Computational fluid dynamics

By - , Build 117

A new BRANZ project combining two different modelling techniques aims to provide more realistic calculations for building scientists.

Figure 2: Temperature distribution in a room as the sun moves from overhead (a) westwards through (b) and (c) shown in a vertical and horizontal plane.
Figure 1: Computational fluid dynamics is used to model air velocity vectors around and between built structures.

In scientific research, computational models are generally less expensive to develop than an experimental programme. A validated model may provide design criteria for an experiment to be set up and monitored for comparison with the predicted modelled system.

This logic translates to real-life building design. One cost-effective way of developing an efficient building design is through the use of computational techniques such as computational fluid dynamics (CFD).

In building science, CFD is the most sophisticated tool available for airflow modelling. It is used in numerous industries, such as medicine and supersonic aircraft design.

In the construction industry, CFD may be used for calculating:

  • airflow over buildings (see Figure 1)
  • smokeflow during fires
  • sizing and placement of plant
  • ventilation paths
  • contamination transport
  • temperature distributions
  • relative humidity.
Figure 1: Computational fluid dynamics is used to model air velocity vectors around and between built structures.

CFD used to study flows

Modelling a building design before construction allows for performance evaluation that will identify the best location of heating or cooling devices, viability of natural ventilation instead of mechanical systems and the effect of solar heating in large glazed areas under clear summer skies.

One strength of using CFD modelling is the ability for a designer to detect and avoid locations of uncomfortable air velocities, temperatures or stagnant air build-up (see Figure 1). However, CFD needs input (boundary conditions) such as heating/cooling loads and surface temperatures, which are assumed by the modeller.

Because CFD modelling is highly sensitive to input supplied by the user, improving the accuracy of the input will provide more realistic results.

Building energy simulation tools

Building energy simulation tools, such as EnergyPlus, are also used to calculate the energy performance and comfort of occupants in buildings. These tools have a well proven track record, and a form of EnergyPlus has been in existence since the early 1980s.

However, these tools differ from CFD models in that they are zonal, meaning each zone comprises one node per room, which is assigned a single temperature (the well mixed approximation). As such, they cannot provide temperature distribution information.

Generally, building energy simulation tools specifically model the energy performance of buildings through simulating the heating, cooling, ventilating and other energy loads. They are commonly used as a design aid, but because air temperature is not uniform throughout any specific room, the well mixed air approximation is a limitation of this type of modelling.

Combining software may mean better results

Industry is moving more towards using CFD in building science. As a result, BRANZ recently started a new research project to develop CFD capability, funded by the Foundation for Research, Science and technology (FRST) and the Building Levy. The result is BRANZ indoor air quality (BRANZIAQ). This is an evolving ‘whole building’ research model that is centred on the coupling of CFD with a building energy simulation tool.

The commercially available PHOENICS was chosen as the CFD software and EnergyPlus, which is available in the public domain, as the building energy simulation tool.

Due to the calculation method, CFD calculates the distribution of temperatures and velocities throughout a room accurately. However, the opposite is true using building energy simulation because surface conditions are known, whereas the distributions in a room are not. Through linking the software packages, the complementary strengths of each may be exploited to produce more realistic results.

So how does it work?

The building energy simulation tool first calculates realistic heat transfer from the walls, ceiling and floor. These results are used in CFD to calculate velocity and temperature distribution. The average room temperature calculated in CFD is compared with the building energy simulation result, and the rate of wall heat transfer is adjusted slightly to give agreement between the programs. Only then may the programs progress to a new time calculation where the process repeats.

Modelling a simple brick building

To demonstrate the coupled system, we modelled a simplified brick building with a flat roof using January weather data for Wellington. The temperature distributions in the vertical and horizontal plane are shown in Figure 2. Plots are spaced at 1¾-hour intervals where the sun is initially at its zenith (Figure 2a). As the sun moves westward, there is a redistribution of temperature (Figures 2b and 2c).

EnergyPlus uses weather data in calculations and supplies the solar heat input to PHOENICS (CFD tool). The thermal redistributions seen in the building can most easily be modelled with PHOENICS. EnergyPlus provides the heat input into the building (which PHOENICS does not) and the resulting wall and roof temperatures are used to drive PHOENICS to calculate the room air temperature distribution.

PHOENICS only has the building structure details and run-time length specification, so without the coupled input, the temperature would remain constant and uniform.

Future development

The completed computational fluid dynamics/building energy simulation coupling provides the foundation for a ‘whole building’ model through linking to other submodels, including biocontaminant, particulate, aerosolisation and appliance models.

The BRANZ project funded over the next 6 years by the Building Levy and FRST to study weathertightness, air quality and ventilation engineering (WAVE – see page 74) will enable these submodels to be developed and incorporated into BRANZIAQ. The result should be a very comprehensive tool for the benefit of industry designers.

Figure 2: Temperature distribution in a room as the sun moves from overhead (a) westwards through (b) and (c) shown in a vertical and horizontal plane.

Download the PDF

More articles about these topics

Articles are correct at the time of publication but may have since become outdated.

Figure 2: Temperature distribution in a room as the sun moves from overhead (a) westwards through (b) and (c) shown in a vertical and horizontal plane.
Figure 1: Computational fluid dynamics is used to model air velocity vectors around and between built structures.

Advertisement

Advertisement