Project · Rooms and HVAC airflow

A real clinic floor. 90 real diffusers. Where does the cold air land?

We took a real building model with its HVAC drawn in, and ran the air on a whole floor. Every diffuser sits where the designer put it and blows its design airflow. The rooms are comfortable at head height. At the ankles, the small rooms are draughty.

An indicative, automated first-pass study. Read what that means below.

Plan of the clinic's second floor coloured by air speed 0.1 m above the floor; bright yellow and green patches show air above 0.25 m/s in several small treatment rooms — Air speed at 0.1 m, ankle height. The bright patches are above 0.25 m/s: 15 % of the floor area.

The result

Comfortable at the head. Draughty at the feet.

23.0 °C

Average air temperature in the occupied rooms. Most of the floor sits between 22.5 and 23.5 °C.

93 %

ADPI (ASHRAE 113) over the floor. 56 of 59 occupied rooms reach 80 % or more.

−0.28

PMV (ISO 7730): a touch on the cool side of neutral, inside the comfortable band.

15 %

Of the floor has air faster than 0.25 m/s at ankle height. That is the finding.

Measured over the 59 occupied rooms, half a metre in from the walls.

The finding

In the small rooms, the jet hits the wall and comes back along the floor.

The treatment rooms on this floor are about 3 × 3.5 m, each with a four-way ceiling diffuser. Cold air leaves the diffuser along the ceiling. In a room that small it reaches the wall before it slows down, drops down the wall and runs back across the floor.

The temperature is right and the averages look good. But a patient on a chair or a dentist at the chairside feels it at the ankles. 15 % of the floor has air faster than 0.25 m/s at 0.1 m. That is what you would check on a design review: diffuser type and size, throw, and where it sits in the room.

The building

A real clinic, with its HVAC drawn in.

Building: A real two-storey clinic, 44,800 sq ft, exported from Revit
HVAC in the model: Ducts, VAV boxes, air handlers and diffusers, with the design airflow on each diffuser
Floor studied: Second floor, 18,500 sq ft
Air volume: 100 rooms and corridors, 14,500 sq ft of air
Supply: 90 real ceiling diffusers at their design flows
Extract: 80 return and exhaust grilles in their real positions
Heat: A heat load in every room

The model is the buildingSMART Medical-Dental Clinic, an openly licensed sample. Credit and licence are at the end of this page.

The run

One automated run, model in, maps out.

Mesh: 1.9 million cells
Solver: OpenFOAM, steady, buoyant, 3,000 iterations
Run time: 5.1 hours on our own machine in Bengaluru
Outputs: Temperature, air speed, ADPI, PMV and draught maps, plus a table for every room

The rooms, doors, diffusers, grilles and their airflows were read from the building model. Nobody drew the floor by hand.

The maps

What the floor looks like.

Plan of the clinic's second floor coloured by air temperature at 1.1 m, from 22 °C (blue) to 24 °C (red); most rooms are between 22.5 and 23.5 °C — Air temperature at 1.1 m, seated head height. Occupied rooms average 23.0 °C; most of the floor is between 22.5 and 23.5 °C.

Plan of the second floor marking each point as comfortable, too warm, too cool or draughty under the ASHRAE 113 ADPI test; most of the floor is comfortable — ADPI check at 1.1 m (ASHRAE 113). Green is comfortable. 93 % over the floor; 56 of 59 occupied rooms at 80 % or more.

Second-floor HVAC layout from the building model: supply ducts in blue, return ducts in red, exhaust in purple, the two air handlers in black — The HVAC we simulated, read straight from the building model: supply (blue), return (red) and exhaust (purple) ducts, VAV boxes and diffusers.

What this is, and what it is not

An indicative first pass, not an engineering sign-off.

Coarser mesh. A whole floor at 1.9 million cells is coarser than the level we validate single rooms at.
Derived loads. Each room's heat load was worked back from its design airflow, not taken from a load calculation.
Simple diffuser model. Each diffuser blows its design flow through its face, without the maker's throw data.
Not validated. Rooms and HVAC airflow is not on sale yet. We release it once it passes our validation cases.

For an engineering-grade study we would add:

A finer mesh on a zone. A group of rooms or one open-plan area meshed at the density we validate single rooms at, not a whole floor at a coarser one.
Your real loads. Room loads from your load calculation, or people and watts per room, in place of loads derived from the design airflows.
Your diffusers' data. The make and model of each diffuser, so the jet leaves the ceiling at the throw the maker publishes.
Envelope for perimeter rooms. Window and wall surface temperatures, so rooms on the façade see the sun and the glass.
A mesh check. One representative room run on two meshes, to show the answer does not change with the grid.

What it already shows is useful: where the jets go, which rooms are slow or warm, and where the draught risk is. All from the design data, before anything is built.

Ducts and pipes: live today

Pressure drop and loss coefficient for a duct run, a transition or a fitting, from one STL. PDF report, 3D viewer and the data files.

Start a duct job

Have a floor you want checked?

Rooms and HVAC airflow is coming soon. Send us an IFC with spaces and the HVAC model, or plans with a diffuser schedule, and tell us what you need to know.

Talk to us

Building model credit.“Medical-Dental Clinic” from BSI (2020) “Medical-Dental Test Files,” buildingSMART International, github.com/buildingsmart-community/Community-Sample-Test-Files, licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). Modified: simulated by Coreframe. The second floor was extracted from the IFC model, its rooms joined into one air volume, and its diffusers and grilles used as the airflow inputs for a CFD run. The maps on this page are Coreframe's results, not part of the original files.