For more than two decades, QuesTek has been developing and improving its proprietary computational technologies, vetted thermodynamic and kinetic databases of elemental properties and calibrated physics-based predictive property models to identify microstructures needed to meet key properties and ensure enhanced performance. As a team we are experienced in a wide range of material systems including iron, aluminum, copper, cobalt, magnesium, molybdenum, nickel, titanium, niobium, tungsten and property / performance prediction (e.g. yielding, fracture, fatigue, creep, corrosion, and oxidation). These models are used to (i) design, discover, and/or optimize materials and thermal processes and (ii) assess a material’s performance capabilities. See below for more details on QuesTek’s material system modeling capabilities.
QuesTek is also a provider of other ICME-based materials solutions, including materials design and materials process optimization.
Atomistic Simulations
Density Functional Theory (DFT) calculations are used to provide key inputs for CALPHAD assessment and ICME model development. DFT calculations also provide critical data for ICME models
Computational Thermodynamics
The evolution of material’s microstructure is modeled as a function of processing parameters and composition using computational thermodynamics tools such as those based off of CALPHAD (Computer Coupling of Phase Diagrams and Thermochemistry)
Mechanistic Models
QuesTek has developed and utilized mechanistic models that 1) enhance understanding of underlying mechanism that governs materials properties and behavior, and enables QuesTek to design new materials. Our models simulate complex mechanism spanning multiple length and time scales, creating reliable and useful tools to perform materials design.
Machine Learning
QuesTek uses machine learning models for various kinds of studies and ICME-aided materials design workflows, including but not limited to Bayesian inference for uncertainty quantification (UQ), Bayesian optimization (BO), regression models based on neural network (NN), random forest (RF), and SISSO.
