Skip to content

Run a simulation

In this lesson you’ll run a thermal study on a design. You’ll start with a basic conduction study, then optionally add fluid coupling and natural convection for a more realistic picture — and see how a fast surrogate can give you a quick estimate before committing to a full solve.

By the end you’ll know how to ask for a thermal result and read it as a field overlay in the app.

Ask your agent:

Run a thermal study on this design, with the heat source on the part and the outer faces held at ambient.

This runs run_thermal, solving for the steady-state temperature throughout the solid. The result is a temperature field across your geometry.

Open the app to see the result. The temperature field is drawn as a color overlay on the model — warmer regions stand out from cooler ones — so you can spot hot spots at a glance and check that heat spreads the way you expect.

A conduction-only study assumes heat just diffuses through the solid. If a fluid (air or coolant) carries heat away, model both together:

Run a conjugate heat-transfer study with the surrounding air as the fluid domain.

This runs simulate_conjugate_heat, solving the solid and the fluid together so you see how the coolant picks up and moves heat.

Step 4 — Add natural convection (optional)

Section titled “Step 4 — Add natural convection (optional)”

If there’s no forced airflow and the part simply sits in still air, buoyancy drives the cooling:

Run a natural-convection study for this part in still air.

This runs simulate_natural_convection, capturing the buoyant flow that rises off warm surfaces.

Full solves are accurate but take time. When you just want a quick number to compare options, ask for a surrogate prediction:

Give me a fast surrogate estimate of the peak temperature.

This runs predict_surrogate, returning a cached, near-instant estimate. Use it to narrow down design choices, then confirm the winner with a full run_thermal study.