Engineering software for the HVAC industry

What this is

A technology demo of one way to approximate airflow in a room. The room is always U-shaped — you can tweak its dimensions within limits. Three ceiling diffusers are placed in it, and for each you can choose the diffuser / throw pattern type, change its flow rate, and slide it a little along its segment. The result gives a feel for how airflow can be simulated under several different methods.

The Process

The pipeline breaks into four questions:
The Room — what is the geometric representation of the volume?
The Throw Pattern Data — formulas, measurement data, or both?
The Calculation — which numerical method, and how much compute do we spend?
The Visualization — what should the user see?

The Room

In principle, the room can be any shape and size — volume is the more precise word, but rooms are the volumes we care about. The demo uses a U-shape with chamfered inside corners: simple enough to keep the controls minimal, complex enough to make the airflow interesting — multiple flow regimes, dead-ends at the leg tops, and exits on the far end walls.

The Throw Pattern Data

The system is designed to take multiple kinds of input. The idea is freedom of choice: if you have validated formulas, plug them in; if you have laboratory measurements, plug those in; if you have both, mix them. The patterns shipped in this demo are generated on the fly from general formulas — but a measured pattern would slot into the same interface.

New Throw Patterns

The same flexibility applies to new products. If you are building diffuser-selection software, your carefully crafted formulas drop straight in. If you are bringing a new diffuser to market, the laboratory measurements you already collect are all that is needed to show how it behaves — the system consumes either, out of the box.

The Calculation

How the simulation is run depends on the budget — how precise you need to be against how fast you need an answer, and how much memory and compute you are willing to spend. The pipeline is built so it can be simplified for speed or extended for accuracy as needed. The demo exposes two of those choices: a quality tier (two voxel-grid resolutions) and a display mode that ranges from the raw throw pattern, through the same pattern squeezed into the room, to the full CFD reconciliation.

The Visualization

Which visualisation looks best is partly a matter of taste. The point that matters most, though, is whether the occupied zone — the part of the room where people actually stand and sit — sees too much air movement. The demo draws streamlines coloured by temperature with width and opacity scaled to air speed, an iso-surface for air speed, an iso-surface for warm-air reach, and a red draught-risk surface that lights up wherever ISO 7730 says people would feel a draught. Together they make the airflow story easy to read at a glance.

Comfort Assessment

The whole reason for simulating airflow is to find out whether the room is comfortable. The demo answers that with the ISO 7730 Draught Rate — the percentage of people predicted to be dissatisfied by draught at each location — evaluated over the standardised occupied zone (the room floor inset from the walls and capped at 1.8 m head height, with a larger setback from exterior walls). You pick a comfort category — the demo reports the worst draught point and the share of the zone that fails the chosen limit. The flow viewer becomes a design tool: move a slider, see whether your layout meets your category.

Comfort Category

Comfort Category — A, B, or C. Picks the ISO 7730 Draught-Rate limit that the draught-risk surface, the hotspots, and the headline verdict are judged against.

Categories:
A (strict, DR ≤ 10 %)
B (standard, DR ≤ 20 %)
C (lenient, DR ≤ 30 %)

Automatic Selection

With the same simulation data available to a higher-level program, diffusers can be selected and adjusted automatically against the airflow outcome — picking the best diffuser type, position and flow rate for a given room without manual iteration. The demo provides the raw material; an outer optimiser drives it.

Room

The Room section adjusts the room itself — leg lengths, connector length, corridor width, ceiling height, and the inside-corner chamfer. The slider ranges are deliberately limited, but wide enough to show how the rest of the pipeline reacts to a changing volume.

Diffusers

Three diffusers sit on the ceiling: one in the connector at the centre of the U, and one along each leg. For each, the pattern dropdown picks the diffuser type — Swirl, 2-way, or Vertical — the offset slider slides it along its segment, and the flow slider sets its supply-air volume in m³/s.

Calculation

Two quality tiers:
Standard — the default. A coarser voxel grid and a shorter solver budget; still a full simulation, just faster.
Accurate — a finer voxel grid and a longer solver budget. More precise, proportionally slower.

The grid resolution adapts to the room size at each tier, so the small under-door exits stay properly resolved at every slider setting.

Display

The Display section controls what you see, in four layers.

Flow mode — how much of the pipeline runs:
Pattern data — the raw throw pattern of each diffuser, drawn unbounded with the room ignored.
Raw patterns — the same patterns squeezed into the room, before the solver runs.
Simulated — the full CFD reconciliation: how the air actually flows.

Thresholds and density

The air-speed iso-surface threshold (defaults to the ISO 7730 0.2 m/s draught level), the iso-temperature threshold, and the number of streamlines drawn (0 hides them).

Show layers

Toggle each visualisation independently: streamlines, the iso-velocity surface, the iso-temperature surface, the red draught-risk surface, the floor draught-rate heat-map, the draught hotspots, and the occupied-zone outline. A Clip to occupied zone toggle restricts the airflow visualisations to that volume.