flow at the grain boundaries. That said, cold welding still takes place, if you wait long enough. Clamp the cleaned parts under high vacuum and wait a few years. $\endgroup$ Jan 25 at 17:09

1 Answer 1


I guess (from having heard a few lectures about solid state inorganics, no real experience on the subject) that complex crystalline structures that form a polycrystalline surface just don't fit together sufficiently, crystallographically, with a second surface. Even if both are polished flat down to the atomic scale.

That is especially true if they are not just metallic, but have some of covalent character bonds in them too. The chance that evem a tiny fraction of domains on both sides would by chance fit together is paractically zero.

For the two pieces to form a continuous phase, a lot of surface atoms have to move around, and they're not going to do that if the jump to the next energetically favourable position is more than one or two atomic diameters.

  • $\begingroup$ So why domains of carbon steel are worse fit together? $\endgroup$
    – sa7
    Jan 26 at 22:33
  • $\begingroup$ They never fit perfectly together, also with plain iron. Unless you take two single crystals and polish and align them to atomic precision. The opposing domain/grain surfaces have to rearrange a bit to form a single, continous grain boundary between them. $\endgroup$
    – Karl
    Jan 28 at 14:50
  • $\begingroup$ Ok, so why carbon steel rearranges/aligns/fits/etc worse than plain iron? This was the question you are trying to answer. $\endgroup$
    – sa7
    Jan 28 at 16:31
  • $\begingroup$ @sa7 Because plain iron has a simple structure, where all lattice positions are more or less equivalent. $\endgroup$
    – Karl
    Jan 28 at 20:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.