Scenario

A process needs a thermal system — a heat exchanger to add or remove duty, or a drying/curing step with a defined window. Guessing the size wastes capital or throttles throughput. Engineers need a grounded sizing estimate from real thermophysical data.

Worked example. A product stream must be cooled from 120 °C to 40 °C, and the coated part must cure without solvent entrapment.

Agent workflow

1. Read the duty: streams, flows, target temperatures.
2. Thermophysical property lookup — pull heat capacities and transfer properties.
3. Heat exchanger sizing — compute duty, LMTD, and required area.
4. Drying / curing process screen — check the film-formation / drying window under realistic conditions.

Demo output

A sizing sheet: required exchanger area and duty, plus a drying/curing window that avoids skinning or solvent entrapment. The magic moment: the sizing shows a smaller exchanger suffices once the correct fluid properties are used instead of rule-of-thumb defaults.

Deterministic vs LLM

• Deterministic — the heat-transfer sizing math, drying model, and thermophysical lookups.
• LLM — the equipment-selection narrative.

Every value is a validated calculation. The agent reads and routes; the engine decides.

Limits

Sizing-level estimates, not a mechanical or detailed thermal design. It scopes the equipment; detailed engineering finalizes it.