Case Stories
Cobalt substitution problem
Cobalt is used in cutting tools that companies such as Sandvik Coromant produce. Cobalt is used as the binder phase in cemented carbides.
The problem these companies might face in the near future is that legislation will not permit the use of this element in the production of these tools because it is cancerogenic – says Erik Holmström, technical area manager for material modelling in Sandvik Coromant.
These tools are made of cemented carbides, a composite of tungsten carbide grains glued by cobalt. The need of finding a replacement for cobalt is then urgent for these companies.
What are the challenges of replacing cobalt in cutting tools?
Well, first one needs to understand what is the property or properties of the material one needs to replace that makes the product good. In this case, cobalt has been used in the cemented carbides since its discovery in the 1900 and not much variation has been made. The reason is because it works and it works well. But why cobalt makes a good glue it is not clear which makes it very challenge since it is not clear which property one needs to mimic.
Traditional approaches call for a search for the substitution among the elements in the periodic table that are chemically similar to the element one wants to replace. In this case that would be iron or nickel. Unfortunately, these are not good enough candidates. One can try by alloying several elements, but which and in which composition? The possibilities are enormous.
Testing many possibilities, sounds like a job for a computer?
Computational modelling can largely help in this search, to narrow down material candidates to a number where it is possible to run experiments in order to determine which materials are the best. Also one can investigate different properties to understand which are the ones we need to mimic to obtain a Co-like behavior.
What was the biggest lesson learned by working in collaboration with a computational modeling group?
The most important lesson is to realize that many properties in materials are governed by how the electrons behave and therefore to be able to model and understand atomic interactions is very relevant. In the case of cobalt, we understood that certain physical properties are important and we could calculate them and compare them in different Co-based alloys to understand how it worked. We could also correlate these properties to functions in the product to gain in performance.
The results were published in Acta Materialia 136, 215-223 (2017).
Was this a work done with the modelling group at Sandvik?
This was a collaborative work between Kungliga Tekniska Högskolan (KTH, the Royal Institute of Technology) and Sandvik Coromant. The important point is that Sandvik does not have access to the computer power required to perform these calculations nor the specialized competence. Therefore we needed to collaborate with researchers at KTH to be able to carry on these large-scale, computationally demanding calculations and be able to understand the results.
