Finding ductile intermetallics: objective and associated parametrization

[dwinston]

@chenweis suggested that a search for ductile materials could be a good goal for this project. I agree. Looking at the elastic-properties data paper, notice the K/G=5 line in Figure 2. The ratio of bulk to shear modulus, aka Pugh's ratio, "has been shown to correlate with ductility in crystalline compounds", and there are only a handful of already-computed materials in our database with "high" Pugh's ratios.

@chenweis also pointed me to a cool letter in Nature Materials from a while ago that not only discusses the potential importance of ductile intermetallic compounds, but also provides guidance towards input-space parameterization that I think could fit well with either evolutionary-algorithm or optimal-learning approaches to finding high-K/G materials.

Let me know what you think.

[HGeerlings]

@dwinston and @chenweis,

I think the intermetallics Nature paper provides a good motivation for the ductile intermetallics problem, and how we can think about setting up design principles in searching for such materials. Given that we can calculate the Reuss-Voigt elastic moduli, searching for high K/G seems like a straightforward yet interesting application.

Obtaining the Pugh ratio from the already existing ~1200 calculated compounds seems trivial, so is it safe to assume this investigation would probe into the larger set of ~8000 materials? Additionally, would our search be confined to only a certain class of materials, such as intermetallics?

Henry

[chenweis]

The scope of materials can be of course expanded beyond intermetallics, but
that depends on the data mining method you are going to use. The
parametrization of materials for data mining can be quite complicated.

On Thu, Jul 9, 2015 at 12:29 PM, HGeerlings notifications@github.com
wrote:

@dwinston https://github.com/dwinston and @chenweis
https://github.com/chenweis,

I think the intermetallics Nature paper provides a good motivation for the
ductile intermetallics problem, and how we can think about setting up
design principles in searching for such materials. Given that we can
calculate the Reuss-Voigt elastic moduli, searching for high K/G seems like
a straightforward yet interesting application.

Obtaining the Pugh ratio from the already existing ~1200 calculated
compounds seems trivial, so is it safe to assume this investigation would
probe into the larger set of ~8000 materials? Additionally, would our
search be confined to only a certain class of materials, such as
intermetallics?

Henry

—
Reply to this email directly or view it on GitHub
#12 (comment)
.