Professional Consulting
Solve a material problem you are facing through professional consulting with Virtual Lab’s material simulation experts.
Virtual Lab professional consulting offers a comprehensive and systematic solution.
In addition to introducing material simulation, experts organize and collect data to analyze. And use machine learning to derive meaningful results.
Design unknown material with minimal cost and time
The most basic structure of atomic units in a material has a decisive influence on the materials properties. By analyzing the structure of the unknown material and predicting its properties, you can design the material rationally.
Analyze the root cause of the problem that occurred in the material
By analyzing the structure and properties of atomic units, you can fundamentally understand problems occurring in materials or physical and chemical phenomena.
Find the variables that lead to optimal performance
You can understand which variables are influencing material design and process optimization. And you can predict the desired goal according to the variable.
Optimize process design using machine learning
By analyzing factory/experimental data and predicting physical properties through machine learning, it is possible to improve yield and optimize processes.
Which calculation answers which question
Consulting problems almost always take one of four shapes, and each calls for a different method.
Root cause of a defect — structure at the atomic scale
When a defect observed macroscopically originates in atomic-scale structure, experiment alone rarely pins down the cause. Calculating defect formation energies, where impurities sit, and bonding at interfaces lets you judge on energetic grounds which structures can actually form. The first step is comparing competing hypothesised structures on one consistent basis and separating the possible from the impossible.
Predicting properties of unknown materials
A composition nobody has synthesised is still tractable computationally. Starting from the electronic structure, predicting band gaps, elastic constants or surface stability narrows the candidate field before synthesis consumes resources. Calculation settings are validated against similar compositions with known measurements first, then extended to the unknown one, and the prediction is delivered with its confidence range attached.
Finding the variables that decide the outcome
What design and process optimisation really need is not one property value but "change this, and how much does that move". Calculating systematically across composition, structure and process conditions produces a sensitivity picture that separates the variables worth controlling in the lab from the ones that can be left alone.
Where machine learning fits
Where plant or experimental data already exists, learning from it is sometimes the faster path than calculating. Collecting and cleaning that data systematically and training a property-prediction model leads into yield improvement and process optimisation. The two are not exclusive — filling sparse regions with calculation and replacing expensive calculation with a learned model is, in practice, the most common combination.
Making the calculation defensible
For a consulting result to carry a downstream decision, the calculation has to be shown to be right on its own terms. We validate the settings against a similar system with known measurements first, run convergence tests on the property of interest, and state which approximations and assumptions were used alongside the result. Predictions come with an uncertainty range wherever one can be established — how much a number moves is more often what a decision needs than the number itself.
Where the questions come from
Batteries and energy materials
Root-cause analysis of capacity fade in electrode materials, ion diffusion pathways, and reactions at the electrolyte interface.
Catalysis and surface science
Adsorption energies and reaction pathways to narrow down active sites, and assessing how a composition change affects activity before it is tried.
Semiconductors and devices
Defect formation energies and electronic structure, quantifying the effect of doping, and interface stability.
Polymers and composites
Molecular dynamics of chain structure and behaviour, mechanical property prediction, and machine-learned property models where data already exists.
Solve various material problems with Virtual Lab consulting
- Semiconductor
- Catalyst
- Thermoelectric
- Polymer
- Steel
- HEA
- Ni-alloy Nanopowders
Meet the Virtual Lab's Consultants
Young-kwang Kim Ph.D.
Professional Consultant
Speciality
- Analyses for materials phenomena using computational techiniques(CALPHAD/MD/DFT/MatCalc/CFD)
- Design and fabrication of new materials using computational techniques and machine learning.
Education
- M. S. / Ph.D. — Dept. of Materials Science and Engineering, Pohang University of Science and Technology (Advisor: Prof. Byeong-Joo Lee)
- B.S. — Dept. of Materials Science and Engineering, Pohang University of Science and Technology
Career
- 2019. 09 – Present — Virtual Lab Professional Consultant / Visiting Researcher, Department of Applied Science, Kyung Hee University
- 2018. 08 – 2019. 08 — Post-doc, Dept. Advanced Aerospace Materials Center, Pohang University of Science and Technology
Jung-a Ryou Ph.D.
B2B Professional Consultant
Speciality
- Computational simulations (DFT/MD)
- Analysis of atomic and electronic structures of organic/inorganic materials
- Research on the development and application of new materials (Semiconductor, Display)
Education
- M. S. / Ph.D. — Dept. of Physics, Sejong University (Advisor: Prof. Suklyun Hong)
- B.S. — Dept. of Physics, Sejong University
Career
- 2020. 12 – Present — Virtual Lab B2B Professional Consultant
- 2018. 01 – 2019. 12 — Post-doc, Korea Advanced Institute of Science and Technology EEWS
- 2014. 10 – 2017. 09 — Post-doc, Quantum Technology Institute, Korea Research Institute of Standards and Science
Materials simulation consulting — frequently asked questions
What kinds of problems does consulting cover?
Root-cause analysis of materials defects, structure analysis and property prediction for unknown materials, and identifying the variables that drive results during design and process optimisation. A problem that arises at the atomic scale is resolved with atomic-scale simulation.
Can a team with no simulation experience commission this?
Yes. We start from deciding which calculations fit the problem, adoption of simulation included, and deliver results in a form your team can act on.
Does it include machine-learning-based process optimisation?
It does. Plant and experimental data are collected and analysed systematically, and machine learning predicts properties to improve yield and optimise the process.
How does consulting differ from adopting Materials Square?
Materials Square is a platform your team runs calculations on. Consulting is our specialists doing the work — problem definition, calculation design, interpretation — and handing over conclusions. Many teams start with consulting and move on to the platform.
Need more information?
Ask Virtual Lab simulation experts for material-related issues. Solve problems by producing optimal results.