Proposal: Materials Modeling

At the time of writing, most of the questions proposed so far fit very well within the scope of existing sites (not only one of them, but two or more). I wonder what makes this proposal unique, so that it will gain enough traction to sustain itself further down the line.

Obviously one would need to attract specialists from these sites (as they know how SE runs), but more importantly people who have never heard of SE. How would that be accomplished, and are there other sources that would be in direct competition?

  • There's a lot of questions here: (1a) how can we be distinguished from ChemistrySE (b) PhysicsSE (c) SciCompSE; (2a) How can we accomplish attracting specialists from other SEs and (2b) from people who have never heard of SE, (3) Are there other sources that would be in direct competition?. I wonder if you would mind if I answer some of these as separate questions, because I worry that my answer will get too long and this can take away from the cogency? – user1271772 Aug 23 at 20:55
  • I had answered a bit about the Computational Science SE on the Chemistry Meta. I would really like to be able to chat with you for a few minutes if you'd be available some time! – user1271772 Aug 23 at 23:06

Nice to meet you Martin! Before making this proposal, I tried to get in touch with you and orthocresol earlier this week on a chat room I made but wasn't able to grab your attention. One year ago we did talk, and you told me that Area51 is deadly. For the last 12 months I've done extensive research before making this proposal. A little bit of that research was revealed in my recent Chemistry Meta post, which I will summarize here:

On Chemistry SE there has been only 296 questions with the materials tag and only one had any of the "computational", or "theory", or "theoretical" tags. This is 0.003% of the questions on Chemistry.

There was also only one question on the Physics SE with both a materials tag and a computational tag. This is 0.0007% of the questions on Physics.

In contrast:
- There has been at least 1007 materials modeling questions about VASP asked on ResearchGate,
- 15,000+ questions asked on the VASP forum in 14 years,
- 2 questions/week on the VASP facebook group, which over 14 years would be ~1456 questions.

The story is similar for LAMMPS, CPMD, CP2K, ABINT, CASTEP, DFTB+, and QE. Why we should have a unified SE for general questions of interest to each of those (related) communities, rather than signing up separately for 7 different forums and asking the same thing in 7 different places, is something I'd like to save for a separate question, so that this answer doesn't become too long.

If pushed to prove it to you, I'm sure I can find enough evidence to convince you that there's more than 100,000 materials modeling questions on the internet right now, all spread across very disparate forums, Q/A sites, and Facebook pages, where users of one medium may not even know about the existence of some of the others.

100k questions would put this SE in the top 15 out of all SE sites, yet neither ChemistrySE nor PhysicsSE have succeeded in attracting this audience as we seem to constitute < 0.005% of the questions on either of those SEs. This is part of why I think we need a Materials Modeling SE: For most of the proposed questions so far, how do I decide whether to ask it on PhysicsSE or ChemistrySE or on some Facebook group or other non-SE website because of fear that not enough Materials Scientists or Materials Engineers will see it? Do I roll a 6-sided die and 1/6 of the time ask on PhysicsSE, 1/6 of the time ask on ChemistrySE, 1/6 of the time ask on a MaterialsModeling facebook group, 1/6 of the time ask on some forum, etc? A similar dilema came up for Quantum Computing, since questions would be valid on Physics, or CS, or Theoretical CS, or Mathematics, or MathOverFlow, or Engineering, or Electrical Engineering, or Information Security, etc. QCSE ended up becoming one of the fastest growing proposals!

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .