Case study · Aerospace & defense

Lockheed Martin cuts heat-exchanger design cycles from weeks to minutes with nTop AI

Conventional CFD for a lattice (TPMS) heat exchanger was slow and brittle: meshing failures, long solves and fragile geometry meant weeks to set up one analysis.

Manufacturer
Lockheed Martin
Use cases
1
Evidence
Vendor-reported
Year
2025

Vendor-reported  This result comes from a vendor's case study. Treat figures as claims until you can confirm them with the manufacturer or in a pilot of your own.

The challenge

Conventional CFD for a lattice (TPMS) heat exchanger was slow and brittle: meshing failures, long solves and fragile geometry meant weeks to set up one analysis.

The approach

Engineers built a parameterized implicit model in nTop, ran mesh-free voxel CFD across a Latin Hypercube sample, and trained a neural network and a Fourier Neural Operator to predict performance and full flow fields for AI-driven inverse design.

1 use case in this implementation

  1. Generative design & engineering simulation

    Neural-network surrogates trained on hundreds of simulations predict heat-exchanger performance in seconds and drive inverse design.

    • reduced design cycles from weeks to minutes
    • Over 400 design points were simulated in just 6–8 hours with zero failures.
    • Inference time: 1 second vs. Simulation time: 3600 seconds

Outcomes as reported by the source. We do not adjust, combine or extrapolate reported figures.

Related case studies

Put AI to work

Find the AI opportunities hiding in your plant.

Genvo, our implementation partner, works with manufacturers to pick the right first project, connect the systems you already run and put AI agents and tools into daily use, with people approving what goes out.

Assessments and consultations are provided by Genvo. How our partnership works.