Alloy Space and Phase Stability
Curate compositions and heat-treatment histories; use CALPHAD and targeted DFT to assess phase fields, ordering, segregation, precipitation, metastability, and sensitivity to uncertain thermodynamic descriptions.
AI for Materials
Refractory Metals and Alloys
Physics-informed alloy and process design for extreme‑temperature, high‑load, and irradiation‑relevant environments.
Start Your ProjectOverview
Design decisions for metals that operate near the limits
Refractory metals and alloys based on W, Mo, Ta, Nb, Re, V, Hf, Zr, and related systems combine high melting points with demanding trade-offs in oxidation resistance, density, ductility, irradiation response, joining, and manufacturability.
CD ComputaBio connects client data with thermodynamic assessment, first-principles calculations, process–structure modeling, and uncertainty-aware machine learning. Projects may begin with a target composition family, legacy test data, candidate crystal structures, powder or feedstock specifications, or component-level requirements. Each study defines temperature, atmosphere, stress state, time, and manufacturing route before ranking candidates. Explore the broader AI for Advanced Metals and Ceramic Materials portfolio for adjacent design and process services.
Core Services
Multiscale analysis matched to the decision
Modules can be commissioned independently or combined into a staged program with explicit go/no-go gates.
High-Temperature Property Modeling
Estimate elastic response, thermal expansion, melting-related properties, creep-relevant descriptors, and temperature-dependent stability using fidelity appropriate to the available data and decision threshold.
Defects, Diffusion, and Interfaces
Calculate vacancy energetics, migration pathways, solute interactions, grain-boundary or coating interfaces, and diffusion trends; introduce finite-temperature treatments when harmonic approximations are insufficient.
Composition–Process Optimization
Link composition and processing variables to microstructure and properties for powder metallurgy, sintering, additive manufacturing, thermomechanical processing, and heat treatment.
AI Screening with Uncertainty
Build interpretable surrogate or active-learning models with grouped validation, calibration, applicability-domain checks, and uncertainty-driven selection of the next calculation or experiment.
Multi-Objective Down-Selection
Rank candidates across temperature capability, density, ductility proxies, oxidation or corrosion constraints, raw-material risk, processing windows, and experimental feasibility using transparent Pareto trade-offs.
Workflow
Integrated project workflow
| Stage | Key Activities | Decision Output |
|---|---|---|
| 1. Scope and operating envelope | Define alloy family, service temperature, atmosphere, load, exposure time, manufacturing route, exclusions, and acceptance tests. | Decision criteria and validation plan |
| 2. Data and model audit | Normalize composition, phase, processing, microstructure, test conditions, units, provenance, and fidelity; map missing or conflicting evidence. | Traceable dataset and gap analysis |
| 3. Feasibility filtering | Apply chemistry, phase-stability, density, cost, oxidation, safety, and manufacturing constraints; establish interpretable baselines. | Feasible design space |
| 4. Coupled modeling | Combine CALPHAD, DFT, atomistic simulation, process models, and ML only where each method resolves a decision-critical uncertainty. | Property and process evidence |
| 5. Uncertainty and sensitivity | Quantify model calibration, data coverage, extrapolation risk, parameter sensitivity, and disagreement among methods. | Confidence tiers and risk register |
| 6. Ranking and validation handoff | Construct Pareto shortlists, select controls, recommend synthesis and characterization conditions, and define update rules for new data. | Candidate portfolio and test matrix |
Deliverables
Decision-ready outputs
Curated Data Package
Composition–process–structure–property records with normalized metadata, quality flags, and provenance.
Thermodynamic and Atomistic Evidence
Phase maps, stability assessments, calculation files, and decision-relevant descriptors.
Validated Model Package
Model cards, cross-validation, calibration, applicability domain, and reproducible settings.
Candidate Portfolio
Ranked compositions, Pareto fronts, sensitivity analysis, confidence tiers, and risk flags.
Process Recommendations
Prioritized manufacturing windows, heat treatments, synthesis conditions, and update rules.
Experimental Handoff
Test matrix, controls, characterization priorities, machine-readable results, and expert report.
Applications
Extreme-environment development programs
Aerospace and Propulsion
Down-select alloys, coatings, and process windows for nozzles, hot structures, thermal protection, and high-heat-flux components.
Fusion and Nuclear Systems
Evaluate tungsten- and refractory-alloy concepts under thermal cycling, irradiation-relevant defect evolution, transmutation chemistry, and joining constraints.
High-Temperature Tooling and Furnaces
Balance creep resistance, dimensional stability, oxidation control, contamination risk, and fabrication constraints for long-duration service.
Additive and Powder Manufacturing
Connect feedstock, composition, thermal history, densification, cracking, porosity, and microstructure to a testable processing window.
Scientific Evidence
Open-access examples informing the workflow
Recent studies demonstrate how machine-learned interatomic potentials can extend DFT-level thermodynamics and defect calculations into high-temperature regimes that are otherwise costly to sample.1,2
Computational screening produces hypotheses, not certified material performance. Oxidation, impurities, grain boundaries, texture, residual stress, irradiation history, processing defects, and component geometry require fit-for-purpose experimental validation.
1 Zhu, L.-F.; Körmann, F.; Chen, Q.; Selleby, M.; Neugebauer, J.; Grabowski, B. Accelerating ab initio melting property calculations with machine learning: application to the high entropy alloy TaVCrW. npj Computational Materials 2024, 10, 274. https://doi.org/10.1038/s41524-024-01464-7. Distributed under Open Access license CC BY 4.0, with modification.
2 Zhang, X.; Divinski, S. V.; Grabowski, B. Ab initio machine-learning unveils strong anharmonicity in non-Arrhenius self-diffusion of tungsten. Nature Communications 2025, 16, 394. https://doi.org/10.1038/s41467-024-55759-w. Distributed under Open Access license CC BY 4.0, with modification.
The cited works are used for scientific context; no endorsement is implied.
Project Strategy
Transparent evidence from screening to experiment
Our modular workflow keeps data provenance, model fidelity, uncertainty, and validation handoffs visible. High-cost calculations are reserved for candidates whose ranking depends on them, while out-of-domain predictions are flagged rather than presented as certainty. To discuss an alloy family, operating envelope, manufacturing route, or internal dataset, please Contact Us or submit the Online Inquiry below.
Related Services
Connected metals and ceramics capabilities
Online Inquiry
Submit your project details below, and our team will respond within 24 hours.
x
Need help getting the data you need?
Talk to our technical team about your project!
I Want To Talk