Ready to Work Together?
Aeroengine Design
Aeroengine Innovation with AI-Driven Engineering Simulation
Intractābilis provides next-generation simulation automation tools that transform how aeroengine design and maintenance teams approach computational modelling. By leveraging domain-trained large language models and intelligent multi-agent systems, we enable aerospace engineers to move from concept to computation empowering faster innovation, reduced costs, and broader team enablement across the aeroengine lifecycle.
Automating CFD Workflows
Computational Fluid Dynamics (CFD) is a critical enabler of innovation across sectors including aerospace, automotive, energy, healthcare, and industrial machinery. Despite its importance, CFD continues to be inaccessible to wider population due to its steep learning curve. Successful deployment demands a deep understanding of numerical methods, programming, and the intricacies of domain-specific software tools. Preparing simulation configurations, often through manual editing of input files and command-line debugging, is not only time-consuming but also prone to error.
Simulation configuration further requires specialist knowledge of meshing, boundary conditions, solver settings, and numerical schemes. This reliance on scarce expertise slows innovation cycles and limits CFD’s benefits to a relatively small pool of practitioners.
To address this challenge, Intractābilis has developed an intelligent multi-agent orchestration system capable of transforming natural language instructions into fully configured, executable CFD simulations. The system coordinates four specialised agents: pre-checker, LLM generator, runner, and corrector within a structured workflow that ensures both syntactic compliance and numerical stability.
By fine-tuning a foundational large language model on domain-specific data and deploying it as the core engine, our approach delivers complete end-to-end automation of CFD workflows, enabling faster execution, greater scalability, and wider accessibility for engineers across disciplines.
Hybrid Classical–Quantum Solvers for Large-Scale CFD in Aeroengine Design
In aeroengine design, high-fidelity transient simulations of multi-stage compressor and turbine geometries require extremely fine unstructured meshes, turbulence models incorporating additional transport equations, and time-accurate marching over thousands of steps. Each design iteration demands the repeated solution of millions of degrees of freedom, making the process both computationally intensive and time-consuming.
Intractābilis has addressed this challenge through the development of a hybrid classical-quantum solution, seamlessly integrated into finite element CFD pipeline. The method focuses on solving the linearised systems arising from the pressure correction equation and, in coupled formulations, the block-structured velocity-pressure system. By applying a generator-function decomposition, the assembled stiffness and coupling matrices are expressed as a sum of unitaries, making them directly compatible with a variational quantum linear solver on NISQ hardware.
This approach has materially shortens the aeroengine design cycle, enabling additional simulation-driven optimisation within the same development window. By reducing the number of costly prototype iterations and increasing confidence in final designs prior to rig testing, it delivers both significant cost savings and a competitive advantage in bringing advanced aeroengine architectures to market.
